PATTERNS OF COEVOLUTION BETWEEN SYMBIOTIC FUNGI, BACTERIA AND SCOLYTINE

By

CRAIG C. BATEMAN

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2018

© 2018 Craig C. Bateman

This dissertation is dedicated to Chris DiFonzo

3

ACKNOWLEDGMENTS

I thank my committee members, Jiri Hulcr, Andrea Lucky, Kirsten Pelz-Stelinski, and Jason Smith for direction and instruction. I thank my senior Hulcr lab members

Caroline Storer, Andrew Johnson, James Skelton, Michelle Jusino, Martin Kostovcik, and Adam Black for their advisement, moral support, and collaboration. I thank my many collaborators at other institutions for their invaluable help in collecting and sharing of specimens and information, including Wang Bo, Li Lili, Liu Guangyu, Wissut

Sittichaya, Park Ji-Hyun, Kim Mee Sook, Park Sangwook, Kim Moo-Sung, Malena

Martinez, Jessenia Rosanna Castro Olaya, Rachel Osborn, Sarah Smith, Matthew

Kasson, Chase Mayers, Anthony Cognato, and Miroslav Kolarik. I thank my parents, siblings, and friends for moral support, including Lynna Durst, Bryan Tarbox, Ashley

Chapman, Verity Salmon, Hans Goertz, Ash Albinson, Anand Roopsind, Kyle Kehus,

Sami Rifai, and Camille Truong.

4

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 7

LIST OF FIGURES ...... 8

ABSTRACT ...... 9

CHAPTER

1 INTRODUCTION ...... 11

2 AMBROSIA PREMNOBIUS CAVIPENNIS (SCOLYTINAE: IPINI) CARRIES HIGHLY DIVERGENT ASCOMYCOTAN AMBROSIA , AFRORAFFAELEA AMBROSIAE GEN. NOV. SP. NOV. (OPHIOSTOMATALES) ...... 18

Methods ...... 21 Quantitative Culturing of Fungi from Beetle Mycangia ...... 21 DNA Extraction, Amplification, Library Preparation, and Sequencing ...... 22 Phylogenetic Analyses ...... 23 Mycangium Presence and Content ...... 24 Fungal Morphology on Agar Medium ...... 26 Results ...... 26 Culturing of Fungi from Beetle Mycangia...... 26 Sequencing and Phylogenetic Analysis ...... 27 Fungal Community Amplicon Analysis ...... 27 Examination of Mycangia ...... 28 ...... 28 Discussion ...... 30

3 PATTERNS IN FUNGAL SYMBIONT FIDELITY AND DIVERSITY ACROSS TEN EVOLUTIONARY ORIGINS OF THE AMBROSIA SYMBIOSIS IN BEETLES (COLEOPTERA, : SCOLYTINAE) ...... 42

Methods ...... 46 Sampling and Dissection ...... 46 DNA Extraction, Amplification, Library Preparation, and Sequencing ...... 47 Data Processing ...... 48 Statistical Analyses ...... 49 Phylogenetic Analyses ...... 50 Results ...... 51 Fungal Community Sequencing ...... 51 Quantitative Culturing ...... 52

5

Beetle-fungus Associations ...... 52 Discussion ...... 54 Symbiont Evolution and Fidelity ...... 54 Predictors of Primary Symbiont Communities ...... 55 Considerations for Fungal Community Sequencing ...... 58 Final Remarks ...... 59

4 INBREEDING AND HAPLODIPLOIDY IN BARK AND AMBROSIA BEETLES (COLEOPTERA: CUCRCULIONIDAE: SCOLYTINAE) IS NOT CORRELATED WITH THE PRESENCE OF INTRACELLULAR BACTERIA ...... 71

Methods ...... 74 Taxon Sampling ...... 74 DNA Extraction and PCR ...... 75 Sample Pooling and High-Throughput Sequencing ...... 76 Data Processing ...... 77 Phylogenetic Analysis of the Beetles ...... 78 Results ...... 79 Sequencing ...... 79 Potential Reproductive Manipulators ...... 79 Discussion ...... 80

5 CONCLUSIONS ...... 92

LIST OF REFERENCES ...... 95

BIOGRAPHICAL SKETCH ...... 108

6

LIST OF TABLES

Table page

2-1 of and Xylariales from used for phylogenetic analyses ...... 38

3-1 Summary of collection information for beetles sampled...... 64

3-2 Summary of associations between beetle species and OTUs with taxonomy assignment. The OTUs included were significantly associated (p<.02) by the Indicator Species Analysis. Beetles marked with * were only represented by a single specimen...... 67

3-3 Prevalence (% of specimens) and abundance (mean colony forming units) of fungi isolated from ambrosia beetle mycangia...... 69

3-4 Results of the permutations-based multivariate analysis of variance (PERMANOVA) showing the effect of fungal symbiont presence on host ecology...... 70

4-1 Taxon sampling with sex-determination systems and inbreeding character states. Pairs of phylogenetically related species differing in reproductive system are shown in alternating background colours...... 85

4-2 Collection localities and number of specimens included for all samples used in miseq community sequencing...... 86

4-3 Taxa used in the 28S phylogenetic analysis of the beetles, with accompanying GenBank accession codes ...... 88

4-4 Results of the permutations-based multivariate analysis of variance (PERMANOVA) showing the effect of intracellular bacteria presence on inbreeding in scolytine beetles...... 89

7

LIST OF FIGURES

Figure page

2-1 Best ML tree from GARLI analysis of 18S, 28S rDNA, and βT (introns 3/4/5 removed) data matrix of Sordariomycetidae with Xylariales outgroup. Values at nodes represent ML bootstrap percentages ...... 34

2-2 A-B. Colony on PDA at 8th day. C-D. Colony on PDA after 14 d. E-F. Aerial hyphae, showing intercalary melanization and denticulate distal hyphae. G-H. Immersed hyphae. I-J. Monillioid hyphae...... 35

2-3 Transverse cross sections of adult female Premnobius cavipennis head showing the content of the paired mycangia ...... 36

2-4 Micro-CT scan of adult female Premnobius cavipennis showing a three- dimensional mycangium structure which opens to the pharynx ...... 37

3-1 Average read abundance of OTUs by taxonomy assignment, per beetle . Fungal taxa only listed and highlighted in color if a significant associate in the Indicator Species Analysis...... 61

3-2 Cophylogeny of ambrosia beetle genera and filamentous ambrosia fungus genera with non-ambrosial outgroups. Cladograms inferred from Johnson et al. (2018) and Vanderpool et al. (2018) ...... 62

3-3 A nonmetric multidimensional scaling (NMDS) ordination of obligate associations between ambrosia beetles and filamentous ambrosia fungi...... 63

4-1 Sampling curve for species richness indicating the completeness of sampling effort. Dotted lines indicate extrapolated data, with colored ranges represented 95% confidence intervals...... 90

4-2 Beetle phylogenetic relationships inferred from 28S rRNA sequence data in RAxML, constrained to the topology of the wider scolytine phylogeny of Johnson et al. (2018) shown adjacent to incidence...... 91

8

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

PATTERNS OF COEVOLUTION BETWEEN SYMBIOTIC FUNGI, BACTERIA AND SCOLYTINE BEETLES

By

Craig C. Bateman

August 2018

Chair: Jiri Hulcr Major: Entomology and Nematology

The ambrosia symbiosis of fungus-farming beetles is one of the most economically and ecologically important symbioses in forests, but it remains poorly characterized across its evolutionary diversity. Here, we use quantitative culturing and high-throughput sequencing to characterize the fungal mutualists and intracellular bacterial endosymbionts. We test three cental hypotheses targeting the invasion ecology, symbiont fidelity, and reproduction of the ambrosia symbiosis. Results indicate the mycangial symbionts from the ambrosia beetle Premnobius cavipennis were retained across multiple introductions to new regions. The fungal symbionts are then reported from 78 ambrosia beetle species, comprising nine de novo evolutionary origins of the symbiosis, and raising the total number of de novo origins to 12 in Fungi. The intracellular bacterial endosymbionts are reported from 14 beetle species, comprising seven inbreeding lineages and their outbreeding sister groups. These results suggest that different clades of ambrosia beetles each have their own patterns in symbiont maintenance. Transmission mechanism, host ecology, and indirect community interactions all are likely to play a role in mediating symbiont community composition.

9

Results from bacterial community sequencing suggest that intracellular bacterial endosymbionts are not causing alternative reproductive systems at a broad scale, but may be causing Paternal Genome Elimination in , and haplodiploidy in

Premnobius. The presented characterization of the ambrosia symbiosis provides a basis for testing interactions experimentally and provides recommendations for predicting and managing ambrosia symbioses as pests.

10

CHAPTER 1 INTRODUCTION

The success of all plant and on earth can be tied to symbiotic interactions with microorganisms like bacteria and fungi. One of the most striking examples dates back perhaps billions of years (Gray et al., 1999), as ancient eukaryotes formed symbioses with microbes that were eventually incorporated into organelles like mitochondria and plastids (Martin et al., 2001). However, microbial symbioses aren’t so rare that they form only once every few millennia. Persistent interactions with microbes can form in a single generation (Hom and Murray, 2014), and their function can depend heavily on context, spanning a continuum of beneficial and harmful functions (Zug and

Hammerstein, 2015; Toju et al., 2017). The pervasion of symbiosis is a relatively recent discovery and has spurned many hypotheses to explain how symbioses persist and function (Casadevall and Pirofski, 2000). Most symbioses are still poorly characterized due to limitations in methodology and tractability of experimental systems, although sequencing methodologies have been a significant advancement.

Contemporary understanding of microbial symbioses has been revolutionized by inventions in sequencing technologies in three defining ways. First, Sanger sequencing allowed for more accurate identification of symbionts and reconstruction of cophylogenies. This advancement helped show that obligate symbioses do not always result from vertical transmission (Genkai-Kato and Yamamaura, 1999; Kikuchi et al.,

2007), and highlighted the role of communities in symbiont evolution (Christian et al.,

2014). Second, community sequencing methods have shown that symbioses are commonplace, and frequently include cryptic diversity (Herre et al., 1999; Prada et al.,

2014), even some taxa that are still only known from indirect observations (Grossart et

11

al., 2016). Finally, high-throughput sequencing (HTS) technologies have revealed that symbiont genomes can change dramatically, including losses and additions to genome size and gene functions (McCutcheon and Moran, 2012). Taken together, sequencing studies repeatedly show that symbiotic interactions have a pervasive effect in evolution.

They also increasingly portray symbioses within a metacommunity framework, and present new questions surrounding the role of host dispersal and indirect effects on symbiont community composition.

As the most successful animal group in terms of species diversity, are ideal model systems for studying the evolution of eukaryotic and microbial symbioses.

All insects are associated with microbes, and their associations can have diverse consequences on phenotype. Some of the most well-studied examples include the bacterial symbionts of aphids. An obligate or primary symbiont of aphids, Buchnera aphidicola, is housed in coadapted insect organs called bacteriocytes, and supplements the imbalanced aphid diet of plant sap (Munson et al., 1991; Brinza et al., 2009). The aphid microbiome also includes other, less frequently encountered symbionts, including

Hamiltonella, which provides protection against parasitoids, and Serratia symbiotica, which can improve an aphid’s tolerance to heat stress (Oliver et al., 2005; Lamelas et al., 2011). This diverse community of symbionts in aphids isn’t unusual. Most insect- microbe symbioses are multi-membered. The hindgut of some cockroaches and termites are composed of an especially diverse community of protists, bacteria, and archea that are capable of degrading lignocellulose and other nutrient-imbalanced substrates (Cruden and Markovetz, 1987; Brune and Ohkuma, 2010). Even as model

12

systems, these symbioses exhibit extensive complexity community structure and mechanisms for symbiont maintenance.

Insects and their microbial partners can mediate their association from diverse communities using either vertical and horizontal transmission. As an example, vertical transmission commonly occurs with intracellular bacteria in the genus Wolbachia, which are maternally inherited and found in at least half of all insect species (Hilgenboecker et al., 2008). Horizontal transmission prevails among fungi associated with fungus-farming insects even though the fidelity of their association can match that of vertically transmitted symbionts (Mueller et al., 2005). Although vertical and horizontal transmission are generally independent of symbiont fidelity, the mechanism mediating each transmission type depends on where the association occurs with the host.

Endosymbionts of insects are most likely to be mediated by their host. For example, most or all insect endosymbionts must bypass the insect innate immune system by resisting effectors like anti-microbial peptides or lacking host immune elicitors (Login et al., 2011; Douglas et al., 2014). Ectosymbionts are more likely to be mediated by microbial community interactions and indirect or stochastic processes (Douglas and

Adair, 2017). But these generalizations don’t always work- in some cases it’s not always clear which organism is the host, which is the symbiont, and where they occur in relation to each other. And in some cases, symbiont communities are so diverse, it has been difficult to establish which are the symbiont taxa, let alone how they function or are maintained. The ambrosia symbiosis is an especially pertinent example.

The ambrosia symbiosis includes wood-boring beetles and fungal mutualists, although the interaction is so diverse, it has been difficult to define its members. The

13

interaction was described even before the term symbiosis was invented (DeBary 1879), when Schmidberger (1836) noticed xylem-boring beetles feeding on an unknown substance, later identified as fungi. It wasn’t for another 80 years that the first beetle mycangium, a specialized organ, was illustrated (Beeson 1916), and the function of mycangia still wasn’t known for another 35 years, when Nunberg (1951) first hypothesized that mycangia were used for storing and transporting fungal tissue to new tree hosts. The discovery of mycangia was crucial because it provided evidence that ambrosia beetles and fungi are coadapted mutualists. Ambrosia fungi also exhibit traits suggesting a shared coevolutionary history with beetles, including swollen conidia that serve as a food “crop” for beetles infesting a nutrient-imbalanced diet of xylem (Batra,

1963). Coevolutionary studies of ambrosia beetles and fungi have progressed slowly beyond this basic conceptual understanding. Ambrosia fungi are morphologically convergent and cryptic, which has made it difficult to simply identify the distribution of symbionts across beetle diversity (Alamouti et al., 2009). Furthermore, the ambrosia symbiosis includes beetles with diverse ecologies that inhabit diverse wood- decomposition communities, making it difficult to couple symbiont associations with other traits. Bacteria have been implicated in the mediating traits and symbiont composition of ambrosia beetles and fungi (García‐Fraile, 2017; Biedermann and Marko

Rohlfs, 2017). Trees, mites, nematodes, and other insects and fungi are also persistent members of tree decomposition communities that interact with ambrosia symbionts, and probably influence coevolutionary patterns (Susoy and Herrmann, 2014; Vissa and

Hofstetter, 2017). Such complex communities have made it difficult to establish the identity and function of organisms participating in the ambrosia symbiosis, but further

14

investigation is warranted because the symbiosis presents a unique scenario to examine how multi-member symbioses evolve.

The ambrosia symbiosis is an ideal model system for exploring how symbiont communities are formed and mediated because the symbiosis has exceptional evolutionary diversity co-occurring with important convergent traits. The ambrosia symbiosis has evolved de novo at least 12 times in one subfamily of Scolytine beetles, comprising over 3,000 species (Jordal and Cognato, 2012; Johnson et al., 2018).

Alternative mating and sex-determination systems have also evolved at least seven times in this group. Such high diversity is useful because it is highly amenable to a comparative framework, providing evolutionary controls to test the influence of environment on coevolutionary outcomes. As a model system, it even has the potential to benefit human agriculture and medicine. Just like fungus farming ants and termites, ambrosia beetles may utilize beneficial microbiomes to maintain a monocultural fungus crop (Mueller and Gerardo, 2002) and understanding their strategy could improve human agricultural practices (Mueller et al., 2005), or provide new antibiotics

(Kaltenpoth, 2009; Barke et al., 2010). Another reason the ambrosia symbiosis is an ideal system to examine the evolution of eukaryotic symbioses is its global relevance.

The symbiosis co-occurs in all biomes with trees and has occasionally caused economic and ecological destruction when transported to non-native regions (Ploetz et al., 2013). Ambrosia fungi, which includes some plant pathogens, can pose the greatest threat, and therefore they need to be well known to make reliable predictions about which beetles are spreading plant pathogens. Furthermore, characterizing the bacterial communities of these beetles could yield novel control strategies (Douglas, 2007).

15

The ambrosia symbiosis requires a more thorough characterization before it can be used as a model system. In the present study, I test the three hypotheses, represented as chapters, that will shed light onto three key aspects of the amborsia beetle ecology and evolution: invasion ecology, symbiont specificiaty, and reproduction:

1. H0: The symbiont composition of invasive beetles has not changed following

introductions into new regions. I study this question using Premnobius

cavipennis.

H1: The symbiont composition of P. cavipennis has changed in identity or

diversity following introductions into new regions.

2. H0: The composition of fungal mutualists in ambrosia beetle mycangia are not

correlated to any beetle traits

H1: The composition of fungal mutualists in ambrosia beetle mycangia are

predicted by either:

i. Mycangium type

ii. Evolutionary origin

iii. Geography

iv. Host ecology

v. Community interactions

3. H0: The presence of intracellular bacteria in scolytine beetles is not correlated

with reproductive mode or sex-determination system

H1: The presence of intracellular bacteria in scolytine beetles is correlated

with inbreeding or alternative reproductive systems

16

I test these hypotheses by using an integrated approach of community metabarcoding, culturing, and histology, and establish the fungal and intracellular bacterial symbionts of ambrosia beetles from broad evolutionary histories (nine de novo origins of fungus farming and seven de novo origins of alternative mating systems). This design has made it possible to rigorously test the correlation between symbiont communities and host traits to determine how symbionts are maintained and interacting with host phenotype. In doing so, I will advance the field of symbiology while providing predictive power able to improve management of harmful invasive ambrosia symbiosis.

17

CHAPTER 2 AMBROSIA BEETLE PREMNOBIUS CAVIPENNIS (SCOLYTINAE: IPINI) CARRIES HIGHLY DIVERGENT ASCOMYCOTAN AMBROSIA FUNGUS, AFRORAFFAELEA AMBROSIAE GEN. NOV. SP. NOV. (OPHIOSTOMATALES)1

Ambrosia beetles (Coleoptera: Curculionidae: Scolytinae and Platypodinae) are diverse, with over 3200 species and at least 13 independently evolved origins (Farrell et al., 2001; Jordal and Cognato, 2012; Kirkendall et al., 2015). Their fungal mutualists, which are cultivated for food inside trees, are poorly understood by comparison. At least seven independent ambrosial origins in fungi are known (Harrington et al., 2010; Endoh et al., 2011; Kasson et al., 2013; Dreaden et al., 2014; Li et al., 2015; Mayers et al.,

2015), but less than 1% of ambrosia beetle symbionts have been resolved using DNA typing. Most of the ambrosia fungi that are known have only been recorded from non- indigenous (exotic) pest beetle populations while the majority of native and economically insignificant beetles remain unstudied. The paucity of information on symbiont identity and ecology has been a hurdle for researchers trying to predict habits of invasive ambrosia communities (Carrillo et al., 2014), test coevolutionary theories

(Mueller and Gerardo, 2002), or generate insights for improving human agriculture

(Mueller et al., 2005).

In the past century, international trade has brought more than 60 non-native scolytine species to the United States, with exotic wood-boring beetles being one of the most common and increasingly-intercepted insects at U.S. ports of travel (Haack et al.,

2014; Smith and Cognato, 2015). The arrival frequency of such imports, and the potential introduction of virulent tree pathogens, make research on exotic ambrosia

1 Printed with permission from Elsiver, article https://doi.org/10.1016/j.funeco.2016.10.008

18

fungi urgent. In some cases, prior data on the identity and ecology of ambrosia fungi could prevent ecological and economic damage. Around 2002, the redbay ambrosia beetle glabratus was introduced to the U.S., and its pathogenic fungal symbiont Raffaelea lauricola began causing mass decline of redbay trees (Persea borbonia) and other lauraceous plants (Fraedrich et al., 2008). Because the fungus was new to science and its interaction with host trees was unknown at the time of its discovery, no preventative action was triggered. In other cases, illumination of symbiont identity and ecological habits can inspire future research. The newly discovered basidiomycete Flavodon ambrosius transported by Ambrosiodmus and Ambrosiophilus ambrosia beetles allows the beetle to reach much larger colonies and persist in wood much longer compared to other ambrosia beetles carrying ascomycetes (Li et al., 2015;

Kasson et al., 2016; Simmons et al., 2016b). Although this symbiosis does not warrant phytosanitary concern, it does have a significant potential for research. In all cases, symbiont identification is the first step to making predictions about the behavior, function and potential impact of the symbiosis.

The subtribe Premnobiini is one example of an entire ambrosia beetle clade whose fungal symbionts have never been identified. Aside from 26 beetle species' descriptions and some notes on gallery construction and inbreeding, very little is known about Premnobiini beetles (Browne, 1961; Wood, 1982; Cognato, 2013). The two constituent genera Premnobius and Premnophilus have evolved and diversified in

Africa. The most well-known species, Premnobius cavipennis, is now globally distributed in tropical biomes (Wood, 1982; Bright and Skidmore, 1997). The African distribution of the majority of Premnobiini suggests that P. cavipennis attained its

19

pantropical distribution by human assistance. Records indicate P. cavipennis has been present in the neotropics since at least 1950 (Browne, 1961). There may have been multiple introductions since the morphologically uniform species has multiple distinct genotypes (Cognato, 2013). It is also possible that P. cavipennis comprise several cryptic species as in the ambrosia beetle Euwallacea fornicatus (O'Donnell et al., 2015), and that a pantropical distribution was attained through ancient dispersals, similar to some Xyleborus ambrosia beetles (Gohli et al., 2016). Sampling of P. cavipennis fungal symbionts could help uncover the beetle's biogeographic history, and help explain its success at colonizing several continents.

The phylogenetic placement of Premnobius within the tribe Ipini also makes it a useful system for examining the transition from phloem feeding to obligate symbiosis.

Premnobiini ambrosia beetles are polyphagous in terms of host trees (Wood, 1982), but may carry highly specific fungal symbionts like most other ambrosia beetles. In contrast, most species from the parental clade Ipini are bark beetles, highly specific to their host trees (Wood, 1982), with little or no known relationship with fungi, and the fungi isolated from them are usually non-specific commensals (Yamaoka et al., 1997; Kopper et al.,

2004). This indicates that Premobiini are an independent origin of the ambrosia symbiosis within a clade (Alamouti et al., 2007, 2009; Lieutier et al., 2016).

A comparative analysis of Premnobiini and its sister taxa, and a contrasting analysis of similarly evolved ambrosia clades, could illuminate how these specific associations evolve. Premnobius is also a useful lineage to investigate symbiont maintenance because beetles in this genus possess a pharyngeal mycangium, which appears to be similar to the pre-oral mycangia found in the more distantly related ambrosia

20

beetles (Schedl, 1962; Francke-Grosmann, 1967). Whether or not the two mycangial types are homologous is not known. If the two mycangia represent a case of morphological convergence, comparisons of their function and structure may provide insights into the mechanisms of control over symbiont communities.

In this study, we sought to identify the fungal symbionts within the mycangium of

P. cavipennis using fungal culturing, community DNA metabarcoding, and histology. To strengthen evidence for the new association, we also updated the description of the pharyngeal mycangium where the fungi are stored, which has only been documented by a single sketch from more than 50 years ago (Schedl, 1962).

Methods

Quantitative Culturing of Fungi from Beetle Mycangia

Seven adult female P. cavipennis specimens used for fungal isolation and culturing were collected from an ethanol-baited light trap at the Montgomery Botanical

Center in Coral Gables, FL on August 12th, 2015. The specimens were stored alive at room temperature with water-moistened sterile paper towels for 2 d after collection.

Immediately preceding fungal isolation from mycangia, live beetles were washed by vortexing for 10 s in a sterile solution of 1 mL water and one drop of Tween 80. A second wash was performed using a solution of only sterile water. To compare fungal communities in different body parts, washed and dried beetles were held with forceps under a dissecting microscope while the head, thorax, and abdomen was separated using a sterile scalpel as previously described (Fraedrich et al., 2008; Kasson et al.,

2013). The three respective body segments were each transferred into 2 mL microcentrifuge tubes containing 1 mL of PBS (phosphate buffer solution), and crushed using sterilized micropestels. Each tube containing a macerated body segment was

21

then serially diluted and plated at concentrations of 1/10, 1/100, and 1/1000 on PDA

(potato dextrose agar) media. Fungi were allowed to grow at 25 °C for 5–7 d before sub-culturing. Representative subcultures of the dominant morphotype recovered from heads were cryopreserved on PDA slants submerged in 15% glycerol stored in the −80

°C Hulcr lab collection at the University of Florida (UF) and the Centraalbureau voor

Schimmelcultures (CBS), Utrecht, The Netherlands.

DNA Extraction, Amplification, Library Preparation, and Sequencing

Four fungal isolates recovered from heads of Florida-collected P. cavipennis were included in DNA studies. Extraction of genomic DNA from fungal cultures was performed by scraping approximately 5–10 μg of mycelium from pure cultures and adding it to 20 μL extraction solution from the Extract-N-Amplify Plant PCR kit (Sigma-

Aldrich). Samples were then incubated at 96 °C for 15 min. Following incubation, 20 μL of 3% BSA (bovine serum albumin) solution was added, vortexed, and spun down. The upper 20 μL of supernatant was used as the PCR template.

PCR amplification for Sanger sequencing was performed on portions of the 18S small subunit and 28S large subunit ribosomal DNA (rDNA) loci using the primer pair

NS1/NS4 (White et al., 1990) and LR0R/LR5 (Vilgalys and Hester, 1990) and on a portion of the β-tubulin (βT) coding gene using Bt2a/Bt2b (O'Donnell and Cigelnik,

1997). Final PCR volumes of 25 μL consisted of 1.25U Taq Polymerase (New England

Biolabs), 200 μM each dNTP, 10 pmol of each primer, and 5% DMSO (V/V). Amplified products were cleaned using the ExoSAP-IT (Affymetrix Inc.) kit according to manufacturer's instructions. Sequencing was performed on ABI PRISM 377XL or

3130XL by the Interdisciplinary Center for Biotechnology Research (ICBR) at the

University of Florida (UF).

22

In addition to Sanger sequencing of live cultures, a complementary approach of community metabarcoding of preserved mycangia was utilized. For DNA extraction from mycangial communities, one cryopreserved P. cavipennis specimen collected during

2014 in Java, and two specimens collected during 2013 in Honduras, were used. The beetle heads were first aseptically removed and then crushed using a micropestle. DNA from each macerated head was extracted using the UltraClean Microbial DNA (MoBio) kit following the manufacturer's protocol. The MiSeq library was constructed via two

PCR steps (O'Donnell et al., 2016). The first-step 28S marker PCR used the template primers LR0R and JH-LSU-369rc (Li et al., 2015), and the following cycling conditions:

95 °C-6 min, (94 °C-30 s, 52 °C-45 s, 72 °C-1:30 min) 25x, 72 °C-5 min. This PCR was repeated using 5 μL PCR amplicon in place of DNA template. A second step PCR using

1 μL of this library was used in a final PCR for eight cycles to anneal Illumina sequencing adapters following the Illumina metabarcoding protocol. DNA concentrations from each library were measured using PicoGreen fluorescent dye from the Quant-iT kit on a Realplex cycler (Eppendorf), and samples were made equimolar at a concentration of 23 ng µL−1 prior to pooling. Pooled libraries were cleaned using the

AMPure XP kit. A 15% PhiX library was added. Paired-end fungal community sequencing was performed at 300 cycles/read in an Illumina MiSeq at the ICBR, at UF.

Metabarcoding sequence data was analyzed in QIIME 1.8.0 (Caporaso et al., 2010), and raw sequence data were deposited in NCBI Sequence Read Archive under accession number SRP072278.

Phylogenetic Analyses

Sequences of 18S rDNA, 28S rDNA, and βT with introns 3/4/5 removed, were aligned in Geneious (Geneious version 9.1.5). The alignment was divided into five

23

partitions for phylogenetic consideration: one partition for each rDNA locus and for each of the three codon positions in the protein encoding βT. The Akaike information criterion in jModelTest 0.1.1 (Guindon and Gascuel, 2003; Posada, 2008) was used to select the nucleotide substitution model for each partition, and analyses continued with a concatenated data matrix. Maximum likelihood (ML) phylogenetic analyses were conducted in GARLI 2.01 (Zwickl, 2006) with the recommended partition parameters to determine the best tree topology and bootstrap support values from 500 search replicates, which were summarized in SumTrees (Sukumaran and Holder, 2010).

Bayesian posterior probabilities (BPP) were estimated with the same partition parameters in an analysis conducted in MrBayes 3.1.2 (Ronquist and Huelsenbeck,

2003), in which two runs of four chains each were executed simultaneously for

5,000,000 generations, with sampling every 500 generations. SumTrees was used to compute BPP from a summary of 7501 trees retained after a burn-in of the first 2500 trees collected. Bootstrap support was assessed in GARLI on the University of Florida

HiPerGator 2.0.

Mycangium Presence and Content

Adult female P. cavipennis specimens first underwent exploratory dissections to investigate the presence of a mycangium or fungal masses suggestive of propagule sequestration. Specimens were collected from ethanol baited traps in Honduras and stored in the Hulcr lab cryopreserved collection until their use. Only heads were examined based on a previous report of the cephalic mycangia (Schedl, 1962), and based on a single, abundant morphotype recovered from all beetle heads used in fungal culturing. Fungal masses found inside the beetle's head were transferred to a

24

microscope slide and observed under oil emersion in a compound microscope to document the presence and morphology of fungal cells.

Five additional female P. cavipennis specimens from the same collection were assessed for the presence of mycangia using a microtome. Heads were prepared as previously described by Li et al. (2015) and Kasson et al. (2016). Heads were orientated with the most anterior end facing downward to permit transverse sectioning and simultaneous bilateral visualization of the mycangia. 10 μm transverse sections were cut using a Microm HM 325 rotary microtome (Walldorf, Germany). Selected slides confirmed by immediate viewing were dried at 60 °C for 24 h, double-stained with Harris hematoxylin and eosin-phloxine by hand, and examined and photographed using a

Nikon Eclipse E600 compound microscope (Nikon Instruments, Melville, New York) equipped with a Nikon Digital Sight DS-Ri1 high-resolution microscope camera and

Nikon NIS-Elements BR 3.2 imaging software.

For a non-destructive 3-D visualization of the mycangia, an adult female was scanned with a Phoenix v|tome|x M (GE's Measurement & Control business, Boston,

USA) at the University of Florida's Nanoscale Research Facility. A 180 kV X-ray tube with a diamond-tungsten target was used with the following settings: 50 kV, 200 mA, a 2 s detector time, averaging of 3 images per rotation and a resulting voxel resolution of

1.89 μm. Raw X-ray data were processed using GE's proprietary datos|x software v 2.3 to produce a series of tomogram images. These Micro-CT image stacks were then viewed, sectioned, measured and analyzed using VG StudioMAX 3.0 (Volume

Graphics, Heidelberg, Germany). Final figures were prepared with Photoshop and

Illustrator (CS5, Adobe).

25

Fungal Morphology on Agar Medium

To determine optimal growth temperature, four subcultures of the dominant fungal morphotype of P. cavipennis were grown concurrently on PDA at 15, 20, 25, 30, and 35 °C. In attempts to induce sporulation, additional culturing was performed at 25

°C on various nutrient agars used in the cultivation of ascomycotan fungi including: malt extract agar (MEA), glucose extract agar (GYEA), carnation leaf agar (CLA),

Ophiostoma selective agar (OSA, Harrington, 1981) with cycloheximide, Sabouraud dextrose agar (SDA), PDA with autoclaved scolytine beetle fragments, PDA with a moist twig fragments, PDA amended with sawdust, and near-ultraviolet light exposure. Three subcultures of isolate Hulcr 9747 were grown on these media for 7–14 d to induce sporulation. Mating tests were performed using four strains (six pairs) on 2% MEA.

Results

Culturing of Fungi from Beetle Mycangia

All seven P. cavipennis macerated heads yielded a single fungal morphotype on

PDA with an average 2200 CFUs (colony forming units) with a range from 1400–3000

CFUs. The morphology of hyphae, conidiophores and conidia resembled

Ophiostomatales but did not conform to any genus known to us. Colony macromorphology on PDA at 25 °C was white, velvety, 60–70 mm after 10 d, turning brown to hazel after 10–14 d (Figure 2-2). Optimal growth of pure cultures of the dominant morphotype on PDA occurred at 25 °C. Despite significant efforts to induce sporulation, no reproductive structures were observed. No fungal growth was recorded at 15 °C, 35 °C, nor on media amended with cycloheximide.

26

Sequencing and Phylogenetic Analysis

Four isolates cultured from P. cavipennis in Florida and one dominant fungus amplified from P. cavipennis mycangia from Java yielded 28S sequences. The 28S sequences had very high G/C content and many repeating regions. All 28S sequences generated from each isolate were different, but >99% similar to each other. A GenBank

BLAST query of 28S sequences from isolate Hulcr 9747 yielded a top match with

Raffaelea sp. PL1001 (accession KJ909293) at 83% similarity and the closest match of the respective 18S sequence was Raffaelea subfusca C2335 (accession KJ909306) at

97% similarity. Phylogenetic analysis of the combined 18S, 28S, and βT sequences revealed the P. cavipennis mycosymbionts represent a monophyletic lineage within the

Ophiostomatales (Figure 2-1).

Fungal Community Amplicon Analysis

The community metabarcoding approach yielded 1,583,464 total reads. After quality filtering, 527,370 reads remained and were included in analysis. When sequence similarity of 99% was set, 289 OTUs were produced. The most prevalent OTU comprised 45.5% of reads, and was identified as the same dominant P. cavipennis mycosymbiont morphotype recovered from culturing, based on phylogenetic analysis

(Java MiSeq LSU, Figure 2-1). A representative sequence was deposited in GenBank

(accession SRP072278). BLAST queries of the representative in GenBank resulted in two accessions with 91% similarity and 100% query coverage: Ophiocordyceps gracilis

(accession HM119587) and Raffaelea arxii (accession EU984298). The next dominant

OTU comprised 15.2% of reads and was identified by a BLAST query as close to an uncultured fungal endophyte in the (accession KP335416.1). The next dominant OTU formed 3.2% of reads most similar to a Phialemonium sp. (accession

27

HM060271); remaining OTUs comprised less than 3% read abundance and were composed of other fungi and .

Examination of Mycangia

Paired fungal sacs were first observed during exploratory dissections within P. cavipennis heads. Each fungal mass was nestled within tissues just posterior to the emarginate notch of the beetle's eyes, adjacent to the pharynx. These paired pharyngeal mycangia were then confirmed in microtome cross-section images (Figure

2-3) and micro-CT scans (Figure 2-4). Serial transverse sections from the heads of all five female P. cavipennis heads confirmed the presence of the paired pharyngeal mycangia.

Taxonomy

3.5.1. Afroraffaelea C Bateman, YT Huang, & DR Simmons, gen. nov.

3.5.1.1. Mycobank MB 816236

Aerial hyphae present, subhyaline, smooth, septate, greatly branched, often perpendicularly; thickened and melanized intercalary; distal hyphae denticulate.

Immersed hyphae subhyaline, smooth, constricted at septa, frequently branched, with tapering distal ends. Monillioid mycelia occurring in peptone-containing medium, fragmented unicellularly or bicellularly. Monophyletic lineage within the

Ophiostomataceae.

3.5.1.2. Type species

Afroraffaelea ambrosiae CC Bateman, YT Huang & DR Simmons.

3.5.1.3. Etymology

28

Prefix Afro-indicating its likely African origin, and ‘raffaelea’ to recognize the ecological similarity to the closely related (though probably not monophyletic) ambrosia fungus genus Raffaelea.

3.5.2. Afroraffaelea ambrosiae C Bateman, YT Huang, & DR Simmons, sp. nov.

3.5.2.1. Mycobank MB 816237

Colony margin effuse, white, velvet, 60–70 mm on PDA at 25 °C after 10 d, turning brown to hazel after 10–14 d. Bottom side of the culture firstly subhyaline to whitish, becoming brown to dark in old culture. Aerial hyphae abundant, subhyaline, smooth, septate, often branched perpendicularly; distal hyphae denticulate; thickened and intercalary melanized with age. Immersed hyphae subhyaline, smooth, septate, somewhat constricted at septa, branched, distally tapering to the end. Monillioid mycelia occurring in peptone-containing medium, fragmented unicellularly or bicellularly. No sporulation or reproductive structures were observed.

3.5.2.2. Type

USA, Florida, Miami-Dade County, Coral Gables (25°39′37.8″N 80°16′47.0″W): isolated from pharyngeal mycangia of P. cavipennis collected at an ethanol-baited light trap, August 12th, 2015 (Holotype-BPI 910155-, Ex-type culture CBS 141678 = CMW

48331 = Hulcr 9747, additional cultures CBS 141679 = CMW 48332 = Hulcr 9759, CBS

141680 = CMW 48333 = Hulcr 9765, CBS 141681 = CMW 48334 = 9772). Genbank accessions: 18S rDNA, KX620931; 28S rDNA, KX620930; βT, KX620929.

3.5.2.3. Etymology

Epithet ambrosiae to acknowledge species' engagement in a symbiosis with an ambrosia beetle.

29

3.5.2.4. Note

A. ambrosiae is characterized by a distinctly fluffy colony, which is a character that has not been recorded for species of Ophiostomatales. No ascocarp was produced in our mating study. Attempts to stimulate the sporulation, e.g. MEA medium with sterilized wood chips, ground-up bark beetle, near-ultraviolet light treatment and other types of media failed to induce sporulation in our isolates. Notably, the monillioid hyphae occurring in peptone-containing media were also observed in microtome sections of beetle mycangia. The fragmentation of monillioid hyphae is a common convergent feature in ambrosia fungi, typically expressed during transport in the mycangium and possibly facilitating beetle feeding.

Discussion

A novel species of fungus described here, Afroraffaelea ambrosiae, is the dominant fungal symbiont of P. cavipennis, based on culturing, community sequencing, and histological methods. No other fungi were recovered in comparable abundance or prevalence by any of the methods used, indicating the relationship between beetle and fungus may be highly specific. This hypothesis is consistent with our confirmation of the fungus from both Floridian and Javan beetle populations. Afroraffaelea represents a divergent branch within the Ophiostomatales, though its placement within the family is robust and does not warrant the description at any higher taxonomic level at this time.

This is the first report describing the identity of fungi associated with Premnobiini beetles, and appears to represent another origin of the ambrosia symbiosis within

Kingdom Fungi. As an independent origin, Afroraffaelea brings the total number of known fungal ambrosia lineages to at least eight. There is still a greater number of known origins among the beetles (thirteen confirmed, Farrell et al., 2001; Jordal and

30

Cognato, 2012; Kirkendall et al., 2015), but most beetle lineages have never had their fungal mutualists identified. As more fungal communities are investigated within the beetles, the number of ambrosial origins within Fungi may actually exceed the number within Coleoptera. Some beetle clades are already known to associate with fungi from multiple ambrosial origins. One example is the Xyleborini, which contains beetles associated with disparate fungal mutualists in the genera Flavodon, Raffaelea, and

Fusarium. Conversely, Raffaelea s.l. ambrosia fungi are widespread across several unrelated beetle tribes including Corthylini, Xyleborini, and Platypodinae beetles.

However, it is likely that these Raffaelea s.l. symbionts will be placed into several other genera, further increasing the number of independent origins of fungal mutualists

(Musvuugwa et al., 2015). Questions remain as to why some groups associate with a greater number of symbiont clades, and how the symbiosis is maintained across these diverse groups.

The biogeographic history of these organisms may show how the association is maintained or lost over time. Phylogenetic analyses indicate that A. ambrosiae may share a common ancestor with Fragosphaeria (Figure 2-1), a genus found in dead and dying trees and loosely associated with galleries of scolytine beetles (Shear, 1923;

Chesters, 1935). However, low bootstrap support at the node uniting Fragosphaeria and

A. ambrosiae indicates an uncertain position within the Ophiostomatales. Explorations of fungi associated with other Premnobiini beetles as well as isolates from P. cavipennis across its invaded range should help resolve the placement of Afroraffaelea within the and may reduce the amount of divergence observed from related taxa.

31

The substantial divergence between Afroraffaelea and other Ophiostomatales is probably a result of long separate evolution, not a methodological artifact. Premnobiini are thought to have evolved in Africa, possibly apart from other obligate

Ophiostomatales-associated beetles (Wood, 1982, 2007; Bright and Skidmore, 1997).

Corthylini, the oldest scolytine ambrosia beetle clade which sometimes carry Raffaelea s.l. symbionts, originated in the Americas, while Xyleborini and Platypodinae associated with Raffaelea s.l. may have originated in Australasia where they are most diverse

(Jordal et al., 2000, 2015; Cognato, 2013; Jordal, 2015). Furthermore, our phylogeny yielded topologies highly similar to studies using a range of conserved (18S, De Beer et al., 2013) to divergent markers (βT, Dreaden et al., 2014; Ando et al., 2016).

The inability of A. ambrosiae to grow in the presence of cycloheximide further distinguishes it from Raffaelea and many other Ophiostomatales, which are known to tolerate the compound (Harrington, 1981). We speculate that the strong divergence, coupled with the absence of this fungus from well-sampled beetles which carry promiscuous other Ophiostomatales symbionts, indicates Afroraffaelea symbionts are highly coadapted to their Premnobius beetle partners, or vice versa.

The confirmation of the same fungus in P. cavipennis mycangia from several non-native regions is significant from the invasion biology perspective. Colonization of non-native regions and integration into non-native ambrosia communities do not seem to change the beetle-fungus association for symbioses that are already specific, such as

Premnobius-Afroraffaelea (this paper) or Xylosandrus-Ambrosiella (Mayers et al., 2015,

Bateman et al., 2015). Horizontal transfer does occur in these groups, but at narrowly diffuse scales within genera, with no initiation of phylogenetically novel relationships.

32

The opposite - mixing between beetle and fungal species in newly colonized regions - appears to be true for promiscuous symbioses, such as Xyleborus-Raffaelea (Carrillo et al., 2014; Simmons et al., 2016a) or Euwallacea-Fusarium (O'Donnell et al., 2015).

Many of these clades have undergone horizontal transfers and symbiont switches with previously non-ambrosial beetles/fungi. The dynamic relationships that can be observed across the ambrosia symbiosis suggests that each clade of beetles or fungi has unique characters that serve to maintain the relationship. More rigorous sampling of underrepresented fungal communities would aid in designing experiments on the maintenance of these symbioses.

33

Figure 2-1. Best ML tree from GARLI analysis of 18S, 28S rDNA, and βT (introns 3/4/5 removed) data matrix of Sordariomycetidae with Xylariales outgroup. Values at nodes represent ML bootstrap percentages ≥70% from a summary of 500 replicates, and branches in bold represent BPP ≥85%.

34

Figure 2-2. A-B. Colony on PDA at 8th day. C-D. Colony on PDA after 14 d. E-F. Aerial hyphae, showing intercalary melanization and denticulate distal hyphae. G-H. Immersed hyphae. I-J. Monillioid hyphae.

35

Figure 2-3. Transverse cross sections of adult female Premnobius cavipennis head showing the content of the paired mycangia. (A) Section showing the entire head with paired mycangia, each between the compound eye and the brain; (B) Compacted mycangial inoculum as an aggregation of monillioid hyphae; and (C) Loose monillioid hyphae from mycangial packet. Bars equal 25 μm. Key: CG, cerebral ganglia; E, compound eye; M, mycangia (paired).

36

Figure 2-4. Micro-CT scan of adult female Premnobius cavipennis showing a three- dimensional mycangium structure which opens to the pharynx, shown from an (A) anterior aspect with scale and (B) lateral aspect.

37

Table 2-1. Species of Sordariomycetidae and Xylariales from used for phylogenetic analyses Genbank accession no. Species 18S rDNA 28S rDNA βT Afroraffaelea ambrosiae Hulcr 9747 KX620931 KX620930 KX620929 Afroraffaelea ambrosiae Hulcr 9759 NA NA NA Afroraffaelea ambrosiae Hulcr 9765 NA NA NA Afroraffaelea ambrosiae Hulcr 9772 NA NA NA Afroraffaelea ambrosiae Java MiSeq NA SRP072278 NA Ascitendus austriacus GQ996542 GQ996539 NA Bombardia bombarda DQ471021 DQ470970 NA Calosphaeria pulchella AY761071 AY761075 NA Camarops microspora DQ471036 AY083821 NA Camarops ustulinoides DQ470989 DQ470941 NA Ceratocystiopsis manitobensis EU984266 DQ294358 DQ296078 Ceratocystiopsis minima NA DQ294361 DQ296081 Ceratocystiopsis minuta NA EU913656 EU913736 Ceratocystiopsis minutabicolor EU984268 DQ294359 DQ296079 Ceratocystiopsis ranaculosa NA DQ294357 DQ296077 Ceratocystiopsis rollhanseniana HQ634834 DQ294362 DQ296082 Ceratomstomella pyrenaica DQ076324 DQ076323 NA Ceratosphaeria lampadophora GU180618 AY346270 NA Chaetosphaeria ciliata GU180614 GU180637 NA Chaetosphaeria curvispora AY502933 GU180636 NA Chrysoporthe cubensis DQ862047 AF408338 NA ostrea DQ471007 DQ470959 NA Diaporthe phaseolorum L36985 U47830 NA Endothia gyrosa DQ471023 DQ470972 NA Esteya vermicola NA EU627684 FJ490553 Fragosphaeria purpurea AF096176 AF096191 NA Fragosphaeria reniformis NA AB189155 NA Gaeumannomyces medullaris FJ176801 FJ176854 NA Gelasinospora tetrasperma DQ471032 DQ470980 NA Graphilbum microcarpum NA GU134186 NA Graphilbum nigrum NA AF137281 NA Graphilbum rectangulosporium AB235159 AB235158 NA Graphilbum sp. NA AY672929 NA Graphilbum sparsum NA AF135575 NA Grosmannia abieticola NA GU134177 GU134148 Grosmannia aenigmatica NA DQ294391 DQ296111 Grosmannia alacris NA JN135313 JN135329 Grosmannia crassivaginata NA DQ294386 DQ296106 Grosmannia galeiformis NA DQ294383 JF280009

38

Table 2-1. Continued Genbank accession no. Species 18S rDNA 28S rDNA βT Grosmannia grandifoliae NA DQ294399 DQ296119 Grosmannia huntii NA DQ294387 DQ296107 Grosmannia laricis NA DQ294393 DQ296113 Grosmannia leptographioides NA DQ294382 DQ296102 Grosmannia penicillata AY858662 DQ294385 DQ296105 Grosmannia piceiperda AY497514 DQ294392 DQ296112 Grosmannia serpens AY497516 JN135314 JN135334 Grosmannia wageneri NA DQ294396 DQ296116 Lanspora coronata DQ470996 U46889 NA Leptographium abieticolens NA AF343701 NA Leptographium bistatum NA AY348304 AY348306 Leptographium chlamydatum NA EU979333 EU979341 Leptographium curviconidium NA HQ406850 HQ406898 Leptographium lundbergii EU984274 DQ294388 DQ296108 Leptographium neomexicanum NA AY553382 AY534930 Leptographium pineti NA DQ062076 DQ062010 Leptographium procerum NA AY553386 NA Leptographium terebrantis NA EU296777 EU296784 Leptographium truncatum NA DQ062052 DQ061986 Leptographium wingfieldii NA AY553398 AY534946 Leucostoma niveum DQ862050 AF362558 NA Magnaporthe grisea AB026819 AB026819 NA Melanconis stilbostoma DQ862043 AF362567 NA Menispora tortuosa AY544723 AY544682 NA crassa X04971 AF286411 NA Ophiostoma araucariae NA DQ294373 DQ296093 Ophiostoma canum NA DQ294372 DQ296092 Ophiostoma floccosum AF139810 AF234836 AY789142 Ophiostoma fusiforme NA DQ294354 AY280461 Ophiostoma AY172021 DQ294381 DQ296101 Ophiostoma karelicum NA EU443756 EU443773 Ophiostoma macrosporum EU984257 AF282873 EU977465 Ophiostoma montium EU984278 DQ294379 AY194963 Ophiostoma multiannulatum NA DQ294366 DQ296086 Ophiostoma palmiculminatum NA DQ316143 DQ821543 Ophiostoma phasma NA DQ316151 DQ821541 Ophiostoma piceae AB007663 AF234837 FJ455597 Ophiostoma piliferum DQ471003 DQ294377 DQ296097 Ophiostoma pluriannulatum NA DQ294365 DQ296085 Ophiostoma protearum NA DQ316145 DQ316163 Ophiostoma pulvinisporum NA DQ294380 DQ296100

39

Table 2-1. Continued Genbank accession no. Species 18S rDNA 28S rDNA βT Ophiostoma quercus AY497515 DQ294376 DQ296096 Ophiostoma seticolle NA AF135578 NA Ophiostoma stenoceras DQ836897 DQ836904 NA Ophiostoma subannulatum NA DQ294364 DQ296084 Ophiostoma tingens EU984258 AF282871 EU977468 Ophiostoma ulmi NA DQ294374 DQ296094 Papulosa amerospora DQ470998 DQ470950 NA Raffaelea albimanens EU170269 EU177452 NA Raffaelea amasae AY858660 EU984295 EU977470 Raffaelea ambrosiae EU170278 EU177453 NA Raffaelea arxii EU170279 EU177454 NA Raffaelea brunnea EU170280 EU177457 EU977460 Raffaelea canadensis EU170270 EU177458 EU977473 Raffaelea ellipticospora KJ909299 EU177446 KJ909298 Raffaelea fusca KJ909300 EU177449 KJ909301 Raffaelea gnathotrichi EU170282 EU177460 NA Raffaelea lauricola NA EU177440 NA Raffaelea lauricola EU257806 KJ909303 KJ909302 Raffaelea montetyi AY497520 EU984301 EU977475 Raffaelea quercivora GQ225703 AB496454 NA Raffaelea rapaneae NA KT182930 NA Raffaelea santoroi EU984261 EU984302 EU977476 Raffaelea scolytodis AM267261 AM267270 NA Raffaelea subalba KJ909304 EU177443 KJ909305 Raffaelea subfusca KJ909306 EU177450 KJ909307 Raffaelea sulcati EU170271 EU177462 NA Raffaelea sulphurea EU170272 EU177463 NA Raffaelea tritirachium EU170273 EU177464 NA Raffaelea vaginata NA KT182932 NA Sordaria fimicola AY545724 AY545728 NA Sphaeronaemella fragariae AY271802 AY271808 NA Sporothrix humicola NA EF139114 EF139100 Sporothrix inflata NA DQ294351 DQ296075 Sporothrix lignivora NA EF139119 EF139104 Sporothrix pallida NA EF139121 EF139110 Sporothrix schenckii M85053 DQ294352 AY280477 Sporothrix variecibatus NA DQ821537 DQ821539 Thyridium vestitum AY544715 AY544671 NA Togninia minima DQ471011 AY761082 NA Togniniella acerosa AY761073 AY761076 NA

40

Table 2-1. Continued Genbank accession no. Species 18S rDNA 28S rDNA βT Outgroup Diatrype disciformis DQ471012 DQ470964 NA Graphostroma platystoma DQ836900 DQ836906 NA Nigrospora oryzae FJ176838 FJ176892 NA Seynesia erumpens AF279409 AF279410 NA Xylaria hypoxylon AY544692 AY544648 NA

41

CHAPTER 3 PATTERNS IN FUNGAL SYMBIONT FIDELITY AND DIVERSITY ACROSS TEN EVOLUTIONARY ORIGINS OF THE AMBROSIA SYMBIOSIS IN BEETLES (COLEOPTERA, CURCULIONIDAE: SCOLYTINAE)

Rarely are mutualism the neat, one-to-one persistent interactions as they’re portrayed when first described. High-throughput sequencing technologies have increasingly revealed unseen diversity in mutualisms. Modern ecological theory is placing these complex interactions within the context of a metacommunity, where persistence of associations is linked to dispersal and community interactions

(Mihaljevic, 2012). The metacommunity framework has subsequently led to discoveries about how symbioses evolve and function under different contexts and provides salient testable hypotheses. For example, within-host mutualisms are more likely to be mediated by hosts because they must bypass immune systems (Adair and Douglas, 2017), and multi-member mutualisms are more likely to be mediated by indirect effects because of cascading changes in nested networks (Skelton et al.,

2016; Guimarães Jr et al., 2017). The ambrosia symbiosis between wood-boring beetles and fungi represents an ideal scenario under which these hypotheses can be tested. Since the symbiosis was first described, understanding of symbiont fidelity has fluctuated between the view of one-to-one mutualism, to a suite of interacting members with diffuse phylogenetic fidelity and multiple evolutionary origins. The debate over symbiont fidelity is still ongoing largely because fungal symbionts have not been identified from the majority of ambrosia beetle evolutionary diversity, making it difficult to predict important functions of symbiotic interactions including pathogen switching or invasion ecology.

The ambrosia symbiosis is characterized by wood-boring beetles (Coleoptera:

Curculionidae: Scolytinae and Platypodinae) which feed on mutualistic fungi inside the woody tissues of trees. Ambrosia beetles primarily inhabit xylem, a relatively

42

nutrient-poor resource for insects, while cultivating fungi from which they feed almost exclusively (Six, 2003). The ambrosia fungi grow inside the xylem and are dispersed to new trees inside specialized beetle organs called mycangia, which have with tremendous diversity in size, shape, and location on the beetle body (Shedl, 1962;

Francke-Grosmann, 1967). This habit has evolved at least twelve times within the beetle subfamilies Scolytinae and Platypodinae (Johnson et al., 2018), and at least ten times within fungi (Vanderpool et al., 2017; Kasson et al., 2013; Mayers et al.,

2015; Kolarik and Kirkendall, 2010; Li et al., 2015; Endoh et al., 2008). Most ambrosia beetles and fungi continue this relationship as harmless saprotrophs, although a few exceptional cases are devastating forest pests (Hulcr and Dunn,

2011).

The composition of ambrosia fungus communities is becoming increasingly important in our understanding and management of non-native ambrosia beetles. In one cryptic beetle species complex, Euwallacea fornicatus s.l., the composition of symbionts has been linked with damaging populations (O’Donnell et al., 2016;

Carrillo et al., 2016), and symbionts can be difficult to detect in areas with multiple non-native introductions and novel symbiont interactions. The identity, fidelity, and function or symbionts in even the most widespread ambrosia beetle pests largely remain confused. In cases such as the black twig borer or the pan-tropical pinhole borer Euplatypus parallelus, only recently have coevolved nutritional mutualists been separated from other fungal associates (Bateman et al.,

2016). In other cases, ambrosia fungi pose a far greater pest than their beetle partners.

Knowing the beetle-fungus associations is especially important where the fungus is a plant pathogen. The ambrosia fungus Raffaelea lauricola is the causal

43

agent of Laurel Wilt, a tree disease that has destroyed redbay (Persia borbonia) and avocado (Persia americana) across the southeastern United States after it was introduced from Asia with its beetle partner (Kendra et al., 2013). The fungus was not known to be a tree pathogen at the time of introduction and was not formally described until six years after being detected in the United States (Harrington et al.,

2008). It’s possible the disease epidemic could have been prevented if the beetle and fungus were known to be harmful. Several other scolytine bark and ambrosia beetles have been successfully eradicated (LaBonte, 2010; Brockerhoff, 2010), but only when they were known to be a pest beforehand, suggesting that future invasions could be prevented with adequate information on symbionts and their ecology. More information is needed to determine how ambrosia fungi might be aiding their beetle partners with invasion, and which symbionts might be destructive in non-native regions.

Although pestiferous ambrosia beetles and fungi are the best known, they represent a minority of species involved in the symbiosis. Out of 3,000 ambrosia beetle species, fewer than 20 species cause economic damages (Kirkendall et al.,

2014), and only the Euwallacea fornicatus sp. complex, Xylosandrus compactus,

Xyleborus glabratus, and possibly Coptoborus ochramactonus attack apparently healthy trees (Ploetz et al., 2015; Drooz, 1985; Stilwell et al., 2014). While most of the “other” thousands of ambrosia beetle-fungus consortia colonize dead or dying trees, they are not ecologically homogeneous. Instead, all species show preferences for trees that fall within specific conditions of stress or decay, size, moisture content, host range, and location within the wood (Wood, 1982; Ghandi et al., 2010; Kasson et al., 2016; Roeper et al., 2017). These differences in environments present a wide

44

range of conditions for symbiont specialization, and a wide range of communities to interact.

The most common hypothesis for how the symbiosis is maintained across many environments and millions of years focuses on transmission mechanism. For most beetles, mycangia are credited with transmitting fungal symbionts through each generation (Francke-Grosmann, 1965; Six, 2003; Kostovcik et al., 2014; Mayers et al., 2015), although for some beetles, the cuticle or gut has been found to be a reliable transmission route (Bateman et al., 2015; Beidermann et al., 2013). The transmission hypothesis relies on sparse and uneven sampling of beetle symbiont communities, even though it is supported by a few studies. Fungal symbionts have only been identified using DNA-based methods from less than 2% of all known ambrosia beetles, and fungi have never been identified from at least 7 homoplasious mycangium types (Hulcr and Stelinski, 2017). The functional roles of symbionts and mycangia would be easier to predict with a more even sampling of the evolutionary and ecological diversity, and more importantly, it would be easier to make better predictions about which symbionts have a high potential for invasion or pathogenicity.

Here, the ambrosia symbiosis is used as a model to address several fundamental questions about multi-species eukaryotic mutualisms. The high- throughput sampling of mutualists from nine evolutionarily independent vector clades sampled across multiple independent biogeographic regions enabled tests of hypotheses that are difficult to test in systems with few origins or narrower range, such as the fungus farming ants and termites. High-throughput sequencing and fungal culturing were used to characterize the diversity and phylogenetic breadth of ten independently evolved ambrosia mutualisms. The questions first asked were: (1)

45

what is the global phylogenetic diversity of ambrosia fungi, (2) what is the total diversity of obligate fungal symbionts within individual beetle mycangia, and (3) what are the fungal symbionts of each independently evolved lineage of ambrosia beetles.

Once symbiont communities were characterized, the multiple evolutionary origins of fungus farming were used to test the hypothesis that beetle mycangia maintain more constrained associations because of strong host control over community composition, and the hypothesis that fungal ecology drives beetle ecology because fungi are directly interacting with plant-based resources. A new sequencing marker and controls were also implemented to improve upon previous high-throughput analyses of the ambrosial communities (Kostovcik et al., 2014).

Methods

Sampling and Dissection

Adult beetles comprising 51 species were collected for fungal sequencing, and 18 species for fungal culturing, from 14 countries (Table 3-1). Specimens were stored alive at ambient temperature with sterile water-moistened paper towels for 0-2 d after collection, and then preserved in >95% ethanol. Sampled galleries were stored with beetles for fungal culturing or preserved immediately in ethanol for community sequencing.

Beetle mycangia were dissected using a sterile scalpel as described previously by Mayers et al. (2018) for Corthylini, Bateman et al. (2016) for

Xylosandrus, and and Bateman et al. (2015) for all others. Where the location of mycangia were not known (Corthyloxiphus, Sueus, Hypothenemus concolor, H. birmanus), beetles were separated into head, thoracic, and abdominal segments, and each was used separately. Mycangia or body segments were then each transferred into 2mL microcentrifuge tubes containing 1mL of Buffer ATL (Qiagen)

46

for community sequencing, or 1mL phosphate buffer solution for culturing, and then crushed using sterilized a micropestel. Fungal culturing proceeded from this step by serial dilution as described by Bateman (2015).

DNA Extraction, Amplification, Library Preparation, and Sequencing

Beetle specimens and galleries sampled for Illumina miseq sequencing were cryopreserved at -80°C in 95% ethanol prior to sequencing. Dissected beetle mycangia or body segments in 180 uL Buffer ATL and 20uL protinase K (Qiagen blood and tissue extraction kit, Valencia, CA) were incubated for 2 hours at 56°C, and subsequently underwent DNA extraction following manufacturers protocols.

Culturing and genomic DNA extractions from pure fungal cultures were performed as described in Bateman et al. (2015).

The 28S marker was chosen instead of the more commonly used ITS locus because many Ophiostomatales ambrosia fungi have high G/C content at the ITS locus, which makes them difficult to amplify, uncompetitive in pooled amplicons, and poorly populated in sequence databases (Dreaden et al., 2014). The 28S marker also continues to be useful in phylogenies of ambrosia fungi and their relatives, even more-so than ITS (de Beer et al., 2014; O’Donnell et al. 2015; Simmons et al., 2016).

Miseq libraries were constructed using a nested PCR approach. The first PCR targeted a 500-600 bp region within the 28S locus using the template primers LR0R

(Vilgalys and Hester 1990) and JH-LSU-369rc (Li et al. 2015) and the following cycling conditions: 95 °C-4 min, (94 °C-30 s, 55 °C-40 s, 72 °C-2:00 min)×38, 72 °C-

5 min. PCR reactions contained a final volume of 25 μL: 1μ of template DNA, 10mM of each primer, 1.25U of Polymerase (Takara ExTaq), PCR buffer (15 mM MgCl2),

250 μM each dNTP, and 5% DMSO (Dimethyl sulfoxide, V/V). A second PCR was then performed to attach Nextera sequencing adapters and indices to the amplicons

47

(Illumina manual). The composition of the second PCRs were the same, with 1μL of

PCR product from the first PCR replacing the template DNA, and using the following cycling conditions: 95 °C-6 min, (94 °C-30 s, 55 °C-45 s, 72 °C-1:30 min) 8x, 72 °C-5 min.

Multiple positive and negative controls were used to test for laboratory contamination and PCR bias during library prep, and barcode switching during sequencing. Negative controls were replicated in triplicate for DNA extraction and

PCR reagents, and for open-air tubes exposed to the laboratory where mycangium dissections took place. Positive controls consisted of 16 pure cultures of ambrosia fungi or close relatives (Figure 3-1). Two mock community positive controls were comprised of DNA from several cultures pooled at equimolar concentrations prior to sequencing. Mock community 1 contained DNA from Ambrosiella nashakimae, A. roeperi, A. xylebori, Phialophoropsis sp., Fusarium sp. (AFC), F. solani FSSC-45a,

Geosmithia sp., Raffaelea subfusca, R. lauricola, Leptographium sp., Grossmannia sp., Candida sp. 1., Candida sp. 2., Flavodon ambrosius, and Cyberlindnera sp, while mock community 2 contained DNA from A. xylebori, Candida sp., Geosmithia sp., F. solani FSSC-45a, and R. subfusca. For each library, DNA concentrations of amplicons were quantified using Qubit (Invitrogen), and then pooled in equimolar amounts at a concentration of 17.9 ng/μL. Pooled libraries were cleaned using ELF fractionation and AmPure prior and sequenced at 300 cycles/read in an Illumina

Miseq platform at the Interdisciplinary Center for Biotechnology Research (ICBR),

University of Florida (UF).

Data Processing

Demultiplexed Miseq sequence data were processed, quality filtered, and grouped into OTUs using the Amptk pipeline (v.0.9.3, Downloaded 06/13/2017).

48

First, primer sequences were removed, and paired-end sequences were trimmed to

300 bp, with a minimum length of 100bp. Sequences representing OTUs were then picked using the dada2 algorithm, which infers sequence OTU assignments rather than by clustering by a fixed percent (Callahan et al., 2015). The dada2 algorithm was chosen because it is sensitive to sequences down to a single base pair, and because the 28S locus is highly conserved in most fungi. Representative OTU sequences were then passed through a fixed filter for 0.05% index-bleed, with a minimum of 10 reads/OTU. The RDP LSU (28S) taxonomy database was downloaded from RDP on 2017-06-01 and amended to reflect current changes in taxonomy of ambrosia fungi in the Ophiostomatales after Vanderpool et al. (2017), and in the Microascales after Mayers et al. (2015). Taxonomy was assigned to each

OTU using default parameters including a hybrid USEARCH (v9.2.64) and vsearch

(v2.3.4) assignment. Data were visualized in the R package ‘phyloseq’ (McMurdie and Holmes 2013) and colors edited in Photoshop CC (Adobe) to ensure accessibility for those with protanopia and deuteranopia.

Statistical Analyses

Multi-level pattern analysis (MPA) (De Cáceres et al., 2010), an expanded type of indicator species analysis (Hill et al. 1975) was used to infer significant associations between OTU prevalence and beetle species and genera. The use of the MPA allowed an explicit test of which mycangium community members appeared in association with each beetle species at a frequency greater than expected by chance and has successfully detected symbiont communities of bark and ambrosia beetles (Skelton et al. 2018). This method also appeared ideal for the broad sampling across ambrosia beetle diversity. The MPA was performed using the multilevel pattern function in the indicspecies package in R. The OTUs that were

49

significantly associated with beetle species in the MPA (p<.02) were subsequently included in a phylogenetic analysis if the initial amptk taxonomy assignment yielded an Ophiostomatalean or Microascalean identification.

Nonparametric multidimensional scaling (NMDS) ordination was used to visualize differences in significantly associated OTUs inferred from the MPA, supplemented with obligate associations to filamentous fungi reported in the literature. Symbiont reports from the literature were only included if symbionts were identified using DNA-sequencing and phylogenetic analysis due to high morphological crypsis of many ambrosia fungi (Alamouti et al. 2009). Symbiont associations were not included if the fungal genus was only known from two or fewer beetle specimens, or if associations occurred in <60% frequency within the given report. The NMDS was conducted on a Jaccard distance matrix in the vegan package in R (Oksanen et al., 2013). A permutational multivariate analysis of variance (PERMANOVA) was then conducted using the adonis function in R with

9999 permutations, to test for associations between symbiont composition and host traits.

Phylogenetic Analyses

Generic taxonomy assignments made by the amptk pipeline were verified or updated using phylogenetic analysis. Amptk-assigned OTUs were only included in the phylogenetic analysis if they were found to be a significant associate (p<.02) by the multi-level pattern analysis, and if they were assigned within the two fungal orders that contain most known ambrosia fungi, Ophiostomatales and Microascales.

Sequences were aligned and manually inspected in Geneious (v9.1). Nucleotide substitution models were chosen for Ophiostomatales OTUs based on the Bayesian information criterion (BIC) in jModeltest 0.1.1 (Guindon & Gascuel 2003, Posada

50

2008). Bayesian phylogenetic analyses were performed in MrBayes by running two runs of four chains for 10,000,000 generations, sampling every 500 generations, conducted on the University of Florida computing cluster HiPerGator 2.0.

Results

Fungal Community Sequencing

Miseq sequencing yielded 13,542,594 total reads from 178 barcoded samples. After quality filtering, 8,379,241 reads were used in the dada2 sequence- picking algorithm, which resulted in 732 iSeqs, henceforth referred to as OTUs. All amptk-assigned genera were verified by phylogenetic analyses. Six samples (3%) failed to recover any sequences, and two species, Pityoborus sp. nov and

Eccoptopterus spinosus, were removed from subsequent analyses due to low read count (mean: <30reads/sample). PCR negative controls yielded 0-4 reads/sample, and the open-air negative control yielded 84 reads which grouped into one OTU identified as an Ambrosiella species. All pure cultures (n=16) were successfully amplified and identified to their original culture, and all yielded multiple OTUs, except

Leptographium sp. (Figure 3-1). Variant OTUs within each pure culture were exactly one base pair different from each other, except for Flavodon ambrosius, which yielded one low-abundance (eight reads) Ambrosiella OTU. Mock communities that contained equimolar DNA from several cultured fungal species recovered all included taxa, except Raffaelea lauricola Mock 1, and R. subfusca and

Cyberlindnera sp. from Mock 2. Less abundant reads were also noted from all

Ophiostomatales spp. (Figure 3-1). In cases where symbiont data were previously available, sequence abundance was similar to CFU abundance in culturing studies.

51

Quantitative Culturing

Fungal culturing from beetle galleries and mycangia yielded isolates identified to genera of known ambrosia fungi, including Raffaelea s.s., Dryadomyces,

Ambrosiella, Phialophoropsis, and Fusarium (AFC) (Table 3-3). The frequency of associations to known ambrosia fungi ranged from 20-100%, with a mean of 86%.

Mean CFUs per beetle species beetle species ranged from 676-7000.

Beetle-fungus Associations

At least nine independently evolved lineages of ambrosia fungi described from previous culture-dependent studies were recovered by community sequencing and culturing, including Raffaelea s.s., Raffaelea lauricola-group, Dryadomyces,

Afroraffaelea, Ambrosiella, Merdithiella, Phialophoropsis, Fusarium, and

Ambrosiozyma. A tenth known lineage that includes ambrosial Geosmithia spp. was not detected, although this lineage is not well distinguished at the 28S locus and their known Neotropical beetle associates were not sampled. These ten linages of ambrosia fungi probably do not represent the total diversity of ambrosia fungi due to incomplete sampling of the vector beetle diversity.

The multilevel pattern analysis recovered significantly associated OTUs, some of which were recovered in high frequency and read abundance from ambrosia beetles with no previously identified symbiont, and including some associations that may represent new evolutionary origins of the symbiosis (Table 3-2, Figure 3-1).

Specifically, two lineages identified as Ophiostomatales spp. were significantly associated with Camptocerus s.l. beetles. One of these fungal lineages, referred to here as Ophiostomatales sp. 1, formed a monophyletic group at a basal node of

Raffaelea s.l. and was uniquely associated to the aterrimus clade in Camptocerus.

The second lineage, referred to as Ophiostomatales sp. 2 associated with the

52

auricomus clade in Camptocerus, likely represents an eleventh lineage of independently evolved ambrosia fungi, as it was distinct from other ambrosia fungi genera. The third new fungus, representing potentially a twelfth ambrosia fungus clade, was an OTU that could only be identified as “Sordariomycete sp.” associated to Xyloterinus. This fungus was illustrated in Abrahamson and Norris (1969) and has recently been confirmed separately by culturing (Mayers et al., in press).

This sampling represents a marked increase in knowledge of symbionts across the diversity of ambrosia beetles. Besides confirming the identities of known fungal symbionts in 27 beetle species, fungal symbionts were newly characterized from 51 previously unstudied ambrosia beetle species. Fungal symbionts are reported for the first time from 33 ambrosia beetle species and 14 beetle genera, including Camptocerus s.l., Corthyloxiphus, Dryocoetoides, Arixyeborus, Beaverium,

Webbia, Microperus, Debus, Dactylipalpus, Hypothenemus dolosus, Indocryphalus, and Sueus (Table 3-1, Figure 3-1). In addition, fungal associates are reported for the first time from putatively ambrosial, phloem-boring beetles, including Hypothenemus birmanus and Hylocurus langstroni.

Although the genus-level cophylogeny of the two associates does not depict one-to-one, reciprocal pairwise divergence across all de-novo evolutionary origins, the within-clade associations remain persistent (Figure 3-2). At a genus level, the vast majority of associations with fungal genera were depauperate. Most ambrosia beetle mycangia were significantly associated with only one fungal genus in the multilevel pattern analysis, with an average of 1.86 symbiont genera per beetle species (Table 3-1).

53

Discussion

Symbiont Evolution and Fidelity

Our synthesis of specificity data from multiple evolutionary origins provides a foundation for predicting the biological interactions of primary ambrosia symbionts.

Matched cladograms summarize the most persistent beetle-fungus pairs, each suggesting interactions at varying scales of diffuse coevolution (Figure 3-2). By establishing each genus pair as a model, we provide an experimental basis for testing interactions between the most persistent associates. From a pest management perspective, horizontal transmission of pestiferous symbionts can be expected to occur most within each genus pair.

Across the diversity of the ambrosia symbiosis, we found that specific one-to- one associations were surprisingly consistent between beetle and fungus genera, despite an abundance of studies reporting shared and low fidelity associations.

Existing reports of symbiont sharing may be the result of uneven sampling of the ambrosia beetle and fungus diversity, with highly targeted investigations of a few fungus clades associated with pests and pathogens (Raffaelea s.l.). Beetles obligately associated with Raffaelea s.l. (some Xyleborini, Corthylini, and

Platypodinae genera) genuinely can carry multiple species of fungi, but in all instances, only one fungal lineage appears obligate (Harrington et al. 2010, Ploetz, et al. 2017, Mayers et al. 2018, Li et al. 2018). Thus, most reports of symbiont sharing appear to involve ambrosia fungi either in facultative roles, or in associations shared with closely related beetle species. Future work needs to integrate multi- locus sequence typing data to more rigorously test whether these potentially more promiscuous associations occur with shared monophyletic lineages, or distinct, unshared lineages within the same genus.

54

Symbiont fidelity data are informative for predicting the incidence of symbionts across space and time, but they are even more informative in the context of host- symbiont physiology and ecology. In the next section, we integrate these broad-scale patterns in specificity with existing literature to discuss multiple mechanisms that were implicated in symbiont maintenance.

Predictors of Primary Symbiont Communities

Our comparison of at least 15 different mycangia suggests that the type of mycangium corresponds to different degrees of fidelity. We found support for a hypothesis provided by Harrington et al. (2014) and Kostovcik et al (2014) which proposed that larger, more complex mycangia have specific associations to

Microascales fungi, compared to more diverse associations associated with smaller, head mycangia. These associations were repeated among polyphyletic lineages containing superficially similar mycangia, suggesting that this pattern is not entirely the result of evolutionary origin. Future studies need to investigate the physiological basis of these different glandular mycangia (Six 2003). Specifically, the composition of gland cells and abundance of antimicrobial peptide transport molecules should be tracked with the ontogeny and colonization of beetle mycangia (Batra 1963, Francke-

Grosmann 1967, Abrahamson 1969, Bracewell and Six 2016). The mechanisms behind each mycangium are likely to be different, but commonalities may exist between immune pathways in homologous gland types. As an alternative hypothesis, mycangia may not be so selective, but instead function successfully in selective environments. Some mycokleptic ambrosia beetles, which don’t have mycangia (Hulcr and Cognato 2010), demonstrate how horizontal transmission can consistently yield a phylogenetically narrow symbiont taxon with such high fidelity that it resembles vertical transmission (Bracewell and Six 2016, Skelton et al. 2018).

55

It is possible that all ambrosia fungi are horizontally recruited, often with mycangia, with varying degrees of transmission success.

The discovery of multiple mechanisms controlling symbiont fidelity re-frames the interaction as a tripartite host symbiosis in which beetles, fungi, and trees are all involved in mediating symbiont composition. Beetles have already been implicated in mediating communities by using mycangia. Fungi are mediating communities by their abilities to colonize distinct microhabitats. For example, Meredithiella and

Ambrosiella were associated with beetles that inhabit a distinct substrate of smaller- diameter wood, which has lower humidity and higher temperature variation, while other ambrosia fungi were more frequently found with beetles preferring in wetter and less temperature-variable tree trunks (Figure 3-3, Table 3-4). Our results indicate that each ambrosia fungal lineage is adapted to distinct ecological niches that are defined by physical properties of the wood substrate. Consequently, the limits of these fungal niches may determine the niche breadth of their obligate beetle vectors. For example, when fungi are specific to wetter or drier habitats they may limit successful colonization of their beetle partners to that substrate. Further studies on the metabolism of ambrosia fungi are needed to determine how they are adapted to colonize diverse substrates of trees and beetle mycangia. Trees, as a third host, appear to be mediating communities that are attempting to establish in living trees, which are more actively defended. For example, Entomocorticium is associated with beetles that obligately attack living pines (Drooz 1985, Paine et al. 1997) but is never found in the mycangia of ambrosia beetles infesting the same tree species at a later stage of tree death (Trypodendron, Xyleborus spp, Platypidinae spp.) (Figure 3-3).

With more complete records of substrate characteristics, tree species, and tree health, it’s likely that fungi and trees specialize on even more distinct communities

56

that cannot be distinguished with current information on tree hosts. Taken together, these observations suggest that beetles, fungi, and trees are all controlling dispersal using within-host processes.

Although our test was explicitly analyzing primary symbionts, we did find that facultative associates and indirect community members may also be modifying symbiont composition. True yeasts (Saccharomycetales: Ambrosiozyma, Pichia,

Candida, Ogatea) are common members of saprotroph communities that have many proposed direct and indirect interactions with bark and ambrosia beetles (Six et al.

2012, Davis et al. 2013, Davis 2015, Dohet et al. 2016, Ranger et al. 2018). In the present study, yeasts were significantly associated with 13 beetle species, making it the third most widespread fungal order associated with beetles in the community sequencing dataset (Table 3-1). The broad distribution of yeast associations suggests they have a very diffuse association to ambrosia beetles as persistent members of the wood decay community. This may be especially true for ambrosia fungi Meredithiella and Phialophoropsis, which are more viable in culture amended with yeast extract (Mayers et al. 2018). More sensitive markers and dissection methods are needed to determine exactly which yeast genera are associated with ambrosia beetles, and where they are most abundant. Future studies also need to integrate evidence from bacteria and other wood decay community members. The evolution of multiple ambrosial origins in Ophiostomatales suggests that cycloheximide tolerance may predispose them as symbionts in the presence of competing or beneficial bacteria (Harrington 1981). These interactions are likely to be especially important in more diverse microhabitats.

57

Considerations for Fungal Community Sequencing

Miseq sequencing was successful in generating high quality data on fungal communities, evident from significant improvements in sequence recovery and controls compared to past high-throughput sequencing studies of ambrosia fungi.

Positive controls indicated there was relatively little PCR bias in detection or read number, and although there was a bias against Ophiostomatales, detection was far improved compared to previous studies using the ITS marker, where these dominant ambrosia fungi were either under-reported or not reported at all (Kostovcik et al.

2015, Miller et al. 2016, Malacrinò et al. 2017). In cases where symbiont data were previously available, sequence abundance was similar to CFU abundance in culturing studies. For example, the polyphagous shot-hole borer, Euwallacea sp. nr. fornicatus, has one of the most well-characterized multi-member fungal communities consisting of Fusarium, Graphium, and Paracremonium (Freeman et al. 2016, Lynch et al. 2017), all of which were recovered in high read abundance (Figure 3-1). For the 13 beetle genera whose symbionts are first reported here, results should be confirmed by culturing.

Nearly all fungi included in the present study showed evidence of intergenomic variation at the 28S locus. All pure cultures (n=16) yielded multiple

OTUs, except Leptographium sp. (Figure 3-1). Variant OTUs within each pure culture were exactly one base pair different from each other. The single-bp variant

OTUs do not appear to be the result of contamination or barcode switching because some variant OTUs were unique in the entire dataset, and they were more abundant than would be expected from sequencing error (Ross et al. 2013, Linder et al.

2018??). Similar patterns were observed in beetle mycangia that were expected to have a single genotype of fungus. For example, the number of Ambrosiella OTUs

58

recovered per beetle mycangium was higher than has been demonstrated by culture-based studies that also used rDNA markers for identification (Mayers et al.

2015, Ito et al. 2017, Beidermann et al. 2018). It is possible that these OTUs represent low-abundance propagules from congeneric ambrosia fungi, and they likely do in some samples, but this pattern cannot be expected in every case because it was observed in all pure cultures, and because of existing evidence of rDNA xenologs or paralogs in close relatives of Microascales and ambrosia fungi (Fourie et al. 2015, O’Donnell and Cigelnik 1997, O’Donnell et al.

1998). It is likely that the persistent, single-bp-variant OTUs were not observed in past high-throughput sequencing studies of ambrosia fungi because OTUs the dada2 OTU-picking algorithm was used. The dada2 algorithm can distinguish abundant sequences down to a single base pair difference (Callahan et al. 2016).

Most culture-independent sequencing of fungi, including ambrosia fungi, has been done using the ITS marker and clustering sequence similarity at 97% which would combine inter/intra-genomic variant sequences that differ in one or a few bases to a single OTU (Lindahl et al. 2013, Kostovcik et al. 2015, Malacrinò et al. 2017).

Final Remarks

In the present study, an unprecedented span of evolutionary diversity was examined in the ambrosia symbiosis, helping to solidify understanding of the global diversity, identity, and fidelity of fungal symbionts. These observations suggest that the symbiosis has evolved at least twelve times in Fungi, including Raffaelea s.s., R. lauricola-group. Dryadomyces, Ophiostomataceae clade 2, Ambrosiella,

Meredithiella, Phialophoropsis, Sordariomycete sp. clade, Fusarium (AFC),

Geosmithia, Ambrosiozyma, and Flavodon. The synthesis of these data portrays all beetles with highly specific associations at varying taxonomic scales. The traits

59

associated with symbiont identity and fidelity were for the most part tied to a tripartite host symbiosis that includes beetles, fungi, and trees. These observations fall in line with contemporary views of microbiomes, which depict within-host processes and community interactions as drivers of symbiont composition.

Integrating tree hosts into ambrosia symbiosis research is only the first step in moving the ambrosia symbiosis towards a community ecology framework. Living and stressed trees are limiting symbiont communities, but it’s not yet clear how other taxa, including yeasts, bacteria, mites, and nematodes, are all together linked to symbiont persistence and fitness. Future research needs to include all taxa and utilize a variety of statistical approaches, including neutral models and multilevel pattern analysis, to distinguish which community members are directly or indirectly influencing symbiont persistence and function. Eventually, experimental work will be critical in validating the function of community interactions.

60

Figure 3-1. Average read abundance of OTUs by taxonomy assignment, per beetle genus. Fungal taxa are listed and highlighted in color only if they were significantly associated in the multi-level pattern analysis.

61

Figure 3-2. Phylogenetic associations of ambrosia beetle genera and filamentous ambrosia fungus genera. Non-ambrosial outgroups are included to demonstrate the origins of the respective beetles and fungi. Cladograms inferred from Johnson et al. (2018) and Vanderpool et al. (2018). Relationships as depicted in the present study (P) and literature (L). Dashed lines indicate support only from high-throughput sequence (miseq) data. Phloem-inhabiting bark beetles noted by asterisk*.

62

Figure 3-3. A nonmetric multidimensional scaling (NMDS) ordination of obligate associations between ambrosia beetles, filamentous ambrosia fungi, and the predominant environments where the beetle-fungus pairs are typically sampled.

63

Table 3-1. Summary of collection information for beetles sampled. Num. Num. Data Beetle taxa Location Collection method/tree host beetles galleries type Ambrosiophilus atratus Mecklenburg, NC, USA unknown 3 m Anisandrus sayi Giles county, VA, USA unknown 3 m Arixyleborus abruptus Oro Province, Papua New Guinea Winkler soil extractor 2 m Arixyleborus scabripennis Oro Province, Papua New Guinea Winkler soil extractor 3 m Beaverium insulindicus Okinawa, Castanopsis 3 2 m Camptocerus aterrimus Peru unknown 4 m Camptocerus auricomus Colón, Costa Rica unknown 4 m Camptocerus niger Ecuador (Smith 45) unknown 2 m Captocerus noel Peru unknown 2 m mutilatus Nantou County, Taiwan () Alpha-copaene+ ethanol trap 3 m Coptoborus ochramactinus Cotopaxi Province, Ecuador ethanol trap OR gallery collected 5 c Coptoborus sp. 1 Cotopaxi Province, Ecuador ethanol trap OR gallery collected 3 c Coptoborus vespatorius Cotopaxi Province, Ecuador ethanol trap OR gallery collected 3 c Corthyloxiphus sp. npv. Ecuador unknown 2 m Corthylus calamarius Costa Rica Poaceae 2 m Corthylus papulans Francisco Morazán, Honduras ethanol trap 3 m Cyclorhipidion bodoanum Buncombe county, NC, USA unk 3 m Cyclorhipidion pelliculosum South Jeolla Province, South Korea Fagaceae 4 c Cyclorhipidion sp. 1 Gyeonggi Province, South Korea Fagaceae 2 c Cyclorhipidion sp. 2 South Jeolla Province, South Korea Fagaceae 1 c Cyclorhipidion spurlinum Oro Province, Papua New Guinea Winkler soil extractor 3 m Dactyliopalpus Bokuro-Abaa, Ghana trunk 1 m Debus defensus Gyeonggi Province, South Korea Fagaceae 4 2 c Debus emarginatus Enga, Papua New Guinea Winkler soil extractor 1 m Western Province, Papua New Debus fallax Guinea Winkler soil extractor 1 m Dryocetoides sp. 1 Cotopaxi Province, Ecuador ethanol trap OR gallery collected 5 c Dryocetoides sp. 2 Cotopaxi Province, Ecuador ethanol trap OR gallery collected 5 c

64

Table 3-1. Continued Num. Num. Data Beetle taxa Location Collection method/tree host beetles galleries type Dryocoetoides capucinus Atlantida, Honduras ethanol trap 1 m Eastern Highland Province, Papua Eccoptopterus spinosus New Guinea Lauraceae 3 m Euwallacea fornicatus East Java, branch 3 m Euwallacea fornicatus (PSHB) Fang District, Litchi chinensis 1 1 m Euwallacea fornicatus (PSHB) Khao Khitchakut District, Thailand Durio 1 1 m Euwallacea posticus Atlantida, Honduras wood 3 m Euwallacea posticus Cotopaxi Province, Ecuador ethanol trap OR gallery collected 4 c Euwallacea similis Dali District, Taiwan (China) unknown 3 m Hadrodemius globus Chiang Mai Province, Thailand Eucaltyptus 1 m Hadrodemius globus Chiang Mai Province, Thailand unknown host 2 m Hadrodemius pseudocomans Chiang Mai Province, Thailand Shorea 2 m Hylocurus langstoni Montgomery County, VA, USA unknown 3 m Hypothenemus birmanus East Java, Indonesia branch 4 2 m Hypothenemus concolor Bia District, Ghana branch 2 m Hypothenemus dolosus Yunnan Province, China ethanol trap 1 m Hypothenemus dolosus Broward County, FL, USA ethanol trap 2 m Hypothenemus dolosus Guyana (6879) ethanol trap 3 m Indocryphalus pubipennis Jeju, South Korea Machilus thunbergii 3 1 m Microperus alpha Nantou County, Taiwan (China) unknown 3 m Western Province, Papua New Microperus parva Guinea Winkler soil extractor 3 m Pityoborus comatus Highlands County, FL, USA light+ethanol trap 1 m Pityoborus comatus Alachua County, FL, USA light+ethanol trap 1 m Montsinery-Tonnegrande, French Pityoborus sp. nov. Guiana Protium sp. 2 m Premnobius cavipennis Atlantida, Honduras ethanol trap 1 m Scolytoplatpus blanfordii Nantou County, Taiwan (China) conophthorin+ ethanol trap 3 m

65

Table 3-1. Continued Num. Num. Data Beetle taxa Location Collection method/tree host beetles galleries type Scolytoplatpus parvus Yunnan Province, China ethanol trap 3 m Nakhon Si Thammarat Province, Sueus niisimai Thailand Nephelium lappaceum 1 m Sueus niisimai Nantou County, Taiwan (China) twig 3 m Theoborus coartatus Cotopaxi Province, Ecuador ethanol trap OR gallery collected 1 c Trypodendron lineatum Whitetop Mountain, VA, USA unknown 2 m Trypodendron sp. Fairbanks, AK, USA Populus 3 1 m Trypodendron sp. nov. Gyeonggi Province, South Korea Pinus 1 c Webbia cornutus Chiang Mai Province, Thailand Shorea 3 m Xyleborinus andrewesi Highlands County, FL, USA ethanol trap 3 m Xyleborinus saxeseni Alachua County, FL, USA ethanol trap 3 m Xyleborus affinis Venus, FL, USA ethanol trap 1 m Xyleborus affinis Atlantida, Honduras ethanol trap 1 m Xyleborus affinis East Java, Indonesia ethanol trap 1 m Xyleborus affinis Los Ríos, Ecuador ethanol trap 7 c Xyleborus bispinatus Highlands county, FL, USA ethanol trap 3 m Xyleborus ferrugineus Alachua County, FL, USA light trap 3 m Xyleborus ferrugineus Los Ríos, Ecuador ethanol trap 6 c Xyleborus impressus Alachua County, FL, USA light trap 3 m Xyleborus princeps Los Ríos, Ecuador ethanol trap 1 c Xyleborus sp. Los Ríos, Ecuador ethanol trap 4 c Xyleborus tribulatus Los Ríos, Ecuador ethanol trap 6 c Xyleborus volvulus Los Ríos, Ecuador ethanol trap 4 c Xylosandrus brevis Gyeonggi Province, South Korea ethanol trap 4 c Xylosandrus compactus Nantou County, Taiwan (China) ethanol trap 1 m Xylosandrus compactus Nantou County, Taiwan (China) quorciverol + ethanol trap 1 m Xylosandrus compactus Alachua County, FL, USA ethanol trap 1 m Xyloterinus politus Whitetop Mountain, VA, USA unknown 1 m Xyloterinus politus Buncombe County, NC, USA unknown 2 m

66

Table 3-2. Summary of associations between beetle species and OTUs with taxonomy assignment. The OTUs included were significantly associated (p<.02) by the Multilevel Pattern Analysis. Beetles marked with * were only represented by a single specimen. Beetle tribe Beetle species Fungal associate Captocerus noel Ophiostomataceae 1 (2) Saccharmycetales (2) Geosmithia Camptocerus aterrimus Ophiostomataceae 1 (2) Camptocerus auricomus Ophiostomataceae 2 (6) Saccharmycetales Fusarium Camptocerus niger Ophiostomataceae 2 Corthylini Corthylus papulans Ambrosiella/Meredithiella (10) Fungi sp. Saccharmycetales Corthylus calamarius Ambrosiella/Meredithiella (8) Beauveria (2) Leuconectria Corthyloxiphus sp. nov. Ambrosiella (5) Cladosporium Raffaelea lauricola group Pityoborus comatus Entomocoticium Raffaelea s.s. (2) Saccharmycetales Ipini Premnobius cavipennis* Afroraffaelea Scolytoplatypodini Scolytoplatypus parvus Ambrosiella (4) Scolytoplatypus blanfordii Ambrosiella (7) Cryphalini Hypothenmeus concolor* Cladosporium Geosmithia Saccharomycetales Hypothenmeus birmanus Microascales/Ambrosiella? Cladosporium Geosmithia Wallemia Hypothenmeus dolosus Geosmithia Hylocurus langstrongi Saccharomycetales Diatrypaceae Cladosporium Geosmithia Dactyliopalpus* Raffaelea lauricola-group Sueus niisimai Ambrosiella Xyleborini Microperus alpha* Cladosporium (5) Microperus parva Raffaelea lauricola Saccharmycetales Xyleborinus andrewesi Raffaelea Xyleborinus saxeseni Bionectria Hypocreales

67

Table 3-2. Continued Beetle tribe Beetle species Fungal associate Xyleborini Xyleborinus saxeseni Geosmithia (2) Raffaelea s.s. (4) Fusicolla Ambrosiophilus atratus Flavodon ambrosius Arixyleborus abruptus Dryadomyces (4) Arixyleborus scabripennis Dryadomyces (4) Cyclorhipidion bodoanum Dryadomyces Cyclorhipidion spurlinum Saccharomycetales (6) Geosmithia Dryocoetoides capucinus* Raffaelea lauricoa-group Beaverium insulindicus Flavodon ambrosius Debus fallax Dryadomyces Debus emarginatus Dryadomyces Euwallacea fornicatus (PSHB) Paracremonium Hypocreales (2) Fusarium (7) Fusarium AFC (2) Graphium Fusicolla Euwallacea fornicatus Fusarium AFC (2) Graphium Fusarium (4) Euwallacea posticus Saccharomycetales Fusarium (3) Hypocreales Euwallacea similis Tremellales Webbia cornutus Dryadomyces Saccharmycetales Xyleborus affinis Raffaelea s.s. Xyleborus ferrugineus Raffaelea s.s. Xyleborus bispinatus Raffaelea s.s. Xyleborus impressus Raffaelea s.s. (2) Anisandrus sayi Ambrosiella Cnestus mutilatus Ambrosiella (8) Hadrodemius globus Ambrosiella Hadrodemius pseudocomans Ambrosiella (8) Xylosandrus compactus Ambrosiella xylebori Xyloterini Xyloterinus politus Sordariomycete Trypodendron lineatum Phialophoriopsis Trypodendron sp. nov Phialophoriopsis (4) Saccharmycetales Esteya Ophiostoma ips Indocryphalus pubipennis Ambrosiella (7)

68

Table 3-3. Prevalence (% of specimens) and abundance (mean colony forming units) of fungi isolated from ambrosia beetle mycangia. Beetle host Fungal taxon Isolate Prevalence Mean CFUs Coptoborus vespatorius Fusarium (AFC) ES3 100% 2550 Coptoborus n/a ochramactinus Fusarium (AFC) RO40 100% Coptoborus sp. 1 Fusarium (AFC) ES45 75% 4667 Cyclorhipidion 3000 pelliculosum Dryadomyces JH12660 100% Cyclorhipidion sp. 1 Dryadomyces JH12626 100% 1074 Cyclorhipidion sp. 2 Dryadomyces JH12621 100% 676 Debus defensus Raffaelea s.s. JH12648 80% 2250 Dryocetoides sp. 1 Raffaelea s.s. ES20 100% 4800 Dryocetoides sp. 2 Raffaelea s.s. ES34 100% 2700 Euwallacea posticus Fusarium (AFC) ES68 100% 7000 Theoborus coartatus Fusarium (AFC) ES107 100% 1000 Trypodendron sp. nov. Phialophoropsis JH12645 20% n/a Xyleborus affinis Raffaelea s.s. MM2 100% 1966 Xyleborus ferrugineus Raffaelea s.s. MM74 33% 2033 Xyleborus sp. Raffaelea s.s. MM30 100% 5050 Xyleborus tribulatus Raffaelea s.s. MM16 50% 4000 Xyleborus volvulus Raffaelea s.s. MM6 100% 2800 Xylosandrus brevis Ambrosiella JH12581 83% 4050

69

Table 3-4. Results of the permutations-based multivariate analysis of variance (PERMANOVA) showing the effect of fungal symbiont presence on host ecology. Source df Sum of squares F R2 p-value Fungal taxon 8 10.796 10.527 0.46992 0.001 Residuals 95 12.178 Total 103 22.974

70

CHAPTER 4 INBREEDING AND HAPLODIPLOIDY IN BARK AND AMBROSIA BEETLES (COLEOPTERA: CUCRCULIONIDAE: SCOLYTINAE) IS NOT CORRELATED WITH THE PRESENCE OF INTRACELLULAR BACTERIA

Bacterial endosymbionts are widespread among insects and many have deep influence on insect host biology. At least half of all insect species are associated with intracellular bacteria in the order Rickettsiales (Hilgenboecker et al., 2008), which contains the most well-studied bacterial endosymbionts. These associations can have diverse effects on insect fitness ranging from mutualism to parasitism. Bacterial endosymbionts have demonstrated influence on their insect hosts by providing protection against natural enemies (Oliver et al., 2003; Scarborough et al., 2005;

Moreira et al., 2009) and tolerance to heat stress (Montllor et al., 2002), provisioning nutrients (Douglas and Prosser, 1992; Brownlie et al., 2009; Sabree et al., 2009), and modifying host reproduction or sex determination (Rousset et al., 1992; Dedeine et al., 2001; Werren et al., 2008). Testing for symbiont function can be exceedingly difficult because interactions can be highly context dependent, or beneficial and harmful functions exist in tandem (Zug and Hammerstein, 2014). As a result, the function of bacterial endosymbioses have not been tested at all for the majority of insect hosts. Only a fraction of all known insect-associated endosymbionts have been characterized using experimental evidence.

Some of the best-known endosymbionts of insects are intracellular bacteria in the genus Wolbachia. This genus became prominent for its ability to influence host reproduction and sexuality, a capacity that is being exploited for novel control methods for insect pests (Douglas, 2007; Wang et al., 2017), including vectors of human disease (Hoffmann et al., 2011). Wolbachia can alter host reproduction by inducing parthenogenesis, killing or feminizing genetic males, and causing cytoplasmic incompatibility (CI) between sperm and egg (Werren et al. 2008 for

71

review). Endosymbiont-induced changes to reproductive functions can be persistent and can have drastic consequences on the evolutionary trajectory of a lineage. For example, Nasonia wasps appear to have been driven towards speciation by

Wolbachia infections causing CI to segregate populations (Bordenstein et al., 2001).

Changes in reproductive functions can also result in a novel fitness advantage for which a species becomes dependent. In the wasp Trichogramma pretiosum,

Wolbachia infections are required for arrhenotokous (haplodiploid) reproduction in an otherwise thelytokous host (Lindsey et al., 2016). In many other cases, endosymbiont bacteria like Wolbachia have been suggested to either directly cause or propel species to evolve haplodiploidy or inbreeding (Normark, 2004; Engelstädter and Hurst, 2006).

Endosymbionts like Wolbachia could precipitate a lineage’s transition towards alternative reproductive or sex-determination systems because they are most commonly transmitted by maternal inheritance. In these cases, there is a fitness advantage for Wolbachia to benefit female survival over males. Haplodiploidy can produce females with twice the genetic fitness as males (Hamilton, 1967; Smith,

2000), and create a mechanism to purge deleterious mutations during a selection event (Goldstein, 1994). At least ten insect clades examined by Normark (2004) have evolved an alternative sex determination system, most of which occur in conjunction with sibling competition or bacterial symbioses. Given the abundance of cases where Wolbachia and other endosymbiots of the insect reproductive tract often occur in insects with non-standard reproduction, it is tempting to hypothesize that the endosymbiont presence is routinely the trigger of such evolutionary changes.

A few case studies are devoted to testing this correlative hypothesis (Kawasaki et al., 2017). However, a proper test of such a hypothesis requires a comparative

72

design, where the presence and absence of endosymbionts is tested in regularly and irregularly reproducing sister clades. In addition, for a strong inference, the clade pairs should be replicated. There are very few insect taxa where such a design is possible.

Bark and ambrosia beetles in the subfamily Scolytinae are one taxon with an especially high diversity in sex determination systems, including diploid, haplodiploid, and paternal genome elimination (PGE). The confirmed haplodiploid species are in the tribes Xyleborini and Dryocoetini (Raffa et al. 2015) which exhibit female-biased sex ratios, high rates of inbreeding, and drastically smaller, flightless males

(Kirkendall, 1983). Other beetles in the genus Hypothenemus (Cryphalini) are functionally haplodiploid by way of PGE (Borsa and Kjellberg, 1995). These species exhibit similar female-biased sex ratios, high rates of inbreeding, and much smaller, flightless males (Kirkendall, 1983). Other scolytine lineages are suspected of evolving haplodiploidly de novo, including Premnobiini and Hyorhynchini, because they exhibit similar phenotypes, but karyotyping has not yet been done (Kirkendall et al., 2015). It has been suggested that Wolbachia could be responsible for these alternative sex determination systems and dimorphic phenotypes, but it has not been rigorously tested. Previous studies have shown that antibiotic treatments may reduce fecundity in three species of scolytines (Xyleborus ferrugineus, Coccotrypes dactyliperda and H. hampei), but no gain of function has been demonstrated from specific endosymbiont taxa (Peleg and Norris, 1973; Zchori-Fein et al., 2006; Mariño et al., 2017). The only comparative analysis of these evolutionarily independent alternative mating systems thus far utilized several haplodiploid genera of scolytines, but they all belonged in the same clade (Xyleborini) with a single origin of haplodiploidy, thus effectively serving as pseudoreplicates (Kawasaki et al., 2017). In

73

addition, only one genus of bacteria was targeted (Kawasaki et al., 2017). A correctly replicated analysis with phylogenetically independent clades is still missing.

I sought to more robustly test the association between endosymbiont bacteria and sex-linked traits in scolytines. I used universal bacterial primers and MiSeq community sequencing to characterize bacterial communities from five alternative mating systems in Scolytinae beetles, and their diploid-diploid sister taxa.

Specifically, I asked:

Are genera of endosymbiotic bacteria with known ability to manipulate a host’s reproductive system present in bark beetles?

If manipulative endosymbiotic bacteria are present in bark beetles, is their presence correlated with inbreeding mating system?

Are intracellular symbiont communities within beetles more similar between phylogenetically more closely related beetles, than between more distantly related beetles?

Methods

Taxon Sampling

The sampling strategy aimed to both sample widely across phylogenetically diverse scolytine tribes, and to select pairs of taxa within tribes (or between closely related tribes) such that each pair consisted of one ‘inbreeding’, and one

‘outbreeding’ species. This permitted the testing of association between the presence of Wolbachia, or other known endosymbiotic reproductive manipulator bacteria, and reproductive strategy in the beetles. I expanded upon the sampling of

Kawasaki et al. (2016), who investigated the presence of Wolbachia in bark beetles, but selected haplodiploid species that belonged to only a single lineage of ambrosia beetles (the tribe Xyleborini), limiting the strength of inferences made regarding the

74

correlation between endosymbiont infection and mating system transformation.

Instead of the highly derived Xyleborini used by Kawasaki et al. (2016), Coccotrypes was selected to represent the clade, as Coccotrypes is most likely the genus in which the haplo-diploid nature of the entire Coccotrypes-Xyleborini clade originated.

Five ethanol-preserved female specimens of each of 14 species (in 13 genera) were selected from among those available in the frozen collection of bark and ambrosia beetles held in the School of Forest Resources and Conservation at the University of Florida. Female specimens were selected because Wolbachia is asymmetrically inherited along the maternal line, and therefore, infected beetles would more likely contain the bacteria within their reproductive organs. Wherever possible, species representatives were selected from diverse geographical localities, to prevent geographic autocorrelation. A total of 70 bark beetle specimens belonging to seven pairs of phylogenetic related taxa differing in reproductive system were studied, as summarized in Table 4-1. Locality data for collected specimens are provided in Table 4-2.

The positive control was represented by two samples of the Asian citrus psyllid (Diaphorina citri) infected with Wolbachia, provided by Dr. Kirsten Pelz-

Stelinski (University of Florida).

DNA Extraction and PCR

To enable the taxonomic identification of endosymbiotic bacteria, a metagenomic sequencing library of indexed PCR-amplified target bacterial 16S rRNA was created from DNA extractions obtained from scolytine beetle abdomens.

The abdomen (containing the female reproductive tract) was separated from each specimen under a dissecting microscope and used in individual subsequent DNA extractions. Care was taken to remove the majority of the hindgut from the abdomen

75

before further processing, to lessen the abundance of DNA from gut-associated prokaryotes. All DNA extractions were performed using DNeasy Blood & Tissue spin column kits (Qiagen), according to the manufacturer’s instructions, with an extended overnight incubation of the samples in 180 μl tissue lysis buffer and 20 μl proteinase-

K solution at 56°C.

Initial PCR amplification was undertaken using a universal primer pair, 515F and 806R, designed to amplify bacterial 16S rRNA (Bates et al 2011), and modified by the addition of Illumina-specific sequencing primer sequences to their 5’ end, according to an established Illumina protocol (Nextera DNA Library Prep Reference

Guide 2018). All samples were amplified in 25 μl PCR reactions using 12.5 μl Premix

Taq™ DNA Polymerase mix (TaKaRa), 1 μl of each primer, and 2 μl of template

DNA, under the following cycling conditions: 95°C, 1 min, 33× (95°C, 30 s; 50°C, 1 min; 72°C, 1 min), 72°C, 5 min. Reaction success was determined on agarose gels containing SYBR Green I nucleic acid gel stain (Invitrogen).

For each sample, a second PCR was undertaken to attach a unique pair of index sequences. This was achieved using 2 μl of the products of the first reaction as template, and 1 μl of sample-specific 8 bp Illumina forward and reverse dual index primers (Nextera XT Index Kit v3). The cycling conditions were as follows: 95°C, 3 min, 8× (95°C, 30 s; 55°C, 30 s; 72°C, 30 s), 72°C, 5 min. The resulting indexed library was subsequently pooled and sequenced on the Illumina MiSeq platform.

Sample Pooling and High-Throughput Sequencing

The concentration of every amplified and indexed sample was measured with a Qubit 3.0 fluorometer, using a dsDNA high sensitivity assay kit (ThermoFisher).

Approximately equimolar aliquots of all samples were then pooled together in a 1.5 ml microcentrifuge tube, to ensure even sequencing depth across them, with a final

76

DNA concentration of 40 ng/μl. Prior to sequencing, the pooled sample was purified using the Agencourt AMPure XP system (Beckman Coulter) to remove unincorporated dNTPs, primers, primer dimers, salts, and other contaminants. The pooled sample was sequenced on a single paired-end run of an Illumina Miseq in 2

X 250 cycles, at the Interdisciplinary Center for Biotechnology Research (University of Florida).

Data Processing

The resulting sequenced reads were first demultiplexed in silico, to obtain the sequences from each beetle specimen separately, then had their associated Illumina indexes and adapters removed, before being analysed in the AMPtk pipeline. The

AMPtk package enabled us to utilize the dada2 algorithm to infer ‘iseqs,’ similar to

OTUs, and subsequently assign them to taxonomic names based upon a database of known sequences. This step also removes chimeric and substitution errors, and was set to limit the number of reads per sample to five. A fixed .5% filter was also passed to account for potential index bleeding, according to rates in past Miseq sequencing studies (Palmer et al., 2017; MacConaill et al., 2018).

A customized database was employed, which consisted of the Ribosomal

Database Project (RDP) database (Cole et al., 2014) that provides “quality- controlled, aligned, and annotated Bacterial and Archaeal 16S rRNA sequences”, with additional 686 Wolbachia 16S sequences added from GenBank. Mapping of

OTUs onto the custom taxonomic sequence database was done using the

USEARCH global alignment search algorithm (Edgar, 2010).

Sequence abundance data were not used in subsequent analyses due to numerous documented biases in amplification and sequencing (Nguyen et al., 2014;

Palmer et al., 2017). Instead, OTU incidence data formed the primary source of

77

information on bacterial communities for statistical analyses. Sampling curves were generated in R (v 3.4.3) using the iNEXT package to determine completeness of beetle sampling effort. The sampling curves were calculated for species richness and Simpson’s diversity using parameters described by Johnson et al. (2016). To test for significant differences in the presence of intracellular bacteria among inbreeding and outbreeding beetles, a PERMANOVA was conducted on a Jaccard matrix using the Adonis function in R (Oksanen et al., 2013).

Phylogenetic Analysis of the Beetles

To statistically test whether bacterial communities were more similar between phylogenetically more closely related beetles than between more distantly related beetles, a phylogenetic tree of the taxa studied was created to obtain phylogenetic distance data. The tree was constructed using nuclear 28S rRNA sequences available on GenBank (Benson et al., 2013) for all the taxa studied (except for the unidentified Cosmoderes sp. and Sphaerotrypes ulmi in the dataset, for which sequences belonging to the congeneric C. monilicollis and Sphaerotrypes sp. were respectively used instead). The sequences used and their GenBank accession codes are shown in Table 4-3. The downloaded sequences were aligned using the

MAFFT version 7 online server (Katoh et al., 2002), incorporating the FFT-NS-I slow iterative refinement strategy with the following parameter values: nucleotide scoring matrix 200PAM/k=2, gap open penalty = 1.53, offset value = 0.0. Alignments were thereafter checked manually in Geneious (v9.1) for quality. The datamatrix was analysed under a maximum likelihood (ML) optimality criterion using RAxML (v7.6.6)

(Stamatakis, 2014) running on the CIPRES web-based server (Miller et al., 2010) to search for the best-scoring tree. A topological constraint was imposed in the analysis, consisting of a pruned tree, containing only the taxa studied here derived

78

from the more comprehensive genomic-level scolytine phylogeny constructed

(Johnson et al., 2018). To assess nodal support, a rapid bootstrap analysis with

1000 iterations was run in parallel with tree-building.

Results

Sequencing

The final DNA concentration of the pooled sample was 49.5 ng/ μl, which was submitted for sequencing in a volume of 300 μl. Sufficient high-quality sequences were obtained for 67 of the 70 samples to allow for downstream analyses, with a total of 8,572,756 reads passing quality control across all samples. Sequencing depth varied by sample from a minimum of 23,737 reads (Premnobius cavipennis

CG022-13154) to a maximum of 71,4458 reads (Xyloterinus politus CG031-1939).

Sample accumulation curves are provided in Figure 4-1.

Potential Reproductive Manipulators

A total of 1974 different iseqs or “OTUs” were identified following taxonomic assignment and filtering in AMPtk. Of these, only four correspond to potential reproductive manipulators including Wolbachia (OTUs 7, 16, 1259, and 1852), two to

Rickettsia (OTUs 4 and 1576), and one to Arsenophonus (OTU 1057). The OTUs representing Wolbachia, Rickettsia and Arsenophonus were present in 33% of 70 samples, and all except two species (Monarthrum fasciatum and Dendroctonus frontalis) included in the sample selection. Samples that included reads identified to potential reproductive manipulators are visualised in Figure 4-2 adjacent to the reconstructed Maximum Likelihood beetle phylogeny. The presence of Wolbachia,

Rickettsia or Arsenophonus is not more likely in the inbreeding species than it is in outbreeding species (p= 0.889, Table 4-4). Beetle species can be broadly categorized into three groups with respect to their endosymbiont prevalence:

79

Species where 100% of specimens produced sequences assigned to one or more of endosymbiont genera. Two inbreeding species: Hypothenemus dissimilis,

Premnobius cavipennis. One outbreeding species: Taphrorychus bicolor.

Species where 80% or fewer specimens produced sequences assigned to an endosymbiont genus. Four inbreeding species: Coccotrypes dactyliperda (80%),

Cryptocarenus seriatus (80%), Sueus niisimai (80%), and Xyloterinus politus (60%),

Five outbreeding species: Sphaerotryoes ulmi (80%), Ips avulsus (60%),

Trypodendron retusum (60%), Cosmoderes sp. (40%), and Dendroctonus terebrans

(20%).

Species where no specimens produced any sequences assigned to an endosymbiont genus. Two species: Dendroctonus frontalis (inbreeding) and

Monarthrum fasciatum (outbreeding).

Discussion

The presence of intracellular bacterial taxa does not predict the reproductive system of Scolytine beetles. Miseq community sequencing showed broad associations to intracellular bacteria that are capable of modifying host reproduction, including Wolbachia, Rickettsia, and Arsenophonus (Figure 3-2) but without any discernible correlation (Table 4-4). This analysis shows the power of a comparative approach in an evolutionarily diverse group such as bark beetles. Skewed sex ratio, inbreeding, and haplodiploidy, were not associated with endosymbiont genera, suggesting they are not causative at a broad scale. All except two beetle species were associated with at least one genus of intracellular bacteria. These findings challenge the hypothesis that suggests intracellular bacteria are what cause or propel some species towards alternative reproductive systems (Normark, 2004;

80

Engelstädter and Hurst, 2006), and suggests a greater diversity of functions in

Scolytine beetles.

Despite no association to reproductive system overall, some Scolytine beetle lineages are still likely to be affected by reproductive manipulators. One beetle genus, Hypothenemus, displayed potentially obligate associations in which the same

Wolbachia sequences were recovered from every specimen, and found from multiple collection locations (Figure 4-1). This association has been previously documented in one species, the coffee berry borer, H. hampei which has a parental genome elimination (PGE) sex determination system (Borsa and Kjellberg, 1995). Vega et al.

(2002) hypothesized that Wolbachia may cause paternal chromosomes to remain condensed during meiosis and in male somatic tissue. The physiological mechanism for sex determination needs to be better understood in Hypothenemus, for example, by tracking the endosymbionts through spermatogenesis and zygote formation.

Genomic evidence of horizontal gene transfer may also suggest long-term associations between the host and other members of the microbiome (Hernandez-

Hernandez, 2017; Acuña et al., 2012). Two other beetle species, P. cavipennis and

T. bicolor, also had 100% infection rates and warrant further examination to investigate their potential involvement in sex-determination or other traits. However,

T. bicolor is regularly outbreeding, does not have biased sex ratios, and is presumably diploid, suggesting its association to Wolbachia may have a different function.

Facultative associations seem to form the majority of interactions between intracellular bacteria like Wolbachia and scolytine beetles. Most beetles had incomplete infection rates, which could not be tied to geographic sampling location.

Even Coccotrypes, which has been previously shown to harbor abundant Wolbachia

81

in Israel, showed no infection in the specimens used in the present study, but instead recovered abundant Rickettsia from two locations sampled. The incomplete infection across populations suggests beetles aren’t reliant on a single endosymbiont taxon.

Most sex-ratios in Scolytinae are very consistent (Kirkendall, 1983; Biedermann,

2010), making it unlikely that transient microbes would be consistently responsible.

It’s also possible that, as Normark (2004) postulates, perhaps these traits only

“originated” from intracellular symbioses. In some cases, endosymbionts may have put their hosts on a trajectory towards haplodiploidly, only to be lost or inconsistent as a symbiont in extant populations. It would be useful to reconstruct the evolutionary history of these bacteria to uncover phylogenetic patterns in symbiont fidelity and host phenotype.

Associations with intracellular bacteria may not be apparent due to limitations in single-marker community sequencing. The 16S marker is useful because it has successfully been used to distinguish bacterial endosymbionts from multiple genera

(Jousselin et al., 2016; Doudoumis et al., 2017). The wsp marker, by contrast, provides more phylogenetically informative sequence data for species in the genus

Wolbachia, but does not amplify taxa from other genera (Van Meer et al. 1999).

Furthermore, for reliable identification to a species or strain level, some bacterial endosymbionts like Wolbachia require multi-locus sequence typing

(Paraskevopoulos et al., 2006). Many cryptic endosymbiont taxa are still being discovered, largely due to limitations of the 16S marker (Zchori-Fein et al., 2004).

Future studies would benefit from the use of multiple markers, microsatellites, or genomic sequencing, to determine the exact identity of these bacteria.

It is not clear what intracellular bacteria are doing in scolytines that have no changes in reproductive system. One hypothesis is that some endosymbionts are

82

mutualists supplementing an imbalanced nutrient diet (Hosokawa et al., 2010). Many

Scolytinae genera feed on substrates with imbalanced nutrients, including xylem, phloem, and seeds (Wood, 1982; Six, 2003). Other insects have overcome low or imbalanced nutrient situations using endosymbionts like Wolbachia (Hosokawa et al.

2010). Genomic analyses might be useful if protein-coding genes have responded in areas of vitamin synthesis or horizontal gene transfer, especially with outbreeding beetles that engaged in potentially obligate associations (T. bicolor), but these analyses might be less insightful if symbionts are facultative or transient. In most clear-cut mutualisms with bacteria like Wolbachia, associations are more often obligate (Aanen and Hoekstra, 2007; Zug and Hammerstein, 2015). An alternative hypothesis is that many endosymbiont genera are mostly commensal or harmful in scolytines. Ideally, an axenic or improved gnotobiotic culture of these beetles will be used to test the function of interactions. Scolytinae will be an especially interesting group to investigate, because they often occur in phloem, which might represent a revivor for endosymbionts that are capable of being transmitted through the host plant tissues (Chrostek et al., 2017). Wolbachia infections could also result from interactions with nematodes (Taylor and Hoerauf, 1999), parasitoids, mites (Cardoza et al. 2008, Popa et al. 2012), or other organisms that are closely associated with the environments of scolytines.

The present study moves the field forward by showing that endosymbiont taxa are not always associated with reproductive mode in Scolytinae. Although the genera Wolbachia, Rickettsia, and Arsenophonus are well known for their affects on host reproduction, their presence is inconsistent for most scolytines, suggesting a symbiosis that is usually facultative and may be diverse in function. Future studies would greatly benefit from more experimental studies in this system, especially using

83

axenic culture. Comparative studies may be most insightful in clades with clear-cut obligate symbioses like in Hypothenemus, and in outbreeding beetles with no known endosymbiont associations, like Monarthrum. It’s likely that the endosymbiont associations observed here are only scratching the surface in characterizing these interactions.

84

Table 4-1. Taxon sampling with sex-determination systems and inbreeding character states. Pairs of phylogenetically related species differing in reproductive system are shown in alternating background colours.

Tribe Species Sex determination system Inbreeding?

Ipini Premnobius cavipennis Haplodiploid? Yes Ipini Ips avulsus Diploid? No Xyloterini Xyloterinus politus Haplodiploid? Yes Xyloterini Trypodendron retusum Diploid? No Hyorhynchini Sueus niijimai Haplodiploid? Yes Diamerini Sphaerotrypes ulmi unknown No Hylurgini Dendroctonus frontalis Haplodoploid? Yes Hylurgini Dendroctonus terebrans Diploid No Hypothenemus Paternal Genome Cryphalini Yes dissimilis Elimination Cryphalini Cosmoderes sp. unknown No Coccotrypes Dryocoetini Haplodiploid Yes dactyliperda Dryocoetini Taphrorychus bicolor Diploid? No Corthylini: seriatus Haplodiploid? Yes Pityopthorina Corthylini: Monarthrum fasciatum Diploid? No Corthylina

85

Table 4-2. Collection localities and number of specimens included for all samples used in miseq community sequencing. Species Locality State/Province Country n Premnobius Hollywood FL USA cavipennis 5 Ips avulsus Gainesville FL USA 5 Trypodendron Fairbanks AK USA retusum 1 Trypodendron WVU forest WV USA retusum 1 Trypodendron Fairbanks AK USA retusum 3 Xyloterinus politus Cassopolis MI USA 4 Xyloterinus politus Jefferson N.F. VA USA 1 Sueus niijimai Guiyang Guizhou China 1 Sueus niijimai Huaxi, Guiyang Guizhou China 2 Sueus niijimai Guiyang Guizhou China 1 Sueus niijimai Nantou Taiwan 1 Sphaerotrypes ulmi Huaxi, Guiyang Guizhou China 5 Dendroctonus Long Island NY USA terebrans 1 Dendroctonus Alligator River NC USA terebrans 1 Dendroctonus Gainesville FL USA terebrans 3 Dendroctonus frontalis Ponte Vedra Beach FL USA 2 Dendroctonus frontalis Jacksonville FL USA 1 Dendroctonus frontalis Austin Cary Forest FL USA 2 Hypothenemus Gainesville FL USA dissimilis 2 Hypothenemus Franklin PA USA dissimilis 1 Hypothenemus Fairfield OH USA dissimilis 1 Hypothenemus Wannang Madang PNG dissimilis 1 Cosmoderes sp. Blawan, Sempol East Java Indonesia 1 Jombang, Cosmoderes sp. East Java Indonesia Bangunrejo 4 Coccotrypes Lenya N. P. Tanintharyi dactyliperda 2 Coccotrypes Gainesville FL USA dactyliperda 3 Taphrorychus bicolor Nizbor Bohemia Czechia 5 Cryptocarenus USA seriatus 1

86

Table 4-2. Continued Species Locality State/Province Country n Cryptocarenus Lake Wales S.F. FL USA seriatus 1 Cryptocarenus Colt creek FL USA seriatus 1 Cryptocarenus Collier Florida USA seriatus 1 Cryptocarenus Orlando FL USA seriatus 1 Monarthrum fasciatum Gainesville FL USA 2 Monarthrum fasciatum Christiansberg VA USA 3

87

Table 4-3. Taxa used in the 28S phylogenetic analysis of the beetles, with accompanying GenBank accession codes Species GenBank Accession code Cosmoderes monilicollis I13692 Sphaerotrypes sp. I13692 Premnobius cavipennis I13693 Dendroctonus frontalis I02535 Dendroctonus terebrans AF308386 Coccotrypes dactyliperda I14147 Sueus niisimai I14148 Hypothenemus dissimilis I13625 Taphrorychus bicolor JX263716 Trypodendron retusum H04532 Xyloterinus politus I13630 Monarthrum fasciatum I13665 Cryptocarenus seriatus I10214 Ips avulsus EU090297

88

Table 4-4. Results of the permutations-based multivariate analysis of variance (PERMANOVA) showing the effect of intracellular bacteria presence on inbreeding in scolytine beetles. Source df Sum of squares F R2 p-value Endosymbiont taxon 3 0.178 0.363 0.054 0.889 Residuals 19 3.109 Total 22 3.287

89

Figure 4-1. Sampling curve for species richness indicating the completeness of sampling effort. Dotted lines indicate extrapolated data, with colored ranges represented 95% confidence intervals.

90

Figure 4-2. Beetle phylogenetic relationships inferred from 28S rRNA sequence data in RAxML, constrained to the topology of the wider scolytine phylogeny of Johnson et al. (2018) shown adjacent to incidence of the target bacterial genera. Bootstrap values shown at nodes. Scale shows nucleotide substitution rate per million years. HD= Haplodiploid, HD?=proposed haplodiploid, PGE= Paternal Genome Elimination. Brown-colored (lighter shaded) bars indicate presence within a beetle with an alternative mating system.

91

CHAPTER 5 CONCLUSIONS

In the present study, I provide novel observations on the most influential community members of this poorly characterized symbiosis: fungi and bacteria.

Comparative analyses depicted high fidelity associations to fungi, and mixed fidelity to intracellular bacteria, each with unique correlations to beetle traits and host ecology.

These results support three overarching conclusions:

(1) Invasive ambrosia beetle Premnobius cavipennis appears to be associated with a single fungal taxon, Afroraffaelea, with high fidelity, across multiple introductions into non-native regions

(2) the composition of symbionts in the ambrosia symbiosis is under a tripartite host control between trees, beetles, and fungi, and

(3) bacterial endosymbionts are not causing beetles to depend on alternative reproductive systems at a broad scale, but are likely to be influential in a few cases where they appear obligate.

The benefit of integrative methodologies was critical in addressing biases and supporting novel symbiont associations. For example, culturing provided abundance data and multi-locus sequence data from specific isolates, which complemented the community sequencing approach that avoids culture bias but generates only incidence data for all community members. However, both methods were not universally available, which highlighted several limitations of each method. Community sequence data were unable to identify fungi to species, which limited the ability to test co-cladogenesis at that level. Amplification and sequencing biases also made it infeasible to determine the relative abundances of different OTUs or species, which would normally provide support

92

for taxa acting as nutritional mutualists. For those species in which only community sequence data are available, culturing should follow to enable morphological species descriptions and more phylogenetically informative sequence data.

Although complementary methodologies were useful in detecting correlations from a comparative approach, they don’t address causality in a truly experimental framework. In other words, the gain or loss of function was not explicitly tested on these organisms in a controlled setting. The comparative approach was chosen because the field of the ambrosia symbiosis is still in its infancy. From an experimental point of view, the nutritional basis of the symbiosis has not been tested in any ambrosia symbiosis where all taxa are known and controlled. The comparative framework was most useful for identifying which community members are the more likely to be symbionts, and which functions that symbionts are most likely contributing to. Future studies need to integrate the information on symbiont identities and distributions from the present study to target their function in an experimental setting.

In the present study, I move the field of symbiosis and insect-fungus-farming research forward by providing a fundamental characterization of the fungal mutualists and bacterial endosymbionts in the ambrosia symbiosis. These results are significant because they build the capacity to continue research on the ambrosia symbiosis as a model system for eukaryotic symbioses. Future comparative analyses will certainly be strengthened by experimental evidence. The significance of these results can also be extended to the management of pestiferous bark and ambrosia beetles. It may be easier to predict which beetles are carrying, or capable of carrying, plant pathogens with these data on symbiont community composition and maintenance. It is also possible

93

that future control strategies for pestiferous ambrosia symbioses could exploit these relationships. These results are intended to serve as a guide for future functional studies on the symbiosis. Hopefully they represent a leap forward in a system that holds great potential for future research.

94

LIST OF REFERENCES

Aanen, D.K. and Hoekstra, R.F., 2007. The evolution of obligate mutualism: if you can’t beat’em, join’em. Trends in Ecology & Evolution, 22(10), pp.506-509.

Abrahamson, L.P. and Norris, D.M., 1969. Symbiontic interrelationships between microbes and ambrosia beetles IV. Ambrosial fungi associated with Xyloterinus politus. Journal of invertebrate pathology, 14(3), pp.381-385.

Abrahamson, L.P., 1970. Physiological interrelationships between ambrosia beetles and their symbiotic fungi.

Acuña, R., Padilla, B.E., Flórez-Ramos, C.P., Rubio, J.D., Herrera, J.C., Benavides, P., Lee, S.J., Yeats, T.H., Egan, A.N., Doyle, J.J. and Rose, J.K., 2012. Adaptive horizontal transfer of a bacterial gene to an invasive insect pest of coffee. Proceedings of the National Academy of Sciences, 109(11), pp.4197-4202.

Barke, J., Seipke, R.F., Grüschow, S., Heavens, D., Drou, N., Bibb, M.J., Goss, R.J., Douglas, W.Y. and Hutchings, M.I., 2010. A mixed community of actinomycetes produce multiple antibiotics for the fungus farming ant Acromyrmex octospinosus. BMC biology, 8(1), p.109.

Bateman, C., Šigut, M., Skelton, J., Smith, K.E. and Hulcr, J., 2016. Fungal associates of the Xylosandrus compactus (Coleoptera: Curculionidae, Scolytinae) are spatially segregated on the insect body. Environmental entomology, 45(4), pp.883-890.

Bates, S.T., Berg-Lyons, D., Caporaso, J.G., Walters, W.A., Knight, R. and Fierer, N., 2011. Examining the global distribution of dominant archaeal populations in soil. The ISME journal, 5(5), p.908.

Batra, L.R., 1963. Ecology of ambrosia fungi and their dissemination by beetles. Transactions of the Kansas Academy of Science (1903-), 66(2), pp.213-236.

Beaver, R.A. and Liu, L.Y., 2013. A synopsis of the pin-hole borers of Thailand (Coleoptera: Curculionidae: Platypodinae). Zootaxa, 3646(4), pp.447-486.

Beaver, R.A., Sittichaya, W. and Liu, L.Y., 2014. A synopsis of the scolytine ambrosia beetles of Thailand (Coleoptera: Curculionidae: Scolytinae). Zootaxa, 3875(1), pp.1-82.

Beaver, R.C., Shih, H.T. 2003. Checklist of Platypodidae (Coleoptera: Curculionoidea) from Taiwan. Plant Prot. Bull 1, pp.75-90.

Beeson, C.F.C., 1916. Ambrosia Beetles or Pin-hole and Shot-hole Borers. Indian Forester, 42(4), pp.217-223.

Benson, D.A., Cavanaugh, M., Clark, K., Karsch-Mizrachi, I., Lipman, D.J., Ostell, J. and Sayers, E.W., 2012. GenBank. Nucleic acids research, 41(D1), pp.D36-D42.

95

Biedermann, P.H., 2010. Observations on sex ratio and behavior of males in Xyleborinus saxesenii Ratzeburg (Scolytinae, Coleoptera). ZooKeys, (56), p.253.

Biedermann, P.H. and Rohlfs, M., 2017. Evolutionary feedbacks between insect sociality and microbial management. Current opinion in insect science, 22, pp.92-100.

Bogar, L.M., Dickie, I.A. and Kennedy, P.G., 2015. Testing the co‐invasion hypothesis: ectomycorrhizal fungal communities on Alnus glutinosa and Salix fragilis in New Zealand. Diversity and Distributions, 21(3), pp.268-278.

Bordenstein, S.R., O'hara, F.P. and Werren, J.H., 2001. Wolbachia-induced incompatibility precedes other hybrid incompatibilities in Nasonia. Nature, 409(6821), p.707.

Borsa, P. and Kjellberg, F., 1996. Experimental evidence for pseudo-arrhenotoky in Hypothenemus hampei (Coleoptera: Scolytidae). Heredity, 76(2), p.130.

Brinza, L., Viñuelas, J., Cottret, L., Calevro, F., Rahbé, Y., Febvay, G., Duport, G., Colella, S., Rabatel, A., Gautier, C. and Fayard, J.M., 2009. Systemic analysis of the symbiotic function of Buchnera aphidicola, the primary endosymbiont of the pea aphid Acyrthosiphon pisum. Comptes Rendus Biologies, 332(11), pp.1034-1049.

Brockerhoff, E.G., Liebhold, A.M., Richardson, B. and Suckling, D.M., 2010. Eradication of invasive forest insects: concepts, methods, costs and benefits. New Zealand Journal of Forestry Science, 40(Suppl).

Brownlie, J.C., Cass, B.N., Riegler, M., Witsenburg, J.J., Iturbe-Ormaetxe, I., McGraw, E.A. and O'Neill, S.L., 2009. Evidence for metabolic provisioning by a common invertebrate endosymbiont, Wolbachia pipientis, during periods of nutritional stress. PLoS pathogens, 5(4), p.e1000368.

Brune, A. and Ohkuma, M., 2010. Role of the termite gut microbiota in symbiotic digestion. In Biology of termites: a modern synthesis (pp. 439-475). Springer, Dordrecht.

Callahan, B.J., McMurdie, P.J., Rosen, M.J., Han, A.W., Johnson, A.J.A. and Holmes, S.P., 2016. DADA2: high-resolution sample inference from Illumina amplicon data. Nature methods, 13(7), pp.581-583.

Campbell, A.S., Ploetz, R.C., Dreaden, T.J., Kendra, P.E. and Montgomery, W.S., 2016. Geographic variation in mycangial communities of Xyleborus glabratus. Mycologia, 108(4), pp.657-667.

Cardoza, Y.J., Moser, J.C., Klepzig, K.D. and Raffa, K.F., 2008. Multipartite symbioses among fungi, mites, nematodes, and the spruce beetle, Dendroctonus rufipennis. Environmental Entomology, 37(4), pp.956-963.s

96

Casadevall, A. and Pirofski, L.A., 2000. Host-pathogen interactions: basic concepts of microbial commensalism, colonization, infection, and disease. Infection and immunity, 68(12), pp.6511-6518.

Christian, N., Whitaker, B.K. and Clay, K., 2015. Microbiomes: unifying animal and plant systems through the lens of community ecology theory. Frontiers in microbiology, 6, p.869.

Chrostek, E., Pelz-Stelinski, K., Hurst, G.D. and Hughes, G.L., 2017. Horizontal transmission of intracellular insect symbionts via plants. Frontiers in microbiology, 8.

Cole, J. R., Q. Wang, J. A. Fish, B. Chai, D. M. McGarrell, Y. Sun, C. T. Brown, A. Porras-Alfaro, C. R. Kuske, and J. M. Tiedje. 2014. Ribosomal Database Project: data and tools for high throughput rRNA analysis Nucl. Acids Res. 42(Database issue):D633- D642.

Cruden, D.L. and Markovetz, A.J., 1987. Microbial ecology of the cockroach gut. Annual Reviews in Microbiology, 41(1), pp.617-643.

Davis, T.S., Crippen, T.L., Hofstetter, R.W. and Tomberlin, J.K., 2013. Microbial volatile emissions as insect semiochemicals. Journal of chemical ecology, 39(7), pp.840-859.

Davis, T.S., 2015. The ecology of yeasts in the bark beetle holobiont: a century of research revisited. Microbial ecology, 69(4), pp.723-732.

De Bary, A., 1879. Die erscheinung der symbiose. Verlag von Karl J. Trübner.

De Beer, Z.W., Duong, T.A., Barnes, I., Wingfield, B.D. and Wingfield, M.J., 2014. Redefining Ceratocystis and allied genera. Studies in Mycology, 79, pp.187-219.

De Cáceres, M., Legendre, P. and Moretti, M., 2010. Improving indicator species analysis by combining groups of sites. Oikos, 119(10), pp.1674-1684.

Dedeine, F., Vavre, F., Fleury, F., Loppin, B., Hochberg, M.E. and Boulétreau, M., 2001. Removing symbiotic Wolbachia bacteria specifically inhibits oogenesis in a parasitic wasp. Proceedings of the National Academy of Sciences, 98(11), pp.6247-6252.

Doudoumis, V., Blow, F., Saridaki, A., Augustinos, A., Dyer, N.A., Goodhead, I., Solano, P., Rayaisse, J.B., Takac, P., Mekonnen, S. and Parker, A.G., 2017. Challenging the Wigglesworthia, Sodalis, Wolbachia symbiosis dogma in tsetse flies: Spiroplasma is present in both laboratory and natural populations. Scientific Reports, 7(1), p.4699.

Douglas, A.E. and Prosser, W.A., 1992. Synthesis of the essential amino acid tryptophan in the pea aphid (Acyrthosiphon pisum) symbiosis. Journal of Insect Physiology, 38(8), pp.565-568.

Douglas, A.E., 2007. Symbiotic microorganisms: untapped resources for insect pest control. TRENDS in Biotechnology, 25(8), pp.338-342.

97

Dreaden, T.J., Davis, J.M., De Beer, Z.W., Ploetz, R.C., Soltis, P.S., Wingfield, M.J. and Smith, J.A., 2014. Phylogeny of ambrosia beetle symbionts in the genus Raffaelea. Fungal biology, 118(12), pp.970-978.

Drooz, A.T., 1985. Insects of eastern forests. USDA For. Serv. Misc. Publ, 1426, p.608.

Edgar, R.C., 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics, 26(19), pp.2460-2461.

Endoh, R., Suzuki, M. and Benno, Y., 2008. Ambrosiozyma kamigamensis sp. nov. and A. neoplatypodis sp. nov., two new ascomycetous yeasts from ambrosia beetle galleries. Antonie van Leeuwenhoek, 94(3), pp.365-376.

Endoh, R., Suzuki, M., Okada, G., Takeuchi, Y. and Futai, K., 2011. Fungus symbionts colonizing the galleries of the ambrosia beetle Platypus quercivorus. Microbial ecology, 62(1), pp.106-120.

Engelstädter, J. and Hurst, G.D.D., 2006. Can maternally transmitted endosymbionts facilitate the evolution of haplodiploidy?. Journal of evolutionary biology, 19(1), pp.194- 202.

Farrell, B.D., Sequeira, A.S., O'Meara, B.C., Normark, B.B., Chung, J.H. and Jordal, B.H., 2001. The evolution of agriculture in beetles (Curculionidae: Scolytinae and Platypodinae). Evolution, 55(10), pp.2011-2027.

Fourie, A., Wingfield, M.J., Wingfield, B.D. and Barnes, I., 2015. Molecular markers delimit cryptic species in Ceratocystis sensu stricto. Mycological Progress, 14(1), p.1020.

Francke-Grosmann, H., 1967. Ectosymbiosis in wood-inhabiting insects. Associations of invertebrates, birds, ruminants, and other biota, pp.141-205.

Gandhi, K.J., Cognato, A.I., Lightle, D.M., Mosley, B.J., Nielsen, D.G. and Herms, D.A., 2010. Species composition, seasonal activity, and semiochemical response of native and exotic bark and ambrosia beetles (Coleoptera: Curculionidae: Scolytinae) in northeastern Ohio. Journal of economic entomology, 103(4), pp.1187-119

García‐Fraile, P., 2017. Roles of bacteria in the bark beetle holobiont–how do they shape this forest pest?. Annals of Applied Biology.

Genkai-Kato, M. and Yamamura, N., 1999. Evolution of mutualistic symbiosis without vertical transmission. Theoretical population biology, 55(3), pp.309-323.

Gharabigloozare, Y., 2015. Raffaelea spp. from five ambrosia beetles in the genera Xyleborinus and Cyclorhipidion (Coleoptera: Curcurlionidae: Scolytinae: Xyleborini).

Goldstein, D.B., 1994. Deleterious mutations and the evolution of male haploidy. The American Naturalist, 144(1), pp.176-183.

98

Grossart, H.P., Wurzbacher, C., James, T.Y. and Kagami, M., 2016. Discovery of dark matter fungi in aquatic ecosystems demands a reappraisal of the phylogeny and ecology of zoosporic fungi. Fungal Ecology, 19, pp.28-38.

Guimarães Jr, P.R., Pires, M.M., Jordano, P., Bascompte, J. and Thompson, J.N., 2017. Indirect effects drive coevolution in mutualistic networks. Nature, 550(7677), p.511.

Hamilton, W.D., 1967. Extraordinary sex ratios. Science, 156(3774), pp.477-488.

Harrington, T.C., 1981. Cycloheximide sensitivity as a taxonomic character in Ceratocystis. Mycologia, pp.1123-1129.

Harrington, T.C., Fraedrich, S.W. and Aghayeva, D.N., 2008. Raffaelea lauricola, a new ambrosia beetle symbiont and pathogen on the Lauraceae. Mycotaxon, 104(2), pp.399- 404.

Harrington, T.C., Aghayeva, D.N. and Fraedrich, S.W., 2010. New combinations in Raffaelea, Ambrosiella, and Hyalorhinocladiella, and four new species from the redbay ambrosia beetle, Xyleborus glabratus. Mycotaxon, 111(1), pp.337-361.

Hernandez-Hernandez, E.M., Fernández-Medina, R.D., Navarro-Escalante, L., Nuñez, J., Benavides-Machado, P. and Carareto, C.M., 2017. Genome-wide analysis of transposable elements in the coffee berry borer Hypothenemus hampei (Coleoptera: Curculionidae): description of novel families. Molecular Genetics and Genomics, 292(3), pp.565-583.

Herre, E.A., Knowlton, N., Mueller, U.G. and Rehner, S.A., 1999. The evolution of mutualisms: exploring the paths between conflict and cooperation. Trends in Ecology & Evolution, 14(2), pp.49-53.

Hom, E.F. and Murray, A.W., 2014. Niche engineering demonstrates a latent capacity for fungal-algal mutualism. Science, 345(6192), pp.94-98.

Hilgenboecker, K., Hammerstein, P., Schlattmann, P., Telschow, A. and Werren, J.H., 2008. How many species are infected with Wolbachia?–a statistical analysis of current data. FEMS microbiology letters, 281(2), pp.215-220.

Hill, M.O., Bunce, R.G.H. and Shaw, M.W., 1975. Indicator species analysis, a divisive polythetic method of classification, and its application to a survey of native pinewoods in Scotland. The Journal of Ecology, pp.597-613.

Hoffmann, A.A., Montgomery, B.L., Popovici, J., Iturbe-Ormaetxe, I., Johnson, P.H., Muzzi, F., Greenfield, M., Durkan, M., Leong, Y.S., Dong, Y. and Cook, H., 2011. Successful establishment of Wolbachia in Aedes populations to suppress dengue transmission. Nature, 476(7361), p.454.

99

Hosokawa, T., Koga, R., Kikuchi, Y., Meng, X.Y. and Fukatsu, T., 2010. Wolbachia as a bacteriocyte-associated nutritional mutualist. Proceedings of the National Academy of Sciences, 107(2), pp.769-774.

Hsiau, P.T. and Harrington, T.C., 2003. Phylogenetics and adaptations of basidiomycetous fungi fed upon by bark beetles (Coleoptera: Scolytidae). SYMBIOSIS- REHOVOT-, 34(2), pp.111-132.

Hulcr, J. and Cognato, A.I., 2010. Repeated evolution of crop theft in fungus‐farming ambrosia beetles. Evolution, 64(11), pp.3205-3212.

Hulcr, J. and Dunn, R.R., 2011. The sudden emergence of pathogenicity in insect– fungus symbioses threatens naive forest ecosystems. Proceedings of the Royal Society of London B: Biological Sciences, 278(1720), pp.2866-2873.

Hulcr, J., Rountree, N.R., Diamond, S.E., Stelinski, L.L., Fierer, N. and Dunn, R.R., 2012. Mycangia of ambrosia beetles host communities of bacteria. Microbial ecology, 64(3), pp.784-793.

Hulcr, J. and Cognato, A.I., 2013. Xyleborini of New Guinea, a taxonomic monograph (coleoptera: curculionidae: scolytinae). Entomological Society of America.

Ito, M. and Kajimura, H., 2017. Landscape‐scale genetic differentiation of a mycangial fungus associated with the ambrosia beetle, (Blandford)(Curculionidae: Scolytinae) in Japan. Ecology and evolution, 7(22), pp.9203- 9221.

Johnson, A.J., Kendra, P.E., Skelton, J. and Hulcr, J., 2016. Species diversity, phenology, and temporal flight patterns of Hypothenemus pygmy borers (Coleoptera: Curculionidae: Scolytinae) in South Florida. Environmental entomology, 45(3), pp.627- 632.

Johnson, A.J., McKenna, D.D., Jordal, B.H., Cognato, A.I., Smith, S.M., Lemmon, A.R., Lemmon, E.L.M. and Hulcr, J., 2018. Phylogenomics clarifies repeated evolutionary origins of inbreeding and fungus farming in bark beetles (Curculionidae, Scolytinae). Molecular phylogenetics and evolution.

Jones, K.G., Dowd, P.F. and Blackwell, M., 1999. Polyphyletic origins of yeast-like endocytobionts from anobiid and cerambycid beetles. Mycological research, 103(5), pp.542-546.

Jousselin, E., Clamens, A.L., Galan, M., Bernard, M., Maman, S., Gschloessl, B., Duport, G., Meseguer, A.S., Calevro, F. and Coeur D'Acier, A., 2016. Assessment of a 16S rRNA amplicon Illumina sequencing procedure for studying the microbiome of a symbiont‐rich aphid genus. Molecular ecology resources, 16(3), pp.628-640.

Kaltenpoth, M., 2009. Actinobacteria as mutualists: general healthcare for insects?. Trends in microbiology, 17(12), pp.529-535.

100

Kasson, M.T., Wickert, K.L., Stauder, C.M., Macias, A.M., Berger, M.C., Simmons, D.R., Short, D.P., DeVallance, D.B. and Hulcr, J., 2016. Mutualism with aggressive wood- degrading Flavodon ambrosius (Polyporales) facilitates niche expansion and communal social structure in Ambrosiophilus ambrosia beetles. Fungal Ecology, 23, pp.86-96.

Katoh, K., Misawa, K., Kuma, K.I. and Miyata, T., 2002. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic acids research, 30(14), pp.3059-3066.

Kawasaki, Y., Schuler, H., Stauffer, C., Lakatos, F. and Kajimura, H., 2016. Wolbachia endosymbionts in haplodiploid and diploid scolytine beetles (Coleoptera: Curculionidae: Scolytinae). Environmental microbiology reports, 8(5), pp.680-688.

Kendra, P.E., Montgomery, W.S., Niogret, J. and Epsky, N.D., 2013. An uncertain future for American Lauraceae: A lethal threat from redbay ambrosia beetle and laurel wilt disease (A review). American Journal of Plant Sciences, 4(03), p.727.

Kikuchi, Y., Hosokawa, T. and Fukatsu, T., 2007. Insect-microbe mutualism without vertical transmission: a stinkbug acquires a beneficial gut symbiont from the environment every generation. Applied and environmental microbiology, 73(13), pp.4308-4316.

Kim, K.H., Choi, Y.J., Seo, S.T. and Shin, H.D., 2009. Raffaelea quercus-mongolicae sp. nov. associated with Platypus koryoensis on oak in Korea. Mycotaxon, 110(1), pp.189-197.

Kirkendall, L.R., Biedermann, P.H. and Jordal, B.H., 2015. Evolution and diversity of bark and ambrosia beetles. Bark beetles: biology and ecology of native and invasive species. San Diego, California: Academic Press, pp.85-156.

Klepzig, K.D. and Wilkens, R.T., 1997. Competitive interactions among symbiotic fungi of the southern pine beetle. Applied and Environmental Microbiology, 63(2), pp.621- 627.

Kolařík, M. and Kirkendall, L.R., 2010. Evidence for a new lineage of primary ambrosia fungi in Geosmithia Pitt (: Hypocreales). Fungal Biology, 114(8), pp.676- 689.

Kolařík, M. and Jankowiak, R., 2013. Vector affinity and diversity of Geosmithia fungi living on subcortical insects inhabiting Pinaceae species in Central and Northeastern Europe. Microbial ecology, 66(3), pp.682-700.

Kostovcik, M., Bateman, C.C., Kolarik, M., Stelinski, L.L., Jordal, B.H. and Hulcr, J., 2015. The ambrosia symbiosis is specific in some species and promiscuous in others: evidence from community pyrosequencing. The ISME journal, 9(1), pp.126-138.

101

LaBonte, J.R., 2010, December. Eradication of an exotic ambrosia beetle, Xylosandrus crassiusculus (Motschulsky), in Oregon. In 21st US Department of Agriculture Interagency Research Forum on Invasive Species 2010 (p. 41).

Lamelas, A., Gosalbes, M.J., Manzano-Marín, A., Peretó, J., Moya, A. and Latorre, A., 2011. Serratia symbiotica from the aphid Cinara cedri: a missing link from facultative to obligate insect endosymbiont. PLoS genetics, 7(11), p.e1002357.

Li, Y., Simmons, D.R., Bateman, C.C., Short, D.P., Kasson, M.T., Rabaglia, R.J. and Hulcr, J., 2015. New fungus-insect symbiosis: culturing, molecular, and histological methods determine saprophytic Polyporales mutualists of Ambrosiodmus ambrosia beetles. PloS one, 10(9), p.e0137689.

Lindahl, B.D., Nilsson, R.H., Tedersoo, L., Abarenkov, K., Carlsen, T., Kjøller, R., Kõljalg, U., Pennanen, T., Rosendahl, S., Stenlid, J. and Kauserud, H., 2013. Fungal community analysis by high‐throughput sequencing of amplified markers–a user's guide. New Phytologist, 199(1), pp.288-299.

Lindsey, A.R., Werren, J.H., Richards, S. and Stouthamer, R., 2016. Comparative genomics of a parthenogenesis-inducing Wolbachia symbiont. G3: Genes, Genomes, Genetics, 6(7), pp.2113-2123.

Login, F.H., Balmand, S., Vallier, A., Vincent-Monégat, C., Vigneron, A., Weiss-Gayet, M., Rochat, D. and Heddi, A., 2011. Antimicrobial peptides keep insect endosymbionts under control. Science, 334(6054), pp.362-365.

Lynch, S.C., Twizeyimana, M., Mayorquin, J.S., Wang, D.H., Na, F., Kayim, M. et al., 2016. Identification, pathogenicity and abundance of Paracremonium pembeum sp. nov. and Graphium euwallaceae sp. nov.—two newly discovered mycangial associates of the polyphagous shot hole borer (Euwallacea sp.) in California. Mycologia, 108(2), pp.313-329.

MacConaill, L.E., Burns, R.T., Nag, A., Coleman, H.A., Slevin, M.K., Giorda, K., Light, M., Lai, K., Jarosz, M., McNeill, M.S. and Ducar, M.D., 2018. Unique, dual-indexed sequencing adapters with UMIs effectively eliminate index cross-talk and significantly improve sensitivity of massively parallel sequencing. BMC genomics, 19(1), p.30.

Madden, A.A., Epps, M.J., Fukami, T., Irwin, R.E., Sheppard, J., Sorger, D.M. and Dunn, R.R., 2018. The ecology of insect–yeast relationships and its relevance to human industry. Proc. R. Soc. B, 285(1875), p.20172733.

Malacrinò, A., Rassati, D., Schena, L., Mehzabin, R., Battisti, A. and Palmeri, V., 2017. Fungal communities associated with bark and ambrosia beetles trapped at international harbours. Fungal Ecology, 28, pp.44-52.

Mariño, Y.A., Verle Rodrigues, J.C. and Bayman, P., 2017. Wolbachia Affects Reproduction and Population Dynamics of the Coffee Berry Borer (Hypothenemus hampei): Implications for Biological Control. Insects, 8(1), p.8.

102

Mayers, C.G., McNew, D.L., Harrington, T.C., Roeper, R.A., Fraedrich, S.W., Biedermann, P.H., Castrillo, L.A. and Reed, S.E., 2015. Three genera in the Ceratocystidaceae are the respective symbionts of three independent lineages of ambrosia beetles with large, complex mycangia. Fungal biology, 119(11), pp.1075- 1092.

McCutcheon, J.P. and Moran, N.A., 2012. Extreme genome reduction in symbiotic bacteria. Nature Reviews Microbiology, 10(1), p.13.

McMurdie, P.J. and Holmes, S., 2013. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PloS one, 8(4), p.e61217.

Miller, M.A., Pfeiffer, W. and Schwartz, T., 2010, November. Creating the CIPRES Science Gateway for inference of large phylogenetic trees. In Gateway Computing Environments Workshop (GCE), 2010 (pp. 1-8).

Montllor, C.B., Maxmen, A. and Purcell, A.H., 2002. Facultative bacterial endosymbionts benefit pea aphids Acyrthosiphon pisum under heat stress. Ecological Entomology, 27(2), pp.189-195.

Moreira, L.A., Iturbe-Ormaetxe, I., Jeffery, J.A., Lu, G., Pyke, A.T., Hedges, L.M., Rocha, B.C., Hall-Mendelin, S., Day, A., Riegler, M. and Hugo, L.E., 2009. A Wolbachia symbiont in Aedes aegypti limits infection with dengue, Chikungunya, and Plasmodium. Cell, 139(7), pp.1268-1278.

Munson, M.A., Baumann, P. and Kinsey, M.G., 1991. Buchnera gen. nov. and Buchnera aphidicola sp. nov., a taxon consisting of the mycetocyte-associated, primary endosymbionts of aphids. International Journal of Systematic and Evolutionary Microbiology, 41(4), pp.566-568.

Nguyen, N.H., Smith, D., Peay, K. and Kennedy, P., 2015. Parsing ecological signal from noise in next generation amplicon sequencing. New Phytologist, 205(4), pp.1389- 1393.

Normark, B.B., 2004. Haplodiploidy as an outcome of coevolution between male-killing cytoplasmic elements and their hosts. Evolution, 58(4), pp.790-798.

Nunberg, M., 1951. Contribution to the knowledge of prothoracic glands of Scolytidae and Platypodidae (Coleoptera). In Annales Musei Zoologici Polonici (Vol. 14, No. 18). nakł. Państwowego Muzeum Zoologicznego.

O'Donnell, K. and Cigelnik, E., 1997. Two divergent intragenomic rDNA ITS2 types within a monophyletic lineage of the fungus Fusarium are nonorthologous. Molecular phylogenetics and evolution, 7(1), pp.103-116.

O'Donnell, K., Cigelnik, E. and Nirenberg, H.I., 1998. Molecular systematics and phylogeography of the Gibberella fujikuroi species complex. Mycologia, pp.465-493.

103

O’Donnell, K., Sink, S., Libeskind-Hadas, R., Hulcr, J., Kasson, M.T., Ploetz, R.C., Konkol, J.L., Ploetz, J.N., Carrillo, D., Campbell, A. and Duncan, R.E., 2015. Discordant phylogenies suggest repeated host shifts in the Fusarium–Euwallacea ambrosia beetle mutualism. Fungal Genetics and Biology, 82, pp.277-290.

Oliver, K.M., Russell, J.A., Moran, N.A. and Hunter, M.S., 2003. Facultative bacterial symbionts in aphids confer resistance to parasitic wasps. Proceedings of the National Academy of Sciences, 100(4), pp.1803-1807.

Oliver, K.M., Moran, N.A. and Hunter, M.S., 2005. Variation in resistance to parasitism in aphids is due to symbionts not host genotype. Proceedings of the National Academy of Sciences of the United States of America, 102(36), pp.12795-12800.

Palmer, J.M., Jusino, M.A., Banik, M.T. and Lindner, D.L., 2017. Non-biological synthetic spike-in controls and the AMPtk software pipeline improve fungal high throughput amplicon sequencing data. bioRxiv, p.213470.

Paraskevopoulos, C., Bordenstein, S.R., Wernegreen, J.J., Werren, J.H. and Bourtzis, K., 2006. Toward a Wolbachia multilocus sequence typing system: discrimination of Wolbachia strains present in Drosophila species. Current microbiology, 53(5), pp.388- 395.

Ploetz, R.C., Hulcr, J., Wingfield, M.J. and de Beer, Z.W., 2013. Destructive tree diseases associated with ambrosia and bark beetles: black swan events in tree pathology?. Plant Disease, 97(7), pp.856-872.

Popa, V., Déziel, E., Lavallée, R., Bauce, E. and Guertin, C., 2012. The complex symbiotic relationships of bark beetles with microorganisms: a potential practical approach for biological control in forestry. Pest management science, 68(7), pp.963- 975.

Prada, C., McIlroy, S.E., Beltrán, D.M., Valint, D.J., Ford, S.A., Hellberg, M.E. and Coffroth, M.A., 2014. Cryptic diversity hides host and habitat specialization in a gorgonian‐algal symbiosis. Molecular ecology, 23(13), pp.3330-3340.

Raffa, K.F., Gregoire, J.C. and Lindgren, B.S., 2015. Natural history and ecology of bark beetles. In Bark Beetles (pp. 1-40).

Roeper, R.A., Bunce, M.A., Harlan, J.E. and Bowker, R.G., 2017. Observations of Xyleborus affinis Eichhoff (Coleoptera: Curculionidae: Scolytinae) in Central Michigan. The Great Lakes Entomologist, 48(3), p.2.

Ross, M.G., Russ, C., Costello, M., Hollinger, A., Lennon, N.J., Hegarty, R., Nusbaum, C. and Jaffe, D.B., 2013. Characterizing and measuring bias in sequence data. Genome biology, 14(5), p.R51.

104

Rousset, F., Bouchon, D., Pintureau, B., Juchault, P. and Solignac, M., 1992. Wolbachia endosymbionts responsible for various alterations of sexuality in . Proc. R. Soc. Lond. B, 250(1328), pp.91-98.

Sabree, Z.L., Kambhampati, S. and Moran, N.A., 2009. Nitrogen recycling and nutritional provisioning by Blattabacterium, the cockroach endosymbiont. Proceedings of the National Academy of Sciences, 106(46), pp.19521-19526.

Scarborough, C.L., Ferrari, J. and Godfray, H.C.J., 2005. Aphid protected from pathogen by endosymbiont. Science, 310(5755), pp.1781-1781Schedl, W., 1962. Ein Beitrag zur Kenntnis der Pilzübertragungsweise bei xylomycetophagen Scolytiden (Coleoptera). In Ein Beitrag zur Kenntnis der Pilzübertragungsweise bei xylomycetophagen Scolytiden (Coleoptera) (pp. 363-387). Springer Berlin Heidelberg.

Schmidberger, J., 1836. Naturgeschichte des Apfelborkenkäfers Apate dispar. Beiträge zur Obstbaumzucht und zur Naturgeschichte der den Obstbäumen schädlichen Insekten, 4, pp.213-230.

Scott, J.J., Oh, D.C., Yuceer, M.C., Klepzig, K.D., Clardy, J. and Currie, C.R., 2008. Bacterial protection of beetle-fungus mutualism. Science, 322(5898), pp.63-63.

Silva, W.D., Mascarin, G.M., Romagnoli, E.M. and Bento, J.M.S., 2012. Mating behavior of the coffee berry borer, Hypothenemus hampei (Ferrari)(Coleoptera: Curculionidae: Scolytinae). Journal of insect behavior, 25(4), pp.408-417.

Simmons, D.R., Wilhelm de Beer, Z., Huang, Y.T., Bateman, C., Campbell, A.S., Dreaden, T.J., Li, Y., Ploetz, R.C., Black, A., Li, H.F. and Chen, C.Y., 2016. New Raffaelea species (Ophiostomatales) from the USA and Taiwan associated with ambrosia beetles and plant hosts. IMA fungus, 7(2), pp.265-273.

Six, D.L., 2003. Bark beetle-fungus symbioses. Insect symbiosis, 1, pp.97-114.

Skelton, J., Doak, S., Leonard, M., Creed, R.P. and Brown, B.L., 2016. The rules for symbiont community assembly change along a mutualism–parasitism continuum. Journal of Animal Ecology, 85(3), pp.843-853.

Smith, N.G., 2000. The evolution of haplodiploidy under inbreeding. Heredity, 84(2), pp.186-192.

Stamatakis, A., 2014. RAxML version 8: a tool for phylogenetic analysis and post- analysis of large phylogenies. Bioinformatics, 30(9), pp.1312-1313.

Stilwell, A.R., Smith, S.M., Cognato, A.I., Martinez, M. and Flowers, R.W., 2014. Coptoborus ochromactonus, n. sp.(Coleoptera: Curculionidae: Scolytinae), an emerging pest of cultivated balsa (Malvales: Malvaceae) in Ecuador. Journal of economic entomology, 107(2), pp.675-683.

105

Susoy, V. and Herrmann, M., 2014. Preferential host switching and codivergence shaped radiation of bark beetle symbionts, nematodes of Micoletzkya (Nematoda: Diplogastridae). Journal of evolutionary biology, 27(5), pp.889-898.

Taylor, M.J. and Hoerauf, A., 1999. Wolbachia bacteria of filarial nematodes. Parasitology Today, 15(11), pp.437-442.

Toju, H., Yamamichi, M., Guimarães Jr, P.R., Olesen, J.M., Mougi, A., Yoshida, T. and Thompson, J.N., 2017. Species-rich networks and eco-evolutionary synthesis at the metacommunity level. Nature ecology & evolution, 1(2), p.0024.

Úbeda, F. and Normark, B.B., 2006. Male killers and the origins of paternal genome elimination. Theoretical population biology, 70(4), pp.511-526.

Van Meer, M.M.M., Witteveldt, J. and Stouthamer, R., 1999. Phylogeny of the endosymbiont Wolbachia based on the wsp gene. Insect molecular biology, 8(3), pp.399-408.

Vanderpool, D., Bracewell, R.R. and McCutcheon, J.P., 2017. Know your farmer: Ancient origins and multiple independent domestications of ambrosia beetle fungal cultivars. Molecular ecology. van de Peppel, L.J.J., Aanen, D.K. and Biedermann, P.H.W., 2017. Low intraspecific genetic diversity indicates asexuality and vertical transmission in the fungal cultivars of ambrosia beetles. Fungal Ecology.

Vannini, A., Contarini, M., Faccoli, M., Valle, M.D., Rodriguez, C.M., Mazzetto, T., Guarneri, D., Vettraino, A.M. and Speranza, S., 2017. First report of the ambrosia beetle Xylosandrus compactus and associated fungi in the Mediterranean maquis in Italy, and new host–pest associations. EPPO Bulletin, 47(1), pp.100-103.

Vega, F.E., Benavides, P., Stuart, J.A. and O’Neill, S.L., 2002. Wolbachia infection in the coffee berry borer (Coleoptera: Scolytidae). Annals of the Entomological Society of America, 95(3), pp.374-378.

Vilgalys, R., and M. Hester. 1990. Rapid genetic identification and mapping of enzymatically amplified ribosomal DNA from several Cryptococcus species. Journal of bacteriology 172:4238-4246.

Vissa, S. and Hofstetter, R.W., 2017. The Role of Mites in Bark and Ambrosia Beetle- Fungal Interactions. In Insect Physiology and Ecology. InTech.

Wang, N., Stelinski, L.L., Pelz-Stelinski, K.S., Graham, J.H. and Zhang, Y., 2017. Tale of the Huanglongbing disease pyramid in the context of the citrus microbiome. Phytopathology, 107(4), pp.380-387.

106

Wood, S.L., 1982. The bark and ambrosia beetles of North and Central America (Coleoptera: Scolytidae), a taxonomic monograph. The bark and ambrosia beetles of North and Central America (Coleoptera: Scolytidae), a taxonomic monograph.

Wood, S.L., 1986. A reclassification of the genera of Scolytidae (Coleoptera). Great Basin Naturalist Memoirs, 10(1), p.2.

Wood, S.L. and Bright Jr, D.E., 1992. Index for Scolytidae. Great Basin naturalist memoirs, 13(1), p.16.

Zchori-Fein, E., Perlman, S.J., Kelly, S.E., Katzir, N. and Hunter, M.S., 2004. Characterization of a ‘Bacteroidetes’ symbiont in Encarsia wasps (Hymenoptera: Aphelinidae): proposal of ‘Candidatus Cardinium hertigii’. International Journal of Systematic and Evolutionary Microbiology, 54(3), pp.961-968.

Zug, R. and Hammerstein, P., 2014. Bad guys turned nice? A critical assessment of Wolbachia mutualisms in arthropod hosts. Biological Reviews, 90(1), pp.89-111.

107

BIOGRAPHICAL SKETCH

Craig spent his childhood birding and exploring the Great Lakes. He first began work in entomology as an undergraduate advised by Christina DiFonzo in her field crops entomology lab at Michigan State University (MSU). As Craig’s interests shifted to insect microbiology, he began research on mosquito-microorganism interactions under the guidance of Michael Kaufman and Edward Walker, and received his Bachelor of

Science from MSU in 2012. Craig then began research on the ambrosia symbiosis with

Jiri Hulcr at the University of Florida (UF). In 2014, he received his Master of Science in forestry at UF after studying the invasion ecology of ambrosia beetles and fungi. In

2018, he then received and his Ph.D. in entomology and nematology from UF for his work on the evolution of fungal and bacterial symbioses of bark and ambrosia beetles.

108