<<

Nova Southeastern University NSUWorks

HCNSO Student Theses and Dissertations HCNSO Student Work

7-21-2017 Bacterial Communities Associated with Healthy and Diseased cervicornis () Using High-Throughput Sequencing Charles Walton Nova Southeastern University, [email protected]

Follow this and additional works at: https://nsuworks.nova.edu/occ_stuetd Part of the Diseases Commons, Environmental and Microbial Commons, Marine Commons, and Atmospheric Sciences and Meteorology Commons, and the Other Genetics and Genomics Commons Share Feedback About This Item

NSUWorks Citation Charles Walton. 2017. Bacterial Communities Associated with Healthy and Diseased Acropora cervicornis (Staghorn Coral) Using High- Throughput Sequencing. Master's thesis. Nova Southeastern University. Retrieved from NSUWorks, . (449) https://nsuworks.nova.edu/occ_stuetd/449.

This Thesis is brought to you by the HCNSO Student Work at NSUWorks. It has been accepted for inclusion in HCNSO Student Theses and Dissertations by an authorized administrator of NSUWorks. For more information, please contact [email protected].

HALMOS COLLEGE OF NATURAL SCIENCES AND OCEANOGRAPHY

BACTERIAL COMMUNITIES ASSOCIATED WITH HEALTHY AND DISEASED ACROPORA CERVICORNIS (STAGHORN CORAL) USING HIGH-THROUGHPUT SEQUENCING

BY

CHARLES J. WALTON

SUBMITTED TO THE FACULTY OF NOVA SOUTHEASTERN UNIVERSITY OCEANOGRAPHIC CENTER IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE WITH A SPECIALTY IN:

MARINE BIOLOGY AND COASTAL ZONE MANAGEMENT

NOVA SOUTHEASTERN UNIVERSITY

JULY 21, 2017

Thesis of CHARLES J. WALTON

Submitted in Partial Fulfillment of the Requirements for the Degree of

Masters of Science:

Marine Biology and Coastal Zone Management

Nova Southeastern University Halmos College of Natural Sciences and Oceanography

July 2017

Approved:

Thesis Committee

Major Professor :______David S. Gilliam, Ph.D.

Committee Member :______Jose V. Lopez, Ph.D.

Committee Member :______Aurelien Tartar, Ph.D.

TABLE OF CONTENTS Acknowledgements ...... II

Abstract ...... III

List of Figures ...... V

List of Tables ...... VII

List of Appendix Figures ...... VIII

List of Appendix Tables ...... IX

Introduction ...... 1

Methods ...... 7 Sample Collection ...... 7 DNA Extraction ...... 8 16S rRNA Gene Amplification and Sequencing ...... 8 Quality Checking and Taxonomic Assignments ...... 9 Microbial Community Statistical Analyses ...... 12 Community Diversity (Alpha Diversity) ...... 12 Community Composition (Beta Diversity) ...... 13 Taxonomic Differential Abundance ...... 15

Results ...... 17 Microbial Community Diversity (Alpha diversity) ...... 17 Microbial Community Composition (Beta diversity) ...... 19 All Sample Types ...... 19 Tissue ...... 25 Taxonomic Differential Abundance ...... 31

Discussion ...... 46

Conclusions ...... 54

Citations ...... 56

Appendix: Figures ...... 78

Appendix: Tables ...... 85

I Acknowledgements Thank you to my advisor Dr. David S. Gilliam for the opportunity to work in the Coral Restoration and Monitoring lab and for having the confidence in my abilities to take on a project beyond the of the laboratory. Additionally, thank you Dave for the all of the excellent field and travel opportunities that allowed me to experience great places and meet many wonderful people. Thank you to Dr. Jose V. Lopez for his guidance and allowing me to use his laboratory for my research. I also thank Dr. Aurelien Tarter for his assistance in the project development and excellent and insightful comments.

Thank you to the Nova Southeastern University President’s Faculty Research and Development Grant for providing funding for this research.

Thank you to the members of the CRRAM lab past and present for their assistance with the fieldwork. Thank you to members of the Microbiology and Genetics laboratory for their assistance during my lab work and accepting me into their lab. A special thank you to Drs. Andia Chaves-Fonnegra and Cole Easson for their long chats and guidance throughout my research and their excellent comments.

A huge thank you to my wife Heather for her patience, understanding, and support throughout this experience. Finally, thank you to my family for their never ending support.

II Abstract Coral diseases were first noted in the 1960s and 1970s and have had major impacts globally on community structures. In the Caribbean, a major outbreak of has been considered responsible for the drastic decline of Caribbean Acroporids since the 1970s. In addition to white band disease, another more recently described condition known as rapid tissue loss (RTL) has had major impacts on Acropora cervicornis populations, specifically offshore Broward County Southeast . While these diseases have contributed to the population decline, determining their etiologies has been elusive.

Coral diseases have been characterized by shifts in their microbial counterparts within many levels of the coral host. While some coral diseases have had specific pathogens identified, research has not been able to determine pathogens for most. Evidence points toward bacterial causes for many diseases, but due to the complexity of the coral and the interaction with the environment, elucidating the causes has proven difficult. Many studies have examined the of specific diseases and determined some potential pathogens or at least taxa playing important roles in the disease, although none have looked at RTL. Recognizing the local affect of RTL on A. cervicornis, this study set out to gain a baseline understanding of the healthy and RTL affected of A. cervicornis.

16S rRNA gene sequencing was used to examine the microbiome of completely healthy colonies, healthy regions of diseased colonies, and the disease margin of diseased colonies. Analysis of four microbial diversity metrics revealed marked increases in diversity with respect to declining health states. Additionally, community dissimilarity analysis and analysis of differentially abundant taxa exhibited distinct microbial community structures due to coral health. Several highly abundant (Rickettsiales, Rhodobacteraceae) and a few low abundance (Bdellovibrionales) taxa were identified as primary drivers of the differences. Additionally, Piscirickettsiaceae, a known pathogen, was consistently associated with RTL and warrants further investigation. All of the taxa identified with in RTL have been associated with other Acroporid and non- Acroporid diseases throughout the Caribbean and the rest of the world. The consistent

III association of similar taxa for coral diseases around the world, including those found in this study, supports the recent ideas of non-specific primary pathogens.

While most disease studies, coral and otherwise, aim to determine a single pathogen for a single disease, this study and others suggest there could be a multitude of organisms responsible for the disease. Therefore understanding the interactions of the coral holobiont and the environment is important to understanding . While this study reveals significant changes in the bacterial community associated with RTL as well as some potential pathogens, the relationships appear complex and perhaps at a functional level rather than merely taxonomic. Furthermore, this study did not examine viruses, fungi, or , which could be possible pathogens. Therefore, to further develop an understanding of RTL and many other coral diseases it will be necessary to consider additional none-bacterial members of the holobiont as well as the bacterial functions and taxa coupled with the roles of environmental factors.

Keywords: Coral, 16S rRNA gene, Microbiome, Rapid Tissue Loss, White Syndrome, Microbial Diversity, , 454 Pyrosequencing, Acropora cervicornis

IV List of Figures Figure 1. Coral Holobiont Concept...... 3

Figure 2. WBD vs. RTL...... 6

Figure 3. Healthy (A) and Diseased (B) Acropora cervicornis Colonies...... 10

Figure 4. Rapid Tissue Loss Affected Acropora cervicornis ...... 11

Figure 5. Rapid Tissue Loss Disease Margin...... 11

Figure 6. Community Diversity...... 18

Figure 7. Tissue Diversity by Treatment...... 19

Figure 8. Bray-Curtis Dissimilarity Dendrogram of All Samples...... 22

Figure 9. Bray-Curtis Dissimilarity nMDS Plot of All Samples...... 23

Figure 10. Bray-Curtis Dissimilarity nMDS Plot by Site...... 23

Figure 11. Weighted UniFrac Distance Dendrogram of All Samples...... 24

Figure 12. Weighted UniFrac PCoA of All Samples...... 25

Figure 13. Bray-Curtis Dissimilarity Dendrogram of Tissue...... 28

Figure 14. Bray-Curtis Dissimilarity nMDS of Tissue...... 29

Figure 15. Weighted UniFrac Distance Dendrogram of Tissue...... 30

Figure 16. Weighted UniFrac Distance PCoA of Tissue...... 31

Figure 17. Relative Abundance of All Samples...... 32

Figure 18. Tissue Relative Abundance...... 34

Figure 19. Relative Abundance...... 35

Figure 20. Non-Proteobacteria Relative Abundance...... 36

V Figure 21. Proteobacteria Relative Abundance Without ...... 37

Figure 22. Healthy vs. Disease Differential Abundance...... 41

Figure 23. Healthy-on-Disease vs. Disease Differential Abundance...... 42

Figure 24. Significantly Differentially Abundant Orders...... 43

VI List of Tables Table 1. Tissue Mean Microbial Diversity...... 18

Table 2. SIMPER Results...... 27

Table 3. Counts of Differentially Abundant Taxa...... 38

Table 4. OTU Differential Abundance...... 44

Table 5. Association of Taxa with Health State...... 45

VII List of Appendix Figures Appendix Figure 1. Rarefaction Curves...... 78

Appendix Figure 2. Sediment Diversity by Site...... 79

Appendix Figure 3. Water Diversity by Site...... 79

Appendix Figure 4. Gap Statistic for Sample Type Clustering (BCD)...... 80

Appendix Figure 5. Gap Statistic for Sample Type Clustering (W-UniFrac)...... 80

Appendix Figure 6. Gap Statistic for Tissue Treatment Clustering (BCD)...... 81

Appendix Figure 7. Gap Statistic for Tissue Treatment Clustering (W-UniFrac)...... 81

Appendix Figure 8. Relative Abundance Taxa Summary...... 82

Appendix Figure 9. Proteobacteria Relative Abundance Without Alphaproteobacteria (free scales)...... 83

Appendix Figure 10. Healthy vs. Healthy-on-Disease Differential Abundance...... 84

VIII List of Appendix Tables Appendix Table 1. Dunn’s Test Summary...... 85

Appendix Table 2. Tissue ART-ANOVA and lsmeans post hoc...... 86

Appendix Table 3. BCD PERMANOVA of All Samples (Free)...... 87

Appendix Table 4. BCD PERMANOVA of All Samples (within Site)...... 87

Appendix Table 5. BCD Pairwise PERMANOVA of All Samples...... 87

Appendix Table 6. W-UniFrac PERMANOVA of All Samples (Free)...... 88

Appendix Table 7. W-UniFrac PERMANOVA of All Samples (within Site)...... 88

Appendix Table 8. W-UniFrac Pairwise PERMANOVA of All Samples...... 88

Appendix Table 9. BCD Pairwise PERMANOVA of Tissue...... 89

Appendix Table 10. W-UniFrac Pairwise PERMANOVA of Tissue...... 89

Appendix Table 11. Total Phyla Relative Abundance (RA)...... 90

Appendix Table 12. Class Level Relative Abundance by Type...... 91

Appendix Table 13. Tissue Phyla Relative Abundance...... 95

Appendix Table 14. Tissue Order Relative Abundance...... 96

Appendix Table 15. Differentially Abundant Taxa...... 99

Appendix Table 16. Taxa Relative Abundance Pairwise Comparisons...... 101

IX Introduction Globally coral reefs have been continuously declining for decades. The widespread decline has been driven by multiple natural and anthropogenic stressors including but not limited to predation, physical disturbance, competition, , enrichment, , and climate change (Zaneveld et al. 2016, Gardner et al. 2003, Bruno et al. 2007, Bruno and Valdivia 2016, Vega Thurber et al. 2009, Hoegh-Guldberg 1999, Yakob and Mumby 2011, Randall and van Woesik 2015, Bruno et al. 2003, Burge et al. 2014, Harvell et al. 2007, Harvell et al. 2002, Harvell et al. 2009, Maynard et al. 2015, Maynard et al. 2010, Green and Bruckner 2000). These stressors and their interactions have resulted in the decline of overall reef health as well as associated and disease.

Coral diseases were first documented in the 1960s and 1970s (Squires 1965, Antonius 1973, Garrett and Ducklow 1975, Dustan 1977) and have been heavily concentrated in the Caribbean (Green and Bruckner 2000). Coral diseases have dramatically increased in range and prevalence in the past few decades and are threatening and reshaping reefs worldwide (Aronson and Precht 2001, Weil and Rogers 2011, Bruno 2015, Bruckner 2016, Ward and Lafferty 2004, Harvell et al. 2007, Harvell et al. 2002, Harvell et al. 1999, Peters 1997, Goreau et al. 1998, Hayes and Goreau 1998, Weil 2004, Willis, Page, and Dinsdale 2004, Sweet and Bateman 2015). The number of coral diseases reported globally has massively increased since initial reports with about 20 coral diseases currently described (Sutherland et al. 2016, Sutherland, Porter, and Torres 2004, Weil and Rogers 2011, Mouchka, Hewson, and Harvell 2010, Lesser et al. 2007, Sussman et al. 2008, Weil 2004), but in some cases as many as 40 are reported when considering conditions or “types” (different presentations of the same disease) (Bruckner 2016, Riegl et al. 2009, Bruckner 2009). The rapid proliferation of coral diseases throughout the world demonstrates the dire need for understanding the pathology and etiology of diseases in order to mount an effort to combat them.

Since the first reports of coral diseases, research has steadily increased and evolved from traditional methods of disease research. Characterizing, isolating, and culturing pathogens

1 for fulfillment of Koch’s postulates has proven difficult as the majority of and possible pathogens are nonculturable (Richardson 1998, Rosenberg et al. 2007, Koch 1882, Evans 1976, Giovannoni and Rappe 2000). As a result, modern methods using molecular techniques have been employed to investigate diseases and their potential pathogens in a wide variety of marine organisms. Clone-based 16S rRNA gene sequencing (Barneah et al. 2007, Bourne et al. 2007, Casas et al. 2004, Closek et al. 2014, Cook 2006, Cook et al. 2013, Cooney et al. 2002, Croquer et al. 2013, Frias-Lopez et al. 2004, Frias-Lopez et al. 2002, Godwin et al. 2012, Kimes et al. 2013, Pantos et al. 2003, Sato, Willis, and Bourne 2010, Sekar et al. 2006, Smith et al. 2015, Sweet and Bythell 2012, 2015, Sweet, Croquer, and Bythell 2014, Wilson et al. 2012, Kellogg et al. 2012), along with the more recent development of high-throughput 16S rRNA gene sequencing and high-throughput microarrays (Apprill, Hughen, and Mincer 2013, Apprill, Weber, and Santoro 2016, Cardenas et al. 2012, Garcia et al. 2013, Gignoux-Wolfsohn and Volmer 2015, Kellogg et al. 2013, Kellogg et al. 2012, Lesser and Jarett 2014, Pollock et al. 2016, Pratte and Richardson 2016, Roder, Arif, Bayer, et al. 2014, Roder, Arif, Daniels, et al. 2014, Rodriguez-Lanetty et al. 2013, Sato, Willis, and Bourne 2013, Shaver et al. 2016, Sunagawa et al. 2009) have greatly advanced coral disease research by providing large amounts of data at relatively low costs.

As suggested by Hernandez-Agreda et al. (2016) and Neave et al. (2017) corals are ‘metaorganisms’ consisting of the coral animal, , viruses, Bacteria, , Fungi, and endolithic , collectively known as the coral holobiont (Rohwer et al. 2002, Knowlton and Rohwer 2003, Rosenberg et al. 2007, Wegley et al. 2007, Reshef et al. 2006). The complex relationships within the holobiont exist in a mutualistic balance (Figure 1), but when this balance is disrupted coral bleaching and/or disease can occur (Rohwer et al. 2002, Knowlton and Rohwer 2003, Ainsworth and Gates 2016, Reshef et al. 2006). Disruption to the homeostatic state of the holobiont can be a result of biological, chemical, or physical changes, both naturally occurring or anthropogenically driven (Thompson et al. 2015, Ainsworth and Gates 2016, Ainsworth, Thurber, and Gates 2010). These environmental changes therefore drive changes in the holobiont community and function. Using the previously discussed molecular techniques the bacterial component of the holobiont can be studied. High-throughput sequencing methods allow

2 for the comparison of targeted community structures of corals that appear healthy and those that have been affected by a disturbance resulting in bleaching or disease. Many current studies focus on the bacterial and archaeal communities of the holobiont because through the evolution of disease research, most evidence points to bacterial causes as illustrated by Sutherland et al. (2016).

Figure 1. Coral Holobiont Concept. The coral holobiont model proposed by Rohwer et al. (2002). The model suggests mutualistic relationships between the coral, their symbiotic algae, and the associated microbial community consisting of bacteria, fungi, , and endolithic algae. The question mark indicates Rohwer et al.’s (2002) suggestion of the relationship, which further work has supported. Additionally, recent studies have found viruses as important members of this relationship. Coral disease prevalence and severity has been on the rise worldwide. While many are being impacted by a multitude of diseases, one disease, white-band disease (WBD), has caused massive destruction throughout the Caribbean dating back to 1977 (Gladfelter et al. 1977, Dustan 1977). WBD is thought to be responsible for the massive Caribbean-wide die off of two Acropora species (Aronson and Precht 2001) resulting in these species being listed as critically endangered under the ICUN Red List (Aronson et al. 2008a, b) and threatened under the US Act (Hogarth 2006).

3 Research has continued to explore the potential causes and of WBD (Gladfelter 1982, Hemond and Vollmer 2010, Miller et al. 2014, Peters 1983, Randall and van Woesik 2015, Ritchie and Smith 1998, Ritchie and Smith 1995, Williams and Miller 2012, Kline and Vollmer 2011).

Acroporids, once the primary corals contributing to reef structure throughout the Caribbean, have been significantly impacted by WBD (Aronson and Precht 2001). More recently, especially off southeast Florida and the , another disease, rapid tissue loss (RTL) has had a large impact with increased prevalence and mortality (Aronson and Precht 2001, Williams and Miller 2005), Gilliam unpublished data). Williams and Miller (2005) described and coined the term ‘rapid tissue loss’ because the disease they observed presented differently from the two accepted descriptions of WBD and quickly progressed into necrosis. White-band disease is described (on A. palmata) as “a sharp line of advance where distally located, brown zooxanthella-bearing coral tissue is cleanly and completely removed from the skeleton, leaving a sharp white zone, about 1 cm wide that grades proximally into algal successional stages” (Gladfelter 1982), whereas WBD Type II has a bleaching margin preceding the tissue loss margin (Ritchie and Smith 1998) (Figure 2A). The disease signs Williams and Miller (2005) observed were “not consistent with either type of WBD” and were therefore termed ‘rapid tissue loss’. Rapid tissue loss is characterized by “acute tissue loss manifested as irregular, multifocal lesions along the branches with apparently healthy tissue remaining in between” (Williams and Miller 2005) (Figure 2B). While the signs are different from WBD, it should be noted that Williams and Miller (2005) do not believe RTL is a new disease as their description of RTL is similar to that described in Curaçao in 1981 (Bak and Criens 1981). While RTL appears to have been initially described in 1981, little research has been conducted into the etiology and transmission of RTL aside from the work of Williams and Miller (2005). The limited research may be a result of Bak and Criens’ (1981) disease description coinciding with the first descriptions of WBD and the widespread knowledge of WBD therefore resulting in their disease observations being believed as another account WBD. However, as Williams and Miller (2005) stated, RTL presents differently from WBD. Based on these observations as well as personal

4 observations of RTL and WBD off Broward County Florida, the two should be treated as different diseases until the etiologies are known.

Recognizing the distinct differences in disease presentation between WBD and RTL coupled with the documented rise in prevalence, this study examined the microbiome associated with healthy and RTL affected Acropora cervicornis. The branching morphology of A. cervicornis and RTL presenting on some branches while others on the same colony appeared healthy, necessitated examining the microbiomes of completely healthy colonies, the healthy branches of diseased colonies, and the disease margins of the RTL affected colonies. The objectives of this study were to: (1) compare the bacterial and archaeal diversity from all three A. cervicornis tissue health states as well as water and sediment samples (alpha diversity) at two different sites, (2) compare the community composition (dissimilarity) between A. cervicornis tissue in different health states and the environmental samples at two sites (beta diversity), (3) determine which taxa drive the differences, if any, between health states, and (4) determine which, if any, taxa or operational taxonomic units (OTUs) significantly change in abundance between health states, as those with increased abundance associated with disease samples could potentially be RTL pathogen(s). It is important to note that while this study does not determine the specific pathogen(s) of RTL, it can provide candidate pathogens and will provide baseline information and guidance for further work focusing on pathogen identification.

5

Figure 2. WBD vs. RTL. (A) White band disease has distinctly different lesions from (B) rapid tissue loss.

6 Methods

Sample Collection Two high-density A. cervicornis patches, BCA and Scooter (Walker et al. 2012), were selected for sample collection, as on-going monitoring revealed a high prevalence of rapid tissue loss (RTL) (Gilliam unpublished). At each site, three completely healthy colonies (Figure 3A), showing no signs of disease lesions, and three diseased (Figure 3B) colonies with a minimum of three diseased branches were identified via SCUBA. Following colony selection samples were collected under Florida Fish and Wildlife Conservation Commission Special Activity License-11-1327-SRP as follows.

For each colony type, tissue samples as well as surrounding water and sediment were collected. For each healthy colony three separate branches were collected, further designated as “H” colonies. Diseased colonies were partitioned into two sample types, the visually healthy regions of the diseased colony (HD) and the disease lesion (D) (Figure 4). From each diseased colony six separate branches were sampled, three HD and three D. The HD samples were taken from separate branches showing no signs of disease over the entire branch length. The D samples were taken at the disease lesion identified by stark white skeleton surrounded by patches of sloughing tissue as described by Williams and Miller (2005) (Figure 5).

Using angled cutting pliers, tissue samples were collected by clipping approximately 5 cm of the branch for H and HD samples and 5 cm including the tissue loss margin for D samples. The clipped fragments were placed in sterile individually labeled Nasco Whirl- Pak® bags. The fragments were taken to the boat, gently rinsed with sterile seawater, placed in new Whirl-Pak® bags and frozen in a cooler containing a dry ice ethanol (Et- OH) slurry. Water samples were collected by opening a sterile 1 L Nalgene™ within 0.5 m of each colony and were kept on ice on the boat. Sediment samples were collected from within 0.5 m of the base of each colony by briefly opening a sterile 50 mL Falcon™ tube and scooping it through the sediment. Sediment samples were frozen on the boat in

7 the dry ice Et-OH slurry. In total 54 tissue fragments, 12 water, and 18 sediment samples were collected.

Upon returning to the lab, all samples in the dry ice Et-OH slurry were transferred to -80° for storage. The water samples were filtered through 0.22 μm Sterivex™ filter units and the filter units were stored at -80°C.

DNA Extraction Prior to DNA extraction, water filters and sediment samples were thawed to pre-process for extraction protocols. Water filters were sectioned into thirds using sterile dissection scissors. One third was further cut into fine pieces and placed into the bead tubes for DNA extraction and the other two-thirds were placed in microcentrifuge tubes and stored at -80°C for future use. Approximately 50 mg of sediment was transferred to the bead tube for DNA extraction. The coral tissue samples were not thawed for pre-processing. Tissue samples were kept frozen and homogenized in liquid nitrogen using a sterile mortar and pestle. Once crushed to a fine powder approximately 50 mg of powder was transferred to the bead tubes for DNA extraction. Remaining powder was stored in microcentrifuge tubes at -80°C. Once samples were in the bead tubes, the extraction steps were identical for all sample types using the MO BIO PowerLyzer Powersoil DNA isolation kit according the manufacturer’s standard protocol (MO BIO, Carlsbad, CA, USA).

16S rRNA Gene Amplification and Sequencing The V4 regions of the 16S rRNA genes were amplified using the universal bacteria and archaea barcoded primer pair 515F and 806R (Caporaso et al. 2011, Bates et al. 2011, Walters et al. 2011). PCR reactions were performed in duplicate with each reaction containing 3 µL each of forward and reverse primer, 1 µL template DNA, 2.4 µL MgCl2, 2.4 µL dNTPs, 0.15 µL Taq polymerase (High Fidelity Taq, TaKARa Otsu, Shiga, Japan), and 3 µL buffer. Reactions were held at 94°C for 3 minutes for an initial denaturation step, followed by amplification for 30 cycles at 94°C for 45 s, 50°C for 60 s, 72°C for 90 s, and a final elongation step of 10 minutes at 72°C. PCR products were

8 visualized on a 1% agarose gel containing gel red. Successful reactions indicated by a clear band were pooled (2 - 25µL reactions per sample) and stored at -20°C until they were shipped for sequencing. Pooled reactions were sent to Advanced Genetic Technologies Center (AGTC) at the University of Kentucky where they were purified, quantified using PicoGreen dsDNA reagent (Invitrogen, Carlsbad, CA), and then sequenced on a 454 GS FLX sequencer (Roche). Due to complications in sequencing, some samples at AGTC (17 samples) needed to be re-sequenced and AGTC was not able to perform the sequencing. These samples were sent from AGTC to the Genomic Analysis and Sequencing Core Facility at Johns Hopkins Bloomberg School of Public Health (JHU). One sample arrived without any product in the microcentrifuge tube. DNA from this sample was re-amplified and sent to JHU. All 17 samples at JHU were purified with the Agencourt AmPure bead kit, PicoGreen quantified, quality assessed using an Agilent BioAnalyzer and then sequenced on a 454 GS Junior sequencer (Roche).

Quality Checking and Taxonomic Assignments Raw sequences were processed in the open source software, Quantitative Insights Into Microbial Ecology (QIIME) version 1.8 (Caporaso, Kuczynski, et al. 2010), using MacQIIME. Sequences were de-multiplexed and quality filtered to keep only sequences with mean quality scores > 25 and read lengths > 200 bp. Following quality filtering, denoising and chimera checking was completed using AmpliconNoise (Quince et al. 2011). AmpliconNoise is a multi-step denoising and chimera checking process employing PyroNoise and SeqNoise for denoising and Perseus for chimera removal. Operational taxonomic units (OTU) were assigned using the UCLUST (Edgar 2010) algorithm in MacQIIME, where OTUs were clustered at ≥ 97% similarity. Subsequently, a representative sequence for each OTU was selected and was assigned using the RDP Classifier 2.2 (Wang et al. 2007) against the Greengenes 13_8 database (McDonald et al. 2012, Werner et al. 2012). Sequences were then aligned to the Greengenes core reference alignment database (DeSantis et al. 2006) using PyNAST (Caporaso, Bittinger, et al. 2010) and the alignments were filtered. Following taxonomic assignment, a phylogenetic tree based on representative sequences was constructed using FastTree (Price, Dehal, and Arkin 2010) for use in analyses.

9

Figure 3. Healthy (A) and Diseased (B) Acropora cervicornis Colonies. Note that healthy colonies could have areas of old mortality provided it was less than 15% and they could not have any signs of recent mortality. Disease colonies were selected with a minimum of three diseased branches and three apparently healthy branches.

10

Figure 4. Rapid Tissue Loss Affected Acropora cervicornis Colony. The yellow-circled area indicates a healthy-on- diseased (HD) portion of a colony and the red-circled area indicates the area sampled as diseased. Note the HD circled area is just one of many on the colony and is to illustrate the definition of an HD.

Figure 5. Rapid Tissue Loss Disease Margin. The red-circled area indicates an example region used for disease sample.

11 Finally, an OTU table was generated and summarized to look for low read samples. Generally, about 1000 reads per sample is considered minimum depth for 454 studies. Examination of the reads per sample as well as rarefaction curves, which provide insight into sequencing depth (Appendix Figure 1), for these data revealed multiple samples falling below 1000 reads per sample. In order to discard low read samples, but not discard so many as to affect statistical power it was determined that samples with < 704 reads should be discarded from the analyses. The 704 read threshold was determined via cor.test in R (R Core Team 2014) checking for correlations between richness/diversity and sequencing depth. Removing samples with < 704 removed the significant correlations with sequencing depth. Therefore samples with < 704 reads were removed from the OTU table, which was then converted to a tab-delimited table for statistical analysis in the open-source statistical software R (R Core Team 2014).

Microbial Community Statistical Analyses For tissue samples, OTU profiles were taxonomically pooled by summing OTUs from each A. cervicornis branch within a colony for each health state, in equal weight, to capture the whole colony profile. This pooling method provides a whole colony representation rather than branch level and has been used in other coral studies as well as human microbiome studies (Vega Thurber et al. 2009, Li et al. 2012, Turnbaugh et al. 2009). Following pooling of the branches, taxa that were not present as two or more reads in at least two samples were removed as they are most likely a result of sequencing and/or PCR errors.

Community Diversity (Alpha Diversity) An OTU table for all samples was used to look for differences between sample types with tissue separated by treatment and separate OTU tables were created for each sample type (sediment, water, tissue). OTU abundances for each table were then converted to relative abundance using the R package phyloseq. Phyloseq was used to calculate the following diversity metrics: OTU Richness (S), the Shannon index (H’), Inverse Simpson’s index (D), and the Chao1 richness estimate.

12 All Sample Types Alpha diversity metrics were compared for all sample types accounting for treatment within the tissue samples using a Kruskal-Wallis rank sum test. This was used since none of the metrics met the assumptions for a one-way analysis of variance (ANOVA). Dunn’s test, correcting for multiple comparisons via Benjamini-Hochberg corrections, was performed using the R package dunn.test (Dinno 2016) to determine which sample types were different.

Sediment Shannon and Inverse Simpson’s indices were compared among sediment samples by site and colony treatment (health state of colony that sediment sample was collected near) using a two-way ANOVA. OTU Richness and Chao1 did not meet the assumptions of the ANOVA and were therefore compared using an aligned rank transform for nonparametric ANOVA (ART-ANOVA) using the R package ARTool (Kay and Wobbrock 2016, Wobbrock et al. 2011).

Water Two-way ANOVAs were used to compare OTU Richness, the Shannon index, and Chao1 Richness estimates for water samples by site and colony treatment. Inverse Simpson’s index was compared using an ART-ANOVA as described for the sediment.

Tissue OTU Richness, Shannon index, Inverse Simpson’s index, and Chao1 were compared among tissue samples by site and treatment (healthy, healthy-on-disease, and disease) using ART-ANOVAs, as all metrics failed to meet the assumptions for an ANOVA. Pairwise contrast comparisons were conducted for significant effects using least-squares means in the lsmeans R package (Lenth 2016).

Community Composition (Beta Diversity)

All Sample Types The OTU table, sample data, phylogenetic tree, and taxonomy table from QIIME were loaded into the phyloseq package in R (McMurdie and Holmes 2013) for dissimilarity

13 and distance calculations as well as plotting. OTUs were filtered to only include those with at least two reads found in a minimum of two samples in order to control for noise resulting from PCR and/or sequencing errors. OTU abundance was converted to relative abundance (RA) to reduce the potential for false positives in the analysis (McMurdie and Holmes 2014). To look at differences between the environmental samples and tissue, Bray-Curtis dissimilarity (BCD) and weighted UniFrac (Lozupone and Knight 2005, Lozupone et al. 2007) were calculated for the RA normalized OTU table. Weighted UniFrac was used in addition to BCD because it takes microbial phylogenetic distance into account and weights the effect based on abundance. Non-metric multidimensional scaling (nMDS) plots were generated to provide a graphical representation of the Bray- Curtis dissimilarity using the phyloseq, vegan, and ggplot2 packages in R (McMurdie and Holmes 2013, Oksanen et al. 2016, Wickham 2009). Weighted UniFrac distances were visualized using principle coordinate analysis (PCoA) plots generated similarly to the BCD nMDS plots. Hierarchical clustering (Johnson 1967) and partitioning around medoids (PAM) (Reynolds et al. 2006) were used to determine the optimal number of clusters for both BCD and weighted UniFrac. The gap statistic (Tibshirani, Walther, and Hastie 2001) was used to further support the clustering. The gap statistic and PAM were performed using the cluster package in R (Maechler et al. 2016). For both BCD and weighted UniFrac, Permutational Multivariate Analysis of Variance using distance matrices (PERMANOVA) was then used to look for community compositional differences by sample type, accounting for tissue treatment, site, and the interaction of the two. Following controlling for site, pairwise PERMANOVAs were performed to determine specific differences resulting from type * treatment interactions using the pairwise.perm.manova function in the R package RVAideMemoire (Hervé 2017). PERMANOVAs (9999 permutations) were performed using the adonis function in the R package vegan (McArdle and Anderson 2001, Oksanen et al. 2016).

Tissue Tissue samples were separated from the water and sediment samples to examine the community level differences due to health state (treatment) and site. Health states were compared pairwise (H vs. D, H vs. HD, and HD vs. D) because HD and D samples were from different parts of the same colony and are therefore not independent samples. An

14 OTU table was created for each treatment pair, OTUs were filtered just as they were in the analysis of all samples (at least two reads in at least two samples), and then converted to relative abundance (RA). Bray-Curtis dissimilarity and weighted UniFrac were calculated for each RA OTU table and then community level comparisons were performed. Cluster analyses (hierarchical, PAM, and gap statistic) were performed to determine the optimal number of clusters. Community composition comparisons were then made using PERMANOVA. For H vs. D and H vs. HD, treatment and site factors as well as their interaction were tested with 9999 permutations. Because the branches for HD vs. D were from the same colonies the model formula ~ Treatment * Site + Colony, with permutations constrained by colony to account for repeated measures, was used. The percent contribution of specific OTUs to the BCD between treatments was calculated using similarity percentage analysis (SIMPER) (Clarke 1993, Oksanen et al. 2016). Tissue BCD and weighted UniFrac were visualized using nMDS and PCoA plots, respectively, generated using phyolseq, vegan, and ggplot2.

Taxonomic Differential Abundance OTU tables for all samples and tissue only, were converted to relative abundance. Relative abundances of taxa were then summarized and plotted to show taxonomic composition of each type using phyloseq and ggplot2.

For the tissue OTU table of raw counts, taxa of the same type were agglomerated at the , class, order, family, and levels to look for differentially abundant taxa. Differential abundance of taxa was tested at each level as well as the OTU level by treatment using the DESeq function in the R package DESeq2 (Love, Huber, and Anders 2014, McMurdie and Holmes 2014). DESeq2 provides increased sensitivity with smaller datasets with even sizes (Weiss et al. 2017) and provides detection of autocorrelation between taxa. The DESeq function had to be implemented stepwise due to the presence of zeros in the count data, which results in geometric means of zero in the size factor estimation step of generalized linear model (GLM) fitting. Therefore, a modified geometric mean calculation was used taking the mean of only counts greater than or equal to one (McMurdie and Holmes 2014). After implementing the modified geometric mean, the data were fit to a Negative Binomial GLM and significant

15 differences (p-value < 0.05 with FDR adjustments) in OTU abundance were determined using the Wald test (Anders and Huber 2010, Love, Huber, and Anders 2014).

In addition to the detection of differentially abundant OTUs via DESeq2, relative abundances of taxa were compared by health state for tissue. Relative abundance comparisons complement DESeq2 differential abundance testing because DESeq2 uses a negative binomial model, which assumes taxon pairs are negatively correlated (Mosimann 1962). The agglomerated OTU tables were converted to relative abundance and taxa with mean RA < 0.1% were filtered from the data. Comparisons were made by treatment using the nonparametric Kruskal-Wallis test. To further examine taxon differences between tissue treatments pairwise comparisons were performed using the Wilcoxon rank sum test. The Wilcoxon rank sum test was implemented using the mt function in phyloseq and p-values were corrected via the Bejamini-Hochberg false discovery rate (fdr) method with a p-value less than 0.05 considered significant.

16 Results A total of 54 tissue branches (18 H, 18 HD, and 18 D), 12 water samples, and 36 sediment samples were sequenced and of those, 44 tissue branches, 11 water samples, and 36 sediment samples met quality standards. These samples resulted in a total of 461,725 reads with 200,694 from sediment, 169,348 from tissue (40,228 H, 44,351 HD, and 84,769 D), and 91,683 from water. After pooling tissue branches by colony as previously described, the 65 (12 water, 36 sediment, and 18 tissue) total sample sequence data revealed 5,021 unique OTUs (97% sequence similarity) ranging in length from 200 – 544 base pairs (bp). Tissue branches alone yielded 30.4% (1,528/5,021) of the total OTUs, 44 of which were unique compared to water and sediment samples.

Microbial Community Diversity (Alpha diversity) Among all sample types there were significant differences due to sample type for observed species (S) (OTU richness) (Kruskal-Wallis (KW): p = 2.42 x 10-8), Chao1 richness estimate (KW: p = 7.68 x 10-10), Shannon diversity (H’) (KW: p = 2.79 x 10-10), and Inverse Simpson’s diversity (D) (KW: p = 2.96 x 10-10) (Appendix Table 1). Dunn’s test revealed that sediment was significantly different from water and all tissue health states for Chao1 (Bejamini-Hochberg (BH) adjusted p < 0.0209), H’ (BH adjusted p < 0.0013), and D (BH adjusted p < 0.0003), but was not different from disease tissue for S (BH adjusted p = 0.1796 diseased tissue, BH adjusted p < 0.0008 other types) (Figure 6 & Appendix Table 1). Comparisons within each sample type by site, health state, and their interactions were performed to determine the main drivers of diversity differences within sample types. Sediment samples did not differ with respect to the health state of the neighboring colony and there was only a significant difference due to site in S (ART- ANOVA: p = 0.0162; Appendix Figure 2). Water samples were not different for any diversity metrics (KW: p > 0.05; Appendix Figure 3). Site did not have an effect on tissue diversity, but health state did, showing significant differences in S (ART-ANOVA: p = 0.0051), Chao1 (ART-ANOVA: p = 0.0051), H’ (ART-ANOVA: p = 0.0047), and D (ART-ANOVA: p = 0.0145) (Figure 7) primarily due to increased diversity in diseased tissue (Table 1; Appendix Table 2).

17

Figure 6. Community Diversity. Each panel is a different diversity metric, note the different scales. In each panel the points represent individual samples and the boxes represent five summary statistics. The bar in each box is the median diversity, the lower and upper hinges are the 1st and 3rd quartiles, the upper whiskers extend to the largest diversity measure no greater than 1.5*IQR (inter-quartile range), the lower whiskers extend to the smallest diversity measure at most 1.5*IQR, and points beyond the whiskers are outlying samples. For Observed species (S) letters indicate group significance, sample types with the same letters are not significantly different. For Chao1, Shannon (H’), and Inverse Simpson (D) the asterisks indicate samples significantly different from all others. Tissue: H = Healthy, HD = Healthy- on-Disease, and D = Disease.

Table 1. Tissue Mean Microbial Diversity. Mean (±SE) diversity measures of tissue for each treatment. S = Observed Richness, Chao1 richness estimate, H’ = Shannon diversity index, and D = Inverse Simpson diversity index.

Treatment S ± SE Chao1 ± SE H' ± SE D ± SE

Healthy 72.83 ± 20.38 72.83 ± 20.38 1.10 ± 0.57 3.85 ± 2.66

Healthy-on-Disease 103.67 ± 15.14 109.31 ± 19.30 1.38 ± 0.40 2.71 ± 1.07

Disease 466.00 ± 100.55 476.74 ± 104.43 3.69 ± 0.30 10.61 ± 1.85

18

Figure 7. Tissue Diversity by Treatment. Each panel is a different diversity metric for tissue samples grouped by health state, note the different scales. See Figure 6 for description of points and box and whiskers. Asterisks for Observed species, Chao1, and Shannon indicate treatments that are significantly different. For Inverse Simpson, treatments with the same letter are not significantly different.

Microbial Community Composition (Beta diversity)

All Sample Types To determine if A. cervicornis tissue microbial composition was different from the surrounding environment, two dissimilarity metrics were used. Bray-Curtis dissimilarity was used to examine the microbiome community composition taking abundance into account and weighted UniFrac was used to assess the composition accounting for microbial phylogenetic distance as well as OTU abundance. Unsupervised learning methods were employed to determine distinct clusters and the relationship with sample type. For BCD, hierarchical clustering identified three major clusters, which can be broken into more, but less obvious groups (Figure 8). Using partitioning around medoids (PAM) and the gap statistic with the one-standard error (SE) criteria as recommended by Tibshirani et al. (2001), seven discrete clusters are optimal. Seven is a large number of clusters relative to the sample size (65). Therefore, a less strict criterion of two standard errors (SE), which Tibshirani et al. (2001) note can be applied because one SE is

19 arbitrary, resulted in three discrete clusters (Appendix Figure 4). The three discrete clusters indicated that most of the community difference was due to sample type, although one cluster contained both sediment and water samples. Higher numbers of clusters, approaching the seven resulting from the one SE standard, led to individual samples being their own cluster, which could be a result of overfitting the cluster model.

For further understanding of the factors driving clustering, BCD was visualized via nMDS and statistically tested using a permutation multivariate analysis of variance (PERMANOVA). Tissue, regardless of treatment, did not overlap with environmental samples; sediment, while highly variable especially between sites, showed only slight overlap with water, which clustered tightly on its own (Figure 9). Significant differences were observed for type, site, type * site interaction, and type * treatment interaction, when comparing BCD with 9999 free permutations (PERMANOVA: p < 0.05; Appendix Table 3). To better understand the differences due to type, treatment, and the type * treatment interaction the factors were examined by blocking the permutations for site to provide within site differences. These results revealed that within sites, sample type had the strongest effect with some effect due to the type * treatment interaction (PERMANOVA: p < 0.05; Figure 10; Appendix Table 4). Additional analysis using pairwise PERMANOVAs showed that while the type * treatment interaction had an effect, primarily due to type, with treatment only different within tissue (Pairwise PERMANOVA: p < 0.01 (false discovery rate (fdr) adjusted); Appendix Table 5). Therefore, isolating the effects and pairwise analysis indicated that type was the main effect with co-variation of treatment with type as well as site by type.

While the BCD results are informative, weighted UniFrac also takes microbial phylogeny into account, meaning more or less related taxa affect the distance measures, and is therefore generally more robust than BCD for clustering of microbiome data. Similar to BCD, hierarchical clustering exhibited three distinct clusters (Figure 11. ), but the gap statistic and PAM, using either one SE or two SEs, provided four discrete clusters (Appendix Figure 5). The four clusters resulted primarily from sample type, consisting of one cluster of sediment, one of water and sediment, and two separate tissue clusters, likely due to treatment. Further visualization and analysis of weighted UniFrac via PCoA

20 and PERMANOVA supported three major clusters due to sample type and a fourth within tissue. As seen with BCD, tissue did not overlap with water or sediment and with the variation decomposed to just two-axes, the first axis explained 43.3% of the variation and the first two combined explained 67.3% (Figure 12). Significant differences were found due to type, site, treatment, and the following interactions: type * site, type * treatment, and site * treatment (PERMANOVA (free permutations): p < 0.05; Appendix Table 6). Controlling for site resulted in significant differences due primarily to treatment (PERMANOVA: p = 0.005) and some due to the type * treatment interaction (PERMANOVA: p = 0.010; Appendix Table 7). Again, as seen with BCD, pairwise PERMANOVAs exhibited differences driven by type, while treatment and/or site only had effects within sample types (Appendix Table 8). The results from both BCD dissimilarity and weighted UniFrac revealed A. cervicornis tissues were in fact different from the environment, which helped guide the within tissue focus related to the presence of RTL.

21

Figure 8. Bray-Curtis Dissimilarity Dendrogram of All Samples. The hierarchical clustering dendrogram of Bray-Curtis dissimilarity for all samples shows clustering of samples based on their microbial community composition. The colors of the sample names indicate the true sample type and the branch color represents the clustering group type.

22

Figure 9. Bray-Curtis Dissimilarity nMDS Plot of All Samples. Overall microbiome composition of A. cervicornis tissue and environmental samples. Treatment for water and sediment refers to the health state of the adjacent coral colony. The nMDS is based on Bray-Curtis dissimilarity. Eighty percent confidence ellipses were drawn for treatment (green, orange, and red).

Figure 10. Bray-Curtis Dissimilarity nMDS Plot by Site. Overall microbiome composition of A. cervicornis tissue and environmental samples split by sites. Eighty percent confidence ellipses are drawn for type (black dashes). Note the spread of sediment within Scooter, which accounts for the site differences.

23

Figure 11. Weighted UniFrac Distance Dendrogram of All Samples. The hierarchical clustering dendrogram of all samples based on weighted UniFrac distance shows clustering based on microbial community composition. The colors of the sample names indicate the true sample type and the branch color represents the clustering group type.

24

Figure 12. Weighted UniFrac PCoA of All Samples. Overall microbiome community composition using weighted UniFrac phylogentic distance for A. cervicornis tissue and environmental samples by type and treatment displayed via PCoA. Eighty percent confidence ellipses drawn for type (black dashes) and treatment (green, orange, and red).

Tissue Examination of the microbial communities of A. cervicornis tissue with respect to treatment was conducted in a similar manner to that for all sample types. The relative abundance of OTUs in tissue was used to generate two dissimilarity/distance matrices, BCD and weighted UniFrac. Partitioning around medoids (PAM) and the gap statistic for BCD suggested three optimal clusters when using the one SE standard or two optimal clusters using two SE, which are evident in the hierarchical clustering (Figure 13; Appendix Figure 6). Principal coordinate analysis of A. cervicornis tissue BCD plotted via nMDS showed two clusters due to health state, one consisted of diseased tissue, and the other healthy and healthy-on-disease (Figure 14.). Pairwise PERMANOVAs for health states were performed using the equation BCD ~ treatment * site for healthy vs. disease and healthy vs. healthy-on-disease. Comparisons for healthy-on-disease vs. disease were made using the equation BCD ~ treatment * site + colony to account for the repeated measures due to HD and D being from the same colonies. Significant

25 differences were found for H vs. D and HD vs. D due to treatment; site had no effect on tissue community composition (PERMANOVA: p < 0.05; Appendix Table 9).

A similarity percentage analysis (SIMPER) was used to determine the taxa driving the dissimilarity between tissue health states. An Alphaproteobacteria from the order Rickettsiales accounts for the majority of the dissimilarity between health-states, contributing to > 20% of the dissimilarity of healthy and healthy-on-disease compared to disease (Table 2). Alpha- and contributed the most to the BCD of health states, primarily due to members of the family Rhodobacteraceae, after Rickettsiales. The total dissimilarity between healthy and disease and healthy-on-disease and disease was 69.57% and 68.54%, respectively. Fifteen OTUs contributed > 0.5% of the divergence of healthy from disease tissue and healthy-on-disease from disease tissue; 11 of the 15 OTUs are shared between comparisons. While BCD was not significantly different between healthy and healthy-on-disease tissue, there was a total dissimilarity of 35.90% with nine OTUs accounting for > 0.5% of the dissimilarity.

26 Table 2. SIMPER Results. OTUs contributing > 0.5% to the dissimilarity of health states in A. cervicornis. Note: Orders in () cannot be classified to a lower taxonomic level.

OTU # Family (Order) Mean % Relative Abundance % Dissimilarity Contribution by Health State Healthy Healthy-on-Disease Disease H vs. D HD vs. D H vs. HD 150441 (Rickettsiales) 79.8669 70.7121 29.5700 26.7369 21.2168 13.6775 359953 Rhodobacteraceae 0.0165 0.0000 8.5273 4.2554 4.2636 - 816411 Rhodobacteraceae 0.0607 0.1915 4.9164 2.4278 2.3768 - 112750 Chlorobiaceae 0.0000 0.0031 4.0803 2.0402 2.0407 - 836059 Colwelliaceae 0.0240 0.0067 3.3464 1.6612 1.6698 - 2670730 Pseudoalteromonadaceae 0.0534 0.3006 1.9510 0.9488 0.8828 - 4436041 Bacteriovoracaceae 1.0424 0.2652 0.7842 0.6764 - 0.5205 4363665 Rhizobiaceae 1.3327 0.3953 0.0000 0.6663 - 0.7981 1145177 Rhodobacteraceae 0.0391 0.0107 1.3447 0.6620 0.6688 - 4449458 Moraxellaceae 1.2557 0.4856 0.0292 0.6292 - 0.7672 350486 Rhodobacteraceae 0.0000 0.0000 1.2401 0.6201 0.6201 - 2556149 Oceanospirillaceae 0.0165 0.0076 1.2211 0.6051 0.6080 - NCR.OTU4450 Cenarchaeaceae 1.1637 3.4092 0.3981 0.5763 1.7526 1.8802 4327205 Alteromonadaceae 0.7027 - 0.5744 0.5389 - - 303714 Ferrimonadaceae 0.0000 0.0000 1.0399 0.5200 0.5200 - 1995363 Staphylococcaceae 0.0041 3.8931 - - 1.9625 1.9459 346431 Rhodobacteraceae 0.1962 2.5251 - - 1.2595 1.2930 NR.OTU28 (Acidimicrobiales) 0.2418 1.7386 - - 0.9200 0.9128 NR.OTU30 Rhodobacteraceae 0.8908 1.2578 - - 0.7168 0.8011

27

Figure 13. Bray-Curtis Dissimilarity Dendrogram of Tissue. The hierarchical clustering dendrogram of A. cervicornis tissue based on Bray-Curtis dissimilarity shows clustering as a result of microbial community composition. The sample label color represents the sample health state.

28

Figure 14. Bray-Curtis Dissimilarity nMDS of Tissue. A. cervicornis tissue microbiome composition according to health state (healthy, healthy-on-disease, and disease). The nMDS was constructed based on Bray-Curtis dissimilarity. Eighty percent confidence ellipses are shown for site (gray dashes) and treatment (green, orange, and red). Weighted UniFrac distance was used to look at clustering by health state incorporating microbial phylogenies. PAM and the gap statistic using two SE indicate two, possibly three, optimal clusters (Appendix Figure 7). Examination of the hierarchical cluster dendrogram shows two clusters with an outlying sample, BCA.HD2.T, creating a third cluster consisting of this single sample (Figure 15). Principal coordinate analysis was conducted based on weighted UniFrac distances to determine whether microbiome composition differed by health state; this approach accounted for microbial phylogeny as well as abundance. Clustering was evident as a result of tissue health state and significant differences were observed between healthy and diseased tissue as well as healthy-on- disease and diseased tissue (PERMANOVA: p < 0.05; Figure 16; Appendix Table 10).

29

Figure 15. Weighted UniFrac Distance Dendrogram of Tissue. The hierarchical clustering dendrogram of A. cervicornis tissue based on weighted UniFrac distance shows the clustering by health state resulting from the microbial community composition. The sample label color represents the sample health state.

30

Figure 16. Weighted UniFrac Distance PCoA of Tissue. A. cervicornis tissue microbiome according to health state (healthy, healthy-on-disease, disease). The PCoA was conducted on the weighted UniFrac distance. Eighty percent confidence ellipses are drawn for site (gray dashes) and treatment (green, orange, red).

Taxonomic Differential Abundance Proteobacteria was the most abundant phylum in all sample types combined, accounting for 62.54 ± 17.32% (mean ± SD) of the relative abundance (Figure 17; Appendix Table 11). Of the OTUs belonging to Proteobacteria, Alphaproteobacteria were most abundant in tissue, regardless of health state (Appendix Table 12). Sediment was split with Alpha- and Gammaproteobacteria being nearly equal in relative abundance (Figure 17; Appendix Table 12). Similarly, water samples revealed nearly equal dominance of Alphaproteobacteria and Flavobacteriia, followed closely by Gammaproteobacteria (Figure 17 & Appendix Figure 8; Appendix Table 12). At the order level, Rhodobacterales, an Alphaproteobacteria, was more abundant in sediment, water, and diseased tissue compared to healthy and healthy-on-diseased tissue (Figure 17).

31

Figure 17. Relative Abundance of All Samples. Relative abundances of the top 50 most abundant OTUs found in tissue (by health state), sediment and water samples. Columns indicate phyla level classification in the samples and colors indicate order.

32

Examination of tissue samples only reveals Proteobacteria as the most dominant phyla in all three-health states (H = 94.37%, HD = 84.97%, D = 81.92%) (Figure 18). In healthy tissue, after Proteobacteria, Bacteroidetes (1.35%) and Firmicutes (1.35%) are the next most abundant bacterial phyla, followed by the archaeal phylum, Crenarchaeota (1.17%). Healthy-on-disease tissue is similar with Firmicutes (4.8%) and Crenarchaeota (4.2%) as second and third most abundant phyla (Figure 18; Appendix Table 13). Proteobacteria relative abundance decreased in diseased tissue relative to H and HD tissue, with Chlorobi (8.19%) and Bacteroidetes (4.53%), being the second and third most abundant phyla, respectively. Lower-level examination within Proteobacteria revealed that all three-health states were dominated by the order Rickettsiales (H = 79.94%, HD = 70.86%, D = 29.99%) (Figure 19). There is a drastic decline in the Rickettsiales abundance in disease tissue compared to H and HD tissue, with increases in Rhodobacterales (22.46%) and Chlorobiales (12.24%) observed (Figure 19,Figure 20; Appendix Table 14). Figure 21 provides better visualization of lower abundance members of Proteobacteria allowing the increase in Gammaproteobacteria and Deltaproteobacteria to be seen. Within Gammaproteobacteria, an increase is seen in Alteromonadales, Oceanospirillales, and Vibrionales between H, HD, and D. Other important taxa that exhibited changes between H, HD, and D at low abundances are Bdellovibrionales, Myxococcales (Deltaproteobacteria), and Saprospiraceae (Bacteroidetes) as well as Legionellales (Epsilonproteobacteriai) (Figure 20; Appendix Figure 9). Finally, a decreasing change from H to HD and to D was seen in the Gammaproteobacteria, Pseudomonadales (Figure 20).

33

Figure 18. Tissue Relative Abundance. Relative abundance of the top 50 OTUs found in A. cervircornis tissue. Each panel is a health state with the bars represtening phyla and colors indicating order.

34

Figure 19. Proteobacteria Relative Abundance. Relative abundance of Proteobacteria present in A. cervicornis tissue. Each panel is a class of Proteobacteria, each bar is a tissue health state and colors represent taxa orders.

35

Figure 20. Non-Proteobacteria Relative Abundance. Relative abundance of non-Proteobacteria in A. cervicornis tissue. Each panel is a class within phyla other than Proteobacteria, each bar is a tissue health state, and colors represent taxa orders. Other = sum of RA for OTUs present in < 0.1% RA.

36

Figure 21. Proteobacteria Relative Abundance Without Alphaproteobacteria. Relative abundance of Proteobacteria with Alphaproteobacteria removed to allow better visualization of lower abundance Proteobacteria. Each panel is a class of Proteobacteria, colored by taxa order, with bars representing A. cervicornis tissue health states.

37

Analyses were conducted to determine if any of the taxa that showed changes in abundance were statistically differentially abundant between tissue health states. Two methods were used, first DESeq2 differential abundance testing, which operates under the assumption that taxa pairs are negatively correlated. The second method was direct comparison of relative abundance via the Kruskal-Wallis non-parametric ANOVA and Wilcoxon rank sum tests for pairwise comparisons. All comparisons were conducted for each agglomerated taxonomic level as well as the OTU level. In total, 32 unique taxa were differentially abundant across all taxa levels. Of these significantly different taxa, 28 are more abundant in diseased tissue than healthy or healthy-on-disease. Table 3 shows the counts of differentially abundant taxa and OTUs at each level as well as those unique to each health state comparison.

Table 3. Counts of Differentially Abundant Taxa. Counts of differentially abundant taxa or OTUs are shown at each taxonomic level. Counts between levels are not additive between levels, i.e. differences at the order level may be from the same class. Numbers in parentheses indicate differences unique to the comparison.

Comparison Phylum Class Order Family Genus OTU H vs. D 0 5 (4) 17 (7) 10 (4) 1 (0) 18 (3) HD vs. D 1 (1) 4 (3) 15 (5) 15 (9) 3 (2) 17 (2) H vs. HD 1 (1) 0 0 0 0 0

The majority of differentially abundant taxa were found at the order and family levels with 32 and 25 differences, respectively. Additionally, most of these taxa were the same between comparisons, H vs. D and HD vs. D. Taxa from the phylum Proteobacteria constituted most of the differentially abundant taxa, primarily from the classes, Deltaproteobacteria and Gammaproteobacteria (Figure 22; Figure 23). Other notable taxa exhibiting differential abundance belong to the classes: Alpha- and Betaproteobacteria, Flavobacteriia, Planctomycetia, Verrucomicrobiae, and Fibrobacteria (Figure 24; Appendix Table 15). Examination of the differentially abundant taxa between H and D tissue at the order level revealed nine Proteobacteria significantly more abundant in D tissue, three Deltaproteobacteria, five Gammaproteobacteria, and one Alphaproteobacteria (Figure 22). One taxon was more abundant in H tissue, an Alphaproteobacteria belonging to the order Rickettsiales. Rickettsiales was also significantly more abundant in HD tissue compared to D tissue

38 (Figure 23). Similar to H tissue, comparison of HD to D tissue revealed significantly different taxa belonging primarily to the Deltaproteobacteria and Gammaproteobacteria classes (Figure 23; Appendix Table 15). Lower-level (order) analysis within the Proteobacteria for H and HD vs. D displayed significant differences in abundances of: Burkholderiales (Betaproteobacteria); Alteromonadales, unclassified HTCC2188, Legionellales, Oceanospirillales, Thiohalorhabdales, , and Vibrionales (Gammaproteobacteria); Desulfobacterales, unclassified GMD14H09, Myxococcales, Spirobacillales (Deltaproteobacteria); Rhodobacterales, and the previously mentioned Rickettesiales (Alphaproteobacteria) (Figure 24). Comparisons were made between H and HD tissue as well, revealing only a significant difference in the abundance of a single phylum, Actinobacteria, and no differences were exhibited at lower taxonomic levels (Appendix Figure 10).

In addition to examination of agglomerated taxonomic levels, health states were tested for differentially abundant OTUs. The OTU level resulted in 20 unique, differentially abundant OTUs, 15 of which were the same OTUs in H vs. D and HD vs. D comparisons. For H vs. D comparisons, 18 OTUs were significantly different, whereas 17 were different in HD vs. D. Similar to the agglomerated taxa results, most of the OTU level differences were OTUs belonging to Proteobacteria (19 OTUs). Two OTUs were not members of the Proteobacteria phylum, but belonged to Fibrobacteres, specifically Fibrobacteria; Ucp1540 (unclassified beyond the order level). Of the 19 OTUs belonging to Proteobacteria, 10 were Alphaproteobacteria, either belonging to the orders Rhodobacterales (9) or Rhizobiales (1). Additionally, seven belonged to Gammaproteobacteria and one each in Deltaproteobacteria and Epsilonproteobacteria (Table 4). Of the differentially abundant OTUs, all were more abundant in disease tissue.

In addition to DESeq differential abundance testing, direct comparisons of relative abundances were made using the Kruskal-Wallis nonparametric ANOVA test for overall health state differences and Wilcoxon rank sum tests for pairwise comparisons between health states. The relative abundance analysis results corroborated most the DESeq differential abundance results, with a few exceptions. Surprisingly, the order Rickettsiales was not identified as significantly different between any of the health states. Additionally,

39 the phylum Planctomycetes was detected as significantly more abundant via the Kruskal-Wallis test, but pairwise comparisons with FDR adjustments do not support this (Table 5; Appendix Table 16). While the Kruskal-Wallis results showed two phyla and four classes with significantly different RAs, pairwise comparisons did not exhibit any significant differences for H vs. D, HD vs. D, or H vs. HD at those same taxonomic levels. Comparing RA for H vs. HD there are no significant differences at any taxonomic level. Similar to the DESeq results the majority of RA differences were observed at the order and family levels (Table 5), specifically for HD vs. D pairwise comparisons (Appendix Table 16). At the OTU level, RA results differed from DESeq results with no OTUs identified as significantly different between health states.

40

Figure 22. Healthy vs. Disease Differential Abundance. Differential abundance of order level taxa for healthy tissue compared to diseased tissue. Each point is an order within the class indicated on the y-axis, colored by phylum. Values along the x-axis indicate the log2fold change of taxa where points < 0 are more abundant in disease tissue. The shape of the points indicates the statistical significance of taxa at FDR adjusted level of 0.05 (circles > 0.05, triangles  0.05). The size of the points represents the base mean abundance scaled by log10.

41

Figure 23. Healthy-on-Disease vs. Disease Differential Abundance. Differential abundance of order level taxa for healthy-on-disease tissue compared to diseased tissue. Each point is an order within the class indicated on the y-axis, colored by phylum. Values along the x-axis indicate the log2fold change of taxa where points < 0 are more abundant in disease tissue. The shape of the points indicates the statistical significance of taxa at FDR adjusted level of 0.05. The size of the points represents the base mean abundance scaled by log10.

42

Figure 24. Significantly Differentially Abundant Orders. Only taxonomic orders determined to be significantly different (fdr adjusted p < 0.05) are shown. Shape indicates the comparison, H vs. D or HD vs. D with taxonomic class indicated by color. Sizes of the points indicate the mean total abundance of the taxon for the comparison on a log10 scale. Points to the left (< 0) are significantly more abundant in disease tissue.

43

Table 4. OTU Differential Abundance. Differentially abundant OTUs between healthy (H) and disease (D), healthy-on-disease (HD) and disease, and healthy and healthy-on-disease tissue determined by DESeq analysis. All OTUs with a FDR < 0.05 (padj) are included in the table.

OTU baseMean log2FoldChange lfcSE stat pvalue padj Kingdom Phylum Class Order Family Genus Comparison NR.OTU82 7.7662 -6.70363 1.81806 -3.68725 2.27E-04 3.81E-03 Bacteria Fibrobacteres Fibrobacteria Ucp1540 H vs. D 836059 42.2941 -6.79006 1.53943 -4.41076 1.03E-05 4.60E-04 Bacteria Proteobacteria Gammaproteobacteria Alteromonadales Colwelliaceae Thalassomonas H vs. D 312009 6.1052 -6.52480 1.81649 -3.59198 3.28E-04 4.00E-03 Bacteria Proteobacteria Gammaproteobacteria Legionellales Francisellaceae H vs. D 2556149 14.9344 -6.51668 1.65993 -3.92588 8.64E-05 2.32E-03 Bacteria Proteobacteria Gammaproteobacteria Oceanospirillales Oceanospirillaceae Oleibacter H vs. D 358464 3.9452 -4.82753 1.76631 -2.73312 6.27E-03 4.67E-02 Bacteria Proteobacteria Gammaproteobacteria Oceanospirillales Oceanospirillaceae H vs. D 355163 5.9074 -5.98162 1.87322 -3.19324 1.41E-03 1.45E-02 Bacteria Proteobacteria Deltaproteobacteria Spirobacillales H vs. D 2670730 18.7532 -5.82615 1.38644 -4.20223 2.64E-05 8.85E-04 Bacteria Proteobacteria Gammaproteobacteria Vibrionales Pseudoalteromonadaceae H vs. D 4366904 9.1616 -6.34769 1.73624 -3.65600 2.56E-04 3.81E-03 Bacteria Proteobacteria Gammaproteobacteria H vs. D 4394104 3.7192 -4.26531 1.53594 -2.77701 5.49E-03 4.64E-02 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae H vs. D 816411 60.9533 -6.05661 1.60344 -3.77726 1.59E-04 3.54E-03 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae H vs. D 1146588 1.7602 -4.75369 1.71535 -2.77126 5.58E-03 4.64E-02 Bacteria Proteobacteria Alphaproteobacteria Rhizobiales Cohaesibacteraceae H vs. D 1145177 23.3653 -4.68669 1.70171 -2.75411 5.89E-03 4.64E-02 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae H vs. D 4350099 12.0953 -6.46906 1.78690 -3.62027 2.94E-04 3.94E-03 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae H vs. D 4327730 11.5347 -6.49229 1.77398 -3.65973 2.52E-04 3.81E-03 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae H vs. D 239119 3.0936 -5.08906 1.80320 -2.82223 4.77E-03 4.56E-02 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae H vs. D 359953 139.8235 -9.05873 1.68579 -5.37359 7.72E-08 1.03E-05 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae H vs. D 350486 16.4196 -7.89489 1.73722 -4.54456 5.50E-06 3.69E-04 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae H vs. D 513711 11.3014 -6.49859 1.86473 -3.48500 4.92E-04 5.50E-03 Bacteria Proteobacteria Epsilonproteobacteria Campylobacterales Campylobacteraceae Arcobacter H vs. D 836059 42.2941 -6.87233 1.54661 -4.44349 8.85E-06 2.71E-04 Bacteria Proteobacteria Gammaproteobacteria Alteromonadales Colwelliaceae Thalassomonas HD vs. D NR.OTU82 7.7662 -6.61744 1.81806 -3.63984 2.73E-04 6.27E-03 Bacteria Fibrobacteres Fibrobacteria Ucp1540 HD vs. D 312009 6.1052 -6.43635 1.81649 -3.54328 3.95E-04 7.27E-03 Bacteria Proteobacteria Gammaproteobacteria Legionellales Francisellaceae HD vs. D 2556149 14.9344 -5.57483 1.64479 -3.38939 7.00E-04 9.21E-03 Bacteria Proteobacteria Gammaproteobacteria Oceanospirillales Oceanospirillaceae Oleibacter HD vs. D 358464 3.9452 -5.22737 1.77152 -2.95078 3.17E-03 2.24E-02 Bacteria Proteobacteria Gammaproteobacteria Oceanospirillales Oceanospirillaceae HD vs. D 355163 5.9074 -5.91556 1.87322 -3.15797 1.59E-03 1.48E-02 Bacteria Proteobacteria Deltaproteobacteria Spirobacillales HD vs. D 4366904 9.1616 -4.75554 1.71613 -2.77109 5.59E-03 3.43E-02 Bacteria Proteobacteria Gammaproteobacteria HD vs. D 330888 7.4607 -5.80439 1.77400 -3.27193 1.07E-03 1.23E-02 Bacteria Proteobacteria Gammaproteobacteria Oceanospirillales Oleiphilaceae HD vs. D 369920 4.7084 -5.13806 1.79956 -2.85517 4.30E-03 2.83E-02 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae HD vs. D 1146588 1.7602 -4.56697 1.71535 -2.66241 7.76E-03 4.20E-02 Bacteria Proteobacteria Alphaproteobacteria Rhizobiales Cohaesibacteraceae Cohaesibacter HD vs. D 1145177 23.3653 -5.38934 1.70866 -3.15413 1.61E-03 1.48E-02 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae HD vs. D 4350099 12.0953 -5.51187 1.77957 -3.09731 1.95E-03 1.50E-02 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae HD vs. D 4327730 11.5347 -5.51483 1.76604 -3.12271 1.79E-03 1.50E-02 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae HD vs. D 239119 3.0936 -4.92010 1.80340 -2.72824 6.37E-03 3.66E-02 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae HD vs. D 359953 139.8235 -9.71035 1.70021 -5.71128 1.12E-08 1.03E-06 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae HD vs. D 350486 16.4196 -7.78837 1.73722 -4.48324 7.35E-06 2.71E-04 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae HD vs. D 513711 11.3014 -6.43663 1.86521 -3.45089 5.59E-04 8.57E-03 Bacteria Proteobacteria Epsilonproteobacteria Campylobacterales Campylobacteraceae Arcobacter HD vs. D

44

Table 5. Association of Taxa with Health State. Median relative abundance of taxa associated with A. cervicornis tissue health analyzed using Kruskal-Wallis (KW) nonparametric ANOVA. DF = degrees of freedom, H = Healthy, HD = Healthy-on-Disease, and D = Disease, FDR = false discovery rate. Analysis was performed for only taxa present in > 0.1% RA.

Taxonomic Taxon KW DF Median Abundance (%) FDR adjusted Level χ2 H HD D P valuea Phylum Bacteria; Fibrobacteres 12.64 2 0.00 0.04 0.82 2.20E-02 Bacteria; Planctomycetes 11.66 2 0.00 0.00 0.12 2.20E-02 Class Fibrobacteres; Fibrobacteria 12.64 2 0.00 0.00 0.12 1.56E-02 Firmicutes; Clostridia 10.47 2 0.00 0.00 0.62 3.07E-02 Planctomycetes; Planctomycetia 12.39 2 0.00 0.02 0.64 1.56E-02 Verrucomicrobia; Verrucomicrobiae 13.36 2 0.00 0.00 0.28 1.56E-02 Order Alphaproteobacteria; Rhodobacterales 9.40 2 1.25 0.73 20.76 2.93E-02 Clostridia; Clostridiales 10.47 2 0.00 0.00 0.57 2.49E-02 Deltaproteobacteria; Desulfobacterales 13.71 2 0.00 0.00 0.26 1.07E-02 Deltaproteobacteria; Myxococcales 13.19 2 0.00 0.00 1.17 1.07E-02 Deltaproteobacteria; Spirobacillales 9.71 2 0.00 0.00 0.25 2.76E-02 Fibrobacteria; Ucp1540 12.64 2 0.00 0.00 0.12 1.08E-02 Gammaproteobacteria; HTCC2188 9.69 2 0.00 0.00 0.22 2.76E-02 Gammaproteobacteria; Legionellales 13.71 2 0.00 0.00 0.46 1.07E-02 Gammaproteobacteria; Oceanospirillales 8.57 2 0.44 0.49 4.16 4.13E-02 Gammaproteobacteria; Otherb 13.10 2 0.00 0.00 0.71 1.07E-02 Gammaproteobacteria; Thiotrichales 11.31 2 0.00 0.00 0.49 1.84E-02 Gammaproteobacteria; Vibrionales 9.73 2 0.36 0.04 2.87 2.76E-02 Planctomycetia; Pirellulales 12.96 2 0.00 0.00 0.54 1.07E-02 Verrucomicrobiae; Verrucomicrobiales 13.36 2 0.00 0.00 0.28 1.07E-02 Family [Saprospirales]; Saprospiraceae 8.50 2 0.00 0.01 0.62 4.57E-02 Alteromonadales; Colwelliaceae 12.57 2 0.02 0.00 3.24 1.26E-02 Alteromonadales; Ferrimonadaceae 12.64 2 0.00 0.00 0.28 1.26E-02 Alteromonadales; OM60 9.26 2 0.00 0.02 0.75 3.82E-02 Clostridiales; Lachnospiraceae 12.64 2 0.00 0.00 0.25 1.26E-02 Flavobacteriales; Cryomorphaceae 9.20 2 0.00 0.03 0.34 3.82E-02 Legionellales; Francisellaceae 12.64 2 0.00 0.00 0.27 1.26E-02 Myxococcales; Otherb 12.96 2 0.00 0.00 0.71 1.26E-02 Oceanospirillales; Oceanospirillaceae 9.09 2 0.13 0.14 2.90 3.82E-02 Oceanospirillales; Oleiphilaceae 9.71 2 0.00 0.00 0.63 3.65E-02 Pirellulales; Pirellulaceae 12.96 2 0.00 0.00 0.54 1.26E-02 Rhizobiales; Hyphomicrobiaceae 8.66 2 0.00 0.00 0.27 4.47E-02 Rhodobacterales; Rhodobacteraceae 9.09 2 0.73 1.21 20.38 3.82E-02 Spirobacillales; Otherb 9.71 2 0.00 0.00 0.25 3.65E-02 Thiotrichales; Piscirickettsiaceae 11.31 2 0.00 0.00 0.46 2.14E-02 Ucp1540; Otherb 12.64 2 0.00 0.00 0.12 1.26E-02 Verrucomicrobiales; Verrucomicrobiaceae 13.36 2 0.00 0.00 0.28 1.26E-02 Vibrionales; Pseudoalteromonadaceae 10.89 2 0.01 0.14 2.00 2.40E-02 Genus Colwelliaceae; Thalassomonas 12.57 2 0.02 0.00 3.24 1.56E-02 Cryomorphaceae; Other 9.20 2 0.00 0.03 0.26 3.97E-02 Ferrimonadaceae; Ferrimonas 12.64 2 0.00 0.00 0.28 1.56E-02 Francisellaceae; Other 12.64 2 0.00 0.00 0.27 1.56E-02 Hyphomicrobiaceae; Other 9.55 2 0.00 0.00 0.27 3.97E-02 Hyphomonadaceae; Other 9.71 2 0.00 0.00 0.26 3.97E-02 Lachnospiraceae; Other 12.64 2 0.00 0.00 0.16 1.56E-02 Oceanospirillaceae; Oleibacter 9.22 2 0.00 0.06 1.17 3.97E-02 Oleiphilaceae; Other 9.71 2 0.00 0.00 0.63 3.97E-02 OM60; Other 9.26 2 0.00 0.02 0.74 3.97E-02 Pirellulaceae; Other 12.96 2 0.00 0.00 0.54 1.56E-02 Piscirickettsiaceae; Other 11.31 2 0.00 0.00 0.45 2.56E-02 Pseudoalteromonadaceae; Other 11.13 2 0.01 0.11 1.61 2.56E-02 Rhodobacteraceae; Other 8.78 2 0.73 1.17 20.18 4.61E-02 aOnly taxa with FDR < 0.05 are shown bTaxon could not be classified at lower levels. It is significant at lower taxonomic levels.

45 Discussion The presence of rapid tissue loss on colonies of Acropora cervicornis resulted in substantially different microbial communities relative to healthy colonies. Rapid tissue loss margins exhibited significantly higher diversity compared to healthy and healthy-on- disease A. cervicornis (Figure 7), which is consistent with other coral diseases (Apprill, Hughen, and Mincer 2013, Bourne et al. 2007, Closek et al. 2014, Croquer et al. 2013, Gignoux-Wolfsohn and Volmer 2015, Pantos and Bythell 2006, Pantos et al. 2003, Roder, Arif, Bayer, et al. 2014, Roder, Arif, Daniels, et al. 2014, Séré et al. 2013). It is worth noting that while there was no significant difference in Inverse Simpson diversity for H compared to D tissue, it is likely due to an outlier in H (Figure 7). This single sample is likely confounding the difference between H and D tissue and requires further investigation. Weighted UniFrac distance and Bray-Curtis dissimilarity analysis further support the differences between H and D tissue (Figure 14, Figure 16). SIMPER analysis, DESeq differential abundance testing, and comparisons of relative abundance reveal multiple taxa contributing to the shifts from healthy to RTL affected. Additionally, comparisons between sites did not reveal any site specificity independent of health state, which differs from a previous study of the A. cervicornis microbiome in the presence of WBD, where site specificity was apparent, especially in healthy A. cervicornis (Gignoux- Wolfsohn and Volmer 2015). The results of this study indicate that the healthy and disease microbiomes are conserved between sites and the changes are due to the presence of RTL, although it should be noted that sites with greater geographic separation may have a stronger effect. Shifts in the microbiome of A. cervicornis with the presence of RTL are exhibited through the decrease of Alphaproteobacteria and increases in Gammaproteobacteria, a class known to contain many pathogenic bacteria (Broszat et al. 2014, Williams et al. 2010). Additionally, many of the taxa and OTUs responsible for the differentiation of healthy, healthy-on-disease, and disease A. cervicornis have been associated with disease or control mechanisms of pathogens and diseases in previous studies and warrant further discussion.

Rickettsiales dominated all tissue health states in terms of relative abundance and was the largest contributor to BCD dissimilarity and was differentially more abundant in H and

46 HD compared to D. Rickettsia-like organisms (RLOs) have long been thought to be potential pathogens of Caribbean Acroporids, particularly with respect to WBD (Peters 1983, Miller et al. 2014). Casas et al. (2004) also showed a member of the order Rickettsiales, coral-associated Rickettsiales (CAR1), associated with WBD, but concluded it was most likely not a pathogen due to the consistent association with healthy A. cervicornis as well as being present in other healthy non-Acroporids. Similar to the conclusions of Casas et al. (2004) and Gignoux-Wolfsohn and Volmer (2015) the results of this study do not support Rickettsiales as the pathogen of RTL. This conclusion results from the consistent association and high relative abundance of Rickettsiales in healthy tissue and subsequent decrease in healthy-on-disease and diseased tissue. Miller et al. (2014) also found a Rickettsiales-like bacterium associated with healthy tissue in their study, which they attributed to the mucocytes already being infected, but not exhibiting signs of disease. While this is possible, it seems highly unlikely, as that would imply that all the healthy A. cervicornis from this study and those studied by Casas et al. (2004) and Gignoux-Wolfsohn and Volmer (2015) were actually diseased. Members of Rickettsiales are obligate intracellular symbionts, many of which are known pathogens, but there are members of Rickettsiales that exist in non-pathogenic relationships (Boscaro et al. 2013, Kellen, Hoffmann, and Kwock 1981, Schulz et al. 2016, Weinert et al. 2009, Kurlovs et al. 2014). Due to Rickettsiales being obligate intracellular symbionts, the observed declines in Rickettsiales abundance could simply be a result of the loss of . As RTL moves through a colony, coral tissue dies removing available host cells for Rickettsiales to occupy. Additionally, Peters (1983) and Miller et al. (2014) identified RLOs as potential pathogens via histological and microscopic methods and they were therefore not able to differentiate pathogenic RLOs from symbiotic ones. Recognizing the limitation of identification, it is important to point-out that there are RLOs that are not members of the order Rickettsiales or even the Alphaproteobacteria class. One such family, identified in this study with significantly increased relative abundance in disease tissue, was Piscirickettsiaceae (Table 5). Piscirickettsiaceae is a family in the class, Gammaproteobacteria, and order Thiotrichales, consisting of six genera (Fryer and Lannan 2007). RLOs thought to be from the order Rickettsiales were identified as potential pathogens in coho salmon (Fryer et al. 1990), but it was later determined that

47 these RLOs are actually Piscirickettsiaceae (Fryer and Hedrick 2003, Mauel and Miller 2002). RLOs from this family are now known to be the cause of piscirickettsiosis, a disease affecting salmonid (Fryer and Hedrick 2003). Piscirickettsiosis can present as skin lesions that progress to shallow ulcers. While comparing lesions on fish to those on corals is difficult, it is possible that the presentation of RTL is similar to that of piscirickettsiosis. The potential similarity in disease presentation between RTL and piscirickettsiosis coupled with Piscirickettsiaceae absent in healthy or healthy-on-disease tissue, but present and significantly more abundant in disease tissue could indicate a member of the Piscirickettsiaceae family as a possible pathogen in RTL. Also, given that Piscirickettsiaceae were initially identified as RLOs belonging to Rickettsiales, it is not out of the question to think that the RLOs identified histologically by Peters (1983) and Miller et al. (2014) could in fact be members of the Piscirickettsiaceae family. Additionally, Piscirickettsiaceae has a known association with other coral diseases. Sweet and Bythell (2015) identified Piscirickettsiaceae as one of 15 potential pathogens of white syndrome (WS) in from . The association of Piscirickettsiaceae with WS and RTL, diseases with similar presentations, as well as the similarities with piscirickettsiosis make Piscirickettsiaceae worthy of further investigation with respect to coral disease. Moreover, the continued presence of Rickettsiales in both healthy and diseased Acroporid tissue suggests that the members of Rickettsiales are more likely non-pathogenic symbionts. Recognizing that Rickettsiales is an order level classification consisting of many members at lower taxonomic levels, it is important to further understand their roles in disease dynamics. The decline in Rickettsiales with the presence of RTL and the increase of other taxa indicate a more complex microbial relationship than Koch and Hill’s “one pathogen = one disease” concept, if RTL is caused by bacteria. Additionally, higher taxonomic resolution of Rickettsiales is necessary to completely understand the organisms present and determine their roles in the healthy and disease microbiome. This study was not designed to look at other possible causes of RTL such as viruses, fungi, and protists. Therefore it is possible an organism from one of the previously mentioned groups is causing the disease and the shift in bacterial community is the result of tissue mortality causing a transition to a saprophytic bacterial community.

48 Many coral disease studies, specifically those on WS, have consistently found increased abundance of Rhodobacteraceae family members in diseased samples and this study is no exception (Sunagawa et al. 2009, Cardenas et al. 2012, Séré et al. 2013, Roder, Arif, Bayer, et al. 2014, Pratte and Richardson 2016, Pollock et al. 2017, Mouchka, Hewson, and Harvell 2010). Rhodobacteraceae increased significantly with the presence of RTL, similar to the increase seen by Pollock et al. (2017) in WS lesions of in the . Rhodobacteraceae are extremely common aquatic bacteria that exhibit high diversity both ecologically and metabolically (Pujalte et al. 2014). The majority of coral disease studies identifying Rhodobacteraceae attribute the increase in abundance to the family generally exhibiting opportunistic lifestyles rather than primary pathogens (Buchan, Gonzalez, and Moran 2005, Elifantz et al. 2013). The presence of Rhodobacteraceae in healthy, healthy-on-disease, and disease tissue with increasing abundance with respect to disease state supports the findings of these other studies, due to the high environmental abundance. Rhodobacteraceae is likely not a primary pathogen, but appears to play a role in RTL as well as many other coral diseases. Therefore, further understanding of the functional role members of the Rhodobacteraceae family play in disease is important. Additionally, this further supports the idea that complex community changes are occurring and there is not a single bacterial pathogen responsible RTL.

Other differentially abundant taxa and OTUs associated with RTL were Flavobacteriales, Alteromonadales, Oceanospirillales, and Vibrionales. Similar to Rhodobacteriales, members of all of these orders have been associated with other coral diseases. Gignoux- Wolfsohn and Volmer (2015) identified 27 Flavobacteriales OTUs associated with WBD in A. cervicornis and suggested one of the OTUs may be the primary pathogen of WBD. Similar to Piscirickettsiaceae, some Flavobacteria are known fish pathogens, which supports Gignoux-Wolfsohn and Volmer (2015) hypothesis that they are possible coral pathogens (Bullock, Hsu, and Shotts Jr 1986, Starliper 2011). Results from this study also showed increased abundance of Flavobacteriales, among several other families also identified by Gignoux-Wolfsohn and Volmer (2015). Furthermore, Gignoux-Wolfsohn and Volmer (2015) suggested the relationship of Pseudomonadales as an antagonist to Flavobacteria columnare shown in fish disease (Tiirola et al. 2002) to be possible in

49 WBD affected corals due to changes in Pseudomnadales exhibited in their WBD study. Similar to Gignoux-Wolfsohn and Volmer (2015), Pseudomonadales abundance decreased with RTL as Flavobacteriales increased, although not significantly. The similarities between RTL and WBD call into question the differentiation between these diseases and other WS. Many studies of WS as well as this study and those of WBD identify similar disease associated microbial communities around the world. Furthermore, Alteromonadales, significantly more abundant in RTL tissue, are common in the microbiomes of other coral diseases such as WBD (Gignoux-Wolfsohn and Volmer 2015), white plague (Sunagawa et al. 2009, Roder, Arif, Daniels, et al. 2014, Roder, Arif, Bayer, et al. 2014), and yellow band (Croquer et al. 2013). In fact, Thalassomonas loyana was identified as the causative agent in white plague-like disease on an reef (Thompson et al. 2006), again emphasizing the complexities of corals diseases. In addition to the previously mentioned fish pathogens, Francisellaceae is a family containing a bacterium of Francisella known to cause mortality in Norwegian cod (Nylund et al. 2006). Francisellaceae were significantly more abundant in RTL affected A. cervicornis than healthy and healthy-on-disease. Francisellaceae, contains only a single genus in which many members are known pathogens (Colquhoun et al. 2014). To further complicate matters, similar to Piscirickettsiales being RLOs and appearing similar to Rickettsiales histologically, Francisella spp. induce granulomatous infections much like RLOs (Nylund et al. 2006). The significant change in abundance of Francisellaceae in RTL suggests that it could also be playing a role in the disease, combined with the previously mentioned taxa.

Another genera, Vibrionales, has been documented to have increased abundances with other coral diseases and exhibits this trend with RTL. Many members belonging to the order Vibrionales, specifically the Vibrionaceae family, have been attributed to coral diseases (Ushijima et al. 2012, Cervino et al. 2008, Ben-Haim et al. 2003). In this study Vibrionaceae were not significantly differentially abundant, but did exhibit increased abundance with RTL. Oceanospirillales identified here and by Gignoux-Wolfsohn and Volmer (2015) associated with RTL and WBD, were also identified in white patch syndrome in the Western Indian (Séré et al. 2013). Although these disease associations exist with Oceanospirillales, it is unlikely they are pathogens as

50 belong to Oceanospirillales and are commonly associated with healthy corals around the world (Morrow et al. 2012, Carlos, Torres, and Ottoboni 2013, Neave et al. 2017, Glasl, Herndl, and Frade 2016). Additionally, Oceanospirillales and more importantly Endozoicomanos were shown to decrease in abundance with lesions in Porites allowing other bacteria such as Verrucomicrobiales, Rhodobacteriaceae, Colwelliaceae (Alteromonadales), and Flavobacteriaceae to increase in abundance (Glasl, Herndl, and Frade 2016, Morrow et al. 2012). This relationship further supports the complex dynamics of the microbiome communities, especially in the presence of disease.

Examination of changes to the most abundant taxa and OTUs is an important step in understanding the microbiome and the association with disease, but there are also low abundance or rare members that may play an important role as well. Bdellovibrionaceae, Myxocccales, and Saprospiraceae, are composed of known predatory bacteria and all were observed in low abundance. These predatory bacteria are primarily known to consume Gram-negative prey, specifically Vibrio spp. (Welsh et al. 2015). Although recently, Iebba et al. (2014) demonstrated that some species may consume Gram-positive prey as well. Sweet and Bythell (2015) observed Bacteriovorax (Bdellovribrionaceae) in white syndrome of A. muricata and suggested their importance in maintaining a healthy microbial population. These predatory bacteria are likely a control mechanism for healthy populations and changes to their population allow for infection (Chen et al. 2011). Not only are these predatory bacteria control mechanisms for the microbial populations, but they are subject to control by their surrounding environment (Chen et al. 2011, Patin et al. 2016, Welsh et al. 2015, Vega Thurber et al. 2009). The influence of the environment as a control on these populations, which in turn act as a control for the microbial community illustrates an important idea that an or interaction approach is necessary to understanding coral disease.

The similarities between RTL, WBD, white plague, and other white syndromes around the world show that there may not be a “single pathogen – single disease” concept for corals. Rather, diseases likely result from the breakdown of complex microbial interactions coupled with the interactions of the coral host and other members of the coral

51 holobiont (Archaea, Fungi, viruses, Symbiodinium spp., etc.). These breakdowns can lead to disease by non-specific pathogens or even a consortium of organisms similar to (Voss and Richardson 2006, Sekar et al. 2006). The coral microbiome is one component of the complex coral holobiont consisting of interactions within the microbiome as well as between other members of the holobiont and the surrounding environment. The positive and negative correlations of taxa with the disease, coupled with the importance of low abundance bacterivores highlight the need to take a systems level approach to understanding disease. Multiple studies have suggested that many of the highly abundant taxa in disease samples are likely secondary infections (Certner and Vollmer 2015, Hunt and Sharp 2014, Lesser and Jarett 2014, Lin et al. 2016, Lokmer et al. 2016, Meyer, Paul, and Teplitski 2014, Muller and van Woesik 2012, Muller and van Woesik 2014, Precht et al. 2016, Quistad et al. 2016, Randall et al. 2016, Rosenberg, Kellogg, and Rohwer 2007a, Rosenberg, Kellogg, and Rohwer 2007b, Sweet and Bythell 2016, Wright et al. 2016, Zvuloni et al. 2015). While this is likely true, there is some breakdown of a control mechanism allowing primary infection followed by secondary infection. Given the similarities between many of the white diseases across , it appears as though a single, specific, pathogen may not be responsible for each disease separately. As Sutherland et al. (2016) discussed, pathogens may be non-specific, and as environments and communities change they are able to infect. This has been termed the shifting etiologies hypothesis. In the review of many coral diseases worldwide, Sutherland et al. (2016) highlighted similarities between some diseases globally and recognized the inability to isolate the same pathogen from the same disease over time. Therefore, perhaps non-specific pathogens are responsible for the disease under the appropriate conditions. Taking this hypothesis and coupling it with the idea of the “disease triangle” put forth by Bruno (2015) introduces a systematic approach. The “disease triangle” concept is that “several factors are necessary – but alone are not sufficient – for an outbreak to occur” (Bruno 2015). From Bruno’s statement it can be inferred that the host, pathogen, and environment all play a role in infection (Bruno 2015). So, with non-specific pathogens present, environmental changes can incite changes to the host and existing microbiome allowing for infection by any opportunistic pathogens. Further tying these points together is the idea presented by Sweet and Bulling

52 (2017) applying the “pathobiome” concept to corals (Ryan 2013, Vayssier-Taussat et al. 2014). The “pathobiome” is a term used to describe the dynamics of the microbiome in response to stress and the onset of disease (Ryan 2013, Vayssier-Taussat et al. 2014, Sweet and Bulling 2017). This connection of the microbiome with the environment is important, especially in the face of climate change and other local stressors that have been linked to increases in disease (Harvell et al. 2002, Bruno et al. 2003, Aronson and Precht 2006, Bruno et al. 2007, Harvell et al. 2009, Mouchka, Hewson, and Harvell 2010, Yakob and Mumby 2011, Williams and Miller 2012, Burge et al. 2014, Lokmer and Mathias Wegner 2015, Randall and van Woesik 2015, Bruno 2015, Maynard et al. 2015, Zvuloni et al. 2015, Precht et al. 2016). The results from this study support the idea of a “pathobiome” approach as there are many taxa exhibiting changes in abundance, positive and negative, with the presence of RTL and are consistent with other diseases around the world. Recognizing the importance of low abundance bacterivores that are impacted by changes to their environment and the susceptibility of the coral host to environmental changes contributes to the stress component of the “pathobiome” concept. This environmental connection further highlights the complex and dynamic relationship between the environment, the coral host, the structuring of the microbiome and other members of the holobiont.

53 Conclusions Examination of the microbiome communities of the environment and Acropora cervicornis with the presence of rapid tissue loss revealed: (1) differences in diversity between the environment and coral as well as differences in the coral due to disease, (2) changes in the community composition with RTL as well as community differences between the coral and the environment, and (3) changes in the relative abundance of specific taxa and OTUs with RTL. Community and taxonomic changes from healthy to RTL affected colonies were consistent with changes observed in WBD as well as other white syndromes around the world. The significant decrease of Rickettsiales abundance between health states combined with increases in a multitude of other, often disease associated OTUs suggested a dynamic community change with RTL. Additionally, the changes to the low abundance bacterivores supported the idea of the community interaction as a whole.

Often studies seek a single pathogen as the cause of disease, but as continued research has shown coral diseases appear more complex. The coral holobiont is a community of many partners interacting harmoniously to create a healthy environment for all members, but disturbance can break down the relationship (McFall-Ngai et al. 2013). For example, changes to the environment can cause changes to the coral and the habitat it provides (Harvell et al. 2007). This results in changes that affect a control mechanism within the microbiome leading to a total breakdown of the community, therefore allowing the proliferation of pathogens leading to coral disease. This research supports the ideas that there is not one specific pathogen for a disease, but perhaps changing etiologies due to non-specific opportunistic pathogens influenced by environmental changes. Therefore, future research would benefit by using the “pathobiome” approach to understand the stressors triggering changes leading to the alterations of the microbiome resulting in disease, possibly due to a multitude of organisms. Additionally future work should also examine viruses, fungi, and protists as possible pathogens. The role of viruses in the coral holobiont is largely unknown, but recent work has shown the potential importance of these particles in the holobiont (Pollock et al. 2014, Soffer et al. 2014).

54 As climate change progresses and local stressors increase, changes to the reef are going to continue to occur. The stress from the environmental changes are compounding, continuing to stress the system will lead to continued increases in disease. While further investigations are needed to fully understand the dynamic “environment-coral-holobiont” relationship, it is imperative to recognize the interconnectivity and address controllable components of the interactions.

55 Citations Ainsworth, T. D., R. V. Thurber, and R. D. Gates. 2010. "The future of coral reefs: a microbial perspective." Trends Ecol Evol 25 (4):233-40. doi: 10.1016/j.tree.2009.11.001. Ainsworth, T., and R. D. Gates. 2016. "Corals' microbial sentinels." Science 352 (6293):1518-1519. doi: 10.1126/science.aad9957. Anders, S., and W. Huber. 2010. "Differential expression analysis for sequence count data." Genome Biol 11 (10):R106. doi: 10.1186/gb-2010-11-10-r106. Antonius, A. 1973. "New observations on coral destruction in reefs." Tenth Meeting of the Association of Island Marine Laboratories of the Caribbean. Apprill, A., K. Hughen, and T. Mincer. 2013. "Major similarities in the bacterial communities associated with lesioned and healthy corals." Environ Microbiol 15 (7):2063-72. doi: 10.1111/1462-2920.12107. Apprill, A., L. G. Weber, and A. E. Santoro. 2016. "Distinguishing between Microbial Habitats Unravels Ecological Complexity in Coral Microbiomes." mSystems 1 (5). doi: 10.1128/mSystems.00143-16. Aronson, R., A. Bruckner, J. Moore, B. Precht, and E. Weil. 2008a. "Acropora cervicornis." The IUCN Red List of Threatened Species 2008: e.T133381A3716457. Aronson, R., A. Bruckner, J. Moore, B. Precht, and E. Weil. 2008b. "Acropora palmata." The IUCN Red List of Threatened Species 2008: e.T133006A3536699. Aronson, Richard B., and William F. Precht. 2001. "White-band disease and the changing face of Caribbean coral reefs." Hydrobiologia (460):25-38. Aronson, Richard B., and William F. Precht. 2006. "Conservation, precaution, and Caribbean reefs." Coral Reefs 25 (3):441-450. doi: Doi 10.1007/S00338-006- 0122-9. Bak, Rolf P.M., and Stanley R. Criens. 1981. "Survival after fragmentation of colonies of Madracis mirabilis, Acropora palmata, and A. cervicornis () and the subsequent impact of a coral disease." Proceedings of the Fourth Internation Coral Reef Symposium 2:222-227.

56 Barneah, O., E. Ben-Dov, E. Kramarsky-Winter, and A. Kushmaro. 2007. "Characterization of black band disease in Red stony corals." Environ Microbiol 9 (8):1995-2006. doi: 10.1111/j.1462-2920.2007.01315.x. Bates, S. T., D. Berg-Lyons, J. G. Caporaso, W. A. Walters, R. Knight, and N. Fierer. 2011. "Examining the global distribution of dominant archaeal populations in soil." ISME J 5 (5):908-17. doi: 10.1038/ismej.2010.171. Ben-Haim, Y., F. L. Thompson, C. C. Thompson, M. C. Cnockaert, B. Hoste, J. Swings, and E. Rosenberg. 2003. "Vibrio coralliilyticus sp. nov., a temperature-dependent pathogen of the coral Pocillopora damicornis." Int J Syst Evol Microbiol 53 (Pt 1):309-15. doi: 10.1099/ijs.0.02402-0. Boscaro, V., M. Schrallhammer, K. A. Benken, S. Krenek, F. Szokoli, T. U. Berendonk, M. Schweikert, F. Verni, E. V. Sabaneyeva, and G. Petroni. 2013. "Rediscovering the genus Lyticum, multiflagellated symbionts of the order Rickettsiales." Sci Rep 3:3305. doi: 10.1038/srep03305. Bourne, David, Yuki Iida, Sven Uthicke, and Carolyn Smith-Keune. 2007. "Changes in coral-associated microbial communities during a bleaching event." ISME J 2 (4):350-363. doi: 10.1038/ismej.2007.112. Broszat, Melanie, Heiko Nacke, Ronja Blasi, Christina Siebe, Johannes Huebner, Rolf Daniel, and Elisabeth Grohmann. 2014. "Wastewater Irrigation Increases the Abundance of Potentially Harmful Gammaproteobacteria in Soils in Mezquital Valley, Mexico." Applied and Environmental Microbiology 80 (17):5282-5291. Bruckner, Andrew. 2009. "The global perspective of incidence and prevalence of coral diseases." Coral Health and Disease in the Pacific: Vision for Action:90. Bruckner, Andrew W. 2016. "History of Coral Disease Research." In Diseases of Coral, edited by Cheryl M. Woodley, Craig A. Downs, Andrew W. Bruckner, James W. Porter and Silvia B. Galloway. Hoboken, New Jersey: John Wiley & Sons, Inc. Bruno, J. F., and A. Valdivia. 2016. "Coral reef degradation is not correlated with local human population density." Sci Rep 6:29778. doi: 10.1038/srep29778. Bruno, John F. 2015. "Marine biology: The coral disease triangle." Nature Climate Change 5 (4):302-303. doi: 10.1038/nclimate2571.

57 Bruno, John F., Laura E. Petes, C. Drew Harvell, and Annaliese Hettinger. 2003. "Nutrient enrichment can increase the severity of coral diseases." Ecology Letters 6 (12):1056-1061. doi: Doi 10.1046/J.1461-0248.2003.00544.X. Bruno, John F., Elizabeth R. Selig, Kenneth S. Casey, Cathie A. Page, Bette L. Willis, C. Drew Harvell, Hugh Sweatman, and Amy M. Melendy. 2007. "Thermal Stress and Coral Cover as Drivers of Coral Disease Outbreaks." PLoS Biol 5 (6):e124. Buchan, A., J. M. Gonzalez, and M. A. Moran. 2005. "Overview of the marine roseobacter lineage." Appl Environ Microbiol 71 (10):5665-77. doi: 10.1128/AEM.71.10.5665-5677.2005. Bullock, GL, Ta-Chun Hsu, and EB Shotts Jr. 1986. "Columnaris disease of fishes." US Fish & Wildlife Publications 129. Burge, C. A., C. Mark Eakin, C. S. Friedman, B. Froelich, P. K. Hershberger, E. E. Hofmann, L. E. Petes, K. C. Prager, E. Weil, B. L. Willis, S. E. Ford, and C. D. Harvell. 2014. "Climate change influences on marine infectious diseases: implications for management and society." Ann Rev Mar Sci 6:249-77. doi: 10.1146/annurev-marine-010213-135029. Caporaso, J Gregory, Justin Kuczynski, Jesse Stombaugh, Kyle Bittinger, Frederic D Bushman, Elizabeth K Costello, Noah Fierer, Antonio Gonzalez Peña, Julia K Goodrich, Jeffrey I. Gordon, Gavin A Huttley, Scott T Kelley, Dan Knights, Jeremy E Koenig, Ruth E. Ley, Catherine A. Lozupone, Daniel McDonald, Brian D Muegge, Meg Pirrung, Jens Reeder, Joel R Sevinsky, Peter J. Turnbaugh, William A. Walters, Jeremy Widmann, Tanya Yatsunenko, Jesse Zaneveld, and Rob Knight. 2010. "QIIME allows analysis of high-throughput community sequencing data." Nature Methods 7 (5):335-336. doi: 10.1038/nmeth0510-335. Caporaso, J. G., K. Bittinger, F. D. Bushman, T. Z. DeSantis, G. L. Andersen, and R. Knight. 2010. "PyNAST: a flexible tool for aligning sequences to a template alignment." Bioinformatics 26 (2):266-7. doi: 10.1093/bioinformatics/btp636.

58 Caporaso, J. Gregory, Christian L. Lauber, William A. Walters, Donna Berg-Lyons, Catherine A. Lozupone, Peter J. Turnbaugh, Noah Fierer, and Rob Knight. 2011. "Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample." Proceedings of the National Academy of Sciences 108 (Supplement 1):4516-4522. doi: 10.1073/pnas.1000080107. Cardenas, A., R. Lm Rodriguez, V. Pizarro, L. F. Cadavid, and C. Arevalo-Ferro. 2012. "Shifts in bacterial communities of two Caribbean reef-building coral species affected by white plague disease." ISME J 6 (3):502-12. doi: 10.1038/ismej.2011.123. Carlos, Camila, Tatiana T. Torres, and Laura M. M. Ottoboni. 2013. "Bacterial communities and species-specific associations with the mucus of Brazilian coral species." Scientific Reports 3. doi: 10.1038/srep01624. Casas, Veronica, David I. Kline, Linda Wegley, Yanan Yu, Mya Breitbart, and Forest Rohwer. 2004. "Widespread association of a Rickettsiales-like bacterium with reef-building corals." Environmental Microbiology 6 (11):1137-1148. doi: 10.1111/j.1462-2920.2004.00647.x. Certner, R. H., and S. V. Vollmer. 2015. "Evidence for Autoinduction and Quorum Sensing in White Band Disease-Causing Microbes on Acropora cervicornis." Sci Rep 5:11134. doi: 10.1038/srep11134. Cervino, J. M., F. L. Thompson, B. Gomez-Gil, E. A. Lorence, T. J. Goreau, R. L. Hayes, K. B. Winiarski-Cervino, G. W. Smith, K. Hughen, and E. Bartels. 2008. "The Vibrio core group induces yellow band disease in Caribbean and Indo--building corals." J Appl Microbiol 105 (5):1658-71. doi: 10.1111/j.1365- 2672.2008.03871.x. Chen, H., R. Athar, G. Zheng, and H. N. Williams. 2011. "Prey bacteria shape the community structure of their predators." ISME J 5 (8):1314-22. doi: 10.1038/ismej.2011.4. Clarke, K. R. 1993. "Non-parametric multivariate analyses of changes in community structure." Australian Journal of Ecology 18:117-143.

59 Closek, C. J., S. Sunagawa, M. K. DeSalvo, Y. M. Piceno, T. Z. DeSantis, E. L. Brodie, M. X. Weber, C. R. Voolstra, G. L. Andersen, and M. Medina. 2014. "Coral transcriptome and bacterial community profiles reveal distinct Yellow Band Disease states in Orbicella faveolata." ISME J 8 (12):2411-22. doi: 10.1038/ismej.2014.85. Colquhoun, Duncan J., Pär Larsson, Samuel Duodu, and Mats Forsman. 2014. "The Family Francisellaceae." In The Prokaryotes: Gammaproteobacteria, edited by Eugene Rosenberg, Edward F. DeLong, Stephen Lory, Erko Stackebrandt and Fabiano Thompson, 287-314. Berlin, Heidelberg: Springer Berlin Heidelberg. Cook, G. M., J. P. Rothenberger, M. Sikaroodi, P. M. Gillevet, E. C. Peters, and R. B. Jonas. 2013. "A comparison of culture-dependent and culture-independent techniques used to characterize bacterial communities on healthy and white plague-diseased corals of the annularis species complex." Coral Reefs 32 (2):375-388. doi: 10.1007/s00338-012-0989-6. Cook, Geoffrey M. 2006. "A Comparative Analysis of Bacterial Communities Associated with Apparently Healthy and Disease Corals Inhabiting the Wider Caribbean Region." Ph.D., Environmental Science and Policy, George Mason University. Cooney, Rory P., Olga Pantos, Martin D. A. Le Tissier, Michael R. Barer, Anthony G. O'Donnell, and John C. Bythell. 2002. "Characterization of the bacterial consortium associated with black band disease in coral using molecular microbiological techniques." Environ Microbiol 4 (7):401-413. Croquer, A., C. Bastidas, A. Elliott, and M. Sweet. 2013. "Bacterial assemblages shifts from healthy to yellow band disease states in the dominant reef coral Montastraea faveolata." Environ Microbiol Rep 5 (1):90-6. doi: 10.1111/j.1758- 2229.2012.00397.x. DeSantis, T. Z., P. Hugenholtz, N. Larsen, M. Rojas, E. L. Brodie, K. Keller, T. Huber, D. Dalevi, P. Hu, and G. L. Andersen. 2006. "Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB." Appl Environ Microbiol 72 (7):5069-72. doi: 10.1128/AEM.03006-05. Dinno, A. 2016. “dunn.test: Dunn's Test of Multiple Comparisons Using Rank Sums R package version 1.3.2.”

60 Dustan, Phillip. 1977. "Vitality of reef coral populations off Key Largo, Florida: recruitment and mortality." Environmental 2 (1):51-58. Edgar, R. C. 2010. "Search and clustering orders of magnitude faster than BLAST." Bioinformatics 26 (19):2460-1. doi: 10.1093/bioinformatics/btq461. Elifantz, H., G. Horn, M. Ayon, Y. Cohen, and D. Minz. 2013. "Rhodobacteraceae are the key members of the microbial community of the initial formed in Eastern Mediterranean coastal seawater." FEMS Microbiol Ecol 85 (2):348-57. doi: 10.1111/1574-6941.12122. Evans, Alfred S. 1976. "Causation and Disease: The Henle-Koch Postulates Revisited." The Yale Journal of Biology and Medicine 49:175-195. Frias-Lopez, J., A. L. Zerkle, G. T. Bonheyo, and B. W. Fouke. 2002. "Partitioning of Bacterial Communities between Seawater and Healthy, Black Band Diseased, and Dead Coral Surfaces." Applied and Environmental Microbiology 68 (5):2214- 2228. doi: 10.1128/aem.68.5.2214-2228.2002. Frias-Lopez, Jorge, James S. Klaus, George T. Bonheyo, and Bruce W. Fouke. 2004. "Bacterial Community Associated with Black Band Disease in Corals." Appl Environ Microbiol 70 (10):5955-5962. Fryer, JL, and RP Hedrick. 2003. ": a Gram negative intracellular bacterial pathogen of fish." Journal of fish diseases 26 (5):251-262. Fryer, JL, CN Lannan, LH Garcés, JJ Larenas, and PA Smith. 1990. "Isolation of a rickettsiales-like organism from diseased coho salmon (Oncorhynchus kisutch) in Chile." Fish pathology 25 (2):107-114. Fryer, John L, and Catharine N Lannan. 2007. "Family II. Piscirickettsiaceae fam. nov." Bergey's Manual® of Systematic Bacteriology: Volume 2: The Proteobacteria, Part B: The Gammaproteobacteria 2:180. Garcia, Gizele D., Gustavo B. Gregoracci, Eidy de O. Santos, Pedro M. Meirelles, Genivaldo G. Z. Silva, Rob Edwards, Tomoo Sawabe, Kazuyoshi Gotoh, Shota Nakamura, Tetsuya Iida, Rodrigo L. de Moura, and Fabiano L. Thompson. 2013. "Metagenomic Analysis of Healthy and White Plague-Affected Mussismilia braziliensis Corals." Microb Ecol 65:1076-1086. doi: 10.1007/s00248-012-0161- 4).

61 Gardner, Toby A., Isabelle M. Côté, Jennifer A. Gill, Alastair Grant, and Andrew R. Watkinson. 2003. "Long-Term Region-Wide Declines in Caribbean Corals." Science 301 (5635):958. Garrett, Peter, and Hugh Ducklow. 1975. "Coral diseases in Bermuda." Nature 253 (5490):349-350. Gignoux-Wolfsohn, Sarah, and Steven V. Volmer. 2015. "Identification of Candidate Coral Pathogens on White Band Disease-Infected Staghorn Coral." PLoS ONE 10 (8). doi: 10.1371/journal.pone.0134416. Giovannoni, S., and M. Rappe. 2000. "Evolution, diversity, and molecular ecology of ." In Microbial ecology of the oceans: ecological and applied microbiology, edited by D. L. Kirchman. New York: Wiley. Gladfelter, W. B. 1982. "White-band disease in Acropora palmata: Implications for the structure and growth of shallow reefs." Bulletin of Marine Science 32 (2):639- 643. Gladfelter, WB, EH Gladfelter, RK Monahan, JC Ogden, and RF Dill. 1977. "Environmental studies of Buck Island Reef National Monument, St." Croix, USVI. National Park Service Rept. Glasl, B., G. J. Herndl, and P. R. Frade. 2016. "The microbiome of coral surface mucus has a key role in mediating holobiont health and survival upon disturbance." ISME J. doi: 10.1038/ismej.2016.9. Godwin, Scott, Elizabeth Bent, James Borneman, and Lily Pereg. 2012. "The Role of Coral-Associated Bacterial Communities in Australian Subtropical White Syndrome of Turbinaria mesenterina." PLoS ONE 7 (9):e44243. doi: 10.1371/journal.pone.0044243. Goreau, Thomas J, James Cervino, Maya Goreau, Raymond Hayes, Marshall Hayes, Laurie Richardson, Garriet Smith, Kalli DeMeyer, Ivan Nagelkerken, and Jaime Garzon-Ferrera. 1998. "Rapid spread of diseases in Caribbean coral reefs." Rev Biol Trop 46 (Suppl 5):157-171. Green, Edmund P., and Andrew W. Bruckner. 2000. "The significance of coral disease epizootiology for coral reef conservation." Biological Conservation 96:347-361.

62 Harvell, C. D., K. Kim, J. M. Burkholder, R. R. Colwell, P. R. Epstein, D. J. Grimes, E. E. Hofmann, E. K. Lipp, A. D. M. E. Osterhaus, R. M. Overstreet, J. W. Porter, G. W. Smith, and G. R. Vasta. 1999. "Emerging Marine Diseases--Climate Links and Anthropogenic Factors." Science 285 (5433):1505-1510. doi: 10.1126/science.285.5433.1505. Harvell, C. Drew, Eric Jordán-Dahlgren, Susan Merkel, Eugene Rosenberg, Laurie Raymundo, Garriet Smith, Ernesto Weil, and Bette Willis. 2007. "Coral disease, environmental drivers, and the balance between coral and microbial associates." Oceanography 20 (1):172-195. Harvell, C. Drew, Charles E. Mitchell, Jessica R. Ward, Sonia Altizer, Andrew P. Dobson, Richard S. Ostfeld, and Michael D. Samuel. 2002. "Climate Warming and Disease Risks for Terrestrial and Marine Biota." Science 296 (5576):2158- 2162. doi: 10.1126/science.1063699. Harvell, Drew, Sonia Altizer, Isabella M. Cattadori, Laura Harrington, and Ernesto Weil. 2009. "Climate change and wildlife diseases: When does the host matter the most?" Ecology 90 (4):912-920. doi: 10.1890/08-0616.1. Hayes, Raymond L, and Nora I Goreau. 1998. "The significance of emerging diseases in the tropical coral reef ecosystem." Rev. Biol. Trop 46 (Supl 5):173-185. Hemond, Elizabeth M., and Steven V. Vollmer. 2010. " and Connectivity in the Threatened Staghorn Coral (Acropora cervicornis) in Florida." PLoS ONE 5 (1):e8652. Hernandez-Agreda, A., W. Leggat, P. Bongaerts, and T. D. Ainsworth. 2016. "The Microbial Signature Provides Insight into the Mechanistic Basis of Coral Success across Reef Habitats." MBio 7 (4). doi: 10.1128/mBio.00560-16. Hervé, Maxime. 2017. “RVAideMemoire: Diverse Basic Statistical and Graphical Functions.” Hoegh-Guldberg, O. 1999. "Climate change, coral bleaching and the future of the world's coral reefs." Marine and Freshwater Research 50:839-866. Hogarth, William T. 2006. Endangered and Threatend Species: Final Listing Determinations for and Staghorn Coral. edited by Department of Commerce.

63 Hunt, John, and William Sharp. 2014. Developing a Comprehensive Strategy for Coral Restoration for Florida. Florida Fish and Wildlife Conservation Commission. Iebba, V., V. Totino, F. Santangelo, A. Gagliardi, L. Ciotoli, A. Virga, C. Ambrosi, M. Pompili, R. V. De Biase, L. Selan, M. Artini, F. Pantanella, F. Mura, C. Passariello, M. Nicoletti, L. Nencioni, M. Trancassini, S. Quattrucci, and S. Schippa. 2014. "Bdellovibrio bacteriovorus directly attacks Pseudomonas aeruginosa and Staphylococcus aureus Cystic fibrosis isolates." Front Microbiol 5:280. doi: 10.3389/fmicb.2014.00280. Johnson, Stephen C. 1967. "Hierarchical clustering schemes." Psychometrika 32 (3):241- 254. doi: 10.1007/bf02289588. Kay, M., and J.O. Wobbrock. 2016. “ARTool: Aligned Rank Transform for Nonparameteric Factorial ANOVAs 0.10.2 (R package).” Kellen, William R., Darlene F. Hoffmann, and Rosalie A. Kwock. 1981. "Wolbachia sp. (Rickettsiales: Rickettsiaceae) a symbiont of the almond moth, Ephestia cautella: Ultrastructure and influence on host fertility." Journal of Pathology 37 (3):273-283. doi: 10.1016/0022-2011(81)90087-2. Kellogg, C. A., Y. M. Piceno, L. M. Tom, T. Z. DeSantis, D. G. Zawada, and G. L. Andersen. 2012. "PhyloChip microarray comparison of sampling methods used for coral microbial ecology." J Microbiol Methods 88 (1):103-9. doi: 10.1016/j.mimet.2011.10.019. Kellogg, Christina A., Yvette M. Piceno, Lauren M. Tom, Todd Z. Desantis, Michael A. Gray, David G. Zawada, and Gary L. Andersen. 2013. "Comparing Bacterial Community Composition between Healthy and White Plague-Like Disease States in Using PhylochipsTM G3 Microarrays." PLoS ONE 8 (11). Kimes, N. E., W. R. Johnson, M. Torralba, K. E. Nelson, E. Weil, and P. J. Morris. 2013. "The Montastraea faveolata microbiome: ecological and temporal influences on a Caribbean reef-building coral in decline." Environ Microbiol 15 (7):2082-94. doi: 10.1111/1462-2920.12130. Kline, David I., and Steven V. Vollmer. 2011. "White Band Disease (type I) of Endangered Caribbean Acroporid Corals is Caused by Pathogenic Bacteria." Sci. Rep. 1.

64 Knowlton, Nancy, and Forest Rohwer. 2003. "Multispecies Microbial Mutualisms on Coral Reefs: The Host as a Habitat." The American Naturalist 162 (S4):S51-S62. Koch, R. 1882. Source Book of Medical History. New York: Dover Publications. Kurlovs, A. H., J. Li, D. Cheng, and J. Zhong. 2014. "Ixodes pacificus ticks maintain embryogenesis and hatching after antibiotic treatment of Rickettsia ." PLoS One 9 (8):e104815. doi: 10.1371/journal.pone.0104815. Lenth, Russell V. 2016. "Least-Squares Means: The R Package lsmeans." Journal of Statistical Software 69 (1):1-33. doi: 10.18637/jss.v069.i01. Lesser, M. P., and J. K. Jarett. 2014. "Culture-dependent and culture-independent analyses reveal no prokaryotic community shifts or recovery of Serratia marcescens in Acropora palmata with ." FEMS Microbiol Ecol 88 (3):457-67. doi: 10.1111/1574-6941.12311. Lesser, Michael P., John C. Bythell, Ruth D. Gates, Ron W. Johnstone, and Ove Hoegh- Guldberg. 2007. "Are infectious diseases really killing corals? Alternative interpretations of the experimental and ecological data." Journal of Experimental Marine Biology and Ecology 346:36-44. Li, Kelvin, Monika Bihan, Shibu Yooseph, and Barbara A. Methé. 2012. "Analyses of the Microbial Diversity across the Human Microbiome." PLoS ONE 7 (6). doi: 10.1371/journal.pone.2118.t001. Lin, Chorng-Horng, Chih-Hsiang Chuang, Wen-Hung Twan, Shu-Fen Chiou, Tit-Yee Wong, Jong-Kang Liu, and Jimmy Kuo. 2016. "Seasonal Changes in Bacterial Communities Associated with Healthy and Diseased Porites Coral in Southern Taiwan." Canadian Journal of Microbiology (ja). Lokmer, A., and K. Mathias Wegner. 2015. "Hemolymph microbiome of Pacific oysters in response to temperature, temperature stress and infection." ISME J 9 (3):670- 82. doi: 10.1038/ismej.2014.160. Lokmer, Ana, M. Anouk Goedknegt, David W. Thieltges, Dario Fiorentino, Sven Kuenzel, John F. Baines, and K. Mathias Wegner. 2016. "Spatial and Temporal Dynamics of Pacific Oyster Hemolymph across Multiple Scales." Frontiers in Microbiology 7. doi: 10.3389/fmicb.2016.01367.

65 Love, Michael I., Wolfgang Huber, and Simon Anders. 2014. "Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2." Genome Biology 15 (12). doi: 10.1186/s13059-014-0550-8. Lozupone, C. A., M. Hamady, S. T. Kelley, and R. Knight. 2007. "Quantitative and qualitative beta diversity measures lead to different insights into factors that structure microbial communities." Appl Environ Microbiol 73 (5):1576-85. doi: 10.1128/AEM.01996-06. Lozupone, C., and R. Knight. 2005. "UniFrac: a new phylogenetic method for comparing microbial communities." Appl Environ Microbiol 71 (12):8228-35. doi: 10.1128/AEM.71.12.8228-8235.2005. Maechler, Martin, Peter Rousseeuw, Anja Struyf, Mia Hubert, and Kurt Hornik. 2016. “cluster: Cluster Analysis Basics and Extensions.” Mauel, Michael J., and Debra L Miller. 2002. "Piscirickettsiosis and pscirickettsiosis-like infections in fish: a review." Veterinary Microbiology 87:279 - 289. Maynard, J. A., K. R. N. Anthony, C. D. Harvell, M. A. Burgman, R. Beeden, H. Sweatman, S. F. Heron, J. B. Lamb, and B. L. Willis. 2010. "Predicting outbreaks of a climate-driven coral disease in the Great Barrier Reef." Coral Reefs 30 (2):485-495. doi: 10.1007/s00338-010-0708-0. Maynard, Jeffrey, Ruben van Hooidonk, C. Mark Eakin, Marjetta Puotinen, Melissa Garren, Gareth Williams, Scott F. Heron, Joleah Lamb, Ernesto Weil, Bette Willis, and C. Drew Harvell. 2015. "Projections of climate conditions that increase coral disease susceptibility and pathogen abundance and virulence." Nature Climate Change 5 (7):688-694. doi: 10.1038/nclimate2625. McArdle, Brian H., and Marti J. Anderson. 2001. "Fitting Multivariate Models to Community Data: A Comment on Distance-Based Redundancy Analysis." Ecology 82 (1):290-297. McDonald, D., M. N. Price, J. Goodrich, E. P. Nawrocki, T. Z. DeSantis, A. Probst, G. L. Andersen, R. Knight, and P. Hugenholtz. 2012. "An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea." ISME J 6 (3):610-8. doi: 10.1038/ismej.2011.139.

66 McFall-Ngai, M., M. G. Hadfield, T. C. Bosch, H. V. Carey, T. Domazet-Loso, A. E. Douglas, N. Dubilier, G. Eberl, T. Fukami, S. F. Gilbert, U. Hentschel, N. King, S. Kjelleberg, A. H. Knoll, N. Kremer, S. K. Mazmanian, J. L. Metcalf, K. Nealson, N. E. Pierce, J. F. Rawls, A. Reid, E. G. , M. Rumpho, J. G. Sanders, D. Tautz, and J. J. Wernegreen. 2013. " in a bacterial world, a new imperative for the life sciences." Proc Natl Acad Sci U S A 110 (9):3229-36. doi: 10.1073/pnas.1218525110. McMurdie, P. J., and S. Holmes. 2013. "phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data." PLoS One 8 (4):e61217. doi: 10.1371/journal.pone.0061217. McMurdie, Paul J., and Susan Holmes. 2014. "Waste Not, Want Not: Why Rarefying Microbiome Data Is Inadmissible." PLoS Comput Biol 10 (4):e1003531. doi: 10.1371/journal.pcbi.1003531. Meyer, J. L., V. J. Paul, and M. Teplitski. 2014. "Community shifts in the surface microbiomes of the coral Porites astreoides with unusual lesions." PLoS One 9 (6):e100316. doi: 10.1371/journal.pone.0100316. Miller, M. W., K. E. Lohr, C. M. Cameron, D. E. Williams, and E. C. Peters. 2014. "Disease dynamics and potential mitigation among restored and wild staghorn coral, Acropora cervicornis." PeerJ 2:e541. doi: 10.7717/peerj.541. Morrow, K. M., A. G. Moss, N. E. Chadwick, and M. R. Liles. 2012. "Bacterial associates of two Caribbean coral species reveal species-specific distribution and geographic variability." Appl Environ Microbiol 78 (18):6438-49. doi: 10.1128/AEM.01162-12. Mosimann, James E. 1962. "On the compound multinomial distribution, the multivariate β-distribution, and correlations among proportions." Biometrika 49 (1):65 - 82. Mouchka, M. E., I. Hewson, and C. D. Harvell. 2010. "Coral-associated bacterial assemblages: current knowledge and the potential for climate-driven impacts." Integr Comp Biol 50 (4):662-74. doi: 10.1093/icb/icq061. Muller, E. M., and R. van Woesik. 2014. "Genetic Susceptibility, Colony Size, and Water Temperature Drive White-Pox Disease on the Coral Acropora palmata." PLoS One 9 (11):e110759. doi: 10.1371/journal.pone.0110759.

67 Muller, Erinn M., and Robert van Woesik. 2012. "Caribbean coral diseases: primary transmission or secondary infection?" Global Change Biology:n/a-n/a. doi: 10.1111/gcb.12019. Neave, M. J., R. Rachmawati, L. Xun, C. T. Michell, D. G. Bourne, A. Apprill, and C. R. Voolstra. 2017. "Differential specificity between closely related corals and abundant Endozoicomonas across global scales." ISME J 11 (1):186-200. doi: 10.1038/ismej.2016.95. Nylund, A., K. F. Ottem, K. Watanabe, E. Karlsbakk, and B. Krossoy. 2006. "Francisella sp. (Family Francisellaceae) causing mortality in Norwegian cod (Gadus morhua) farming." Arch Microbiol 185 (5):383-92. doi: 10.1007/s00203-006-0109-5. Oksanen, Jari F., Gullaume Blanchet, Michael Friendly, Roeland Kindt, Pierre Legendre, Dan McGlinn, Peter R. Minchin, R. B. O’Hara, Gavin L. Simpson, Peter Solymos, M. Henry H. Stevens, Eduard Szoecs, and Helen Wagner. 2016. “vegan: Community Ecology Package.” Pantos, Olga, and John Bythell. 2006. "Bacterial community structure associated with white band disease in the elkhorn coral Acropora palmata determined using culture-independent 16S rRNA techniques." Diseases of Aquatic Organisms 69:79-88. Pantos, Olga, Rory P. Cooney, Martin D. A. Le Tissier, Michael R. Barer, Anthony G. O'Donnell, and John C. Bythell. 2003. "The bacterial ecology of a plague-like disease affecting the Caribben coral Montastrea annularis." Environ Microbiol 5 (5):370-382. Patin, N. V., M. Schorn, K. Aguinaldo, T. Lincecum, B. S. Moore, and P. R. Jensen. 2016. "Effects of actinomycete secondary metabolites on sediment microbial communities." Appl Environ Microbiol. doi: 10.1128/AEM.02676-16. Peters, Esther C. 1983. "Possible Causal Agent of "White-Band Disease" in Caribbean Acroporid Corals." Journal of Invertebrate Pathology 41:394 - 396. Peters, Esther C. 1997. "Diseases of Reef : Corals." In Life and Death of Coral Reefs, edited by Charles Birkeland, 118-122. New York: Chapman & Hall.

68 Pollock, F Joseph, Elisha M Wood-Charlson, Madeleine JH van Oppen, David G Bourne, Bette L Willis, and Karen D Weynberg. 2014. "Abundance and morphology of virus-like particles associated with the coral Acropora hyacinthus differ between healthy and white syndrome-infected states." Marine Ecology Progress Series 510:39-43. Pollock, F. J., N. Wada, G. Torda, B. L. Willis, and D. G. Bourne. 2016. "Repeated sampling of white syndrome-affected corals reveals distinct microbiome at disease lesion fronts." Appl Environ Microbiol. doi: 10.1128/AEM.02799-16. Pollock, F. Joseph, Naohlsa Wada, Gergely Torda, Bette L. Willis, and David G. Bourne. 2017. "White Syndrome-Affected Corals Have a Distinct Microbiome at Disease Lesion Fronts." Appl Environ Microbiol 83 (2). doi: 10.1128/. Pratte, Z. A., and L. L. Richardson. 2016. "Possible links between white plague-like disease, scleractinian corals, and a cryptochirid gall crab." Dis Aquat 122 (2):153-161. doi: 10.3354/dao03074. Precht, W. F., B. E. Gintert, M. L. Robbart, R. Fura, and R. van Woesik. 2016. "Unprecedented Disease-Related Coral Mortality in Southeastern Florida." Sci Rep 6:31374. doi: 10.1038/srep31374. Price, Morgan N., Paramvir S. Dehal, and Adam P. Arkin. 2010. "FastTree 2 – Approximately Maximum-Likelihood Trees for Large Alignments." PLoS ONE 5 (3):e9490. doi: 10.1371/journal.pone.0009490. Pujalte, María J., Teresa Lucena, María A. Ruvira, David Ruiz Arahal, and M. Carmen Macián. 2014. "The Family Rhodobacteraceae." In The Prokaryotes: Alphaproteobacteria and Betaproteobacteria, edited by Eugene Rosenberg, Edward F. DeLong, Stephen Lory, Erko Stackebrandt and Fabiano Thompson, 439-512. Berlin, Heidelberg: Springer Berlin Heidelberg. Quince, C., A. Lanzen, R. J. Davenport, and P. J. Turnbaugh. 2011. "Removing noise from pyrosequenced amplicons." BMC Bioinformatics 12:38. doi: 10.1186/1471- 2105-12-38. Quistad, S. D., J. A. Grasis, J. J. Barr, and F. L. Rohwer. 2016. "Viruses and the origin of microbiome selection and immunity." ISME J. doi: 10.1038/ismej.2016.182. R Core Team. 2014. "R: A Language and Environment for Statistical Computing."

69 Randall, C. J., A. G. Jordan-Garza, E. M. Muller, and R. van Woesik. 2016. "Does Dark- Spot Syndrome Experimentally Transmit among Caribbean Corals?" PLoS One 11 (1):e0147493. doi: 10.1371/journal.pone.0147493. Randall, C. J., and R. van Woesik. 2015. "Contemporary white-band disease in Caribbean corals driven by climate change." Nature Climate Change 5 (4):375-379. doi: 10.1038/nclimate2530. Reshef, L., O. Koren, Y. Loya, I. Zilber-Rosenberg, and E. Rosenberg. 2006. "The coral probiotic hypothesis." Environ Microbiol 8 (12):2068-73. doi: 10.1111/j.1462- 2920.2006.01148.x. Reynolds, A. P., G. Richards, B. de la Iglesia, and V. J. Rayward-Smith. 2006. "Clustering Rules: A Comparison of Partitioning and Hierarchical Clustering Algorithms." Journal of Mathematical Modelling and Algorithms 5 (4):475-504. doi: 10.1007/s10852-005-9022-1. Richardson, Laurie. 1998. "Coral diseases: what is really known?" Trends in Ecology and Evolution 13 (11):438-443. Riegl, B., A. Bruckner, S. L. Coles, P. Renaud, and R. E. Dodge. 2009. "Coral reefs: threats and conservation in an era of global change." Ann N Y Acad Sci 1162:136- 86. doi: 10.1111/j.1749-6632.2009.04493.x. Ritchie, Kim B., and Garriet W. Smith. 1995. "Preferential carbon utilization by surface bacterial communities from water mass, normal, and white-band diseased Acropora cervicornis." Molecular Marine Biology and Biotechnology 4 (4):345- 352. Ritchie, Kim, and G. W. Smith. 1998. "Type-II white-band disease." Rev Biol Trop 46:199-203. Roder, C., C. Arif, T. Bayer, M. Aranda, C. Daniels, A. Shibl, S. Chavanich, and C. R. Voolstra. 2014. "Bacterial profiling of White Plague Disease in a comparative coral species framework." ISME J 8 (1):31-9. doi: 10.1038/ismej.2013.127. Roder, C., C. Arif, C. Daniels, E. Weil, and C. R. Voolstra. 2014. "Bacterial profiling of White Plague Disease across corals and oceans indicates a conserved and distinct disease microbiome." Mol Ecol 23 (4):965-74. doi: 10.1111/mec.12638.

70 Rodriguez-Lanetty, M., C. Granados-Cifuentes, A. Barberan, A. J. Bellantuono, and C. Bastidas. 2013. "Ecological Inferences from a deep screening of the Complex Bacterial Consortia associated with the coral, Porites astreoides." Mol Ecol 22 (16):4349-62. doi: 10.1111/mec.12392. Rohwer, Forest, Victor Seguritan, Farooq Azam, and Nancy Knowlton. 2002. "Diversity and distribution of coral-associated bacteria." Marine Ecology Progress Series 243:1-10. doi: 10.3354/meps243001. Rosenberg, Eugene, C. A. Kellogg, and Forest Rohwer. 2007a. "Coral Microbiology." In Oceanography. Rosenberg, Eugene, Christina A. Kellogg, and Forest Rohwer. 2007b. "Coral Mirobiology." Oceanography 20 (2):146-154. Rosenberg, Eugene, Omry Koren, Leah Reshef, Rotem Efrony, and Ilana Zilber- Rosenberg. 2007. "The role of in coral health, disease, and evolution." Nature 5:355-362. Ryan, E. T. 2013. "The intestinal pathobiome: its reality and consequences among infants and young children in resource-limited settings." J Infect Dis 208 (11):1732-3. doi: 10.1093/infdis/jit509. Sato, Y., B. L. Willis, and D. G. Bourne. 2010. "Successional changes in bacterial communities during the development of black band disease on the reef coral, hispida." ISME J 4 (2):203-14. doi: 10.1038/ismej.2009.103. Sato, Y., B. L. Willis, and D. G. Bourne. 2013. "Pyrosequencing-based profiling of archaeal and bacterial 16S rRNA genes identifies a novel archaeon associated with black band disease in corals." Environ Microbiol. doi: 10.1111/1462- 2920.12256. Schulz, Frederik, Joran Martijn, Florian Wascher, Ilias Lagkouvardos, Rok Kostanjšek, Thijs J. G. Ettema, and Matthias Horn. 2016. "A Rickettsiales symbiont of amoebae with ancient features." Environmental Microbiology 18 (8):2326-2342. doi: 10.1111/1462-2920.12881.

71 Sekar, R., D. K. Mills, E. R. Remily, J. D. Voss, and L. L. Richardson. 2006. "Microbial communities in the surface mucopolysaccharide layer and the black band of black band-diseased Siderastrea siderea." Appl Environ Microbiol 72 (9):5963-73. doi: 10.1128/AEM.00843-06. Séré, Mathieu G., Pablo Tortosa, Pascale Chabanet, Jean Turquet, Jean-Pascal Quod, and Michael H. Schleyer. 2013. "Bacterial Communities Associated with Porites White Patch Syndrome (PWPS) on Three Western Indian Ocean (WIO) Coral Reefs." PLoS ONE 8 (12):e83746. doi: 10.1371/journal.pone.0083746. Shaver, E. C., A. A. Shantz, R. McMinds, D. E. Burkepile, R. L. Vega Thurber, and B. R. Silliman. 2016. "Effects of predation and nutrient enrichment on the success and microbiome of a foundational coral." Ecology. doi: 10.1002/ecy.1709. Smith, D., P. Leary, J. Craggs, J. Bythell, and M. Sweet. 2015. "Microbial communities associated with healthy and White syndrome-affected Echinopora lamellosa in aquaria and experimental treatment with the antibiotic ampicillin." PLoS One 10 (3):e0121780. doi: 10.1371/journal.pone.0121780. Soffer, N., M. E. Brandt, A. M. Correa, T. B. Smith, and R. V. Thurber. 2014. "Potential role of viruses in white plague coral disease." ISME J 8 (2):271-83. doi: 10.1038/ismej.2013.137. Squires, Donald F. 1965. "Neoplasia in a Coral?" Science 148 (3669):503-505. Starliper, Clifford E. 2011. "Bacterial coldwater disease of fishes caused by Flavobacterium psychrophilum." Journal of Advanced Research 2:97 - 108. doi: 10.1016/j.jare.2010.04.001. Sunagawa, S., T. Z. DeSantis, Y. M. Piceno, E. L. Brodie, M. K. DeSalvo, C. R. Voolstra, E. Weil, G. L. Andersen, and M. Medina. 2009. "Bacterial diversity and White Plague Disease-associated community changes in the Caribbean coral Montastraea faveolata." ISME J 3 (5):512-21. doi: 10.1038/ismej.2008.131. Sussman, M., B. L. Willis, S. Victor, and D. G. Bourne. 2008. "Coral pathogens identified for White Syndrome (WS) epizootics in the Indo-Pacific." PLoS One 3 (6):e2393. doi: 10.1371/journal.pone.0002393.

72 Sutherland, K. P., B. Berry, A. Park, D. W. Kemp, K. M. Kemp, E. K. Lipp, and J. W. Porter. 2016. "Shifting white pox aetiologies affecting Acropora palmata in the Florida Keys, 1994-2014." Philos Trans R Soc Lond B Biol Sci 371 (1689). doi: 10.1098/rstb.2015.0205. Sutherland, Kathryn P., James W. Porter, and Cecilia Torres. 2004. "Disease and immunity in Caribbean and Indo-Pacific zooxanthellate corals." Marine Ecology Progress Series 266:273-302. doi: 10.3354/meps266273. Sweet, M., and J. Bythell. 2012. "Ciliate and bacterial communities associated with White Syndrome and Brown Band Disease in reef-building corals." Environ Microbiol 14 (8):2184-99. doi: 10.1111/j.1462-2920.2012.02746.x. Sweet, M., and J. Bythell. 2015. "White syndrome in Acropora muricata: nonspecific bacterial infection and ciliate histophagy." Mol Ecol 24 (5):1150-9. doi: 10.1111/mec.13097. Sweet, M., and J. Bythell. 2016. "The role of viruses in coral health and disease." J Invertebr Pathol. doi: 10.1016/j.jip.2016.12.005. Sweet, M. J., A. Croquer, and J. C. Bythell. 2014. "Experimental antibiotic treatment identifies potential pathogens of white band disease in the endangered Caribbean coral Acropora cervicornis." Proc Biol Sci 281 (1788):20140094. doi: 10.1098/rspb.2014.0094. Sweet, Michael J., and Kelly S. Bateman. 2015. "Diseases in associated with mariculture and commercial fisheries." Journal of Sea Research 104:16-32. doi: 10.1016/j.seares.2015.06.016. Sweet, Michael J., and Mark T. Bulling. 2017. "On the Importance of the Microbiome and Pathobiome in Coral Health and Disease." Frontiers in Marine Science 4. doi: 10.3389/fmars.2017.00009. Thompson, F. L., Y. Barash, T. Sawabe, G. Sharon, J. Swings, and E. Rosenberg. 2006. "Thalassomonas loyana sp. nov., a causative agent of the white plague-like disease of corals on the Eilat coral reef." Int J Syst Evol Microbiol 56 (Pt 2):365- 8. doi: 10.1099/ijs.0.63800-0.

73 Thompson, J. R., H. E. Rivera, C. J. Closek, and M. Medina. 2015. "Microbes in the coral holobiont: partners through evolution, development, and ecological interactions." Front Infect Microbiol 4:176. doi: 10.3389/fcimb.2014.00176. Tibshirani, Robert, Guenther Walther, and Trevor Hastie. 2001. "Estimating the number of clusters in a data set via the gap statistic." Journal of the Royal Statistical Society: Series B 63 (2):411 - 423. Tiirola, Marja, E. Tellervo Valtonen, Päivi Rintamäki-Kinnunen, and Markku S. Kulomaa. 2002. "Diagnosis of flavobacteriosis by direct amplification of rRNA genes." Diseases of Aquatic Organisms 51:93 - 100. Turnbaugh, P. J., M. Hamady, T. Yatsunenko, B. L. Cantarel, A. Duncan, R. E. Ley, M. L. Sogin, W. J. Jones, B. A. Roe, J. P. Affourtit, M. Egholm, B. Henrissat, A. C. Heath, R. Knight, and J. I. Gordon. 2009. "A core gut microbiome in obese and lean twins." Nature 457 (7228):480-4. doi: 10.1038/nature07540. Ushijima, B., A. Smith, G. S. Aeby, and S. M. Callahan. 2012. "Vibrio owensii induces the tissue loss disease Montipora white syndrome in the Hawaiian reef coral Montipora capitata." PLoS One 7 (10):e46717. doi: 10.1371/journal.pone.0046717. Vayssier-Taussat, M., E. Albina, C. Citti, J. F. Cosson, M. A. Jacques, M. H. Lebrun, Y. Le Loir, M. Ogliastro, M. A. Petit, P. Roumagnac, and T. Candresse. 2014. "Shifting the paradigm from pathogens to pathobiome: new concepts in the light of meta-omics." Front Cell Infect Microbiol 4:29. doi: 10.3389/fcimb.2014.00029. Vega Thurber, Rebecca , Dana Willner-Hall, Beltran Rodriguez-Mueller, Christelle Desnues, Robert A. Edwards, Florent Angly, Elizabeth Dinsdale, Linda Kelly, and Forest Rohwer. 2009. "Metagenomic analysis of stressed coral ." Environmental Microbiology 11 (8):2148-2163. doi: 10.1111/j.1462- 2920.2009.01935.x. Voss, Joshua, and Laurie Richardson. 2006. "Nutrient enrichment enhances black band disease progression in corals." Coral Reefs 25 (4):569-576. doi: 10.1007/s00338- 006-0131-8.

74 Walker, B. K., E. A. Larson, A. L. Moulding, and D. S. Gilliam. 2012. "Small-scale mapping of indeterminate arborescent acroporid coral (Acropora cervicornis) patches." Coral Reefs 31 (3):885-894. doi: 10.1007/s00338-012-0910-3. Walters, W. A., J. G. Caporaso, C. L. Lauber, D. Berg-Lyons, N. Fierer, and R. Knight. 2011. "PrimerProspector: de novo design and taxonomic analysis of barcoded polymerase chain reaction primers." Bioinformatics 27 (8):1159-61. doi: 10.1093/bioinformatics/btr087. Wang, Q., G. M. Garrity, J. M. Tiedje, and J. R. Cole. 2007. "Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy." Appl Environ Microbiol 73 (16):5261-7. doi: 10.1128/AEM.00062-07. Ward, Jessica R., and Kevin D. Lafferty. 2004. "The Elusive Baseline of Marine Disease: Are Diseases in Ocean Increasing?" PLoS Biology 2 (4):542-547. Wegley, L., R. Edwards, B. Rodriguez-Brito, H. Liu, and F. Rohwer. 2007. "Metagenomic analysis of the microbial community associated with the coral Porites astreoides." Environ Microbiol 9 (11):2707-19. doi: 10.1111/j.1462- 2920.2007.01383.x. Weil, Ernesto. 2004. "Coral Reef Diseases in the Wider Caribbean." In Coral Health and Disease, edited by Eugene Rosenberg and Yossi Loya, 35-68. Springer. Weil, Ernesto, and Caroline S. Rogers. 2011. "Coral Reef Diseases in the Atlantic- Caribbean." In Coral Reefs: An Ecosystem in Transition, edited by Zvy Dubinsky and Noga Stambler. Springer Science-Business Media B.V. Weinert, L. A., J. H. Werren, A. Aebi, G. N. Stone, and F. M. Jiggins. 2009. "Evolution and diversity of Rickettsia bacteria." BMC Biol 7:6. doi: 10.1186/1741-7007-7-6. Weiss, Sophie, Zhenjiang Zech Xu, Shyamal Peddada, Amnon Amir, Kyle Bittinger, Antonio Gonzalez, Catherine Lozupone, Jesse R. Zaneveld, Yoshiki Vázquez- Baeza, Amanda Birmingham, Embriette R. Hyde, and Rob Knight. 2017. "Normalization and microbial differential abundance strategies depend upon data characteristics." Microbiome 5 (1):27. doi: 10.1186/s40168-017-0237-y. Welsh, R. M., J. R. Zaneveld, S. M. Rosales, J. P. Payet, D. E. Burkepile, and R. V. Thurber. 2015. "Bacterial predation in a marine host-associated microbiome." ISME J. doi: 10.1038/ismej.2015.219.

75 Werner, J. J., O. Koren, P. Hugenholtz, T. Z. DeSantis, W. A. Walters, J. G. Caporaso, L. T. Angenent, R. Knight, and R. E. Ley. 2012. "Impact of training sets on classification of high-throughput bacterial 16s rRNA gene surveys." ISME J 6 (1):94-103. doi: 10.1038/ismej.2011.82. Wickham, Hadley. 2009. ggplot2: Elegant Graphics for Data Analysis: Springer-Verlag New York. Williams, D. E., and M. W. Miller. 2012. "Attributing mortality among drivers of population decline in Acropora palmata in the Florida Keys (USA)." Coral Reefs 31 (2):369-382. doi: 10.1007/s00338-011-0847-y. Williams, Dana E., and Margaret W. Miller. 2005. "Coral disease outbreak: pattern, prevalence and transmission in Acropora cervicornis." Marine Ecology Progress Series 301:119-128. Williams, K. P., J. J. Gillespie, B. W. Sobral, E. K. Nordberg, E. E. Snyder, J. M. Shallom, and A. W. Dickerman. 2010. "Phylogeny of gammaproteobacteria." J Bacteriol 192 (9):2305-14. doi: 10.1128/JB.01480-09. Willis, Bette, Cathie A. Page, and Elizabeth A. Dinsdale. 2004. "Coral Disease on the Great Barrier Reef." In Coral Health and Disease, edited by Eugene Rosenberg and Yossi Loya, 69-104. Springer. Wilson, B., G. S. Aeby, T. M. Work, and D. G. Bourne. 2012. "Bacterial communities associated with healthy and Acropora white syndrome-affected corals from American Samoa." FEMS Microbiol Ecol 80 (2):509-20. doi: 10.1111/j.1574- 6941.2012.01319.x. Wobbrock, Jacob O., Leah Findlater, Darren Gergle, and James J. Higgins. 2011. "The aligned rank transform for nonparametric factorial analyses using only anova procedures." Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, Vancouver, BC, Canada. Wright, Rachel M., Carly D. Kenkel, Carly E. Dunn, Erin N. Shilling, Line K. Bay, and Mikhail V. Matz. 2016. "Higher stress and immunity responses are associated with higher mortality in reef-building coral exposed to a bacterial challenge." doi: 10.1101/059444.

76 Yakob, Laith, and Peter J. Mumby. 2011. "Climate change induces demographic resistance to disease in novel coral assemblages." Proceedings of the National Academy of Sciences 108 (5):1967-1969. doi: 10.1073/pnas.1015443108. Zaneveld, J. R., D. E. Burkepile, A. A. Shantz, C. E. Pritchard, R. McMinds, J. P. Payet, R. Welsh, A. M. Correa, N. P. Lemoine, S. Rosales, C. Fuchs, J. A. Maynard, and R. V. Thurber. 2016. "Overfishing and nutrient pollution interact with temperature to disrupt coral reefs down to microbial scales." Nat Commun 7:11833. doi: 10.1038/ncomms11833. Zvuloni, A., Y. Artzy-Randrup, G. Katriel, Y. Loya, and L. Stone. 2015. "Modeling the Impact of White-Plague Coral Disease in Climate Change Scenarios." PLoS Comput Biol 11 (6):e1004151. doi: 10.1371/journal.pcbi.1004151.

77 Appendix: Figures

Appendix Figure 1. Rarefaction Curves. Rarefaction curves are used for evaluating sequencing depth. Curves reaching a plateau have reached saturation indicating sequence diversity has been covered. Tissue samples (bottom) have plateaued indicating saturation, whereas not all of the water (top) or sediment (middle) have plateaued.

78

Appendix Figure 2. Sediment Diversity by Site. Each panel is a different diversity measure for sediment samples separated by site, as site was the only factor having an effect on sediment diversity. Significant differences in sediment samples between sites were only found for S, indicated by the *.

Appendix Figure 3. Water Diversity by Site. Each panel represents a different diversity measures for water samples separated by site. Site and treatment did not have any significant effects on any measures of diversity of water.

79

Appendix Figure 4. Gap Statistic for Sample Type Clustering (BCD). The points on the curve represent the estimate of the gap statistic based on Bray-Curtis dissimilarity for all samples. The standard error intervals are shown as dashed bars (1-SE) and red bars (2-SE). Based on the 2-standard error criteria it is observed that three clusters are optimal.

Appendix Figure 5. Gap Statistic for Sample Type Clustering (W-UniFrac). The points on the curves represent the estimate of the gap statistic based on weighted UniFrac distance for all samples. The standard error intervals are shown as dashed bars (1-SE) and red bars (2-SE). Based on the 2-standard error criteria it is observed that three clusters are optimal.

80

Appendix Figure 6. Gap Statistic for Tissue Treatment Clustering (BCD). The points on the curves represent the estimate of the gap statistic based on Bray-Curtis dissimilarity for Acropora cervicornis tissue. The standard error intervals are shown as dashed bars (1-SE) and red bars (2-SE). Based on the 2-standard error criteria it is observed that two clusters are optimal.

Appendix Figure 7. Gap Statistic for Tissue Treatment Clustering (W-UniFrac). The points on the curves represent the estimate of the gap statistic based on weighted UniFrac distance for Acropora cervicornis tissue. The standard error intervals are shown as dashed bars (1-SE) and red bars (2-SE). Based on the 2-standard error criteria it is observed that 2 – 3 clusters are optimal.

81

Appendix Figure 8. Relative Abundance Taxa Summary. Relative abundances of the top 50 most abundant OTUs found in tissue (by health state), sediment and water samples. Columns indicate class level classification in the samples and colors indicate order.

82

Appendix Figure 9. Proteobacteria Relative Abundance Without Alphaproteobacteria (free scales). Relative abundance of Protoeobacteria with Alphaproteobacteria removed and variable y-scales between classes (panels) allows visualization of changes in low abundance, but potentially important taxa. Colors represent taxa orders and bars are health states.

83

Appendix Figure 10. Healthy vs. Healthy-on-Disease Differential Abundance. Differential abundance of order level taxa for healthy tissue compared to diseased tissue. Each point is an order within the class indicated on the y-axis, colored by phylum. Values along the x-axis indicate the log2fold change of taxa where points < 0 are more abundant in disease tissue. The shape of the points indicates the statistical significance of taxa a FDR adjusted level of 0.05. The size of the points represents the base mean abundance scaled by log10.

84

Appendix: Tables

Appendix Table 1. Dunn’s Test Summary. Summary of statistical output of Kruskal-Wallis (KW) test followed by Dunn’s test of multiple comparisons using rank sums for each diversity metric by sample type. Chi-squared (χ2) and top p-values are for the overall KW. P-values for comparisons for each metric are adjusted using the Benjimini-Hochberg (BH) method. Codes in comparisons refer to treatment/health state: H = Healthy, HD = Healthy-on-Disease, and D = Disease. Bold p-values are significantly different for the comparison.

Observed Species (S) Chao1 Shannon (H') Inverse Simpson (D)

KW: χ2 & p χ2 = 41.226 p = 2.413 x 10-8 χ2 = 48.428 p = 7.683 x 10-10 χ2 = 50.535 p = 2.792 x 10-10 χ2 = 50.417 p = 2.955 x 10-10 Comparisons Z statistic p-value Z statistic p-value Z statistic p-value Z statistic p-value

Sediment - Tissue (HD) 4.376 0.000 4.591 0.000 4.758 0.000 4.754 0.000

Sediment - Tissue (D) 0.988 0.180 2.392 0.021 3.278 0.001 3.655 0.000

Tissue (HD) - Tissue (D) -2.588 0.010 -1.679 0.078 -1.130 0.162 -0.840 0.287

Sediment - Tissue (H) 4.756 0.000 5.051 0.000 4.798 0.000 4.734 0.000

Tissue (HD) - Tissue (H) 0.290 0.386 0.351 0.363 0.031 0.542 -0.015 0.494

Tissue (D) - Tissue (H) 2.878 0.005 2.031 0.042 1.160 0.176 0.824 0.256

Sediment - Water 3.506 0.001 4.097 0.000 4.194 0.000 3.952 0.000

Tissue (HD) - Water -1.423 0.097 -1.208 0.142 -1.287 0.165 -1.448 0.148

Tissue (D) - Water 1.521 0.092 0.703 0.268 -0.002 0.499 -0.493 0.346

Tissue (H) - Water -1.753 0.066 -1.607 0.077 -1.322 0.186 -1.431 0.127

85

Appendix Table 2. Tissue ART-ANOVA and lsmeans post hoc. (A)Statistical output of Aligned Rank Transform Analysis of Variance (ART-ANOVA) for factors contributing to differences in diversity measures. Diversity measures: S = Observed species, Chao1 richness estimate, H’ = Shannon, D = Inverse Simpson. Statistic: F = F statistic, Df = Degress of freedom, Df.res = Residuals Degrees of freedom, Pr (>F) = p-value. (B) Statistical output of least-squares means (lsmeans) post hoc pairwise comparisons within treatment. Lsmeans = Least-squares means, SE = Standard Error, CL = Confidence level (0.95). Bold p-values = significant. (A) Diversity Factors Statistic Measures Site Treatment Site:Treatment F 1.972 13.057 1.015 Df 1 2 2 S Df.res 7.009 6.593 6.844 Pr (>F) 0.203 0.005 0.411 F 1.981 13.057 0.674 Df 1 2 2 Chao1 Df.res 7.018 6.593 6.758 Pr (>F) 0.202 0.005 0.541 F 0.053 13.567 0.939 Df 1 2 2 H' Df.res 7.055 6.562 6.643 Pr (>F) 0.824 0.005 0.437 F 0.001 8.724 0.465 Df 1 2 2 D Df.res 6.981 6.488 6.808 Pr (>F) 0.974 0.014 0.647

(B) lsmeans Treatment lsmean SE Df lower.CL upper.CL Healthy 5.333 1.645 9.570 1.646 9.021 S Healthy-on-Disease 8.167 1.645 11.590 4.568 11.765 Disease 15.000 1.645 9.570 11.312 18.688 Healthy 5.333 1.645 9.570 1.646 9.021 Chao1 Healthy-on-Disease 8.167 1.645 11.590 4.568 11.765 Disease 15.000 1.645 9.570 11.312 18.688 Healthy 6.167 1.745 9.160 2.231 10.102 H' Healthy-on-Disease 7.333 1.745 11.790 3.525 11.142 Disease 15.000 1.745 9.160 11.064 18.936 Healthy 8.500 2.204 8.610 3.481 13.519 D Healthy-on-Disease 6.500 2.204 11.910 1.695 11.305 Disease 13.500 2.204 8.610 8.481 18.519 Contrasts Contasts estimate SE Df t.ratio p.value Healthy - Healthy-on-Disease -2.833 2.326 6.180 -1.218 0.485 S Disease - Healthy-on-Disease 6.833 1.509 12.000 4.529 0.002 Disease - Healthy 9.667 2.326 9.570 4.155 0.006 Healthy - Healthy-on-Disease -2.833 2.326 6.180 -1.218 0.485 Chao1 Disease - Healthy-on-Disease 6.833 1.509 12.000 4.529 0.002 Disease - Healthy 9.667 2.326 9.570 4.155 0.006 Healthy - Healthy-on-Disease -1.167 2.467 6.190 -0.473 0.886 H' Disease - Healthy-on-Disease 7.667 1.528 11.950 5.019 0.001 Disease - Healthy 8.833 2.467 9.160 3.580 0.014 Healthy - Healthy-on-Disease 2.000 3.116 7.750 0.642 0.802 D Disease - Healthy-on-Disease 7.000 1.609 11.260 4.351 0.003 Disease - Healthy 5.000 3.116 8.610 1.604 0.295

86 Appendix Table 3. BCD PERMANOVA of All Samples (Free). Statistical output of adonis (PERMANOVA) testing differences in community composition between types, sites, and health states based on Bray-Curtis dissimilarity for all sample types. The equation: bray ~ Type * Site * Treatment with 9999 free permutation was used. Factor:Factor indicate interaction terms of the two factors. Bold Pr values are significant.

Factor Df SumsOfSqs MeanSqs F.Model R2 Pr(>F) Type 2 10.1042 5.0521 25.6959 0.42074 0.0001 Site 1 0.7464 0.7464 3.7964 0.03108 0.0012 Treatment 2 0.6373 0.3186 1.6206 0.02654 0.0599 Type:Site 2 0.7157 0.3579 1.8201 0.0298 0.0297 Type:Treatment 2 0.8632 0.4316 2.1952 0.03594 0.0068 Site:Treatment 2 0.5447 0.2724 1.3852 0.02268 0.1349 Type:Site:Treatment 2 0.3769 0.1885 0.9585 0.01569 0.4635 Residuals 51 10.0272 0.1966 0.41753

Total 64 24.0156 1

Appendix Table 4. BCD PERMANOVA of All Samples (within Site). Statistical output of adonis (PERMANOVA) testing differences in community composition between types and treatment within sites based on Bray-Curtis dissimilarity. The equation: bray ~ Type * Treatment with 199 permutations blocked by site was used. Factor:Factor indicates the interaction term of the two factors. Bold Pr values are significant.

Factor Df SumsOfSqs MeanSqs F.Model R2 Pr(>F) Type 2 10.1042 5.0521 23.6237 0.42074 0.005 Treatment 2 0.6349 0.3174 1.4843 0.02644 0.995 Type:Treatment 2 0.8728 0.4364 2.0407 0.03634 0.020 Residuals 58 12.4037 0.2139 0.51649

Total 64 24.0156 1

Appendix Table 5. BCD Pairwise PERMANOVA of All Samples. Matrix of p-values for pairwise PERMANOVA tests for differences in community composition between sample types based on Bray-Curtis dissimilarity. P-values are adjusted for multiple comparisons via the false discovery rate (fdr) method. Bold p-values indicate significant differences.

Tissue (H) Tissue (HD) Tissue (D) Water

Tissue (HD) 0.5120 - - - Tissue (D) 0.0025 0.0056 - - Water 0.0014 0.0014 0.0014 - Sediment 0.0014 0.0014 0.0014 0.0014

87 Appendix Table 6. W-UniFrac PERMANOVA of All Samples (Free). Statistical output of adonis (PERMANOVA) testing differences between types, sites, and health states based on weighted UniFrac distance. The equation: weighted UniFrac ~ type * site * treatment with 9999 free permutations was used. Factor:Factor indicate interaction terms of the two factors. Bold Pr values are significant.

Factors Df SumsOfSqs MeanSqs F.Model R2 Pr(>F) Type 2 0.95325 0.47662 53.318 0.55726 0.0001 Site 1 0.06184 0.06184 6.918 0.03615 0.0002 Treatment 2 0.04143 0.02072 2.317 0.02422 0.0218 Type:Site 2 0.05169 0.02585 2.891 0.03022 0.0077 Type:Treatment 2 0.07418 0.03709 4.149 0.04337 0.0008 Site:Treatment 2 0.04666 0.02333 2.61 0.02728 0.013 Type:Site:Treatment 2 0.02564 0.01282 1.434 0.01499 0.1694 Residuals 51 0.4559 0.00894 0.26652

Total 64 1.71059 1

Appendix Table 7. W-UniFrac PERMANOVA of All Samples (within Site). Statistical output of adonis (PERMANVOVA) testing differences between type and treatment within sites based on weighted UniFrac distance. The equation weighted UniFrac ~ type * treatment with 199 permutations blocked by site was used. Factor:Factor indicate interaction terms of the two factors. Bold Pr values are significant.

Factors Df SumsOfSqs MeanSqs F.Model R2 Pr(>F) Type 2 0.95325 0.47662 43.116 0.55726 0.005 Treatment 2 0.04101 0.02051 1.855 0.02398 1 Type:Treatment 2 0.07517 0.03759 3.4 0.04395 0.01 Residuals 58 0.64116 0.01105 0.37482

Total 64 1.71059 1

Appendix Table 8. W-UniFrac Pairwise PERMANOVA of All Samples. Matrix of p-values for pairwise PERMANOVA tests for differences in community composition between sample types based on Bray-Curtis dissimilarity. P-values are adjusted for multiple comparisons via the false discovery rate (fdr) method. Bold p-values indicate significant differences.

Tissue (H) Tissue (HD) Tissue (D) Water

Tissue (HD) 0.511 - - - Tissue (D) 0.0156 0.0063 - - Water 0.0017 0.0017 0.0029 - Sediment 0.0017 0.0017 0.0017 0.0017

88 Appendix Table 9. BCD Pairwise PERMANOVA of Tissue. Statistical output output of adonis (PERMANOVA) testing differences within A. cervicornis tissue by health state based on Bray-Curtis dissimilarity. H = Healthy, HD = Healthy-on-Disease, D = Disease. Bold Pr values are significant.

Comparison Factor Df SumsOfSqs MeanSqs F.Model R2 Pr(>F) H vs. D

Treatment 1 0.7836 0.7836 5.7139 0.3618 0.0050

Site 1 0.1542 0.1542 1.1244 0.0712 0.2906

Treatment:Site 1 0.1312 0.1312 0.9565 0.0606 0.3930

Residuals 8 1.0971 0.1371 0.5065

Total 11 2.1662 1.0000

H vs. HD

Treatment 1 0.0565 0.0565 0.5492 0.0517 0.5201

Site 1 0.0540 0.0540 0.5252 0.0494 0.5508

Treatment:Site 1 0.1595 0.1595 1.5510 0.1460 0.3052

Residuals 8 0.8224 0.1028 0.7529

Total 11 1.0924 1.0000

HD vs. D

Treatment 1 0.7346 0.7346 4.9831 0.3449 0.0313

Site 1 0.1410 0.1410 0.9567 0.0662 0.0625

Colony2 1 0.0666 0.0666 0.4516 0.0313 0.0625

Treatment:Site 1 0.1560 0.1560 1.0580 0.0732 0.4063

Residuals 7 1.0319 0.1474 0.4845

Appendix Table 10. W-UniFrac Pairwise PERMANOVA of Tissue. Statistical output of adonis (PERMANOVA) testing differences within A. cervicornis tissue by health state based on weighted UniFrac distance. H = Healthy, HD = Healthy-on-Disease, D = Disease. Bold Pr values are significant.

Comparison Factor Df SumsOfSqs MeanSqs F.Model R2 Pr(>F) H vs. D Treatment 2 0.091838 0.045919 3.9415 0.34298 0.0065 Site 1 0.005689 0.005689 0.4883 0.02125 0.7446 Treatment:Site 2 0.030437 0.015219 1.3063 0.11367 0.243 Residuals 12 0.139802 0.01165 0.5221

Total 17 0.267766 1

H vs. HD Treatment 1 0.006613 0.0066131 0.53112 0.04984 0.4687 Site 1 0.005292 0.005292 0.42502 0.03988 0.6416 Treatment:Site 1 0.02117 0.0211704 1.70026 0.15955 0.252 Residuals 8 0.09961 0.0124513 0.75072

Total 11 0.132686 1

HD vs. D Treatment 1 0.06064 0.06064 4.3913 0.3224 0.03125 Site 1 0.010452 0.010452 0.7569 0.05557 0.0625 Colony2 1 0.003564 0.003564 0.2581 0.01895 0.0625 Treatment:Site 1 0.016766 0.016766 1.2141 0.08914 0.3125 Residuals 7 0.096664 0.013809 0.51394

Total 11 0.188085 1

89 Appendix Table 11. Total Phyla Relative Abundance (RA). Mean, standard deviation, median, minimum, and maximum relative abundance at the phylum level for all samples (water, sediment, tissue) combined. Only phyla present in ≥ 0.1% are shown.

Phylum Mean RA SD RA Median RA Min RA Max RA Proteobacteria 62.54% 17.32% 56.42% 39.48% 98.90% Bacteroidetes 15.49% 9.99% 15.93% 0.21% 38.45% 5.13% 5.83% 3.03% 0.02% 23.55% Planctomycetes 3.97% 3.55% 3.10% 0.04% 12.03% Actinobacteria 3.18% 2.12% 3.20% 0.09% 10.38% Firmicutes 2.14% 3.93% 0.87% 0.01% 25.61% Crenarchaeota 2.09% 2.91% 0.84% 0.01% 18.29% Euryarchaeota 1.63% 2.13% 1.09% 0.01% 10.25% Verrucomicrobia 1.08% 0.90% 0.99% 0.06% 4.33% Acidobacteria 1.04% 0.88% 0.81% 0.01% 3.02% Chlorobi 0.85% 4.07% 0.10% 0.01% 23.87% Spirochaetes 0.71% 0.73% 0.39% 0.03% 2.53% Tenericutes 0.67% 1.27% 0.13% 0.01% 5.98% SAR406 0.63% 0.66% 0.44% 0.02% 2.74% Chloroflexi 0.59% 0.63% 0.32% 0.02% 2.63% Gemmatimonadetes 0.50% 0.45% 0.35% 0.01% 1.57% Fusobacteria 0.46% 0.55% 0.23% 0.01% 2.65% KSB3 0.36% NA 0.36% 0.36% 0.36% Fibrobacteres 0.36% 0.67% 0.10% 0.03% 2.27% WS3 0.32% 0.27% 0.25% 0.01% 1.19% Nitrospirae 0.23% 0.24% 0.13% 0.01% 1.04% Lentisphaerae 0.21% 0.24% 0.12% 0.01% 1.03% GN02 0.20% 0.27% 0.08% 0.01% 1.01% WWE1 0.19% 0.28% 0.06% 0.02% 0.90% Caldithrix 0.19% 0.20% 0.11% 0.01% 0.94% SBR1093 0.18% 0.13% 0.16% 0.00% 0.51% Poribacteria 0.17% 0.25% 0.09% 0.02% 0.78% [Caldithrix] 0.14% 0.08% 0.14% 0.08% 0.20% PAUC34f 0.13% 0.19% 0.06% 0.01% 0.88% GN04 0.13% 0.09% 0.10% 0.03% 0.35% Chlamydiae 0.12% 0.10% 0.09% 0.02% 0.53% OP3 0.11% 0.10% 0.06% 0.01% 0.34% Elusimicrobia 0.11% 0.19% 0.04% 0.01% 0.61% OD1 0.10% 0.08% 0.05% 0.03% 0.20%

90 Appendix Table 12. Class Level Relative Abundance by Type. Mean, standard deviation, median, minimum, and maximum relative abundance at the class level for each sample type. Only classes present in ≥ 0.1% are shown.

Class Type Mean RA SD RA Median RA Min RA Max RA Alphaproteobacteria Tissue (H) 84.38% 22.49% 94.35% 39.20% 96.32% Alphaproteobacteria Tissue (HD) 77.22% 20.66% 81.53% 45.12% 97.53% Alphaproteobacteria Tissue (D) 54.33% 11.63% 52.51% 37.19% 71.55% Alphaproteobacteria Water 30.66% 6.19% 30.79% 18.65% 38.65% Flavobacteriia Water 28.30% 4.40% 28.39% 22.03% 35.63% Gammaproteobacteria Sediment 27.10% 5.87% 27.05% 15.26% 37.89% Gammaproteobacteria Water 21.99% 5.18% 21.38% 16.24% 30.54% Gammaproteobacteria Tissue (D) 20.75% 9.27% 18.59% 12.21% 35.06% Alphaproteobacteria Sediment 15.09% 7.33% 12.67% 5.49% 31.38% Chlorobia Tissue (D) 12.02% 16.68% 12.02% 0.23% 23.82% Deltaproteobacteria Sediment 8.43% 3.75% 8.96% 1.56% 18.97% Synechococcophycideae Water 8.33% 6.33% 6.65% 2.32% 21.75% Gammaproteobacteria Tissue (H) 7.74% 11.37% 1.84% 0.52% 29.03% Flavobacteriia Sediment 6.57% 3.62% 5.42% 2.26% 18.99% Gammaproteobacteria Tissue (HD) 5.80% 4.16% 6.77% 0.51% 11.61% Bacilli Tissue (HD) 4.70% 10.25% 0.53% 0.02% 25.61% Deltaproteobacteria Tissue (D) 4.54% 1.95% 4.41% 1.49% 7.08% Thaumarchaeota Tissue (HD) 4.15% 7.93% 0.32% 0.20% 18.29% Acidimicrobiia Sediment 4.15% 1.39% 3.97% 1.38% 6.96% Planctomycetia Sediment 4.14% 2.23% 4.05% 0.38% 8.70% Cytophagia Sediment 3.67% 1.68% 3.65% 1.26% 7.40% [Saprospirae] Sediment 3.50% 1.78% 3.16% 0.58% 7.40% Synechococcophycideae Sediment 3.20% 6.32% 0.42% 0.10% 22.63% Oscillatoriophycideae Sediment 2.81% 2.04% 2.20% 0.22% 7.35% Bacteroidia Sediment 2.65% 4.07% 1.29% 0.10% 23.12% Flavobacteriia Tissue (D) 2.61% 2.09% 1.91% 0.91% 6.37% Thermoplasmata Water 2.49% 1.31% 2.06% 1.09% 5.54% Thaumarchaeota Sediment 2.36% 2.06% 1.82% 0.05% 7.87% [Leptospirae] Tissue (HD) 2.27% NA 2.27% 2.27% 2.27% Clostridia Sediment 2.18% 2.86% 1.06% 0.06% 11.30% Acidimicrobiia Water 1.83% 0.83% 1.83% 0.52% 3.46% Acidimicrobiia Tissue (HD) 1.83% 4.14% 0.14% 0.02% 10.28% Deltaproteobacteria Water 1.80% 1.22% 1.46% 0.10% 3.78% Acidimicrobiia Tissue (D) 1.80% 1.59% 1.80% 0.10% 4.05% Epsilonproteobacteria Tissue (D) 1.73% 1.73% 1.09% 0.40% 4.70% Thermoplasmata Sediment 1.73% 2.71% 0.36% 0.03% 10.25% Epsilonproteobacteria Sediment 1.67% 2.12% 0.94% 0.05% 11.46% Deltaproteobacteria Tissue (H) 1.56% 2.26% 0.71% 0.19% 6.06% Clostridia Tissue (D) 1.49% 2.42% 0.63% 0.07% 6.37% Bacilli Tissue (H) 1.27% 2.08% 0.38% 0.07% 4.99% Cytophagia Tissue (HD) 1.26% 1.72% 0.46% 0.06% 4.27% Actinobacteria Tissue (HD) 1.26% 2.42% 0.34% 0.02% 6.19% Thaumarchaeota Tissue (H) 1.16% 1.33% 0.71% 0.04% 3.71% [Saprospirae] Tissue (D) 1.15% 1.28% 0.78% 0.03% 3.48% Flavobacteriia Tissue (H) 1.06% 2.03% 0.19% 0.09% 4.69%

91 Class Type Mean RA SD RA Median RA Min RA Max RA AB16 Water 0.96% 0.29% 1.00% 0.44% 1.34% Betaproteobacteria Tissue (HD) 0.94% 1.17% 0.63% 0.01% 3.10% Planctomycetia Tissue (D) 0.89% 0.59% 0.79% 0.19% 1.82% Mollicutes Sediment 0.84% 1.38% 0.28% 0.01% 5.98% Verrucomicrobiae Sediment 0.79% 0.53% 0.62% 0.15% 2.53% [Pedosphaerae] Tissue (H) 0.78% NA 0.78% 0.78% 0.78% Deltaproteobacteria Tissue (HD) 0.77% 0.96% 0.27% 0.06% 2.47% Cytophagia Tissue (D) 0.75% 0.68% 0.51% 0.03% 1.98% Fibrobacteria Tissue (D) 0.73% 0.96% 0.13% 0.05% 2.27% Synechococcophycideae Tissue (HD) 0.72% 1.17% 0.25% 0.11% 3.08% Thaumarchaeota Tissue (D) 0.71% 1.29% 0.09% 0.06% 3.31% Spirochaetes Sediment 0.70% 0.68% 0.42% 0.03% 2.53% Gemm-2 Tissue (H) 0.68% NA 0.68% 0.68% 0.68% AB16 Sediment 0.65% 0.83% 0.12% 0.03% 2.74% Phycisphaerae Sediment 0.65% 0.35% 0.58% 0.06% 1.29% OM190 Sediment 0.62% 0.34% 0.55% 0.08% 1.58% Sva0725 Sediment 0.62% 0.43% 0.58% 0.02% 1.62% Fusobacteriia Sediment 0.59% 0.58% 0.35% 0.04% 2.65% Unassigned Tissue (H) 0.59% 0.93% 0.07% 0.05% 1.66% Anaerolineae Sediment 0.58% 0.44% 0.46% 0.03% 1.59% Bacilli Tissue (D) 0.50% 0.53% 0.25% 0.06% 1.21% [Rhodothermi] Sediment 0.50% 0.30% 0.38% 0.05% 1.21% Acidobacteria-6 Tissue (H) 0.49% NA 0.49% 0.49% 0.49% Epsilonproteobacteria Tissue (H) 0.48% 0.24% 0.55% 0.07% 0.73% Flavobacteriia Tissue (HD) 0.46% 0.59% 0.31% 0.02% 1.64% BD1-5 Tissue (H) 0.46% 0.37% 0.32% 0.18% 0.88% Gemm-2 Sediment 0.45% 0.32% 0.33% 0.05% 1.22% Bacteroidia Water 0.44% 0.46% 0.23% 0.06% 1.36% Anaerolineae Tissue (D) 0.42% 0.56% 0.17% 0.04% 1.48% Chlorobia Sediment 0.41% NA 0.41% 0.41% 0.41% Anaerolineae Tissue (H) 0.41% 0.58% 0.10% 0.05% 1.08% Cytophagia Water 0.40% 0.17% 0.39% 0.14% 0.75% Spirochaetes Tissue (H) 0.39% NA 0.39% 0.39% 0.39% Opitutae Sediment 0.39% 0.24% 0.32% 0.04% 1.14% Actinobacteria Tissue (D) 0.38% 0.21% 0.35% 0.17% 0.61% RB25 Sediment 0.38% 0.30% 0.29% 0.03% 1.09% Synechococcophycideae Tissue (D) 0.37% 0.25% 0.28% 0.13% 0.71% Bacilli Sediment 0.37% 0.60% 0.10% 0.02% 1.98% Unassigned Sediment 0.36% 0.24% 0.31% 0.03% 1.15% MAT-CR-H3-D11 Sediment 0.36% NA 0.36% 0.36% 0.36% Acidimicrobiia Tissue (H) 0.35% 0.57% 0.10% 0.07% 1.37% PRR-12 Sediment 0.35% 0.26% 0.28% 0.03% 1.19% Betaproteobacteria Water 0.35% 0.24% 0.25% 0.06% 0.82% SAR202 Tissue (D) 0.35% 0.54% 0.11% 0.03% 1.15% Verrucomicrobiae Tissue (D) 0.35% 0.23% 0.29% 0.05% 0.66% Acidobacteria-6 Tissue (D) 0.34% NA 0.34% 0.34% 0.34% Gloeobacterophycideae Sediment 0.34% 0.23% 0.29% 0.03% 0.89%

92 Class Type Mean RA SD RA Median RA Min RA Max RA BD4-9 Sediment 0.34% NA 0.34% 0.34% 0.34% BD1-5 Tissue (D) 0.34% 0.45% 0.14% 0.05% 1.01% Thaumarchaeota Water 0.33% 0.61% 0.03% 0.01% 1.24% Pla3 Sediment 0.33% 0.27% 0.25% 0.02% 1.22% [Saprospirae] Water 0.32% 0.25% 0.30% 0.05% 0.71% Solibacteres Tissue (H) 0.32% 0.24% 0.32% 0.15% 0.49% Anaerolineae Tissue (HD) 0.31% 0.12% 0.35% 0.17% 0.41% Clostridia Tissue (HD) 0.31% 0.19% 0.31% 0.17% 0.44% Betaproteobacteria Sediment 0.30% 0.25% 0.22% 0.04% 1.11% Fusobacteriia Tissue (H) 0.29% NA 0.29% 0.29% 0.29% Nitrospira Tissue (H) 0.29% NA 0.29% 0.29% 0.29% Sva0725 Tissue (H) 0.29% NA 0.29% 0.29% 0.29% Unassigned Tissue (HD) 0.29% 0.37% 0.15% 0.02% 0.82% Oscillatoriophycideae Tissue (D) 0.28% 0.21% 0.30% 0.05% 0.58% SAR202 Tissue (H) 0.28% 0.43% 0.05% 0.02% 0.78% Phycisphaerae Water 0.28% 0.12% 0.28% 0.09% 0.42% Cytophagia Tissue (H) 0.27% 0.11% 0.29% 0.12% 0.37% A712011 Sediment 0.27% 0.16% 0.29% 0.03% 0.46% [Cloacamonae] Sediment 0.27% 0.32% 0.12% 0.03% 0.90% Synechococcophycideae Tissue (H) 0.26% 0.29% 0.17% 0.02% 0.68% Opitutae Tissue (HD) 0.26% NA 0.26% 0.26% 0.26% Spirochaetes Tissue (D) 0.26% 0.24% 0.21% 0.05% 0.56% Verruco-5 Sediment 0.26% 0.20% 0.20% 0.04% 0.98% Nitrospira Tissue (D) 0.26% 0.42% 0.02% 0.01% 0.74% Unassigned Tissue (D) 0.25% 0.20% 0.23% 0.04% 0.61% Betaproteobacteria Tissue (H) 0.25% 0.24% 0.19% 0.05% 0.68% [Lentisphaeria] Sediment 0.25% 0.25% 0.15% 0.03% 1.03% SAR202 Sediment 0.24% 0.42% 0.13% 0.03% 1.94% Nitrospira Sediment 0.24% 0.25% 0.15% 0.03% 1.04% Gloeobacterophycideae Tissue (D) 0.23% 0.19% 0.18% 0.03% 0.47% Epsilonproteobacteria Water 0.23% 0.19% 0.29% 0.01% 0.50% TA18 Sediment 0.22% 0.15% 0.22% 0.03% 0.70% Solibacteres Tissue (HD) 0.22% 0.27% 0.10% 0.03% 0.52% A712011 Water 0.21% 0.08% 0.22% 0.10% 0.36% VHS-B5-50 Tissue (HD) 0.21% NA 0.21% 0.21% 0.21% Planctomycetia Water 0.20% 0.13% 0.18% 0.06% 0.50% [Pedosphaerae] Water 0.20% 0.10% 0.17% 0.08% 0.35% Caldithrixae Sediment 0.20% 0.20% 0.11% 0.02% 0.94% Clostridia Tissue (H) 0.20% NA 0.20% 0.20% 0.20% Verrucomicrobiae Tissue (H) 0.20% NA 0.20% 0.20% 0.20% [Pedosphaerae] Sediment 0.19% 0.38% 0.12% 0.02% 2.01% Bacteroidia Tissue (D) 0.19% 0.09% 0.16% 0.10% 0.34% C6 Sediment 0.18% 0.20% 0.12% 0.04% 1.01% BME43 Sediment 0.18% 0.13% 0.16% 0.02% 0.42% [Chloracidobacteria] Sediment 0.17% 0.17% 0.10% 0.03% 0.69% Gemm-1 Sediment 0.17% 0.21% 0.07% 0.02% 0.75% [Rhodothermi] Water 0.17% 0.08% 0.18% 0.06% 0.32%

93 Class Type Mean RA SD RA Median RA Min RA Max RA Oscillatoriophycideae Tissue (HD) 0.17% 0.20% 0.17% 0.03% 0.31% Planctomycetia Tissue (HD) 0.17% 0.21% 0.05% 0.04% 0.41% Acidobacteria-6 Sediment 0.16% 0.13% 0.12% 0.04% 0.55% Ignavibacteria Sediment 0.16% 0.20% 0.11% 0.02% 0.88% Epsilonproteobacteria Tissue (HD) 0.16% 0.04% 0.17% 0.10% 0.21% OM190 Water 0.16% 0.12% 0.12% 0.03% 0.47% Gemm-4 Sediment 0.15% 0.13% 0.14% 0.01% 0.42% Mb-NB09 Sediment 0.15% NA 0.15% 0.15% 0.15% Sphingobacteriia Sediment 0.15% 0.11% 0.12% 0.03% 0.41% Bacilli Water 0.15% 0.18% 0.15% 0.02% 0.28% Actinobacteria Sediment 0.15% 0.13% 0.10% 0.02% 0.44% 4C0d-2 Tissue (H) 0.15% 0.11% 0.20% 0.02% 0.22% Betaproteobacteria Tissue (D) 0.14% 0.12% 0.11% 0.02% 0.34% KSB1 Sediment 0.14% 0.08% 0.14% 0.08% 0.20% BD1-5 Sediment 0.14% 0.10% 0.12% 0.01% 0.30% Sva0725 Tissue (HD) 0.13% 0.19% 0.04% 0.03% 0.41% Chlamydiia Sediment 0.13% 0.11% 0.09% 0.02% 0.53% Unassigned Sediment 0.13% 0.10% 0.10% 0.04% 0.48% SAR202 Water 0.13% 0.06% 0.13% 0.03% 0.23% Sphingobacteriia Tissue (D) 0.12% 0.13% 0.05% 0.03% 0.27% Elusimicrobia Sediment 0.12% 0.20% 0.05% 0.02% 0.61% Planctomycetia Tissue (H) 0.12% 0.11% 0.12% 0.04% 0.19% [Brachyspirae] Sediment 0.12% 0.09% 0.09% 0.04% 0.30% OS-K Sediment 0.11% 0.10% 0.09% 0.02% 0.33% Nitrospira Tissue (HD) 0.11% 0.10% 0.10% 0.02% 0.22% BD7-11 Sediment 0.11% 0.15% 0.06% 0.02% 0.61% 4C0d-2 Sediment 0.11% 0.09% 0.10% 0.01% 0.41% Ellin6529 Sediment 0.11% 0.12% 0.05% 0.01% 0.40% Sphingobacteriia Tissue (H) 0.11% 0.12% 0.11% 0.02% 0.20% TK17 Sediment 0.11% NA 0.11% 0.11% 0.11% vadinHA49 Sediment 0.11% 0.09% 0.08% 0.01% 0.38% GN15 Tissue (D) 0.11% NA 0.11% 0.11% 0.11% Phycisphaerae Tissue (D) 0.10% 0.06% 0.08% 0.04% 0.19% GN15 Sediment 0.10% 0.09% 0.07% 0.03% 0.24% Actinobacteria Tissue (H) 0.10% 0.07% 0.11% 0.02% 0.19% Chlamydiia Tissue (HD) 0.10% NA 0.10% 0.10% 0.10% Gloeobacterophycideae Tissue (HD) 0.10% NA 0.10% 0.10% 0.10% Nostocophycideae Tissue (HD) 0.10% NA 0.10% 0.10% 0.10% OM190 Tissue (D) 0.10% 0.06% 0.10% 0.04% 0.16% 5bav_B12 Sediment 0.10% 0.07% 0.10% 0.03% 0.17% Sva0725 Tissue (D) 0.10% 0.10% 0.07% 0.01% 0.20% Rubrobacteria Tissue (H) 0.10% NA 0.10% 0.10% 0.10% ZB2 Sediment 0.10% 0.08% 0.06% 0.03% 0.20% PBS-25 Sediment 0.10% 0.08% 0.06% 0.01% 0.24% At12OctB3 Sediment 0.10% 0.10% 0.06% 0.02% 0.24%

94 Appendix Table 13. Tissue Phyla Relative Abundance. Mean, standard deviation, median, minimum, and maximum relative abundance of phyla present in A. cervicornis tissue by health state. Only phyla present in ≥ 0.1% RA are shown.

Phylum Treatment Mean RA SD RA Median RA Min RA Max RA Proteobacteria Healthy 94.37% 9.32% 98.07% 75.42% 99.03% Proteobacteria Healthy-on-Disease 84.97% 16.71% 92.11% 60.57% 98.75% Proteobacteria Disease 81.92% 6.08% 83.45% 70.43% 86.81% Chlorobi Disease 8.19% 13.96% 0.23% 0.03% 24.30% Firmicutes Healthy-on-Disease 4.81% 10.22% 0.84% 0.02% 25.65% Bacteroidetes Disease 4.53% 3.94% 3.33% 0.87% 11.87% Crenarchaeota Healthy-on-Disease 4.24% 8.12% 0.32% 0.18% 18.72% Actinobacteria Healthy-on-Disease 3.11% 4.31% 0.76% 0.13% 10.52% Spirochaetes Healthy-on-Disease 2.27% NA 2.27% 2.27% 2.27% Actinobacteria Disease 2.07% 1.62% 1.81% 0.46% 4.62% Firmicutes Disease 1.99% 2.36% 1.25% 0.29% 6.63% Bacteroidetes Healthy-on-Disease 1.53% 1.77% 0.70% 0.15% 4.54% Bacteroidetes Healthy 1.35% 2.07% 0.41% 0.17% 5.03% Firmicutes Healthy 1.32% 2.19% 0.38% 0.07% 5.23% Crenarchaeota Healthy 1.17% 1.34% 0.71% 0.04% 3.75% Cyanobacteria Disease 1.00% 0.50% 1.01% 0.19% 1.63% Planctomycetes Disease 0.90% 0.54% 0.82% 0.27% 1.77% Cyanobacteria Healthy-on-Disease 0.77% 1.28% 0.26% 0.09% 3.36% Fibrobacteres Disease 0.75% 0.98% 0.13% 0.05% 2.30% Nitrospirae Disease 0.75% NA 0.75% 0.75% 0.75% Acidobacteria Healthy 0.71% 0.80% 0.71% 0.15% 1.28% Crenarchaeota Disease 0.70% 1.31% 0.06% 0.03% 3.33% Gemmatimonadetes Healthy 0.69% NA 0.69% 0.69% 0.69% Chloroflexi Disease 0.68% 0.98% 0.23% 0.07% 2.58% Verrucomicrobia Healthy 0.53% 0.65% 0.53% 0.07% 0.99% Chloroflexi Healthy 0.52% 0.90% 0.09% 0.02% 1.88% GN02 Healthy 0.46% 0.37% 0.32% 0.18% 0.89% Poribacteria Healthy 0.41% 0.54% 0.41% 0.02% 0.79% Actinobacteria Healthy 0.40% 0.49% 0.20% 0.09% 1.38% Spirochaetes Healthy 0.39% NA 0.39% 0.39% 0.39% Verrucomicrobia Disease 0.38% 0.24% 0.35% 0.08% 0.69% GN02 Disease 0.34% 0.46% 0.15% 0.05% 1.02% PAUC34f Healthy 0.33% 0.49% 0.07% 0.02% 0.89% Acidobacteria Healthy-on-Disease 0.32% 0.32% 0.29% 0.04% 0.66% Fusobacteria Healthy 0.30% NA 0.30% 0.30% 0.30% Nitrospirae Healthy 0.30% NA 0.30% 0.30% 0.30% PAUC34f Healthy-on-Disease 0.28% 0.32% 0.17% 0.02% 0.63% Cyanobacteria Healthy 0.28% 0.30% 0.19% 0.02% 0.79% Chloroflexi Healthy-on-Disease 0.27% 0.18% 0.33% 0.02% 0.42% Spirochaetes Disease 0.25% 0.25% 0.20% 0.03% 0.57% Acidobacteria Disease 0.22% 0.19% 0.16% 0.07% 0.54% SBR1093 Healthy-on-Disease 0.21% NA 0.21% 0.21% 0.21% Verrucomicrobia Healthy-on-Disease 0.16% 0.17% 0.11% 0.02% 0.35% Poribacteria Disease 0.14% 0.01% 0.14% 0.14% 0.15% Planctomycetes Healthy-on-Disease 0.12% 0.13% 0.06% 0.04% 0.32% Nitrospirae Healthy-on-Disease 0.12% 0.10% 0.11% 0.02% 0.22% Gemmatimonadetes Disease 0.11% 0.07% 0.08% 0.04% 0.21% Chlamydiae Healthy-on-Disease 0.11% NA 0.11% 0.11% 0.11% SAR406 Healthy-on-Disease 0.11% NA 0.11% 0.11% 0.11% Chlamydiae Disease 0.10% 0.12% 0.04% 0.03% 0.27%

95 Appendix Table 14. Tissue Order Relative Abundance. Mean, standard deviation, median, minimum, and maximum relative abundance of orders present in A. cervicornis tissue by health state. Only orders present in ≥ 0.1% RA are shown.

Order Treatment Mean RA SD RA Median RA Min RA Max RA Rickettsiales Healthy 79.94% 29.53% 92.54% 20.24% 95.22% Rickettsiales Healthy-on-Disease 70.86% 23.36% 77.89% 29.76% 96.57% Rickettsiales Disease 29.99% 10.19% 27.62% 20.29% 47.62% Rhodobacterales Disease 22.46% 14.88% 20.76% 3.93% 45.54% Chlorobiales Disease 12.24% 16.98% 12.24% 0.23% 24.25% Alteromonadales Disease 8.75% 5.01% 6.62% 4.65% 17.87% Bacillales Healthy-on-Disease 4.71% 10.27% 0.53% 0.02% 25.65% Oceanospirillales Disease 4.70% 2.72% 4.16% 1.69% 9.16% Rhodobacterales Healthy-on-Disease 4.53% 6.09% 1.25% 0.19% 14.49% Cenarchaeales Healthy-on-Disease 4.24% 8.12% 0.32% 0.18% 18.72% Vibrionales Disease 3.15% 1.64% 2.87% 1.05% 5.71% Alteromonadales Healthy 3.13% 6.53% 0.27% 0.07% 14.81% Unassigned Disease 2.55% 1.20% 2.46% 1.01% 4.29% Flavobacteriales Disease 2.52% 2.15% 1.73% 0.71% 6.40% [Leptospirales] Healthy-on-Disease 2.27% NA 2.27% 2.27% 2.27% Pseudomonadales Healthy 2.19% 4.95% 0.18% 0.07% 12.28% Actinomycetales Healthy-on-Disease 1.86% 2.89% 0.52% 0.21% 6.20% Acidimicrobiales Healthy-on-Disease 1.86% 4.24% 0.13% 0.02% 10.52% Xanthomonadales Healthy-on-Disease 1.78% 2.85% 0.21% 0.07% 5.07% Rhizobiales Healthy 1.77% 3.34% 0.53% 0.16% 8.59% Rhodobacterales Healthy 1.76% 2.83% 0.73% 0.19% 7.50% Acidimicrobiales Disease 1.75% 1.58% 1.72% 0.08% 4.01% Campylobacterales Disease 1.75% 1.73% 1.09% 0.40% 4.72% Myxococcales Disease 1.56% 1.35% 1.17% 0.20% 3.84% Alteromonadales Healthy-on-Disease 1.45% 2.61% 0.36% 0.09% 6.73% Clostridiales Disease 1.44% 2.38% 0.57% 0.07% 6.26% Oceanospirillales Healthy 1.43% 2.20% 0.49% 0.02% 5.73% Bdellovibrionales Disease 1.36% 1.31% 1.01% 0.15% 3.33% Bdellovibrionales Healthy 1.33% 1.95% 0.65% 0.11% 5.23% [Saprospirales] Disease 1.27% 1.21% 1.02% 0.27% 3.23% Pseudomonadales Healthy-on-Disease 1.27% 2.29% 0.26% 0.19% 5.36% Cytophagales Healthy-on-Disease 1.25% 1.73% 0.46% 0.06% 4.28% Bacillales Healthy 1.22% 1.97% 0.38% 0.07% 4.74% Enterobacteriales Healthy-on-Disease 1.21% 1.90% 0.22% 0.01% 3.41% Cenarchaeales Healthy 1.17% 1.34% 0.71% 0.04% 3.75% Unassigned Healthy 1.13% 1.96% 0.32% 0.13% 4.64% Flavobacteriales Healthy 1.07% 2.05% 0.17% 0.09% 4.74% Spirobacillales Disease 0.96% 1.31% 0.41% 0.06% 3.20% Chromatiales Healthy 0.92% 1.69% 0.10% 0.02% 3.46% Rhizobiales Disease 0.91% 0.29% 0.96% 0.44% 1.22% HTCC2188 Healthy 0.89% NA 0.89% 0.89% 0.89% Rhizobiales Healthy-on-Disease 0.84% 1.01% 0.42% 0.02% 2.73% Cytophagales Disease 0.84% 0.61% 0.54% 0.41% 1.86% Chromatiales Disease 0.82% 0.97% 0.34% 0.23% 2.72% Burkholderiales Healthy-on-Disease 0.80% 1.14% 0.39% 0.01% 3.03% Ucp1540 Disease 0.75% 0.98% 0.13% 0.05% 2.30% Nitrospirales Disease 0.75% NA 0.75% 0.75% 0.75% Cenarchaeales Disease 0.70% 1.31% 0.06% 0.03% 3.33% Sphingomonadales Healthy-on-Disease 0.69% 1.07% 0.14% 0.02% 2.73% Lactobacillales Disease 0.68% NA 0.68% 0.68% 0.68% Legionellales Disease 0.66% 0.66% 0.46% 0.12% 1.98%

96 Order Treatment Mean RA SD RA Median RA Min RA Max RA Rhodospirillales Disease 0.66% 0.93% 0.21% 0.13% 2.31% Pirellulales Disease 0.65% 0.48% 0.54% 0.11% 1.43% Synechococcales Healthy-on-Disease 0.64% 1.09% 0.21% 0.02% 2.84% Oceanospirillales Healthy-on-Disease 0.63% 0.57% 0.44% 0.05% 1.40% Rhodospirillales Healthy 0.61% 0.80% 0.30% 0.07% 1.78% Vibrionales Healthy 0.60% 1.16% 0.05% 0.02% 2.67% [Pedosphaerales] Healthy 0.59% NA 0.59% 0.59% 0.59% Rhodospirillales Healthy-on-Disease 0.57% 0.65% 0.29% 0.16% 1.53% Bdellovibrionales Healthy-on-Disease 0.53% 0.60% 0.22% 0.02% 1.37% Sphingomonadales Disease 0.52% 0.44% 0.45% 0.08% 1.03% Thiotrichales Disease 0.51% 0.37% 0.49% 0.08% 0.95% SBR1031 Disease 0.50% 0.54% 0.50% 0.12% 0.88% Vibrionales Healthy-on-Disease 0.50% 0.55% 0.36% 0.05% 1.57% Caldilineales Healthy 0.49% NA 0.49% 0.49% 0.49% Campylobacterales Healthy 0.48% 0.24% 0.55% 0.07% 0.73% Desulfobacterales Disease 0.45% 0.46% 0.26% 0.14% 1.37% Unassigned Healthy-on-Disease 0.45% 0.73% 0.11% 0.04% 1.89% Flavobacteriales Healthy-on-Disease 0.41% 0.58% 0.23% 0.02% 1.58% Spirochaetales Healthy 0.39% NA 0.39% 0.39% 0.39% Bacillales Disease 0.39% 0.44% 0.25% 0.06% 1.23% Chromatiales Healthy-on-Disease 0.39% 0.50% 0.26% 0.02% 1.26% Actinomycetales Disease 0.37% 0.23% 0.35% 0.09% 0.61% NB1-j Healthy-on-Disease 0.36% 0.54% 0.12% 0.02% 1.16% Acidimicrobiales Healthy 0.35% 0.58% 0.10% 0.06% 1.38% Xanthomonadales Disease 0.35% 0.41% 0.15% 0.08% 0.82% Sphingomonadales Healthy 0.35% 0.25% 0.30% 0.08% 0.69% iii1-15 Disease 0.34% NA 0.34% 0.34% 0.34% Verrucomicrobiales Disease 0.33% 0.25% 0.28% 0.03% 0.67% Solibacterales Healthy 0.32% 0.25% 0.32% 0.15% 0.49% Clostridiales Healthy-on-Disease 0.31% 0.19% 0.31% 0.17% 0.44% Synechococcales Healthy 0.30% 0.25% 0.19% 0.13% 0.59% BPC015 Healthy 0.30% NA 0.30% 0.30% 0.30% Fusobacteriales Healthy 0.30% NA 0.30% 0.30% 0.30% Lactobacillales Healthy 0.30% NA 0.30% 0.30% 0.30% Nitrospirales Healthy 0.30% NA 0.30% 0.30% 0.30% Sva0725 Healthy 0.30% NA 0.30% 0.30% 0.30% SBR1031 Healthy-on-Disease 0.29% 0.18% 0.35% 0.09% 0.42% NB1-j Healthy 0.28% 0.22% 0.30% 0.05% 0.49% [Marinicellales] Disease 0.27% 0.16% 0.22% 0.15% 0.45% GCA004 Disease 0.27% NA 0.27% 0.27% 0.27% Cytophagales Healthy 0.27% 0.11% 0.30% 0.12% 0.37% HTCC2188 Disease 0.27% 0.25% 0.22% 0.03% 0.68% Opitutales Healthy-on-Disease 0.26% NA 0.26% 0.26% 0.26% Thiohalorhabdales Disease 0.26% 0.25% 0.17% 0.05% 0.66% Chroococcales Disease 0.26% 0.20% 0.27% 0.03% 0.52% Kiloniellales Disease 0.25% 0.18% 0.27% 0.03% 0.45% Spirochaetales Disease 0.25% 0.25% 0.20% 0.03% 0.57% SBR1031 Healthy 0.25% 0.30% 0.10% 0.05% 0.59% Burkholderiales Healthy 0.23% 0.20% 0.19% 0.05% 0.59% GMD14H09 Disease 0.23% 0.11% 0.27% 0.11% 0.38% EC94 Healthy-on-Disease 0.23% 0.35% 0.04% 0.01% 0.63% Gloeobacterales Disease 0.22% 0.19% 0.18% 0.03% 0.48% Solibacterales Healthy-on-Disease 0.22% 0.27% 0.11% 0.03% 0.52% Desulfovibrionales Disease 0.21% 0.19% 0.22% 0.02% 0.40% Pirellulales Healthy-on-Disease 0.21% NA 0.21% 0.21% 0.21%

97 Order Treatment Mean RA SD RA Median RA Min RA Max RA Pseudanabaenales Disease 0.21% 0.21% 0.14% 0.03% 0.57% Thiotrichales Healthy-on-Disease 0.20% 0.16% 0.20% 0.09% 0.32% Pseudanabaenales Healthy-on-Disease 0.20% 0.02% 0.20% 0.19% 0.21% Clostridiales Healthy 0.20% NA 0.20% 0.20% 0.20% Enterobacteriales Healthy 0.20% NA 0.20% 0.20% 0.20% iii1-15 Healthy 0.20% NA 0.20% 0.20% 0.20% Methylococcales Healthy 0.20% NA 0.20% 0.20% 0.20% Spirobacillales Healthy 0.20% NA 0.20% 0.20% 0.20% Verrucomicrobiales Healthy 0.20% NA 0.20% 0.20% 0.20% Synechococcales Disease 0.19% 0.23% 0.09% 0.02% 0.61% Xanthomonadales Healthy 0.18% 0.19% 0.12% 0.02% 0.39% Caulobacterales Healthy 0.16% 0.05% 0.18% 0.09% 0.20% MSBL9 Disease 0.16% NA 0.16% 0.16% 0.16% Sphingobacteriales Disease 0.16% 0.16% 0.16% 0.05% 0.27% Campylobacterales Healthy-on-Disease 0.16% 0.04% 0.16% 0.11% 0.22% Caulobacterales Disease 0.16% 0.22% 0.05% 0.01% 0.41% Bacteroidales Disease 0.15% 0.11% 0.15% 0.03% 0.32% Caldilineales Disease 0.15% 0.16% 0.14% 0.01% 0.30% MLE1-12 Healthy 0.15% 0.11% 0.20% 0.02% 0.22% Caulobacterales Healthy-on-Disease 0.14% 0.01% 0.14% 0.13% 0.15% Sva0725 Disease 0.14% 0.09% 0.18% 0.03% 0.20% Burkholderiales Disease 0.14% 0.13% 0.06% 0.03% 0.34% Planctomycetales Disease 0.14% 0.12% 0.10% 0.04% 0.34% Kordiimonadales Healthy 0.14% 0.09% 0.14% 0.07% 0.20% NB1-j Disease 0.14% 0.12% 0.08% 0.03% 0.31% Halanaerobiales Disease 0.13% 0.03% 0.13% 0.11% 0.16% Unassigned Disease 0.13% 0.07% 0.12% 0.07% 0.22% Sva0725 Healthy-on-Disease 0.13% 0.20% 0.04% 0.02% 0.42% FAC87 Healthy-on-Disease 0.12% 0.14% 0.12% 0.02% 0.22% Nitrospirales Healthy-on-Disease 0.12% 0.10% 0.11% 0.02% 0.22% Sphingobacteriales Healthy 0.11% 0.12% 0.11% 0.02% 0.20% [Marinicellales] Healthy-on-Disease 0.11% NA 0.11% 0.11% 0.11% Arctic96B-7 Healthy-on-Disease 0.11% NA 0.11% 0.11% 0.11% Chlamydiales Healthy-on-Disease 0.11% NA 0.11% 0.11% 0.11% CL500-15 Healthy-on-Disease 0.11% NA 0.11% 0.11% 0.11% Stigonematales Healthy-on-Disease 0.11% NA 0.11% 0.11% 0.11% DRC31 Disease 0.10% 0.10% 0.09% 0.01% 0.22% Actinomycetales Healthy 0.10% 0.07% 0.11% 0.02% 0.19% EC94 Healthy 0.10% NA 0.10% 0.10% 0.10% Legionellales Healthy 0.10% NA 0.10% 0.10% 0.10% Myxococcales Healthy-on-Disease 0.10% 0.07% 0.10% 0.05% 0.15% Rubrobacterales Healthy 0.10% NA 0.10% 0.10% 0.10% Chlamydiales Disease 0.10% 0.12% 0.04% 0.03% 0.27%

98 Appendix Table 15. Differentially Abundant Taxa. Differentially abundant taxa at the phylum, class, order, family, and genus levels between healthy and disease tissue, healthy-on-disease and disease tissue, and healthy and healthy-on-disease tissue identified via DESeq analysis. All taxa with FDR < 0.05 (padj) are included in the table.

OTU baseMean log2FoldChange lfcSE stat pvalue padj Kingdom Phylum Class Order Family Genus Level Comparison NCR.OTU11000 10.62178 -4.29737 1.32640 -3.23988 1.20E-03 2.46E-02 Bacteria Bacteroidetes [Saprospirae] Class H vs. D NCR.OTU11000 10.29125 -4.29032 1.31044 -3.27395 1.06E-03 5.66E-03 Bacteria Bacteroidetes [Saprospirae] [Saprospirales] Order H vs. D 266637 32.31966 -2.12254 0.80279 -2.64396 8.19E-03 2.56E-02 Bacteria Bacteroidetes Flavobacteriia Flavobacteriales Order H vs. D NCR.OTU11000 9.35966 -5.70759 1.53981 -3.70669 2.10E-04 5.30E-03 Bacteria Bacteroidetes [Saprospirae] [Saprospirales] Saprospiraceae Family H vs. D 248079 4.89974 -4.60718 1.56917 -2.93605 3.32E-03 2.14E-02 Bacteria Bacteroidetes Cytophagia Cytophagales Flammeovirgaceae Family H vs. D NR.OTU75 2.38923 -4.47352 1.39326 -3.21082 1.32E-03 2.46E-02 Bacteria Cyanobacteria Oscillatoriophycideae Class H vs. D NR.OTU75 2.17384 -4.32512 1.35483 -3.19238 1.41E-03 5.88E-03 Bacteria Cyanobacteria Oscillatoriophycideae Chroococcales Order H vs. D NR.OTU82 4.35895 -5.07206 1.60832 -3.15364 1.61E-03 2.46E-02 Bacteria Fibrobacteres Fibrobacteria Class H vs. D NR.OTU82 5.67701 -5.25808 1.62885 -3.22808 1.25E-03 5.66E-03 Bacteria Fibrobacteres Fibrobacteria Ucp1540 Order H vs. D 4407268 14.04347 -5.41356 1.50248 -3.60309 3.14E-04 1.92E-02 Bacteria Firmicutes Clostridia Class H vs. D 4407268 17.48753 -5.60782 1.52106 -3.68679 2.27E-04 4.83E-03 Bacteria Firmicutes Clostridia Clostridiales Order H vs. D 4407268 6.68251 -5.81620 1.76535 -3.29464 9.85E-04 1.14E-02 Bacteria Firmicutes Clostridia Clostridiales Lachnospiraceae Family H vs. D 4319491 6.18865 -3.87984 1.26247 -3.07322 2.12E-03 2.58E-02 Bacteria Planctomycetes Planctomycetia Class H vs. D 4319491 4.71332 -3.55792 1.31359 -2.70855 6.76E-03 2.25E-02 Bacteria Planctomycetes Planctomycetia Pirellulales Order H vs. D NCR.OTU6437 5.22253 -4.42412 1.35085 -3.27507 1.06E-03 5.66E-03 Bacteria Proteobacteria Deltaproteobacteria Desulfobacterales Order H vs. D 4301759 5.13887 -3.45201 1.37530 -2.51000 1.21E-02 3.55E-02 Bacteria Proteobacteria Gammaproteobacteria Thiotrichales Order H vs. D 836059 97.65516 -2.84234 0.85550 -3.32244 8.92E-04 5.66E-03 Bacteria Proteobacteria Gammaproteobacteria Alteromonadales Order H vs. D 312009 6.71008 -5.34136 1.30111 -4.10524 4.04E-05 2.02E-03 Bacteria Proteobacteria Gammaproteobacteria Legionellales Order H vs. D 355163 4.72857 -4.26701 1.55923 -2.73662 6.21E-03 2.25E-02 Bacteria Proteobacteria Deltaproteobacteria Spirobacillales Order H vs. D 150441 4167.50163 2.78646 0.80984 3.44077 5.80E-04 4.83E-03 Bacteria Proteobacteria Alphaproteobacteria Rickettsiales Order H vs. D 2670730 34.88286 -2.84673 0.81461 -3.49461 4.75E-04 4.83E-03 Bacteria Proteobacteria Gammaproteobacteria Vibrionales Order H vs. D 359953 312.88786 -3.41156 0.94240 -3.62006 2.95E-04 4.83E-03 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Order H vs. D 544033 17.52699 -4.99352 1.44241 -3.46192 5.36E-04 4.83E-03 Bacteria Proteobacteria Deltaproteobacteria Myxococcales Order H vs. D 365755 2.59714 -4.51443 1.39203 -3.24307 1.18E-03 5.66E-03 Bacteria Proteobacteria Gammaproteobacteria Thiohalorhabdales Order H vs. D 836059 36.07527 -5.30813 1.23441 -4.30015 1.71E-05 9.90E-04 Bacteria Proteobacteria Gammaproteobacteria Alteromonadales Colwelliaceae Family H vs. D 312009 5.06358 -5.48589 1.76479 -3.10852 1.88E-03 1.82E-02 Bacteria Proteobacteria Gammaproteobacteria Legionellales Francisellaceae Family H vs. D 4351327 7.45799 -3.87688 1.31275 -2.95325 3.14E-03 2.14E-02 Bacteria Proteobacteria Gammaproteobacteria Alteromonadales OM60 Family H vs. D 2670730 20.95967 -3.26537 1.10146 -2.96458 3.03E-03 2.14E-02 Bacteria Proteobacteria Gammaproteobacteria Vibrionales Pseudoalteromonadaceae Family H vs. D 303714 12.11850 -6.48966 1.78365 -3.63841 2.74E-04 5.30E-03 Bacteria Proteobacteria Gammaproteobacteria Alteromonadales Ferrimonadaceae Family H vs. D 330888 5.99382 -4.37864 1.66578 -2.62858 8.57E-03 4.97E-02 Bacteria Proteobacteria Gammaproteobacteria Oceanospirillales Oleiphilaceae Family H vs. D 359953 323.81407 -3.65603 1.10675 -3.30340 9.55E-04 1.14E-02 Bacteria Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae Family H vs. D 836059 42.31486 -5.72122 1.36343 -4.19619 2.71E-05 3.53E-03 Bacteria Proteobacteria Gammaproteobacteria Alteromonadales Colwelliaceae Thalassomonas Genus H vs. D 153007 3.01230 -3.83484 1.40557 -2.72833 6.37E-03 2.25E-02 Bacteria Verrucomicrobia Verrucomicrobiae Verrucomicrobiales Order H vs. D 4422405 11.68770 4.18950 1.57073 2.66722 7.65E-03 4.44E-02 Bacteria Actinobacteria Actinobacteria Actinomycetales Micrococcaceae Family HD vs. D

99

OTU baseMean log2FoldChange lfcSE stat pvalue padj Kingdom Phylum Class Order Family Genus Level Comparison NR.OTU5 4.38003 -4.68247 1.62003 -2.89036 3.85E-03 2.81E-02 Bacteria Actinobacteria Acidimicrobiia Acidimicrobiales wb1_P06 Family HD vs. D NCR.OTU11000 10.29125 -3.20850 1.27014 -2.52609 1.15E-02 4.38E-02 Bacteria Bacteroidetes [Saprospirae] [Saprospirales] Order HD vs. D 266637 32.31966 -2.18962 0.79810 -2.74353 6.08E-03 3.14E-02 Bacteria Bacteroidetes Flavobacteriia Flavobacteriales Order HD vs. D NCR.OTU11000 9.35966 -4.53397 1.49059 -3.04172 2.35E-03 2.45E-02 Bacteria Bacteroidetes [Saprospirae] [Saprospirales] Saprospiraceae Family HD vs. D 266637 2.69308 -5.16717 1.77171 -2.91649 3.54E-03 4.84E-02 Bacteria Bacteroidetes Flavobacteriia Flavobacteriales Cryomorphaceae Fluviicola Genus HD vs. D NR.OTU82 5.48444 -5.39544 1.66454 -3.24140 1.19E-03 3.33E-02 Bacteria Fibrobacteres Phylum HD vs. D NR.OTU82 4.35895 -5.17770 1.61471 -3.20658 1.34E-03 1.34E-02 Bacteria Fibrobacteres Fibrobacteria Class HD vs. D NR.OTU82 5.67701 -5.50996 1.63483 -3.37036 7.51E-04 7.13E-03 Bacteria Fibrobacteres Fibrobacteria Ucp1540 Order HD vs. D 343059 1.72570 -4.42412 1.68524 -2.62522 8.66E-03 4.50E-02 Bacteria Firmicutes Clostridia Clostridiales [Acidaminobacteraceae] Family HD vs. D 4407268 6.68251 -6.07384 1.76755 -3.43630 5.90E-04 1.08E-02 Bacteria Firmicutes Clostridia Clostridiales Lachnospiraceae Family HD vs. D 4319491 4.71332 -4.00273 1.33484 -2.99867 2.71E-03 1.93E-02 Bacteria Planctomycetes Planctomycetia Pirellulales Order HD vs. D 4319491 4.76804 -4.17558 1.53234 -2.72497 6.43E-03 4.27E-02 Bacteria Planctomycetes Planctomycetia Pirellulales Pirellulaceae Family HD vs. D 268968 20.78635 3.80505 1.00356 3.79154 1.50E-04 4.49E-03 Bacteria Proteobacteria Betaproteobacteria Class HD vs. D 544033 54.22925 -1.97077 0.65673 -3.00086 2.69E-03 2.02E-02 Bacteria Proteobacteria Deltaproteobacteria Class HD vs. D NCR.OTU6437 5.22253 -4.78405 1.37218 -3.48644 4.89E-04 7.00E-03 Bacteria Proteobacteria Deltaproteobacteria Desulfobacterales Order HD vs. D 836059 97.65516 -2.50475 0.85174 -2.94076 3.27E-03 2.07E-02 Bacteria Proteobacteria Gammaproteobacteria Alteromonadales Order HD vs. D 312009 6.71008 -5.48035 1.31700 -4.16125 3.17E-05 1.80E-03 Bacteria Proteobacteria Gammaproteobacteria Legionellales Order HD vs. D 2556149 63.01290 -2.12781 0.83646 -2.54384 1.10E-02 4.38E-02 Bacteria Proteobacteria Gammaproteobacteria Oceanospirillales Order HD vs. D 355163 4.72857 -5.33587 1.57828 -3.38082 7.23E-04 7.13E-03 Bacteria Proteobacteria Deltaproteobacteria Spirobacillales Order HD vs. D 150441 4167.50163 2.63948 0.80982 3.25934 1.12E-03 9.09E-03 Bacteria Proteobacteria Alphaproteobacteria Rickettsiales Order HD vs. D 4416718 1.96246 -3.71477 1.42367 -2.60928 9.07E-03 3.98E-02 Bacteria Proteobacteria Gammaproteobacteria HTCC2188 Order HD vs. D 268968 18.44346 3.93077 1.00153 3.92475 8.68E-05 2.47E-03 Bacteria Proteobacteria Betaproteobacteria Burkholderiales Order HD vs. D NCR.OTU16264 1.95932 -4.12026 1.43306 -2.87514 4.04E-03 2.30E-02 Bacteria Proteobacteria Deltaproteobacteria GMD14H09 Order HD vs. D 151316 3.17793 -4.21179 1.58593 -2.65573 7.91E-03 4.44E-02 Bacteria Proteobacteria Alphaproteobacteria Rhizobiales Hyphomicrobiaceae Family HD vs. D 836059 36.07527 -6.11512 1.24681 -4.90461 9.36E-07 6.83E-05 Bacteria Proteobacteria Gammaproteobacteria Alteromonadales Colwelliaceae Family HD vs. D 312009 5.06358 -5.73571 1.76632 -3.24726 1.17E-03 1.42E-02 Bacteria Proteobacteria Gammaproteobacteria Legionellales Francisellaceae Family HD vs. D 2556149 28.11412 -2.88140 0.99674 -2.89084 3.84E-03 2.81E-02 Bacteria Proteobacteria Gammaproteobacteria Oceanospirillales Oceanospirillaceae Family HD vs. D 303714 12.11850 -6.76665 1.78738 -3.78579 1.53E-04 5.59E-03 Bacteria Proteobacteria Gammaproteobacteria Alteromonadales Ferrimonadaceae Family HD vs. D 330888 5.99382 -5.98004 1.69632 -3.52530 4.23E-04 1.03E-02 Bacteria Proteobacteria Gammaproteobacteria Oceanospirillales Oleiphilaceae Family HD vs. D 1146588 1.39352 -4.29430 1.64993 -2.60272 9.25E-03 4.50E-02 Bacteria Proteobacteria Alphaproteobacteria Rhizobiales Cohaesibacteraceae Family HD vs. D 513711 15.92934 -4.13589 1.41555 -2.92175 3.48E-03 2.81E-02 Bacteria Proteobacteria Epsilonproteobacteria Campylobacterales Campylobacteraceae Family HD vs. D 836059 42.31486 -6.41042 1.36987 -4.67957 2.87E-06 1.18E-04 Bacteria Proteobacteria Gammaproteobacteria Alteromonadales Colwelliaceae Thalassomonas Genus HD vs. D 513711 18.42427 -4.83213 1.49043 -3.24211 1.19E-03 2.43E-02 Bacteria Proteobacteria Epsilonproteobacteria Campylobacterales Campylobacteraceae Arcobacter Genus HD vs. D 153007 3.00936 -4.88284 1.46329 -3.33688 8.47E-04 1.27E-02 Bacteria Verrucomicrobia Verrucomicrobiae Class HD vs. D 153007 3.01230 -4.99168 1.43208 -3.48561 4.91E-04 7.00E-03 Bacteria Verrucomicrobia Verrucomicrobiae Verrucomicrobiales Order HD vs. D 153007 3.14141 -5.29193 1.61285 -3.28111 1.03E-03 1.42E-02 Bacteria Verrucomicrobia Verrucomicrobiae Verrucomicrobiales Verrucomicrobiaceae Family HD vs. D NCR.OTU4450 37.31494 2.79483 1.02927 2.71536 6.62E-03 3.14E-02 Archaea Crenarchaeota Thaumarchaeota Cenarchaeales Order HD vs. D NR.OTU5 70.08461 -2.91317 0.87020 -3.34771 8.15E-04 2.28E-02 Bacteria Actinobacteria Phylum H vs. HD

0 10

100

Appendix Table 16. Taxa Relative Abundance Pairwise Comparisons. Relative abundance pairwise comparisons between healthy (H) and disease (D), healthy-on-disease (HD) and disease, and healthy and healthy-on-disease tissue determined by Wilcoxon rank sum tests. All taxa with FDR < 0.05 are included in the table.

OTU teststat rawp adjp plower Kingdom Phylum Class Order Family Genus fdr Level Comparison NR.OTU75 -2.88231 0.00216 4.76E-02 1.08E-03 Bacteria Cyanobacteria Oscillatoriophycideae Chroococcales 3.22E-02 Order H vs. D

4319491 -2.88231 0.00216 4.76E-02 1.08E-03 Bacteria Planctomycetes Planctomycetia Pirellulales 3.22E-02 Order H vs. D

NCR.OTU6437 -2.88231 0.00216 4.76E-02 1.08E-03 Bacteria Proteobacteria Deltaproteobacteria Desulfobacterales 3.22E-02 Order H vs. D

544033 -2.88231 0.00216 4.76E-02 1.08E-03 Bacteria Proteobacteria Deltaproteobacteria Myxococcales 3.22E-02 Order H vs. D

312009 -2.88231 0.00216 4.76E-02 1.08E-03 Bacteria Proteobacteria Gammaproteobacteria Legionellales 3.22E-02 Order H vs. D

365755 -2.88231 0.00216 4.76E-02 1.08E-03 Bacteria Proteobacteria Gammaproteobacteria Thiohalorhabdales 3.22E-02 Order H vs. D

4301759 -2.88231 0.00216 4.76E-02 1.08E-03 Bacteria Proteobacteria Gammaproteobacteria Thiotrichales 3.22E-02 Order H vs. D

3713320 -2.88231 0.00216 6.71E-02 1.08E-03 Bacteria Actinobacteria Acidimicrobiia Acidimicrobiales ntu14 4.33E-02 Family HD vs. D

4348329 -2.88231 0.00216 6.71E-02 1.08E-03 Bacteria Actinobacteria Acidimicrobiia Acidimicrobiales koll13 4.33E-02 Family HD vs. D

266637 -2.88231 0.00216 6.71E-02 1.08E-03 Bacteria Bacteroidetes Flavobacteriia Flavobacteriales Cryomorphaceae 4.33E-02 Family HD vs. D

343059 -2.88231 0.00216 6.71E-02 1.08E-03 Bacteria Firmicutes Clostridia Clostridiales [Acidaminobacteraceae] 4.33E-02 Family HD vs. D

1146588 -2.88231 0.00216 6.71E-02 1.08E-03 Bacteria Proteobacteria Alphaproteobacteria Rhizobiales Cohaesibacteraceae 4.33E-02 Family HD vs. D

NCR.OTU6437 -2.88231 0.00216 5.41E-02 1.08E-03 Bacteria Proteobacteria Deltaproteobacteria Desulfobacterales 3.72E-02 Order HD vs. D

544033 -2.88231 0.00216 5.41E-02 1.08E-03 Bacteria Proteobacteria Deltaproteobacteria Myxococcales 3.72E-02 Order HD vs. D

836059 -2.88231 0.00216 6.71E-02 1.08E-03 Bacteria Proteobacteria Gammaproteobacteria Alteromonadales Colwelliaceae 4.33E-02 Family HD vs. D

216695 -2.88231 0.00216 5.41E-02 1.08E-03 Bacteria Proteobacteria Gammaproteobacteria HTCC2188 3.72E-02 Order HD vs. D

312009 -2.88231 0.00216 5.41E-02 1.08E-03 Bacteria Proteobacteria Gammaproteobacteria Legionellales 3.72E-02 Order HD vs. D

2556149 -2.88231 0.00216 5.41E-02 1.08E-03 Bacteria Proteobacteria Gammaproteobacteria Oceanospirillales 3.72E-02 Order HD vs. D

2556149 -2.88231 0.00216 6.71E-02 1.08E-03 Bacteria Proteobacteria Gammaproteobacteria Oceanospirillales Oceanospirillaceae 4.33E-02 Family HD vs. D

153007 -2.88231 0.00216 5.41E-02 1.08E-03 Bacteria Verrucomicrobia Verrucomicrobiae Verrucomicrobiales 3.72E-02 Order HD vs. D

153007 -2.88231 0.00216 6.71E-02 1.08E-03 Bacteria Verrucomicrobia Verrucomicrobiae Verrucomicrobiales Verrucomicrobiaceae 4.33E-02 Family HD vs. D

1 10 101