University of Florida Thesis Or Dissertation Formatting Template

PATHOGEN, HOST, AND ENVIRONMENTAL DYNAMICS: A CASE STUDY OF PANULIRUS ARGUS VIRUS 1 IN CARIBBEAN SPINY LOBSTERS

ABIGAIL S. CLARK

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

UNIVERSITY OF FLORIDA

2017

To Jason, Mom, Dad, and Grandma

ACKNOWLEDGMENTS

This work would not have been successful without the support of several individuals and funding sources. I would first like to thank my supervisory chair, Donald Behringer, and my supervisory co-chair, Thomas Waltzek, without whom this work would not have possible. I am sincerely grateful to them for their leadership, intellect, and endless encouragement. The following personnel assisted with sample collection: Angelo Jason Spadaro, Mark Butler, Joshua

Anderson, Jack Butler, Gaya Gnanalingam, Danielle Puls, Nathan Berkebile, Erica Ross, David

Ousley, Evan Hill, and Alyssa Thompson. Additional help was provided by Tom Matthews,

Casey Butler, and Gabby Renchen of Florida Fish and Wildlife Conservation Commission who supplied postlarval lobsters. Assistance with molecular work was provided by Linda Archer,

Natalie Stilwell, Patrick Thompson, Jason Ferrante, Galaxia Cortés-Hinojosa, Gabriel Diaz,

Anna Swigris, and Jaime Haggard. I would like to acknowledge the following faculty members, students, and support staff: Kuttichantran Subramaniam, Andrew Kane, Ross Brooks, Ruth

Francis-Floyd, James Wellehan, James Colee, Debra Murie, Daryl Parkyn, Larry Tolbert, Justin

Stilwell, Samantha Koda, Mameow Preeyanan, Kamonchai Imnoi, Elizabeth Scherbatskoy,

Nelmarie Landrau Giovannetti, Rachel Henriquez, Jared Freitas, Pedro Henrique Viadanna, and

Maria Jose Robles Malagamba. These individuals were instrumental in various aspects of my research, ranging from laboratory logistics to lobster dissections. Finally, I would like to recognize my committee members, Samantha Wisely, Jessica Moss Small, and Robert Swett for their unending support and mentorship. This project was funded by a Florida Sea Grant Scholars program grant (ASC), a University of Florida Opportunity Seed Fund grant (DCB, JFXW, and

TBW), and National Science Foundation – Biological Oceanography grant OCE-0928398

(DCB).

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 7

LIST OF FIGURES ...... 8

LIST OF ABBREVIATIONS ...... 9

ABSTRACT ...... 10

CHAPTER

1 LITERATURE REVIEW ...... 12

Disease Ecology ...... 12 Environmental Effects on Crustacean Health ...... 15 Crustacean Immune Response ...... 16 Pathogens in Crustaceans ...... 17

2 VALIDATION OF A TAQMAN REAL-TIME QUANTITATIVE PCR ASSAY FOR THE DETECTION OF PANULIRUS ARGUS VIRUS 1 ...... 20

Introduction ...... 20 Materials and Methods ...... 21 In silico TaqMan qPCR Primer and Probe Design ...... 21 Detection of PaV1 DNA Using the qPCR Assay ...... 21 Estimation of the qPCR Assay Slope, Y-Intercept, Correlation Coefficient (R2), Efficiency, Dynamic Range, Analytical Sensitivity, Repeatability, Reproducibility, and Analytical Specificity ...... 22 Estimation of the qPCR Assay Diagnostic Sensitivity and Specificity ...... 23 Results...... 25 In silico TaqMan qPCR Primer and Probe Design ...... 25 Estimation of the qPCR Assay Slope, Y-Intercept, Correlation Coefficient (R2), Efficiency, Dynamic Range, Analytical Sensitivity, Repeatability, Reproducibility, and Analytical Specificity ...... 25 Estimation of the qPCR Assay Diagnostic Sensitivity and Specificity ...... 26 Discussion ...... 26

3 VIABILITY OF PANULIRUS ARGUS VIRUS 1 IN SEAWATER ...... 39

Introduction ...... 39 Materials and Methods ...... 41 Lobster Screening ...... 41 Viral Purification and Quantitation ...... 42

Sampling and Experimental Setup ...... 42 Results...... 44 Viral Inocula and Concentrations ...... 44 PaV1 Viability ...... 45 Discussion ...... 46

4 THE RELATIONSHIP BETWEEN VIRAL LOAD AND INFECTION IN CARIBBEAN SPINY LOBSTERS EXPOSED TO PANULIRUS ARGUS VIRUS 1 ...... 55

Introduction ...... 55 Materials and Methods ...... 58 Lobster Collection and Screening ...... 58 Inoculation ...... 58 Sampling ...... 60 Data Analyses ...... 60 Results...... 60 Viral Load ...... 60 Lobster Mortality ...... 61 Discussion ...... 62

5 SPONGES AND THE SPATIAL EPIDEMIOLOGY OF PANULIRUS ARGUS VIRUS 1 IN CARIBBEAN SPINY LOBSTERS THROUGHOUT THE FLORIDA KEYS, USA ...... 74

Introduction ...... 74 Materials and Methods ...... 77 Sample Collection ...... 77 Data Analyses ...... 78 Results...... 79 PaV1 Prevalence ...... 79 Community Characteristics ...... 79 PaV1 Prevalence vs. Community Characteristics ...... 80 Discussion ...... 81

6 CONCLUSIONS ...... 95

LIST OF REFERENCES ...... 98

BIOGRAPHICAL SKETCH ...... 112

LIST OF TABLES

Table page

2-1 Primers and probe used in the development of the PaV1 TaqMan® qPCR assay...... 29

2-2 Samples used to validate the PaV1 TaqMan® real-time qPCR assay...... 30

2-3 Inter-assay variability (reproducibility) and intra-assay variability (repeatability) of the TaqMan® real-time qPCR assay for PaV1...... 35

4-1 Number of viral copies injected into each lobster at the beginning of experiment 1 and experiment 2...... 65

4-2 Results of a one-way ANOVA performed on the net change in viral count over time.. ...66

4-3 Results from a two-way ANOVA for experiment 1...... 67

4-4 Results from a two-way ANOVA for experiment 2...... 68

4-5 Results from a two-way repeated measures ANOVA...... 69

5-1 Summary of lobster surveys...... 85

5-2 Summary of community characteristics...... 86

5-3 Results of a generalized logistic mixed model...... 87

LIST OF FIGURES

Figure page

2-1 Sequences from 61 PaV1 alleles used for in silico primer and probe design...... 36

2-2 Assay amplification plot of triplicate 10-fold serial dilutions carrying between 107- 100 PaV1 plasmid DNA copies...... 37

2-3 Standard curve generated using triplicate 10-fold serial dilutions carrying between 107-100 PaV1 plasmid DNA copies...... 38

3-1 Experimental setup...... 51

3-2 Viral load detected in lobsters (n = 22) added to inoculated water for experiment 1...... 52

3-3 Viral load detected in lobsters (n = 40) added to inoculated water for experiment 2...... 53

3-4 Histology of PaV1-positive tissues...... 54

4-1 Log-transformed values representing mean number of viral copies per inoculum ...... 70

4-2 Net change in the number of viral copies detected six weeks post-inoculation...... 71

4-3 Regression of percent mortality of lobsters throughout experiments...... 72

4-4 Regression of percent survival of lobsters throughout experiments...... 73

5-1 Linear regression of lobsters infected with PaV1 and the lobster density at each sampling site (n = 20)...... 88

5-2 Distribution of sampling sites (n = 20) and the corresponding prevalence of PaV1 as determined by qPCR...... 89

5-3 Sponge density and species diversity of sampling sites (n = 7) across the lower region of the Florida Keys...... 90

5-4 Sponge density and species diversity of sampling site (n = 7) across the middle region of the Florida Keys...... 91

5-5 Sponge density and species diversity of sampling site (n = 6) across the upper region of the Florida Keys...... 92

5-6 Scree plot generated from the PCA...... 93

5-7 Biplot of the first two principal components...... 94

LIST OF ABBREVIATIONS

ANOVA Analysis of Variance

EtOH Ethanol

PCR Polymerase Chain Reaction

SEM Standard Error of the Mean

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

PATHOGEN, HOST, AND ENVIRONMENTAL DYNAMICS: A CASE STUDY OF PANULIRUS ARGUS VIRUS 1 IN CARIBBEAN SPINY LOBSTERS

Abigail S. Clark

December 2017

Chair: Donald C. Behringer Co-chair: Thomas B. Waltzek Major: Fisheries and Aquatic Sciences

The relationship between pathogens, hosts, and the environment is complex. The ecology of Panulirus argus Virus 1 (PaV1) is no exception. PaV1 is a pathogenic virus that infects the commercially-important Caribbean spiny lobster, Panulirus argus. This virus occurs throughout most of the range of P. argus, spanning the Caribbean Sea and Gulf of Mexico from Panama through the Antilles and North to Florida (USA). In some areas of the Caribbean, PaV1 reaches a mean prevalence of nearly 20%. In Florida Bay, a shallow lagoon situated between the Florida

Keys and mainland Florida, the prevalence of PaV1 varies spatially and temporally, ranging from 0 to 100%. Results from the present study revealed that lobster size and density, rather than habitat characteristics, best explained this variability. Though not significant, more infected lobsters tended to be in less complex habitats (e.g., high seagrass coverage and low sponge abundance). Another factor potentially driving variability in the distribution of PaV1 is the viability of the virus outside of its host (i.e., free virions in seawater). If PaV1 is infectious outside of the host only briefly, then even short distances or physical barriers could cause heterogeneity in prevalence among lobster populations. We determined that free-floating PaV1 was likely viable in seawater for up to 21 d and may, therefore, be able to infect metapopulations

across the greater Caribbean, especially populations predominantly consisting of juvenile lobsters. PaV1 causes clinical infections and mortality in juveniles and subclinical infections in adults. One potential explanation for this discrepancy in prevalence is viral load. We found that small juvenile lobsters had higher viral load of PaV1 relative to large juvenile lobsters, which may explain why PaV1-induced mortality is highest in lobsters of smaller size classes. Levels of

PaV1 were measured in experimental lobsters using a TaqMan real-time quantitative polymerase chain reaction (qPCR) assay that we developed and validated. The qPCR assay had an analytical sensitivity of 100% and an analytical specificity of 84%, and was therefore used to quantify lobster viral count in the present study. This work will collectively help to clarify interactions between PaV1, its host, and the environment.

CHAPTER 1 LITERATURE REVIEW

Disease Ecology

Diseases in the marine environment are diverse and ubiquitous. They infect a variety of marine organisms, ranging from phytoplankton to mammals. The first well-documented outbreak of a marine disease was that of the long-spine sea urchin, Diadema antillarum, in the 1980s

(Lessios et al. 1984). During this same period, there were massive die-offs of turtle grass,

Thalassia testudinum, in Florida (USA) (Robblee et al. 1991). Marine mammals, such as the common bottlenose dolphin, Tursiops truncatus, have also suffered mass mortality events from as early as 1987 (Kuehl et al. 1991). There has since been an increasing number of reported marine diseases and epidemics among many taxa (Harvell et al. 1999, 2004, Ward & Lafferty

2004).

Disease ecology examines the ecological interactions and effects of pathogens and hosts; and in a broader context the effect of disease at the individual, community, and meta-community levels (Park 2012). The “epidemiological triangle” that is oftentimes associated with disease ecology consists of three components: pathogen, host, and environment (Scholthof 2007, Vander

Wal et al. 2014). This interdisciplinary field combines ecology, immunopathology, epidemiology, biology, oceanography, and more recently, phylodynamics. These areas are applied in combination to understand the complex relationship between pathogens, hosts, and the environment.

Interactions between these three components of disease ecology are multifactorial and are, consequently, often difficult to ascertain. The complexities of disease ecology exist due to, and are magnified by, highly variable and interactive pathogen-host properties; properties which are driven by biotic and abiotic conditions. For example, a species can simultaneously host

several different pathogens, making it difficult for researchers to isolate and study any single species of pathogen within the host. On the other hand, a pathogen can exhibit variability in host specificity, further confounding studies of pathogen-host relationships. For example, families of virus within the group of nucleo-cytoplasmic large DNA viruses (NCLDV) infect different host species. Members of iridoviridae are found globally and have a broad host specificity, infecting both invertebrates and fish, whereas mimiviruses only infect Acanthamoeba polyphaga (Raoult et al. 2004, Claverie et al. 2009, Chinchar et al. 2009). Changes in environmental conditions are also taken into consideration when studying elements of disease ecology. Salmon infected by

Tetracapsuloides bryosalmonae, a parasite causing proliferative kidney disease (PKD), suffer impaired aerobic performance and will likely experience increased mortality with rising water temperature (Bruneaux et al. 2017). Both biotic and abiotic factors are taken into account when investigating aspects of disease ecology.

Biotic factors drive the transmission of pathogens among communities. Such factors include, but are not limited to, host behavior and prey-predator interactions. Disease avoidance is an adaptive strategy with which individuals prevent infections (Hart 1990). Otherwise social individuals exhibit behavioral immunity, which reduces the rate at which a pathogen is spread throughout the population (Navarrete & Fessler 2006, Dolan et al. 2014, Butler et al. 2015).

Interacting with infected individuals increases the risk that a healthy individual will become infected and/or be preyed upon (Loehle 1995, Behringer et al. 2006, Behringer et al. 2008). For instance, the female three-spined stickleback, Gasterosteus aculeatus, mate with males that produce parasitic-resistant offspring (Barber et al. 2001). Contrary to this study, it is sometimes beneficial to interact with an infected individual for reproductive or defensive purposes (Hart

1990). Not only can disease influence mate selection and reproductive fitness of fish, but disease

can also increase predation pressures on infected hosts (Crowden & Broom 1980, Hart 1990,

Behringer & Butler 2010). For example, the visual acuity of eyefluke-infested dace, Leuciscus leuciscus, is diminished which demands that dace forage in well-lit areas. This foraging strategy increases the exposure of dace to avian predators. When fish (i.e., intermediate host) are consumed by avian hosts, the flukes are spread from one species to another (Crowden & Broom

1980). Through host behavior, the spread of pathogens is facilitated or impeded throughout the environment.

The environment is dynamic, which presents a suite of challenges for pathogens and susceptible hosts. Fluctuations in temperature and salinity can lead to suppressed immunity and increased host susceptibility (Bowden et al. 2007, Yanjiao et al. 2011). For example, hyposaline conditions can increase pathogenicity of infective agents by increasing host susceptibility

(Takagishi et al. 2009), while hypersaline conditions facilitate the progression of marine diseases such as Dermo disease. Dermo disease, or perkinsosis, is lethal in the eastern oyster Crassostrea virginica and thrives under hypersaline conditions. As salinity increases, oysters experience disease-induced mortality (Powell et al. 1992). Hypo- and hypersalinity can decrease innate immune function, while low temperatures compromise both innate and adaptive immunity (Le

Morvan et al. 1998, Watts et al. 2001, Yanjiao et al. 2011). For example, low water temperature

(~10ºC) causes and exacerbates fungal infections in the channel catfish, Ictalurus punctatus, reducing fish immune response and causing mortality (Bly 1993). A similar response is observed in white sturgeon, Acipenser transmontanus, exposed to White Sturgeon Iridovirus (WSIV). At low temperatures (10ºC), cumulative mortality of fish is highest due to WSIV infection. At higher temperatures (23ºC), mortality is greatest due to the WSIV and to secondary infections from Flavobacterium spp. (Watson et al. 1998). Similarly, juvenile pallid sturgeon,

Scaphirhynchus albus, reared in warm water (23ºC) have greater ranavirus-induced cumulative mortality as compared to those held in cold water (17ºC; Stilwell et al. unpubl. data). Changes in temperature and salinity are merely a few of the factors affecting the prevalence of pathogens.

Environmental Effects on Crustacean Health

Environmental variables can suppress crustacean immune systems and increase the distribution of pathogens. Species experiencing thermal stress, for example, suffer reduced immunity and are at greater risk of infection (Martin et al. 2010). Temperature appears to have an indirect effect on epizootic shell disease (ESD) in the American lobster, Homarus americanus. At high temperatures (20ºC), lobsters experience shorter molt cycles and are thus able to shed their exoskeletons prior to lesion formation (Tlusty & Metzler 2012). At ambient temperatures (9-15ºC), Hematodinium sp., a species of parasitic dinoflagellate, yield greater infections in the blue crab, Callinectes sapidus. Intensity of blue crab Hematodinium infection decreases with temperature and salinity (Messick et al. 1999). More recently, Li et al. (2008) discovered Hematodinium sp. in the mud crab, Scylla serrata, cultured at hyposaline conditions

(< 9ºC).

Ocean acidification plays yet another role in propagating marine diseases. The increasing concentration of CO2 in the atmosphere and oceans has resulted in an increasing acidity of seawater. Calcium carbonate, the structural compound composing coral reefs, the shells of mollusks, and the exoskeletons of crustaceans, erodes when exposed to hyperacidity. If pathogens can tolerate or adapt to increasingly acidic environments, then the opportunity for infection of the immunocompromised host will increase as ocean acidification progresses.

Crustaceans must adapt to withstand changes in ocean acidity, as well as potential invasion by existing or novel pathogens. The Norway lobster, Nephrops norvegicus, has a wide thermal tolerance range, but it cannot survive low pH levels at its upper thermal limit. When crustacean

immunity is compromised, this species and others are at risk of infection by emerging pathogens

(Hernroth et al. 2012).

Crustacean Immune Response

Crustaceans have several lines of defense against pathogenic infection. Behavioral immunity is an effective means by which lobsters prevent infection. For example, the Caribbean spiny lobster, Panulirus argus, avoids cohabitation with conspecifics infected by Panulirus argus Virus 1 (PaV1) (Behringer et al. 2006, Anderson & Behringer 2010). This reduces the likelihood that healthy lobsters will contract, and thus spread, the virus through its various modes of transmission (i.e., waterborne, ingestion, and contact; Butler et al. 2008, Behringer & Butler

2010, Candia-Zulbarán et al. 2015). If infected, the crustacean immune response acts as the next level of defense. There are two types of immune response, innate and adaptive. Innate immunity describes the host’s nonspecific response to invasion by foreign microorganisms. The innate immune system is linked to the recognition of pathogen-associated molecular patterns (PAMPs).

Detection of these conserved structures triggers innate immunity, which consists of cellular and humoral elements (Janeway & Medzhitov 2002, Mogensen 2009, Vazquez 2009, Hauton 2012).

The cellular component of innate immunity includes phagocytic macrophages and granulocytes while the humoral, or non-cellular, component includes chemical and physical barriers (Basset et al. 2003). The American lobster has been used as a model for measuring phagocytic and antimicrobial activity in invertebrates (Mori & Stewart 2006, Brisbin et al. 2015). The red claw crayfish, Cherax quadricarinatus, and the Pacific white shrimp, Litopenaeus vannamei, have served as models for studying granulocytes and coagulation, respectively (Chaikeeratisak et al.

2014, Duan et al. 2014). Unlike innate immunity, adaptive immunity has antigenic properties and the ability to target specific agents. Until recently, adaptive immunity was believed to be absent in invertebrates.

However, there is increasing evidence that invertebrate immune systems recognize pathogens from past infections, a characteristic of adaptive immunity (Arala-Chaves & Sequeira

2000, Kurtz & Franz 2003, Agaisse 2007). Unlike adaptive immunity in vertebrates, this

“alternative adaptive immunity” identifies specific pathogens in invertebrates through the alternative splicing of exons (Kurtz & Armitage 2004, Chou et al. 2009). Alternative splicing has been observed in the copepod, Macrocyclops albidus, parasitized by tapeworms and in L. vannamei exposed to White Spot Syndrome Virus (WSSV) (Kurtz & Franz 2003, Chou et al.

2009). Behavior, innate, and alternative adaptive immunity are all ways through which crustaceans defend themselves against pathogens.

Pathogens in Crustaceans

Crustaceans are susceptible to a variety of pathogens. Over the past few decades, epizootic shell disease (ESD) has infected populations of the American lobster. Bacteria associated with ESD include Aquimarina sp. ‘homaria’, Thalassobius sp., and Candidatus

Kopriimonas aquarianus (Chistoserdov et al. 2012, Quinn et al. 2012). The defining symptom of

ESD is an irregular formation of lesions along the dorsal carapace (Smolowitz et al. 2005). In severe cases, lesions penetrate deep enough into the exoskeleton to expose the tissue beneath, leading to mortality of the host. The presence of A. ‘homaria’ and of similar patterns in lesion formation indicate that the European lobster, Homarus gammarus, are also at risk of ESD infection (Whitten et al. 2014). The effects of ESD have likely extended beyond aquatic health into the fishery since American lobster landings in southern New England have declined during the same period. While environmental conditions and pollutants are likely factors in reduced landings, ESD may also be at play (Castro et al. 2012).

Gaffkaemia (‘red tail disease’) is yet another disease that has deleterious effects on lobster fisheries. This fatal disease is caused by the bacterium, Aerococcus viridans var. homari,

and is found in the American lobster, Homarus americanus, and the European lobster. Heavily infected lobsters eventually develop lethargy and discolored abdomens (Snieszko & Taylor 1947,

Stewart & Arie 1973). Gaffkaemia is prevalent among lobsters held in captivity and, to a lesser extent, in wild populations of the European lobster (Alderman 1996, Stebbing et al. 2012).

Gaffkaemia, a disease endemic in American lobster populations, spread to European waters and lobsters through importation (Stewart et al. 1966, Alderman 1996, Lavallée et al. 2001). In a

Wales holding facility, > 75% of European lobsters were dead or moribund as a result of gaffkaemia (Stebbing et al. 2012). Even though gaffkaemia is one of the most common and lethal diseases associated with clawed lobsters, other bacterial diseases include impoundment shell disease (ISD), diet-induced and/or enzootic shell disease (EnSD), and black spot/trauma induced disease (TSD; Smolowitz et al. 2014).

Not all fatal diseases in crustaceans are bacterial or lesion-forming. WSSV, a pathogen notorious for devastating penaeid shrimp aquaculture farms, has the potential to cause mortality in lobster populations. In the laboratory, this non-specific virus has been used to induce white spot syndrome in the European lobster and in multiple species of spiny lobster (Chang et al.

1998, Syed Musthaq et al. 2006, Bateman et al. 2012). While the American lobster does not naturally host any known viruses, Clark et al. (2013) discovered that WSSV can replicate in

American lobsters.

Hematodinium spp. also exhibits low host specificity. These dinoflagellates have been reported in other crustaceans, including the blue crab, snow crab, Chionoecetes opilio, and

Norway lobster (Newman & Johnson 1975, Field et al. 1992, Taylor & Khan 1995, Shields &

Squyars 2000). Known to cause ‘bitter crab disease,’ Hematodinium sp. was discovered in the

Tanner crab, Chionoecetes bairdi, off of southeastern Alaska. The bitter aftertaste following

consumption of infected tissues was responsible for substantially reducing the market value of

Tanner crabs on several occasions (Meyers et al. 1987, 1990).

A species-specific pathogen, PaV1, infects the Caribbean spiny lobster. No other host has been identified for this lethal virus (Butler et al. 2008). Symptoms of diseased lobsters include lethargy, milky-white hemolymph, and discoloration of the exoskeleton (Shields & Behringer

2004). In Florida, 11% of lobsters caught in commercial traps tested positive for PaV1

(Behringer et al. 2012). However, prevalence of PaV1 in wild populations varies temporarily and geographically, ranging from 0 to 100% (Behringer et al. 2011, Moss et al. 2012). PaV1 variability exists within individual lobsters as well, establishing lethal disease in juveniles and subclinical infections in adults (Butler et al. 2008). Several factors likely give rise to these variations in prevalence and progression, such as the viability of PaV1.

The overall goal of the present study was to clarify the interactions between PaV1, its host, and the environment. To do so, we first designed and validated a real-time quantitative polymerase chain reaction (qPCR) assay. This molecular tool was used to determine the viability of PaV1, to describe the relationship between dosage of PaV1 and response by lobsters, and to identify the drivers of variability in PaV1 prevalence. Here, we present our findings to better understand the disease ecology of PaV1 in Caribbean spiny lobsters.

CHAPTER 2 VALIDATION OF A TAQMAN REAL-TIME QUANTITATIVE PCR ASSAY FOR THE DETECTION OF PANULIRUS ARGUS VIRUS 1

Introduction

Panulirus argus Virus 1 (PaV1) is a pathogenic virus infecting the Caribbean spiny lobster, Panulirus argus. Since its discovery in 2000, a great deal of work has been done to understand its epidemiology (Lozano-Álvarez et al. 2008, Butler et al. 2009, Moss et al. 2013), ecology (Behringer et al. 2006, Anderson & Behringer 2011), and fishery impacts (Behringer et al. 2012, Huchin-Mian et al. 2013). However, PaV1 remains an unclassified DNA virus given that only two small fragments of the genome have been reported (Li et al. 2006, Montgomery-

Fullerton et al. 2007). While the PaV1 genome has not yet been published, this work is in progress (Waltzek et al. unpubl. data).

PaV1-infected lobsters display behavioral abnormalities such as occasional tremors and lethargy that may manifest as individuals unable to right themselves (Shields & Behringer 2004).

Infected lobsters may also display gross lesions including the appearance of milky hemolymph or discoloration of the carapace (Shields & Behringer 2004). Healthy individuals avoid PaV1- infected conspecifics, causing an increase in the number of dens with solitary occupants

(Behringer et al. 2006). Diseased animals display microscopic lesions with infected hemocytes

(e.g., hyalinocytes and semi-granulocytes) and spongy connective tissue (Shields & Behringer

2004, Li et al. 2006) displaying karyomegaly, margination of condensed chromatin, and faint eosinophilic nuclear inclusions (Shields & Behringer 2004). Observation of cytopathic effect within hemocytes has been used to quantify infectious virus using a 50% tissue culture infectious dose assay (TCID50) (Li & Shields 2007). Ultrastructural examination of infected cells reveals nuclear arrays of naked polygonal nucleocapsids (~180 nm in diameter) surrounding a spherical electron dense DNA core (~118 nm in diameter) (Shields & Behringer 2004). Development of a

specific PaV1 fluorescence in situ hybridization assay paired with histopathology and transmission electron microscopy revealed that the virus establishes a viremia with infected cells observed in and around the hepatopancreas, hindgut, foregut, gill, heart, skin, nervous, and ovarian tissues (Li et al. 2006).

Diagnosis of PaV1-infection has historically relied upon the aforementioned gross, microscopic, and ultrastructural pathology. In more recent years, the validation of a specific and sensitive endpoint PCR assay targeting a region of the PaV1 genome that encodes a hypothetical protein has become an increasingly important diagnostic method (Montgomery-Fullerton et al.

2007, Huchin-Mian et al. 2009, Moss et al. 2012, 2013). Here we report the development and partial validation of a PaV1 TaqMan real-time quantitative PCR assay to be used as an expedient and cost-effective molecular diagnostic tool while also providing biologically useful quantification of PaV1 load.

Materials and Methods

In silico TaqMan qPCR Primer and Probe Design

Sequences for 61 PaV1 alleles were downloaded from GenBank. These sequences, which encode for a hypothetical protein, were discovered in spiny lobsters from 8 regions in the

Caribbean Sea (Moss et al. 2013). All of the sequences were first aligned in MAFFT (Katoh &

Toh 2008). A consensus sequence was then generated in BioEdit 7.2.6.1 (Hall 1999) and imported into PrimerExpress v2.0 (Applied Biosystems), which was used to design PaV1 primers and hydrolysis probes (Table 2-1). The default settings were used for each software program.

Detection of PaV1 DNA Using the qPCR Assay

Reaction volumes were 20 μL and consisted of: 0.36 μM of each primer, 0.1 μM of probe, 4 μL of nucleic acid template (≤ 100 ng of total DNA), 10 μL of universal qPCR mix

(TaqMan® Fast Universal PCR Master Mix 2X, Applied Biosystems), and 3 μl of molecular grade water. Samples and standards were run in triplicate on 96-well polypropylene plates

(Olympus Plastics, Genesee Scientific) and sealed with a 50 µm polyolefin film (ThermalSeal

RTS, Excel Scientific). Each plate contained triplicate 10-fold dilution series of linearized plasmid DNA (standard) carrying between 107-100 copies of the PaV1 hypothetical protein sequence targeted by the qPCR. Eukaryotic 18s rRNA endogenous control assays (Applied

Biosystems Assay ID Hs99999901_s1) were run singly alongside each sample. Samples, standards, reagents, and water were loaded and mixed onto the 96-well plates using a QIAgility instrument (Qiagen). Plates were run on a 7500 Fast Real-Time PCR System (Applied

Biosystems) under the following thermocycling conditions: 95°C for 20 s, followed by 40 cycles at 95°C for 3 s, and 60°C for 30 s. To increase signal-to-noise ratio and specificity of the probe, a

MGB quencher was added to the probe. A threshold cycle (Ct) was calculated and interpreted as a positive result for samples if the ROX (passive reference dye) normalized FAM signal exceeded the threshold assigned by the Applied Biosystems software.

Estimation of the qPCR Assay Slope, Y-Intercept, Correlation Coefficient (R2), Efficiency, Dynamic Range, Analytical Sensitivity, Repeatability, Reproducibility, and Analytical Specificity

The following parameters were estimated using serially-diluted plasmid DNA standards from 21 experiments (plates): slope, y-intercept, correlation coefficient (R2), efficiency, dynamic range, analytical sensitivity, repeatability, and reproducibility. The efficiency of each experiment was calculated as 10-1/slope -1 (Bustin et al. 2009) by the Applied Biosystems software. To measure the performance of the qPCR assay, the inter-assay variability (reproducibility) and intra-assay variability (repeatability) were determined. For both metrics, the percent coefficient of variation (CV% = (standard deviation/mean) x 100%)) was calculated using the standard

deviation of each Ct within (repeatability) or among (reproducibility) the 21 experiments

(plates).

The ability of the qPCR to detect only PaV1 (analytical specificity) was assessed by testing a set of 116 positive PaV1 samples (described below under the diagnostic sensitivity and specificity section) as well as a collection of other double-stranded DNA viruses described below. DNA from tissues infected with iridoviruses from the following genera were tested:

Lymphocystivirus (Lymphocystis disease virus, LCDV), Megalocytivirus (Infectious spleen and kidney necrosis virus, ISKNV), and Ranavirus (Frog virus 3, FV3) (Table 2-2). The LCDV sample was derived from a copperband butterflyfish, Chelmon rostratus, displaying gross proliferative fin lesions characteristic of the disease. The presence of LCDV DNA within the sample was confirmed by PCR and Sanger sequencing of the partial viral DNA polymerase gene as described by Hanson et al. (2006). Descriptions of the ISKNV and FV3-infected DNA samples have been previously described (Subramaniam et al. 2016, Tan et al. 2004, respectively). The qPCR assay was also tested against a nimavirus (White spot syndrome virus,

WSSV) and an uncharacterized DNA virus in European green crab, Carcinus maenas, related to

PaV1 (Table 2-2). DNA was extracted from a penaid shrimp containing white lesions typical of

WSSV infections. The presence of WSSV DNA within the sample was confirmed using nested

PCR and Sanger sequencing as previously described (Liang et al. 2011). The HLV-infected tissue DNA was kindly provided by Dr. Grant Stentiford.

Estimation of the qPCR Assay Diagnostic Sensitivity and Specificity

In total, 165 hemolymph or leg tissue samples were acquired from juvenile (< 76 mm carapace length, CL) or adult ( 76 mm CL) Caribbean spiny lobsters collected throughout the

Caribbean Sea in 2008, 2009, or 2015 (Table 2-2). Insulin syringes (27 G x 5/8 1cc) were used to

draw 0.1-0.2 mL of hemolymph from the proximal joint of the fifth periopod. Alternatively, leg tissue samples were taken when hemolymph collection was not logistically possible (e.g., leg sinus too small for phlebotomy) or syringes were not available (e.g., some locations in the

Caribbean). Each sample was then transferred to a 1.5 mL microcentrifuge tube containing 0.9 mL of 95% EtOH. Samples were stored at -20C until processed at the University of Florida

Aquatic Pathobiology Laboratory (APL; Gainesville, FL, USA).

At APL, DNA was extracted from hemolymph or leg samples using either a DNeasy

Blood and Tissue Kit (Qiagen) or Quick-gDNA MiniPrep Kit (Zymo Research) using the manufacturers’ protocols except as noted below. Prior to adding genomic lysis buffer (Zymo

Research), hemolymph samples were centrifuged at room temperature (14,950 x g for 1 min).

Without disturbing the pelleted hemolymph, the 95% EtOH was decanted and discarded.

Samples were dried at room temperature to evaporate any residual EtOH from the pellet. After 1 h, 25 µL of hemolymph were transferred to a clean microcentrifuge tube. For leg tissue samples, the 95% EtOH was poured off and the leg was macerated using a surgical blade. As stated by the manufacturer (Zymo Research), 200 µL of lysis buffer was then added to each sample

(hemolymph or leg), pulse vortexed, and briefly spun. Pestles were used to manually homogenize the leg tissue samples. All samples remained at room temperature for 20 min before being vortexed and spun again. A NanoDrop 2000 Spectrophotometer (ThermoScientificTM) was used to quantify the DNA concentration and purity of each sample prior to storage at -20C.

Samples were subsequently determined to be positive or negative for PaV1 using endpoint PCR

(Montgomery-Fullerton et al. 2007). In total, 116 known PaV1-positive and 49 PaV1-negative samples were used to estimate the diagnostic sensitivity and specificity of the qPCR assay (Table

2-2).

Results

In silico TaqMan qPCR Primer and Probe Design

PrimerExpress v2.0 (Applied Biosystems) delivered a primer/probe set amplifying a 69 bp region of the PaV1 genome (Table 2-1). Examination of the in silico specificity of the PaV1 probe (PaV1P) displayed a single mismatch at position 37 for alleles 33, 34, and 40 (Figure 2-1).

The reverse primer (PaV1R) displayed a single mismatch at position 53 for allele 43.

Estimation of the qPCR Assay Slope, Y-Intercept, Correlation Coefficient (R2), Efficiency, Dynamic Range, Analytical Sensitivity, Repeatability, Reproducibility, and Analytical Specificity

Based on 21 experiments (plates), the mean values (± SEM) calculated for each of the parameters were as follows: slope = -3.35 ± 0.03, Y-intercept = 38.25 ± 0.23, R2 = 0.989 ±

0.0006, and efficiency = 99.19% ± 1.02 (Figure 2-3). Given amplification of the 100 standard occurred in 65.1% (41/63) of the wells, the limit of detection of the assay (analytical sensitivity) was 101 plasmid copies of PaV1 DNA (testing positive in 100% of the wells) (Table 2-3).

Therefore, the dynamic range of the qPCR assay was 107-101 copies of plasmid copies of PaV1

DNA (Figures 2-2, 2-3). The coefficient of variation of this assay ranged from 0.3 – 0.6% for the inter-assay variability (reproducibility) and from 0.05 – 1.75% for the intra-assay variability

(repeatability) (Table 2-3). The qPCR assay reproducibility and repeatability were within acceptable limits (< 5%) (Marancik & Wiens 2013).

The specificity of the in silico primer and probe design were confirmed by the qPCR assay testing (Figure 2-1). The qPCR assay amplified all 118 PaV1 samples previously determined to be positive by endpoint PCR. The following viruses all tested negative using the qPCR assay: ISKNV, LCDV, FV3, WSSV, and HLV. The 18s rRNA internal controls were positive for all samples.

Estimation of the qPCR Assay Diagnostic Sensitivity and Specificity

Testing of 116 known PaV1-positive samples and 49 PaV1-negative samples indicated a diagnostic sensitivity of 100% (116/116) and diagnostic specificity of 84% (41/49) when compared to the endpoint PCR as the gold standard test. Of the 49 samples negative by endpoint

PCR, eight tested positive by the qPCR assay (Table 2-2).

Discussion

In this study, we designed and partially validated a TaqMan real-time quantitative PCR assay for the detection of Panulirus argus Virus 1 (PaV1) in the Caribbean spiny lobster. The in silico analysis revealed very few mismatches when the primers and probe were mapped against the sequences representing the 61 PaV1 alleles recognized within the Caribbean Sea (Moss et al.

2013). The initial assay validation revealed a high correlation coefficient, efficiency, sensitivity, specificity, repeatability, and reproducibility. The qPCR assay always detected 10 plasmid copies of PaV1 DNA and did not amplify other double-stranded DNA viruses indicating its high analytical sensitivity and specificity. It performed well against known PaV1 positive and negative lobster tissue samples indicating its high diagnostic sensitivity (100%) and specificity

(85%). The data presented herein meet the analytical and diagnostic performance criteria outlined by the World Organization for Animal Health (OIE 2016). Future validation efforts are needed to ensure the qPCR assay is reproducible and can be implemented in other laboratories performing PaV1 diagnostics.

The qPCR assay offers several advantages over the currently available endpoint PCR assay including: 1) providing an expedient result without time-consuming agarose gel electrophoresis, 2) providing biologically useful information such as sample viral load, and 3) increased sensitivity and specificity. Of the 48 samples that were confirmed negative for PaV1 by endpoint PCR, eight were positive when tested against the qPCR assay. These seven samples

may represent false positive results reported by the qPCR assay that could be the result of nonspecific binding of the primers and probe to an unknown target. However, the mean Ct for these eight samples was 37 ± 0.24 (SEM), which corresponds to <10 viral copies / 4 µL of extracted lobster tissue DNA. Therefore, it is possible that these samples were below the limit of detection for the endpoint PCR used as the gold standard test. This is supported by the fact that experiments to detect 100 plasmid copies of PaV1 DNA were positive in 16.7% of lanes (1/6 lanes) using the endpoint PCR assay as compared to 65.1% of wells (41/63) using the qPCR assay (data not shown). Although preliminary, these data suggest the qPCR assay has a lower limit of detection relative to the endpoint PCR assay. Thus, the diagnostic specificity of the qPCR assay may actually be higher than the reported 84%.

In balance, the TaqMan real-time quantitative PCR requires substantial capital investment in equipment as compared to the currently available endpoint PCR assay.

Furthermore, Sanger sequencing of amplicons generated by endpoint PCR has been used to elucidate PaV1 epidemiology (Moss et al. 2013). Thus, the partially validated TaqMan qPCR assay presented herein and the aforementioned endpoint PCR assay are both useful diagnostic and research tools.

The presented PaV1 qPCR assay is an expedient and cost-effective tool making it useful for researchers and diagnosticians alike. Future research applications of the qPCR assay include determining viral load in tissues over time to better understand PaV1 tissue tropism over the course of an infection as has been studied in giant tiger prawn, Penaeus monodon, infected with white spot syndrome virus (Jeswin et al. 2015). Furthermore, the qPCR assay could be adapted to detect viral DNA in seawater samples for environmental monitoring. Finally, the qPCR assay, in

conjunction with other available diagnostic modalities, could be used by resource agencies to monitor PaV1 disease in populations of Caribbean spiny lobster.

Table 2-1. Primers and probe used in the development of the PaV1 TaqMan qPCR assay. Primers (PaV1F, PaV1R) and probe (PaV1P) bind within a PaV1 hypothetical protein gene sequence (GenBank accession# EF206313.1). The endpoint PCR product using primers 45aF and 543aR (Montgomery-Fullerton et al. 2007) was cloned into a plasmid to generate the linearized plasmid standards.

Amplicon Melting Position size (nt) Primers/probe temperature in gene including name Primer/probe sequence (°C) (5’-3’) primers 45aF TTCCAGCCCAGGTACGTATC 62.3 45 - 64 499 543aR AACAGATTTTCCAGCAGCGT 58.4 524 - 543

PaV1F CGTTGTACGGAATCGTTATTAAAGC 61.3 256 - 280 69 PaV1R GACACGACCAATTGAAGAAAAACTAC 61.4 299 - 324 PaV1P 6FAM-CCCGTGATGCTTGC-MGB/NFQ 52.9 284 - 297

Table 2-2. Samples used to validate the PaV1 TaqMan® real-time qPCR assay. All PaV1 samples were first tested by conventional PCR (“endpoint PCR result”) to determine if they were PaV1 positive (+) or negative (-). Every sample was then tested using the PaV1 qPCR (“qPCR result”). Additional double-stranded DNA viruses listed below were tested by qPCR to determine the analytical specificity of the assay. NA = not applicable. Endpoint qPCR PCR Sample ID Virus Collection site result result Reference PA15009 PaV1 Florida Keys, USA + + This study PA15010 PaV1 Florida Keys, USA + + This study PA15008 PaV1 Florida Keys, USA + + This study PA15007 PaV1 Florida Keys, USA + + This study PA15015 PaV1 Florida Keys, USA + - This study PA15016 PaV1 Florida Keys, USA + + This study PA15017 PaV1 Florida Keys, USA + + This study PA15018 PaV1 Florida Keys, USA + + This study PA15019 PaV1 Florida Keys, USA + + This study PA15020 PaV1 Florida Keys, USA + + This study PA15021 PaV1 Florida Keys, USA + + This study PA15022 PaV1 Florida Keys, USA + + This study PA15023 PaV1 Florida Keys, USA + + This study PA15024 PaV1 Florida Keys, USA + + This study PA15025 PaV1 Florida Keys, USA + + This study PA15026 PaV1 Florida Keys, USA + + This study PA15028 PaV1 Florida Keys, USA + + This study PA15076 PaV1 Florida Keys, USA + + This study PA15079 PaV1 Florida Keys, USA + + This study PA15078 PaV1 Florida Keys, USA + + This study PA15057 PaV1 Florida Keys, USA + + This study PA15064 PaV1 Florida Keys, USA + + This study PA15065 PaV1 Florida Keys, USA + + This study PA15066 PaV1 Florida Keys, USA + + This study PA15068 PaV1 Florida Keys, USA + + This study PA15069 PaV1 Florida Keys, USA + + This study PA15070 PaV1 Florida Keys, USA + + This study PA15071 PaV1 Florida Keys, USA + + This study PA15072 PaV1 Florida Keys, USA + + This study PA15073 PaV1 Florida Keys, USA + + This study PA15074 PaV1 Florida Keys, USA + + This study PA15075 PaV1 Florida Keys, USA + + This study

Table 2-2. Continued Endpoint qPCR PCR Sample ID Virus Collection site result result Reference PA15077 PaV1 Florida Keys, USA + + This study PA15080 PaV1 Florida Keys, USA + + This study PA15081 PaV1 Florida Keys, USA + + This study PA15082 PaV1 Florida Keys, USA + + This study PA15083 PaV1 Florida Keys, USA + + This study PA15084 PaV1 Florida Keys, USA + + This study PA15085 PaV1 Florida Keys, USA + + This study PA15086 PaV1 Florida Keys, USA + + This study PA15088 PaV1 Florida Keys, USA + + This study PA15089 PaV1 Florida Keys, USA + + This study PA15090 PaV1 Florida Keys, USA + + This study PA15091 PaV1 Florida Keys, USA + + This study PA15092 PaV1 Florida Keys, USA + + This study PA15093 PaV1 Florida Keys, USA + + This study PA15094 PaV1 Florida Keys, USA + + This study PA15095 PaV1 Florida Keys, USA + + This study PA15096 PaV1 Florida Keys, USA - - This study PA15097 PaV1 Florida Keys, USA - - This study PA15098 PaV1 Florida Keys, USA + + This study PA15099 PaV1 Florida Keys, USA + + This study PA15100 PaV1 Florida Keys, USA + + This study PA15101 PaV1 Florida Keys, USA + + This study PA15102 PaV1 Florida Keys, USA + + This study PA15005 PaV1 Florida Keys, USA + + This study PA15006 PaV1 Florida Keys, USA + + This study PA15001 PaV1 Florida Keys, USA + + This study PA15002 PaV1 Florida Keys, USA + + This study PA15003 PaV1 Florida Keys, USA + + This study PA15004 PaV1 Florida Keys, USA + + This study PA15103 PaV1 Florida Keys, USA + + This study PA15104 PaV1 Florida Keys, USA + + This study PA15105 PaV1 Florida Keys, USA + + This study PA15106 PaV1 Florida Keys, USA + + This study PA15107 PaV1 Florida Keys, USA + + This study PA15108 PaV1 Florida Keys, USA + + This study PA15109 PaV1 Florida Keys, USA + + This study

Table 2-2. Continued Endpoint qPCR PCR Sample ID Virus Collection site result result Reference PA15110 PaV1 Florida Keys, USA + + This study PA15111 PaV1 Florida Keys, USA + + This study PA15112 PaV1 Florida Keys, USA + + This study PA15113 PaV1 Florida Keys, USA + + This study PA15114 PaV1 Florida Keys, USA + + This study PA15115 PaV1 Florida Keys, USA + + This study PA15116 PaV1 Florida Keys, USA + + This study PA15117 PaV1 Florida Keys, USA + + This study PA15118 PaV1 Florida Keys, USA + + This study PA15119 PaV1 Florida Keys, USA + + This study PA15120 PaV1 Florida Keys, USA + + This study PA15121 PaV1 Florida Keys, USA + + This study PA15122 PaV1 Florida Keys, USA + + This study PA15123 PaV1 Florida Keys, USA + + This study PA15124 PaV1 Florida Keys, USA + + This study PA15125 PaV1 Florida Keys, USA + + This study PA15126 PaV1 Florida Keys, USA + + This study PA15127 PaV1 Florida Keys, USA + + This study PA15128 PaV1 Florida Keys, USA + + This study PA15129 PaV1 Florida Keys, USA + + This study PA15130 PaV1 Florida Keys, USA + + This study PA15131 PaV1 Florida Keys, USA + + This study PA15132 PaV1 Florida Keys, USA + + This study PA15139 PaV1 Florida Keys, USA + + This study PA15140 PaV1 Dominican Republic + + Moss et al. 2013 PA15157 PaV1 Dominican Republic + + Moss et al. 2013 PA15173 PaV1 Dominican Republic + + Moss et al. 2013 PA15142 PaV1 Dominican Republic + + Moss et al. 2013 PA15153 PaV1 Dominican Republic + + Moss et al. 2013 PA15159 PaV1 Dominican Republic + + Moss et al. 2013 PA15141 PaV1 Puerto Rico + + Moss et al. 2013 PA15147 PaV1 Puerto Rico + + Moss et al. 2013 PA15155 PaV1 Puerto Rico + + Moss et al. 2013 PA15163 PaV1 Puerto Rico + + Moss et al. 2013 PA15164 PaV1 Puerto Rico + + Moss et al. 2013 PA15165 PaV1 Puerto Rico + + Moss et al. 2013

Table 2-2. Continued Endpoint qPCR PCR Sample ID Virus Collection site result result Reference PA15168 PaV1 Puerto Rico + + Moss et al. 2013 PA15170 PaV1 Puerto Rico + + Moss et al. 2013 PA15144 PaV1 Mexico + + Moss et al. 2013 PA15151 PaV1 Mexico + + Moss et al. 2013 PA15160 PaV1 Mexico + + Moss et al. 2013 PA15145 PaV1 Belize + + Moss et al. 2013 PA15149 PaV1 Belize + + Moss et al. 2013 PA15166 PaV1 Belize + + Moss et al. 2013 PA15172 PaV1 Belize + + Moss et al. 2013 PA15146 PaV1 Panama + + Moss et al. 2013 PA15148 PaV1 Cuba + + Moss et al. 2013 PA15154 PaV1 Bahamas + + Moss et al. 2013 PA15156 PaV1 Bahamas + + Moss et al. 2013 PA15161 PaV1 Bahamas + + Moss et al. 2013 PA15171 PaV1 Bahamas + + Moss et al. 2013 PA15175 PaV1 Florida Keys, USA - - This study PA15176 PaV1 Florida Keys, USA - - This study PA15177 PaV1 Florida Keys, USA - - This study PA15178 PaV1 Florida Keys, USA + - This study PA15179 PaV1 Florida Keys, USA - - This study PA15180 PaV1 Florida Keys, USA - - This study PA15181 PaV1 Florida Keys, USA + - This study PA15182 PaV1 Florida Keys, USA - - This study PA15183 PaV1 Florida Keys, USA + - This study PA15184 PaV1 Florida Keys, USA - - This study PA15185 PaV1 Florida Keys, USA + - This study PA15186 PaV1 Florida Keys, USA - - This study PA15187 PaV1 Florida Keys, USA - - This study PA15188 PaV1 Florida Keys, USA - - This study PA15189 PaV1 Florida Keys, USA - - This study PA15190 PaV1 Florida Keys, USA - - This study PA15191 PaV1 Florida Keys, USA - - This study PA15192 PaV1 Florida Keys, USA - - This study PA15193 PaV1 Florida Keys, USA - - This study PA15194 PaV1 Florida Keys, USA - - This study PA15195 PaV1 Florida Keys, USA - - This study

Table 2-2. Continued Endpoint qPCR PCR Sample ID Virus Collection site result result Reference PA15196 PaV1 Florida Keys, USA - - This study PA15197 PaV1 Florida Keys, USA - - This study PA15198 PaV1 Florida Keys, USA - - This study PA15199 PaV1 Florida Keys, USA - - This study PA15200 PaV1 Florida Keys, USA - - This study PA15201 PaV1 Florida Keys, USA + - This study PA15202 PaV1 Florida Keys, USA - - This study PA15203 PaV1 Florida Keys, USA - - This study PA15204 PaV1 Florida Keys, USA - - This study PA15205 PaV1 Florida Keys, USA - - This study PA15206 PaV1 Florida Keys, USA - - This study PA15207 PaV1 Florida Keys, USA - - This study PA15208 PaV1 Florida Keys, USA + - This study PA15209 PaV1 Florida Keys, USA - - This study PA15210 PaV1 Florida Keys, USA - - This study PA15211 PaV1 Florida Keys, USA - - This study PA15212 PaV1 Florida Keys, USA - - This study PA15213 PaV1 Florida Keys, USA - - This study PA15214 PaV1 Florida Keys, USA - - This study PA15215 PaV1 Florida Keys, USA + - This study PA15216 PaV1 Florida Keys, USA - - This study PA15217 PaV1 Florida Keys, USA - - This study PA15218 PaV1 Florida Keys, USA - - This study PA15219 PaV1 Florida Keys, USA - - This study PA15220 PaV1 Florida Keys, USA - - This study WVL17269 WSSV Arizona, USA* - NA Lightner unpubl. data WVL17268 HLV United Kingdom - NA Stentiford unpubl. data TL14009 ISKNV Midwest, USA - NA Subramaniam et al. 2016 PSRV2009 FV3 Missouri, USA - NA Waltzek et al. 2014 FOF15001 LCDV Florida, USA - NA Waltzek unpubl. data * indicates that specimens were experimentally infected at this location

Table 2-3. Inter-assay variability (reproducibility) and intra-assay variability (repeatability) of the TaqMan® real-time qPCR assay for PaV1. A total of 21 experiments (plates) were performed on multiple days to measure reproducibility of the assay. Every plasmid standard (107-100) was run in triplicate. The repeatability of the assay was determined based on the results of the three wells for each plasmid dilution from a single experiment (plate). The table layout was adapted from Marancik and Wiens (2013). Inter-assay variability Standard Mean Standard Coefficient of Number of wells dilutions Ct deviation variation (%) positive (/63) 107 15.3 0.04 0.25 63 106 18.0 0.06 0.32 63 105 21.2 0.07 0.33 63 104 24.7 0.08 0.32 63 103 28.0 0.09 0.31 63 102 31.4 0.14 0.44 63 101 35.0 0.13 0.36 63 100 38.2 0.22 0.58 41 Intra-assay variability Standard Mean Standard Coefficient of Number of wells dilutions Ct deviation variation (%) positive (/3) 107 15.7 0.05 0.32 3 106 18.4 0.06 0.32 3 105 21.6 0.05 0.25 3 104 25.0 0.01 0.05 3 103 28.2 0.05 0.19 3 102 31.6 0.22 0.7 3 101 35.0 0.54 1.53 3 100 38.5 0.67 1.75 2

10 20 30 40 50 60 ....|....|....|....|....|....|....|....|....|....|....|....|....|.... PaV1F PaV1P PaV1R JN786984.1 A33 CGTTGTACGGAATCGTTATTAAAGCTATCCCGTGATGCTTGCAGTAGTTTTTCTTCAATTGGTCGTGTC JN786983.1 A32 ...... G...... JX987106.1 A36 ...... JX987129.1 A60 ...... JX987118.1 A48 ...... JX987122.1 A52 ...... JN786963.1 A12 ...... JN786961.1 A10 ...... JN786959.1 A8 ...... JX987105.1 A35 ...... JX987116.1 A46 ...... JX987121.1 A51 ...... JN786982.1 A31 ...... JX987110.1 A40 ...... A...... JN786979.1 A28 ...... JN786973.1 A22 ...... JN786956.1 A5 ...... JN786957.1 A6 ...... JN786954.1 A3 ...... JX987117.1 A47 ...... JN786958.1 A7 ...... JN786980.1 A29 ...... JN786960.1 A9 ...... JN786964.1 A13 ...... C...... JN786978.1 A27 ...... JN786977.1 A26 ...... JX987128.1 A29 ...... JX987115.1 A45 ...... JX987109.1 A39 ...... JX987114.1 A44 ...... JN786976.1 A25 ...... JN786975.1 A24 ...... JN786974.1 A23 ...... JN786972.1 A21 ...... JN786968.1 A17 ...... JN786965.1 A14 ...... JN786969.1 A18 ...... JX987107.1 A37 ...... JX987108.1 A38 ...... JN786970.1 A19 ...... JN786966.1 A15 ...... JX987120.1 A50 ...... JX987112.1 A42 ...... JX987111.1 A41 ...... JX987119.1 A49 ...... JN786967.1 A16 ...... JN786952.1 A1 ...... JX987123.1 A53 ...... JN786953.1 A2 ...... JN786962.1 A11 ...... JN786955.1 A4 ...... JX987126.1 A57 ...... JX987125.1 A55 ...... JX987124.1 A54 ...... JX987130.1 A61 ...... JX987127.1 A58 ...... JN786981.1 A30 ...... JN786971.1 A20 ...... JX987104.1 A34 ...... T...... JX987103.1 A33 ...... T...... JX987113.1 A43 ...... T......

Figure 2-1. Sequences from 61 PaV1 alleles used for in silico primer and probe design. All sequences were aligned to identify a conserved region (69 bp) of a hypothetical protein gene. The forward primer (PaV1F), probe (PaV1P), and reverse primer (PaV1R) are highlighted in yellow. Positions 1-69 above correspond to positions 256- 324 in a PaV1 hypothetical protein gene sequence (GenBank accession # EF206313.1).

Figure 2-2. Assay amplification plot of triplicate 10-fold serial dilutions carrying between 107- 100 PaV1 plasmid DNA copies. The y-axis shows the normalized receptor signal (∆Rn) and the x-axis shows the threshold cycle (Ct) of each standard dilution. The blue line (0.5) represents the default amplification threshold provided by the Applied Biosystems software.

Figure 2-3. Standard curve generated using triplicate 10-fold serial dilutions carrying between 107-100 PaV1 plasmid DNA copies. The mean qPCR assay parameters (± SEM) averaged over the 21 experiments (plates) are provided in the figure legend. The X- axis shows the log plasmid standard copy number and the Y-axis indicates the corresponding cycle threshold (Ct) value.

CHAPTER 3 VIABILITY OF PANULIRUS ARGUS VIRUS 1 IN SEAWATER

Introduction

Pathogens infecting marine organisms are exposed to a mix of favorable and unfavorable conditions. Marine environments are generally considered open and, therefore, promote high connectivity (McCallum et al. 2003). Similar to the larval and genetic exchange documented in several species of decapod (Albrecht et al. 2014, Rodríguez-Rey et al. 2014), pathogens can be transported by ocean currents across long distances. An iconic example of this phenomenon is the 1980s die-off of the long-spine sea urchin, Diadema antillarum. The pathogen responsible for this marine epizootic decimated sea urchin populations throughout the

Caribbean Sea, spanning from Panama to Florida (USA). The epidemic spread at a rate of 4990 km year -1 (Lessios et al. 1984, Lessios 1988), which introduces another characteristic of the marine environment - in the ocean, pathogens can spread at rapid rates. McCallum et al. (2003) hypothesized that the currents moving along coastlines are particularly apt to be responsible for this type of pathogen transport. between locations. In contrast to terrestrial environments, vectors do not appear to play a role in accelerating the rate of epidemics. In the terrestrial environment, flying vectors are responsible for causing and maintaining epidemics such as calicivirus in wild rabbit populations (Cooke & Fenner 2002). Only a handful of non-flying, marine vectors have been identified to date, including the marine fireworm, Hermodice carunculata, and microsporidian hyperparasites of the new genus and species, Hyperspora aquatica n. gn., n.sp

(Sussman et al. 2003, Stentiford et al. 2017). H. carnuculata is a vector for the coral-bleaching bacteria, Vibrio shiloi (Sussman et al. 2003), while H. aquatica is a proposed vector for the paramyxid Marteilia cochillia in common cockles, Cerastoderma edule (Stentiford et al. 2017).

Given that vectors are uncommon in the marine environment, understanding the mechanisms of

transport, the oceanic connectivity of susceptible host populations, and the viability of free- floating pathogens is extremely important. Changes in environmental conditions affect the viability of viruses in the ocean. A study conducted by Takahashi et al. (2016) revealed that the enteric, non-enveloped norovirus (NoV) remains infective in seawater for up to 5 d in the summer when temperatures are higher. However, NoV persists for up to 7 d during the winter, illustrating that changes in the marine environment alter the viability of viruses. The viability for several marine viruses remains unclear, as is the case with Panulirus argus Virus 1 (PaV1).

PaV1 is a pathogenic virus which infects most life history stages of the Caribbean spiny lobster, Panulirus argus. The virus is host specific and exhibits strong genetic connectivity across much of its Caribbean range (Butler et al. 2008, Moss et al. 2013). More alleles are shared in the north and northeastern regions of the Caribbean than are in the southwestern-most region.

While empirical studies have ruled out vertical transmission to larvae as the connectivity vector

(Behringer & Butler unpubl. data), theoretical modelling work has shown that postlarvae could be vectors of PaV1. Alternatively, models have also shown that the passive transport of PaV1 virions could be a viable means of connecting populations and could explain the observed distribution of PaV1 in the Caribbean (Kough et al. 2015). The dispersal potential and connectivity of PaV1 by waterborne-transmission depends on many factors, but in particular the viability of the virus outside of a host in seawater.

Kough et al. (2015) used a 5 d pelagic period, as estimated ex situ, when modelling the movement patterns of PaV1 in seawater (Butler & Behringer unpubl. data). This viability estimate is consistent with those of other viruses found in the marine environment (Kough et al.

2015), such as NoV, which remains infectious in seawater for a minimum of 5 d, depending on seasonality and salinity (Takahashi et al. 2016). White Spot Syndrome Virus (WSSV), which

have an envelope, is viable in seawater for 12 d (Satheesh Kumar et al. 2013). However, typically non-enveloped viruses, such as PaV1, are capable of persisting longer outside of their hosts than are enveloped viruses (Howie et al. 2008). For example, enveloped viruses, such as influenza A virus (H1N1) and herpes simplex virus type 1 (HSV-1), remain infectious for fewer than 5 d. The non-enveloped viruses, minute virus of mice (MVM) and coxsackievirus B4

(CVB4), can persist for several weeks (Firquet et al. 2015). These differences in viral viabilities exist because enveloped viruses are structurally less stable than non-enveloped viruses. The lipid bilayers of viral envelopes contain glycoproteins which are important in host-cell recognition, but are sensitive to variations in environmental conditions. Once its envelope is damaged (e.g., by sunlight, changes in salinity, etc.), the virus is inactivated and is no longer infectious to the host (Flannery et al. 2013, Takahashi et al. 2016).

The objective of this study was to experimentally determine the viability of PaV1 virions in seawater, which will confirm or permit refinement of the passive particle model used to describe the Caribbean connectivity of PaV1 (Kough et al. 2015). Given that PaV1 is a non- enveloped virus and likely persists in water for longer than enveloped viruses (Shields &

Behringer 2004), it was hypothesized that PaV1 would remain viable for at least 7 d.

Materials and Methods

Lobster Screening

In spring and fall 2016, postlarval lobsters were collected to the west of Long Key, Florida

Keys, USA. Lobsters were obtained approximately seven days following the new moon, using artificial Witham-style collectors (Butler & Herrnkind 1991). Lobsters were transported to and held individually at the University of Florida Fisheries and Aquatic Sciences wet facility

(Gainesville, FL, USA). Following metamorphosis into the early benthic juvenile (EBJ) stage, the fifth periopod was removed from each animal and preserved in 0.9 mL of 95% EtOH for later

molecular analyses at the University of Florida Aquatic Pathobiology Laboratory (Gainesville,

FL). DNA was extracted from each leg tissue sample using Quick-gDNA MiniPrep Kits (Zymo

Research) and techniques described in Chapter 2. Each leg sample was manually homogenized in

200 µL of lysis buffer using an RNase-free disposable pellet pestle. The extracted material was then tested for PaV1 using the qPCR assay described in Chapter 2. Any lobsters which tested positive for PaV1 were not used in the experiment.

Viral Purification and Quantitation

For each experiment, hemolymph was harvested from an infected lobster caught in the

Florida Keys and displaying clinical signs of infection (i.e., moribund and with milk-white hemolymph). Hemolymph was collected from the moribund lobster and stored immediately in a

-20°C freezer. Prior to purifying PaV1, the hemolymph was thoroughly thawed on ice. The thawed hemolymph was placed in an Allegra X-14R Refrigerated Benchtop Centrifuge

(Beckman Coulter, Brea, CA, USA) and spun at 3000 x g for 10 minutes (4°C). The supernatant was then added to a 5-mL serological pipet, passed through a 0.8 µM syringe filter, and transferred to a 15-mL polypropylene conical tube. DNA extraction was performed on 25 µL of the purified sample. Extracted material was diluted 3-fold so not to overload the qPCR assay.

The diluted sample was processed in triplicate alongside an 18s rRNA endogenous control assays (Applied Biosystems Assay ID Hs99999901_s1) and a standard curve (Chapter 2). The qPCR results were ultimately used to calculate the concentration of viral copies in the inoculated water.

Sampling and Experimental Setup

To determine the viability of PaV1, lobsters were introduced to water previously inoculated with a known quantity of viral particles. In each experiment, the maximum volume of purified virus was added to a predetermined volume of water. The volume of water was

determined by the sample size in each experiment. For experiment 1, a 33 L volume of artificial seawater was inoculated with 3.5 mL of purified virus. A separate 25.5 L volume of artificial seawater was not and served as the source of water for experimental controls. For experiment 2,

114 L were inoculated with 4.75 mL of purified virus from a different donor lobster and 18 L were used for controls. In each experiment, the stocks of inoculated and non-inoculated water were aerated in 121.1 L polyethylene vessels for 24 h. Aeration was used to mix and maintain

PaV1 in the water. After 24 h, 1.5 L aliquots of the inoculated water were transferred to empty, 3

L experimental tanks (n = 22 and 76 for experiments 1 and 2, respectively). The same was done with the non-inoculated (control) water (n = 17 and 12 for experiments 1 and 2, respectively). A

5 x 5 cm flat piece of Vexar® plastic mesh screen was placed in each tank to provide structure and shelter for the lobsters. Tanks were then covered with aluminum foil to reduce evaporation

(Figure 3-1A). Since the tanks were inoculated one time, no water changes were performed throughout the course of the study to maintain the viral concentration in every experimental tank.

To maintain the water temperature at 26 ± 1°C, all tanks were placed in baths filled approximately halfway with freshwater (Figure 3-1B). One bath contained the control tanks and a second bath contained the experimental tanks, so that the two groups were kept separate. A water heater was positioned in the sump to heat the recirculating water. Water temperatures inside of the experimental tanks were monitored daily. Every other day, all lobsters were fed brine shrimp, Artemia sp., that had been thawed immediately beforehand. Lobsters were fed to satiation, or the point at which lobsters no longer responded to the shrimp.

Following the inoculation and distribution of water, the healthy lobsters were introduced to each tank over a period of several days. For experiment 1, lobsters were added on the following days post-inoculation: 1, 2, 3, 4, 5, 6, and 7. For experiment 2, lobsters were placed in the tanks

on days 1, 7, 14, and 21. Each tank received and held only one lobster for the duration of each experiment. All lobsters remained in their respective tanks, exposed to the inoculated or non- inoculated water for 14 d following their introduction. For example, lobsters added on day 21 were removed on day 35. After the 14-d exposure period, leg tissue samples were collected and preserved in 95% EtOH as previously described. Leg samples were retested using qPCR techniques. To confirm that PaV1 was infectious, lobster tissues were fixed in Davidson’s fixative for 48 h and subsequently transferred to 70% EtOH for later histological examination of tissues for infectious PaV1 particles. To ensure sufficient penetration of fixative into tissues, the cephalothorax and abdomen were separated before submersion in fixative. Prior to sectioning, lobsters were dissected along the coronal, or frontal, plane using a single blade razor. The embedded dorsal and ventral halves were stained using hematoxylin and eosin (H&E;

Histological Tech Services; Gainesville, FL). The stained tissues were examined for lesions using the Olympus BX53 Telepathology Microscope System and Olympus cellSens imaging software (Olympus Scientific Solutions Americas, Waltham, MA, USA).

Results

Viral Inocula and Concentrations

Using qPCR, the viral count of each inoculum was measured to determine the concentration of viral copies in the inoculated tanks. For experiment 1, adding 3.5 mL of purified

PaV1 to the artificial seawater yielded 1,739 viral copies / µL in each tank. For experiment 2, a total of 4.75 mL of purified virus were added to a different stock of artificial seawater. The concentration of PaV1 for experiment 2 was 100 viral copies / µL of extracted DNA, over 10- fold less than that of experiment 1. Since every tank contained a total of 1.5 L, each lobster was exposed to 2.6 x 109 viral copies in experiment 1 and 1.5 x 108 viral copies in experiment 2.

PaV1 Viability

Hemolymph collected from the donor lobster in experiment 1 had a viral count of 3.8 x 107 viral copies / µL of extracted DNA, while the donor lobster used in experiment 2 had 2.4 x 106 viral copies / µL of extracted DNA.

PaV1 was detected in lobsters over the duration of this study. In experiment 1, lobsters introduced 1 d after inoculation had a mean viral count of 108.2 ± 40.4 viral copies / µL of extracted DNA (n = 4; Figure 3-2). The viral count in infected lobsters steadily decreased through 14 d of exposure post-inoculation. PaV1 was not detected among the control group. In experiment 2, the mean viral count 1 d post-inoculation was 1.61 ± 1.48 viral copies / µL of extracted DNA (n = 10; Figure 3-3). The viral counts in experiment 2 showed no obvious trend overtime. Due to a limited number of animals available for experiment 2, lobsters were not added to the control group on days 14 and 21. However, those added on days 1 and 7 tested negative for PaV1. The total mortality of lobsters was 2.56% (1 / 39; experiment 1) and 35.5%

(16 / 45; experiment 2).

Tissues from a subsample of infected lobsters (n = 22) in experiment 2 were processed for histology. Lobsters added on the following days post-inoculation were examined: days 1 (n = 7),

2 (n = 1), 3 (n = 1), 7 (n = 5), 14 (n = 4), and 21 (n = 4). Tissues showed no evidence of histopathology (Figure 3-4). There were no signs of karyomegaly, emarginated chromatin,

Cowdry-like inclusion bodies, or phagocytic activity within the hepatopancreas or in any other tissue (Li et al. 2008).

Because no lesions were observed in the dissected tissues, DNA was re-extracted from a subset of lobster leg tissues (n = 5) that had already tested positive for PaV1. However, the leg tissues were first rinsed with deionized water for 10 min prior to the DNA extraction (Moss pers. comm.). The objective in doing this was to determine if the viral particles measured previously

by qPCR were inside of the leg tissues rather than on the exterior of the legs. After processing the rinsed leg samples, PaV1 was detected again in all five samples. This indicates that the virus likely infected the lobsters, even though lesions had not yet formed in the tissues.

Discussion

The results of this study reveal that PaV1 may remain viable in seawater for at least 21 d,

16 d longer than previously proposed and modeled (Kough et al. 2015). PaV1 was detected in tissue samples from every sampling period of experiments 1 and 2 via qPCR assay. PaV1 was reliably detected across both experiments, but there were stark differences in viral counts between the two. In general, more viral copies of PaV1 were measured in experiment 1 as compared with experiment 2. The reason for this discrepancy was likely attributable to the differences in viral load of the donor lobsters. Because there were fewer viral copies with which to inoculate tanks in experiment 2, the lobsters were exposed to a lower concentration of PaV1 relative to those in experiment 1.

The qPCR assay was able to detect and enumerate viral copies in the lobsters at all time periods, however, this molecular tool has limitations. Real-time qPCR assays amplify regions of viral DNA that are both infectious and non-infectious. Therefore, it is not possible to differentiate between viral particles that were actively replicating and those that were not. None of the stained tissues exhibited clear signs of infection. Two explanations are proposed for the absence of pathology in PaV1-positive lobsters. The qPCR assay could have merely measured nucleic acid from non-replicating viral particles. In a previous study, cell culture and qPCR techniques were used to measure infectious adenovirus since cell culture confirms viral infectivity (Staggemeier et al. 2017). In the present study, it is alternatively possible that the

PaV1 infection was too light to observe histological changes. The dissected lobsters had < 200 viral copies / µL of extracted DNA. Since this study represents the first comparison made

between the viral count and histology of PaV1-positive lobsters, the reported viral load may have been too low to activate an immune response. Similar findings were observed in juvenile pallid sturgeon. Tissues of fish that tested positive for ranavirus did not always contain inclusion bodies, as determined by in situ hybridization (Stilwell et al. unpubl. data). Another method for confirming viral infectivity is the detection of messenger RNA (mRNA), a molecule necessary in the replication of viruses (Nolan et al. 2006, Chambers et al. 2014). The presence of mRNA provides strong evidence of an active pathogen.

Since PaV1 was detected in lobsters up to 21 d following inoculation, it is possible that

PaV1 remains viable during this time period. Previous projections of PaV1 distribution were based, in part, on the assumption that PaV1 has a viability of 5 d in seawater (Kough et al. 2015).

Prolonged viability of 21 d, therefore, has the potential to increase the distance to which PaV1 is passively dispersed across the Caribbean. In addition, larval or postlarval lobsters in the water column may be at a greater risk of becoming infected by free-floating, active particles of PaV1 in the water. However, larval or postlarval lobsters in the surface may have a refuge from PaV1.

Briones-Fourzán et al. (2012) showed that, while lobsters in shallow habitats are still at risk, the rates of waterborne transmission are lower in shallower habitats. They hypothesized that this may be due to solar UV radiation penetrating the shallow water column and destroying free- floating PaV1 virions. Laboratory UV sterilizers have been demonstrated effective at destroying or at least inactivating PaV1 virions (Behringer et al. 2006), but this remains an untested hypothesis.

The transmission risk potential of PaV1 in the water column is akin to that of aerosolized bacteria for terrestrial organisms. Polymenakou et al. (2008) determined that small particles of bacteria were capable of traversing the eastern Mediterranean Sea due to a dust event originating

in Africa. Approximately 24% of these bacteria were linked to pathogens of plants, animals, and humans. Additional studies have specifically characterized the composition and movement patterns of marine aerosols. Bacterial communities collected over the Pacific Ocean were similar to those found in eastern and northern China, whereas bacteria collected over the Norwegian Sea had a great proportion of marine-derived microorganisms (Xia et al. 2015). Infectious PaV1 is unlikely to be transported as an aerosol, but Caribbean hydrology could cause PaV1 virions to experience a similar degree of movement.

The Caribbean is connected by a network of water currents. The three strongest current systems in this region are the Yucatan Current, the Loop Current, and the Florida Current. Water currents move northwestward across the Caribbean Sea and into the Gulf of Mexico from the equatorial Atlantic Ocean (Centurioni & Niiler 2003). For PaV1 viral particles, there is little exchange between northwestern and southeastern regions of the Caribbean. However, in areas where oceanographic barriers are absent, PaV1 particles are capable of being transported from the Yucatan to the Northern Caribbean via Cuba (Kough et al. 2015). This pathway would explain the strong genetic connectivity of PaV1 across Northern Caribbean (Moss et al. 2013).

When a disease outbreak originates in the Northern Caribbean, lobsters in the Central or

Southern Caribbean are less likely to be infected by waterborne transmission (Kough et al.

2015). However, Kough et al. (2015) noted that it is possible for PaV1 particles to be transported throughout the Caribbean by gyres off of Columbia and Panama. Given that PaV1 likely remains viable longer than modeled (21 d vs. 5 d), the pathogen could hypothetically be spread by these gyres and infect populations of lobsters across the greater Caribbean. This further supports the genetic exchange between Northern Caribbean, Central Caribbean, and Panama (Moss et al.

2013).

One noteworthy drawback in this study was the use of different donor lobsters for each experiment. The donor lobster used in experiment 1 had a higher viral count relative to the lobster used in experiment 2, which may explain the discrepancy in viral count detected in each exposed lobster. The use of different donor lobster across experiments introduces the possibility that the experimental lobsters were also exposed to various alleles and levels of virulence (Moss et al. 2012, 2013). In the future, hemolymph from multiple donor lobsters should be pooled prior to the study to ensure that lobsters in both experiments are exposed to the same alleles and dosages of PaV1.

This study was a first in understanding the environmental viability of PaV1. The next step in corroborating PaV1 infectivity is to design a quantitative (q)RT-PCR assay to detect mRNAs associated with protein production in the replication cycle of PaV1. The detection of mRNA would provide evidence that PaV1 is actively replicating within the host. A similar test was developed for locating late mRNA in human papillomavirus to differentiate between clinically-relevant, active infections from inactive infections (Chambers et al. 2014). The established qPCR assay for PaV1 (Chapter 2), coupled with an mRNA qRT-PCR assay, could be applied to experiments examining PaV1 persistence in a suite of environmental conditions, such as elevated sea surface temperatures. Since disease outbreaks are positively associated with increasing water temperatures (Burge et al. 2014), testing the viability of PaV1 at different temperatures would provide insight into how rising sea surface temperatures may alter the infectivity of PaV1. To test the viability of murine norovirus in different temperatures and salinities, Takahashi et al. (2016) placed the pathogen inside of dialysis tubes, which were then submerged in various aquatic locations. In contrast to Burge et al. (2014), murine norovirus remained viable for longer when exposed to colder seawater temperatures likely due to the

natural decrease in solar UV exposure during winter. To determine the effect of temperature on the viability of PaV1, a similar in situ experiment is recommended.

Figure 3-1. Experimental setup. A) Covered tanks each containing inoculated, or non-inoculated, water and a single lobster. B) All tanks held in temperature-controlled water tables.

Figure 3-2. The viral load detected in lobsters (n = 22) added to inoculated water for experiment 1. PaV1 was not detected in lobsters (n = 17) added to the non-inoculated water. Error bars represent SEM.

Figure 3-3. The viral load detected in lobsters (n = 40) added to inoculated water for experiment 2. PaV1 was not detected in lobsters (n = 5) among the control group. Error bars represent SEM.

Figure 3-4. Histology of PaV1-positive tissues. A) Hepatopancreas dissected from a lobster added to water 1 d post-inoculation. No lesions or eosinophilic inclusions were observed in the hemal sinuses (HS), tubules (T), or lumen of the tubules (L). B) Hepatopancreas dissected from a lobster added to the water 21 d after inoculation. No lesions or abnormalities were present. Reserve inclusions (RI) were identified in a handful of lobsters (n = 4). Both tissues were magnified by 100x.

CHAPTER 4 THE RELATIONSHIP BETWEEN VIRAL LOAD AND INFECTION IN CARIBBEAN SPINY LOBSTERS EXPOSED TO PANULIRUS ARGUS VIRUS 1

Introduction

Dose-response relationships, which determine how a host responds to a known concentration of pathogen, can be divided into two basic model types, dichotomous and time-to- occurrence (Altshuler 1981). A dichotomous response refers to the proportion of subjects that, at the termination of an experiment, are either infected or not. A time-to-occurrence response measures the number of subjects that contract the virus throughout an experiment. The latter model has a temporal component, whereas the former does not (Altshuler 1981).

A number of dose-response models have been developed to understand the relationship between dose and outcome of an illness, especially in the context of human health (Altshuler

1981, Felício et al. 2015, Van Abel et al. 2017). In a recent study, a series of risk assessments were conducted using norovirus dose-response models. The objective was to determine which of the existing models predicts the highest risk of illness when various doses of norovirus were ingested. No single model outperformed the others because risk of infection varied, especially at the low doses (< 10 genomic equivalent copies) (Van Abel et al. 2017). In a separate risk assessment, pathogens (e.g., norovirus, Salmonella spp., etc.) found in foods of non-animal origin (FoNAO) were examined based on seven criteria, including dose-response relationship

(Felício et al. 2015). Pathogens were ranked according to the dose or growth of pathogen necessary to cause disease. Norovirus and Salmonella spp. were ranked the highest because a low dose is sufficient to trigger an illness (Felício et al. 2015).

Dose-response relationships have also been described for non-human subjects. As an example, Dwyer et al. (1997) collected larvae from feral populations of gypsy moths, Lymantria

dispar dispar, and exposed them to different doses of nuclear polyhedrosis virus. After also exposing laboratory-reared larvae to the virus, the two group responses were compared. Feral larvae were less susceptible to infection than were the lab larvae, demonstrating that feral moths are more heterogeneous in their susceptibility to the virus. In a different study, feline immunodeficiency virus (FIV) was studied in cats inoculated with varying doses of infectious

FIV-PPR (strain isolated previously from cats with AIDS-like symptoms). Cats that received the lowest dose of 50 TCID50 had fewer behavioral changes and no signs of lymphadenopathy

(enlargement of lymph nodes). In cats given higher levels (e.g., 250 TCID50), these symptoms were dose-dependent (Hokanson et al. 2000). Few studies have evaluated the dose response in organisms exposed to marine pathogens. Prior et al. (2003) developed a bioassay for Taura

Syndrome Virus (TSV) and White Spot Syndrome Virus (WSSV) in shrimp. When injected into

7 7 shrimp, TSV had a lethal infective dose (LD50) of between 1:6.667 x 10 and 1:7.692 x 10 . For

6 6 WSSV, the LD50 ranged from 1:4.444 x 10 to 1:4.505 x 10 . When challenged with waterborne transmission, the LD50 for TSV was 1:2857. Mortality in shrimp was too low for estimating the

LD50 of waterborne WSSV. In a separate study, dose-response curves were used to develop an adenosine-triphosphate (ATP) content-based assay for measuring proliferation of the parasite

Perkinsus marinus (Shridhar et al. 2013). Aside from these two studies, little attention has been given to the dose-response relationship between marine organisms and pathogens. The present study, which evaluates the dose-response relationship in lobsters exposed to Panulirus argus

Virus 1 (PaV1), is necessary for expanding this area of research.

The Caribbean spiny lobster, Panulirus argus, has significant economic value, accounting for 34,574 t in annual landings and nearly $1 billion across the greater Caribbean

(FAO 2014). Considering the importance of this fishery, it is alarming that up to 17% of lobsters

caught by commercial fishermen are infected by a pathogen (Behringer et al. 2012, Moss et al

2013). Discovered in 2000, PaV1 is the only virus known to naturally infect any species of lobster worldwide (Shields & Behringer 2004). PaV1 infects life history stages of P. argus ranging from postlarvae to adult lobsters. While PaV1 is found in most life history stages, infections vary in severity with respect to lobster size, establishing gross pathology that leads to mortality in juveniles but only causing subclinical infections in adults. Among juveniles, susceptibility to PaV1 transmission is inversely correlated with increasing lobster size (Butler et al. 2008). Clinically diseased juveniles display several visual and behavioral signs including milky-white hemolymph, conspicuous red or pink (“cooked”) discoloration of the exoskeleton, and general lethargy (Shields & Behringer 2004, Lozano-Álvarez et al. 2008). Large, adult lobsters exhibit these gross symptoms very infrequently (<< 1%) (Shields & Behringer 2004).

Therefore, molecular techniques are necessary for diagnosing PaV1 in asymptomatic adults.

The reasons for this discrepancy in PaV1 prevalence across lobster size are unknown. Prior studies have examined viral load as a potential factor in disease progression. Brunner et al.

(2005) determined that dose of ranavirus inoculum had a positive relationship with ranavirus virulence in the tiger salamander, Ambystoma tigrinum. Another study compared viral load and progression of WSSV in tissues of the giant tiger prawn, Penaeus monodon (Jeswin et al. 2015).

WSSV was first detected in haemocytes, followed by pleopods, muscle, and gills of infected prawns. Of these tissues, gills had the highest viral load. Viral load was used in the present study to measure the dose-response relationship between PaV1 progression and lobster size.

The objective of this study was to measure the response of juvenile lobsters to increasingly dilute viral loads of PaV1 and to determine the minimum infectious dose for each size class of juveniles. It was hypothesized that juvenile lobsters would experience a more rapid

progression of PaV1 over time, as determined by viral count, when challenged with a higher viral load. This pattern would be similar in larger juvenile lobsters; however, larger lobsters would have a lower viral count relative to small lobsters.

Materials and Methods

Lobster Collection and Screening

In fall 2014, juvenile lobsters were collected in the Florida Keys, USA and held at the

University of Florida Aquatic Pathobiology Laboratory (Gainesville, FL, USA). Lobsters (25 –

35 mm carapace length, CL) were separated into individual 37.8 L tanks and screened for PaV1.

Using 27 G x 5/8 1cc insulin syringes, between 0.1 and 0.2 mL of hemolymph were removed from each lobster at the proximal joint of the fifth periopod. All hemolymph samples were stored at -20C in microcentrifuge tubes containing 0.9 mL of 95% EtOH. DNA was extracted from each sample using a Quick-gDNA MiniPrep Kit (Zymo Research) and the methods described in

Chapter 2. The DNA extract was tested for PaV1 using a conventional PCR assay and gel electrophoresis (Montgomery-Fullerton et al. 2007, Moss et al. 2012, Chapter 2). Finally, PCR products were run on 1% agarose gels. Any lobster that tested positive for PaV1 was excluded from the study.

In fall 2015, larger juvenile lobsters (40 – 55 mm CL) were collected in the Florida Keys for the second experiment. The lobsters were held and sampled as above.

Inoculation

For each experiment, healthy lobsters were inoculated with hemolymph collected from a clinically-infected lobster. Donors were identified by characteristic white hemolymph (Shields &

Behringer 2004). Prior to inoculation, the viral count in each donor was measured using qPCR to standardize respective viral concentrations across experiments. Phosphate-buffered saline (PBS) was added to the infected hemolymph in order to create serial dilutions: 10-2, 10-4, and 10-6. An

inoculum of 0.2 mL of diluted hemolymph was injected into each lobster at the base of its fifth periopod. Lobsters in the negative control and PaV1-positive, undiluted control groups received

0.2 mL of PBS and 0.2 mL of undiluted donor hemolymph, respectively. All samples of inocula were preserved in 0.9 mL of 95% EtOH and later processed using qPCR. The resulting number of viral copies was used to standardize the inocula used in experiment 2.

In experiment 1, approximately 2.2 mL of infected hemolymph were removed from a donor lobster and used as inocula. Lobsters in the ‘undiluted’ treatment group (n = 8) each received 0.2 mL of hemolymph directly from the donor lobster, while 0.5 mL of the hemolymph was used to prepare dilutions for the remaining treatment groups (10-2, 10-4, and 10-6). Nine lobsters were haphazardly assigned to two of the dilution groups (10-2, 10-4), eight to the third dilution group (10-6), and seven to the control group. Lobsters in the control group were injected with 0.2 mL of PBS.

In experiment 2, 0.24 mL of hemolymph was transferred from a different donor lobster to each experimental lobster (n = 10) in the ‘undiluted’ group. A volume of 0.24 mL was used experiment 2 because this donor had a lower viral load. The volumes were standardized to ensure that lobsters in both experiments received the same concentration of viral particles. Next, 0.38 mL of infected hemolymph was used to prepare serial dilutions (10-2, 10-4, and 10-6), of which

0.2 mL was then delivered to lobsters (n = 9) in each of the dilution groups. Nine lobsters were assigned to the control group.

Lobsters were held individually in 37.8-liter glass aquaria and fed every other day. Their diet consisted of frozen squid, bivalves, and shrimp. Water temperatures were monitored daily and maintained at 26 ± 1°C. Salinity was measured every day and the following parameters were tested once per week: pH, ammonium, nitrate, and nitrite. A 50% water change was performed

on every tank each week.

Sampling

Following inoculation, 0.1 mL of hemolymph was removed from the proximal joint of the fifth periopod once per week and preserved in 0.9 mL of 95% EtOH. Lobsters were sampled over a six-week period after which the experiment was terminated, and the lobsters euthanized and dissected. All samples and inocula were tested for PaV1 using DNA extraction and qPCR techniques described in Chapter 2. To extract DNA from the inocula, the inocula were centrifuged at room temperature (14,950 x g for 1 min). The EtOH was decanted and the pellet dried at room temperature for 1 h. DNA was extracted from approximately 25 mg of dried pellet.

Data Analyses

The viral copies within each inoculum were compared between experiments 1 and 2 using a paired t-test. A one-way ANOVA was used to test the null hypothesis that there were no changes in viral count over time. Two factors, treatment (inoculum) and time (week of sample collection), were used in a two-way ANOVA to determine if these variables had an effect on the response (viral count). The effect of these variables and of lobster size were tested using a two- way repeated measures ANOVA. All statistical analyses were performed in RStudio (RStudio:

Integrated Development for R. RStudio, Inc., Boston, MA, URL: https://www.rstudio.com/).

Results

Viral Load

In experiment 1, the mean size of inoculated lobsters (n = 41) was 29.3 ± 0.43 mm CL (±

SEM). In experiment 2, inoculated lobsters (n = 46) had a mean CL of 49 ± 0.55 mm. The inocula administered in experiments 1 and 2 had a similar number of viral copies within respective treatment groups (Figure 4-1) (paired t-test: t = 0.966, p = 0.344). Therefore, lobsters in corresponding treatment groups of experiments 1 and 2 were injected with a similar

concentration of PaV1 (Table 4-1).

There was a significant difference between the mean net change in viral copies (Figure 4-2) detected in samples collected at the beginning of each experiment and at the end on week six

(Table 4-2) (one-way ANOVA: F9,61 = 4.19, p = 0.0003). The mean difference was significant in experiment 1, where lobsters assigned to the “undiluted” group had higher viral counts relative to those in experiment 2 and in the remaining treatment groups (p < 0.0006, Tukey’s HSD).

The sampling week post-inoculation did not significantly affect the progression of PaV1 in small lobsters (Table 4-3) (two-way ANOVA: F5,198 = 1.96, p = 0.0857). Dose treatment (two- way ANOVA: F4,198 = 5.56, p = 0.0003) and its interaction with time (two-way ANOVA: F20,198

= 4.86, p < 0.0005) had a significant effect (Table 4-3). In experiment 2, time (Table 4-4) (two- way ANOVA: F5,225 = 1.24, p = 0.1597), treatment (two-way ANOVA: F4,225 = 1.66, p = 0.1597) nor their interaction (two-way ANOVA: F20,225 = 0.88, p = 0.6119) had an effect on the number of viral copies measured in large lobsters.

When combining the data collected from experiments 1 and 2, treatment no longer had an effect on viral count (Table 4-5) (two-way repeated measures ANOVA: F4,5 = 1.57; p = 0.313).

Lobster size, however, had a block effect within treatment (two-way repeated measures

ANOVA: F5,5 = 3.23, p = 0.007). Weeks post-inoculation and its interaction with treatment also showed significant effects on viral count (two-way repeated measures ANOVA: F5,5 = 5.9, p <

0.0005 and F20,5 = 2.21, p = 0.002, respectively).

Lobster Mortality and Survival

Mortality was observed throughout the study. In experiment 1, 19.5% of the small lobsters died within six weeks of inoculation (Figure 4-3A). In experiment 2, 10.9% of the large lobsters died (Figure 4-3B). Lobsters in the “undiluted” group (experiment 1) experienced the highest mortality of 75%. In experiment 1, 80.5% of the small lobsters survived throughout the study

(Figure 4-4A). In experiment 2, 89.1% of the large lobsters survived (Figure 4-4B). No clinical signs of PaV1 were observed in either experiment, but all lobsters tested positive for PaV1 by qPCR.

Discussion

This is the first study to measure size-dependent differences in viral counts of PaV1- infected lobsters. Smaller juvenile lobsters had significantly higher viral count when exposed to

“undiluted” hemolymph compared to those exposed to diluted hemolymph. The viral count in small lobsters has a positive relationship with time and PaV1 dose. Conversely, larger juvenile lobsters did not respond to dose or time. While there was an increase in viral count over time and across treatment groups, the change was not significant for larger juveniles. All lobsters appeared to have become infected with PaV1 even at the lowest dose. Therefore, the minimum dose was not determined.

Lobster size had a significant effect on the progression of PaV1, thus supporting our hypothesis and previous findings that PaV1 severity is highest in small lobsters relative to larger lobsters (Butler et al. 2008). All experimental lobsters became infected, even the largest lobsters with the lowest dose. Small lobsters that received undiluted hemolymph at the beginning of experiment 1 had the greatest net increase in viral count and suffered the highest mortality. Total mortality among the small lobsters was 50% greater than among the larger juvenile lobsters.

Large lobsters had lower mortality and viral load in experiment 2 relative to their smaller conspecifics in experiment 1.

Lobsters used in the present study were collected from Florida Bay, a nursery habitat consisting of predominantly juvenile lobsters. Prevalence of PaV1 in this region varies spatially and temporally with some areas reaching a prevalence of 100% (Behringer et al. 2011, Chapter

5). Since large juveniles have a lower viral count, it would stand to reason that perhaps areas

with low PaV1 prevalence are occupied by lobsters of a larger size class. The results from

Chapter 5 support this finding because smaller lobsters were more likely to contract PaV1 than large lobsters. This is further corroborated by the present study in which small lobsters had the highest mortality rate and the lowest survival rate compared to larger lobsters. Small lobsters may be more sensitive to PaV1 and the resulting metabolic exhaustion (Shields & Behringer

2004). If a lobster survives the infection and matures to a larger size class, the lobster may hypothetically have an immune response to subsequent PaV1 infections (Haunton 2012).

Few studies have examined dose-response relationships between pathogens and their invertebrate hosts, but results similar to those in the present study have been found in terrestrial organisms. In calves administered Salmonella tyhimurium, an inverse relationship was observed between age of the calf and mortality from bovine salmonellosis. Calf mortality was also inversely linked to S. tyhimurium dose. Calves challenged with high doses suffered more severe symptoms than those administered low doses (Smith et al. 1979).

In addition to size and age of an individual, virulence of a pathogen can affect the susceptibility to infection. Laramore et al. (2009) discovered that variations in the virulence of

WSSV isolates led to differences in mortality among Pacific white shrimp infected with the isolates. Honey bee larvae, Apis mellifera, responded similarly when challenged with different species of fungi from the genus Aspergillus. fungi. Of the ten isolates tested, three (Aspergillus flavus, Aspergillus nomius, and Aspergillus phoenicis) caused stonebrood disease and subsequent mortality in larvae because some isolates were more virulent than others (Foley et al. 2014). The present study did not control for possible variance in virulence attributable to unknown differences in the virulence of different alleles identified for PaV1 (Moss et al. 2012, 2013).

Hemolymph from a different donor lobster was used in experiments 1 and 2, so it is possible that

different alleles were used and each has a different virulence. In the future, hemolymph from multiple donor lobsters should be pooled and used across all experimental animals.

In this study, hemolymph from a donor lobster infected with PaV1 was injected into experimental lobsters. Therefore, lobster mortality and associated viral counts were likely greater than they would have been had the lobsters been naturally exposed to PaV1. To simulate natural conditions and transmission, the next step would be to repeat this study using other known modes of PaV1 transmission: waterborne, contact with a diseased conspecific, and ingestion of infected tissue (Butler et al. 2008). Knowing the minimum dose of PaV1 required for natural transmission may further explain the adaptive value of behavioral immunity in reducing PaV1 transmission among lobster populations (Butler et al. 2015). The relationship between PaV1 dose and response by lobsters of different size classes is complex, but this study is the first to clarify these complexities.

Table 4-1. Number of viral copies injected into each lobster at the beginning of experiment 1 and experiment 2. Experiment Control 10-6 10-4 10-2 Undiluted 1 0 7.37 x 104 1.01 x 107 3.81 x 109 1.67 x 1011 2 0 8.58 x 104 2.94 x 107 4.5 x 109 1.67 x 1011

Table 4-2. Results of a one-way ANOVA performed on the net change in viral count over time. These values reflect the combined data from both experiments. Source df MS F P Treatment 9 3.14 x 1014 4.19 0.000301 Error 61 7.48 x 1013 Total 50

Table 4-3. Results from a two-way ANOVA for experiment 1. Treatment included five groups: control, 10-6, 10-4, 10-2, and undiluted. Six time points, or weeks post-inoculation, were used for time. Source df MS F P Treatment 4 1.5 x 1014 5.5511 0.0002957 Time 5 5.31 x 1013 1.9635 0.0856591 Treatment x Time 20 1.31 x 1014 4.8571 < 0.0005 Error 198 2.71 x 1013 Total 227

Table 4-4. Results from a two-way ANOVA for experiment 2. Treatment included five groups: control, 10-6, 10-4, 10-2, and undiluted. Time included six sampling periods, or weeks post-inoculation. Source df MS F P Treatment 4 1.52 x 1013 1.6624 0.1597 Time 5 1.14 x 1013 1.2390 0.2917 Treatment x Time 20 8.07 x 1012 0.8808 0.6119 Error 225 9.16 x 1012 Total 254

Table 4-5. Results from a two-way repeated measures ANOVA. Treatment included the five groups: control, 10-6, 10-4, 10-2, and undiluted. Lobster size (small, large) was treated as a block effect within each treatment group. The six sampling periods, or weeks, were represented by time. Source df MS F P Treatment 4 2.27 x 1015 1.567 0.313 Block(Treatment) 5 1.45 x 1015 3.233 0.007 Time 5 2.64 x 1015 5.896 < 0.0005 Treatment x Time 20 9.92 x 1013 2.214 0.002 Error 5 4.48 x 1013 Total 59

Figure 4-1. Log-transformed values representing mean number of viral copies per inoculum delivered to lobsters in experiments 1 (Expt. 1) and 2 (Expt. 2). Each inoculum was processed in triplicate. The diluted and undiluted hemolymph used in a single experiment originated from the same source lobster. Every inoculum was processed in triplicate. Since the SEM were marginal, the error bars are not visible.

Figure 4-2. Net change in the number of viral copies detected six weeks post-inoculation. Values above SEM bars indicate sample size of lobsters alive at the end of each experiment. Data were log-transformed.

Figure 4-3. Regression of percent mortality of lobsters throughout A) experiment 1 and B) experiment 2. R2 value for each trend line is listed in the data series.

Figure 4-4. Regression of percent survival of lobsters throughout A) experiment 1 and B) experiment 2. R2 value for each trend line is listed in the data series.

CHAPTER 5 SPONGES AND THE SPATIAL EPIDEMIOLOGY OF PANULIRUS ARGUS VIRUS 1 IN CARIBBEAN SPINY LOBSTERS THROUGHOUT THE FLORIDA KEYS, USA

Introduction

A number of factors drive the distribution of pathogens across landscapes. In terrestrial ecosystems, pathogen dispersal can be influenced by habitat fragmentation, or the presence (or absence) of geographical barriers (Ericson et al. 1999, Colling & Matthies 2004, McCallum et al.

2003, 2004). For example, spatially-distinct populations of susceptible hosts have a reduced risk of contracting viruses from infected communities in surrounding areas. Perkins and Matlack

(2002) illustrated this point by discovering that the spread of pathogens across commercial forests can be mitigated by isolating plantations from one another. Therefore, fragmented terrain can inhibit pathogen movement. However, Plantegenest et al. (2007) highlighted that, in some cases, pathogens benefit from habitat fragmentation. For example, pathogens may infect hosts residing on the edge of adjacent patches more so than in the center of the patches (Murcia 1995).

In addition to the physical confines of geographical barriers, terrestrial systems also contain flying insect vectors and avian hosts through which the distribution of pathogens is accelerated

(McCallum et al. 2003). For example, the bushfly, Musca vetustissima, and blowfly, Calliphora spp., have contributed to the spread of rabbit haemorrhagic disease across Australia and New

Zealand (Cooke & Fenner 2002). Pathogens, such as West Nile Virus, are distributed by avian hosts at a rate of 1200 km year -1. Even so, the rates of pathogen dispersal across terrestrial environments are generally lower compared to marine systems (McCallum et al. 2003).

Marine environments are generally considered more open (Hellberg et al. 2002, McCallum et al. 2003, 2004, Real & Biek 2007). This open environment gives rise to high connectivity over long distances, and consequently, epidemics spread at a faster rate in marine environments than

they do in terrestrial ones. In addition to the open nature of seascapes, pathogens may be transported across wide spatial ranges by the currents driven along coastlines (McCallum et al.

2003). Lessios et al. (1984) documented the progressive die-offs suffered by the long-spine sea urchin, Diadema antillarum, and hypothesized that the pathogen responsible was likely distributed across the Caribbean by ocean currents. This die-off occurred at a maximum rate of

4990 km year -1 (Lessios et al. 1984, Lessios 1988).

In contrast to terrestrial environments, few vectors have been discovered in the marine environment. While flying insect vectors cause and maintain epidemics on land (Cooke &

Fenner 2002, McCallum et al. 2003), flying vectors do not exist in marine ecosystems. The marine fireworm, Hermodice carunculata, is one of few vectors described from the ocean. This species of worm harbors Vibrio shiloi, a pathogen that causes bleaching in corals (Sussman et al.

2003). Another non-flying, marine vector is the hyperparasite, Hyperspora aquatica n. gn., n.sp

(Sussman et al. 2003, Stentiford et al. 2017), which infects the paramyxid Marteilia cochillia

(Stentiford et al. 2017). These types of vectors are rare, probably because they are largely unnecessary in the marine environment where the water itself is a highly effective means of transport. However, where animal vectors are absent, the viability of the pathogens outside of the host is a key factor contributing to pathogen transmission and spread. When a pathogen remains infectious for extended periods, it can be transported across long distances in the marine environment and infect populations of hosts along the way (Kough et al. 2015).

An example of an at-risk marine species is the Caribbean spiny lobster, Panulirus argus.

This lobster is renowned for its economic value to communities around the Caribbean Sea and in the Florida Keys (USA). With annual landings exceeding 34,574 t and $1 billion (FAO 2014), P. argus is the highest valued marine fishery in the Caribbean. Lobster landings, however, have

decreased in many Caribbean countries (FAO 2014). While many factors may be involved in this decline, researchers and managers suspect that the virus, PaV1, may play a role. PaV1 (Panulirus argus Virus 1) is found in many P. argus populations throughout Caribbean and Florida waters

(Moss et al. 2013). It infects most life history stages of P. argus, but is most lethal to small juvenile lobsters that typically inhabit vegetated nursery habitats. The Florida Keys, where PaV1 prevalence is among the highest in all of the Caribbean (Moss et al. 2013), are predominantly composed of seagrass beds, coral reefs, hard-bottom, and open sand substrate. Hard-bottom habitat, which accounts for 30 – 40% of the nearshore benthos surrounding the Florida Keys, is characterized by a combination of fleshy macroalgae, stony and gorgonian corals, and sponges

(Chiappone & Sullivan 1994, Herrnkind et al. 1997).

Prior work has shown that recently settled spiny lobsters, termed early benthic juveniles

(EBJs), contract PaV1 when deployed in cages as ‘sentinels’ within the hard-bottom areas of

Florida Bay, but not in seagrass beds (Behringer unpubl. data). Thus, PaV1 appears to exhibit variability in prevalence across habitats of different type, community composition, or structural complexity. Variability in PaV1 prevalence is also observed across the Florida Keys. For example, some regions of Florida have a PaV1 prevalence of 100% while others have 0% prevalence (Behringer et al. 2011). The drivers of this discrepancy in prevalence remain unknown.

The objective of this study was to identify the patterns in variability of PaV1 prevalence across the Florida Keys and determine if these patterns could be explained by a suite of habitat characteristics. To identify relationships between disease prevalence across habitat, the following habitat features were described and then compared with the prevalence of PaV1: sponge community characteristics, hard coral community characteristics, gorgonian community

characteristics, seagrass cover, and macroalgal cover.

Materials and Methods

Sample Collection

Hard-bottom sites in the Florida Keys were identified and surveyed between May and

August 2015 (Table 5-1, Figure 5-1). Research vessels towed divers throughout hard-bottom habitat – areas likely to contain lobsters – to identify sampling sites. Sites with a minimum of ~

10 juvenile lobsters were surveyed in order to maximize the number of sampling sites used in this study. Twenty sites were ultimately identified as suitable and surveyed. Ten of these sites belong to an ongoing, annual monitoring program that was begun in 2000 (Behringer et al.

2011). The remaining ten sites were discovered during the 2015 sampling period and had not been sampled for PaV1 previously.

At each site, two SCUBA divers used hand nets to catch lobsters. Each diver was allotted a search time of 30 min (excluding capture and handling time). When lobsters were located, the diver stopped the timer before catching the lobsters. The diver then resumed the timer and continued to search for lobsters. Relative lobster abundance was represented as number of lobsters collected (catch-per-unit-effort, or CPUE) over a 60-min period (Bertelsen et al. 2009).

All captured lobsters were brought onboard the research vessel where the following data were recorded for each animal: sex, carapace length (CL), molt condition, number of injuries (not sustained during capture), and visible signs of PaV1. Between 0.1 ml – 0.2 ml of hemolymph were drawn from the fifth periopod sinus of each lobster using a 27 G x 5/8 1cc insulin syringe.

Occasionally, leg tissue samples were collected in lieu of hemolymph for juveniles too small for the syringe (< 20 mm CL). All samples were preserved in 1.5 ml microcentrifuge tubes containing 0.9 ml of 95% EtOH. Samples were stored in -20C until tested for PaV1. Lobster samples were processed using a TaqMan® real-time quantitative PCR (qPCR) assay described in

Chapter 2. All diagnostic work was done at the University of Florida Aquatic Pathobiology

Laboratory (Gainesville, FL). Prevalence of PaV1 was calculated using the proportion of infected lobsters relative to lobster abundance, multiplied by 100% (Table 5-1).

The divers also characterized the benthic characteristics of each site using 2-m wide belt transects (Tellier et al. 2008, Bertelsen et al. 2009). One diver extended a transect tape 25 m in each of the four cardinal directions from a haphazardly-positioned central point on the site.

Therefore, each of the four surveys covered an area of 50 m2 (total area of 200 m2). After securing the transect tape, the second diver used a digital video camera (Canon IXUS 125 HS) to record a 1-m swath of the bottom along each side of the belt transect. These digital videos were later analyzed for sponge size, density, and species identity.

ImageJ was used to measure sponge size, while the remaining habitat parameters were visually enumerated and then recorded in Microsoft Excel (2017). To accurately measure sponge size, screenshots of every video were taken and used to calibrate the ImageJ scale to each video.

Lobsters typically do not reside in shelters, such as sponges, that are < 20 cm in diameter

(Childress 1995). Therefore, sponges were measured and binned into one of two categories: small (< 20 cm in diameter) or large (≥ 20 cm in diameter; Bertelsen et al. 2009). For every site, mean densities of sponges, hard corals, and gorgonians were calculated using the abundance of each characteristic and the total area surveyed (200 m2). The point-intercept method was incorporated to estimate percent seagrass and macroalgal cover at each site. A total of 360 points within the 200 m2 area (90 points / 50 m2 belt transect) were surveyed for seagrass or algae. The percent cover of each was calculated.

Data Analyses

An average was calculated for each of the following variables: total sponges, large sponges, gorgonians, hard corals, seagrass cover, and algal cover. The seven variables that

described EBJ habitat were included in a principal component analysis (PCA) to reduce the dimensionality of further analyses and to eliminate correlation among variables. Cattell’s scree test (1966) was used to determine the number of principal component (PC) scores incorporated into the model. Any eigenvalues above the “elbow” of the scree plot were incorporated into a generalized logistic mixed model, which tests the probability of eliciting a binary response based on a combination of fixed and random effects. Also included in the model were survey sites (1-

20) and two lobster parameters, lobster size (CL) and mean lobster density. Some sites (n = 6) used in the long-term monitoring program contained concrete blocks from a previous study

(Behringer 2003). The number of blocks, the PC scores, and the lobster parameters were treated as fixed effects, while the sites were treated as random. Disease status of each individual lobster was treated as a binary response variable, where 1 = infected (PaV1 positive) and 0 = healthy

(PaV1 negative). The PCA and logistic regression were completed using RStudio (RStudio:

Integrated Development for R. RStudio, Inc., Boston, MA, URL: https://www.rstudio.com/).

Results

PaV1 Prevalence

A total of 570 lobsters were sampled and subsequently tested for PaV1. On average, 28.5 ±

4.12 lobsters (39.4 ± 0.56 mm CL, SEM) were caught per site during the timed surveys. Lobster densities ranged from 0.015 to 1.52 lobsters / min with an abundance of 9 – 92 lobsters / site

(Table 5-1). Mean PaV1 prevalence was 63.2 ± 5.6 % and ranged from 0% (Cudjoe Bay) to

100% (Don 3). There were variations in disease prevalence across the Florida Keys (Figure 5-1).

Community Characteristics

Of the 3399 sponges measured in this study, 608 were ≥ 20 cm in diameter and 2791 were

< 20 cm. Averaged across all sites, there was a mean abundance of 170 ± 23 sponges (0.042 ±

0.01 sponges / m2), which ranged in density from 0.1 to 1.96 sponges / m2 (Table 5-2). Twelve

sponge species were identified, three of which were the most abundant: the loggerhead

Spheciospongia vesparium, the vase Ircinia campana, and the brown branching Ircinia sp.

(Figures 5-3, 4, 5).

The mean density of hard corals and gorgonians was 0.33 ± 0.19 corals / m2 and 0.006 ±

0.002 corals / m2 (n = 20), respectively. Seagrass (turtle grass, Thalassia testudinum) covered a mean area of 13.01 ± 2.72%, while algae covered 34.8 ± 5.96% (n = 20) (Table 5-2). The predominant algae from each survey site belonged to the genus, Laurencia.

PaV1 Prevalence vs. Community Characteristics

All of the community characteristics measured in this study were included in the PCA

(Table 5-2). From this analysis, a scree plot was generated containing the eigenvalues associated with each of the six principal components (Figure 5-6). The first two axes, or components, had eigenvalues above the “elbow” and so were therefore used to generate PC scores, PC 1 and PC 2.

The four remaining components formed a line towards the right of the graph, illustrating that these components explained little variance. Ultimately, PC 1 explained 78% of variability in the data while PC 2 accounted for 22% (Figure 5-7).

These two sets of PC scores and the lobster parameters (n = 570 and n = 20 for lobster size and density, respectively) were then incorporated into the logistic mixed model. Both lobster size and density had significant effects on the disease status of lobsters (Table 5-3). The smaller a lobster, the more likely it was to be infected (p < 0.001). High densities of lobsters were also more likely to be infected with PaV1 (p = 0.0042) (Figure 5-1). However, PC 1 and PC 2 did not yield significant results (p = 0.2425 and p = 0.3204, respectively). Therefore, the six habitat characteristics did not explain the variance in PaV1 prevalence. Similar results were observed for the cement blocks, which did not have a significant effect on the number of infected lobsters (p =

0.7991) (Table 5-3).

Discussion

In the present study, lobster size and density had a significant effect on the number of

PaV1-infected lobsters. PaV1 cases were associated with small juvenile lobsters and high lobster densities. The virus was detected in lobsters across all but one sampling site, but the probability of infection did not exhibit a significant relationship with any of the habitat characteristics measured (total sponges, large sponges, gorgonians, hard corals, seagrass cover, algal cover) nor with the presence of the concrete block.

Behavioral immunity was likely a driving factor in the results in this study. The total density of sponges or the density of only large sponges did not predict the likelihood of a lobster being infected with PaV1, which would likely be explained by the avoidance behavior exhibited by healthy lobsters (Behringer & Butler 2006, Butler et al. 2015). Lobsters typically avoid aggregating with diseased conspecifics in or under sponges. This likely explains why the relative density of large, shelter-forming sponges was not predictive of the prevalence of PaV1. Even when large sponges are sparse, lobsters still elicit an avoidance response that reduces transmission risk (Dolan et al. 2014, Butler et al. 2015). Considering that shelter availability was not a predictor for PaV1 infection, it is surprising that lobster density contributed to variations in disease prevalence since lobsters typically avoid sheltering with infected conspecifics, regardless of how densely populated an area. In this study, the mean prevalence of PaV1 by qPCR analysis

(63.2 ± 5.6 %) was relatively high with one site reaching 100%. Previous findings revealed that healthy and infected lobsters sometimes occupy the same shelter in areas with high PaV1 prevalence (Briones-Fourzán et al. 2012). One possible explanation for this, beyond contact transmission, is the high proportion of infected lobsters inhabiting a common area. Even if lobsters are not cohabitating (and transmitting the virus through contact), the virus is still likely being shed into the water column by infected lobsters. Given that PaV1 is also spread by

waterborne transmission (Butler et al. 2008), isolated lobsters may still be exposed to the virus, especially small lobsters. In the present study, a large proportion of lobsters in the sampling sites were small juveniles. On average, lobsters were 39.4 ± 0.56 mm in length, which means that many were highly susceptible to PaV1 compared to a population of larger lobsters (Butler et al.

2008, Chapter 4). This is supported by the results of the present study, confirming that small lobsters were more likely to contract PaV1 (Table 5-3).

Though not significant, habitat did have a marginal effect on the number of PaV1-infected lobsters. Of the six variables measured, three (gorgonians, algae, and seagrass) had a positive but non-significant relationship with PaV1 infection. Sites with higher algal coverage, seagrass coverage, and gorgonian abundance tended to contain a higher proportion of infected lobsters.

For areas with more sponges and hard corals, there were fewer PaV1 cases. This trend, though not significant, is consistent with prior work which showed that shelter-limited habitats can have high levels of PaV1 (Briones-Fourzán et al. 2012). In the present study, sites with a high algal and seagrass coverage possibly followed a similar trend because these areas had fewer sponges and hard corals yet higher PaV1 prevalence.

These findings are corroborated and contradicted by prior work. Few studies have evaluated the interactions between marine pathogens and habitat complexity. However, work modeling pathogen transmission across landscapes has shown that habitat heterogeneity can decrease the overall prevalence of a pathogen (Jones et al. 2011, Cross et al. 2013). In the present study, habitat containing relatively high densities of sponges (high-quality habitat) may have had a marginal effect on reducing PaV1 levels whereas algae- and seagrass-dominated habitat did not. When using a simulation model, Cross et al. (2013) determined that areas with low habitat quality experienced a decline in host population size (due to resource limitation) and disease

prevalence. However, the present study showed that PaV1 infections were more common in what may be considered low quality juvenile lobster habitat (e.g., sites containing few sponges and covered predominantly by algae or seagrass). When considering these discrepancies in findings, it is important to recognize that the present study was conducted in the marine environment and the other (Cross et al. 2013) was a terrestrial-based simulation. Few studies have evaluated the direct effect of marine habitats on the prevalence of pathogens, but a number of studies have acknowledged the correlation between habitat loss, disease, and declining populations of host species, such as the eastern oyster, Crassostrea virginica (Wilson et al. 1992, Harvell et al. 1999,

Wilberg et al. 2011). No causal relationship has yet to be discovered between habitat degradation and disease prevalence in marine environments, underscoring the importance of the present study.

While the link between habitat quality and the disease prevalence and distribution of disease remains to be understood, the interaction between environmental change and disease outbreaks has become increasingly clearer. In a recent study, Oliver et al. (2017) described the consequences of elevated temperatures on marine species in the Tasman Sea. During this anomalous warming event, oyster communities suffered the first outbreak Pacific Oyster

Mortality Syndrome (POMS). Coral reefs respond similarly to rising temperatures, as exemplified by Bruno et al. (2007). During warm temperature anomalies, scleractinian corals on the Great Barrier Reef became more susceptible to white syndrome. High densities of corals had the highest frequency of infection, consistent with previous results linking host density and disease prevalence (Easson et al. 2013).

In the present study, elevated levels of PaV1 were associated with increasing lobster densities. Following the cyanobacteria blooms in the early 1990s, sponge communities

throughout Florida Bay suffered massive die-offs (Butler et al. 1995). Juveniles that would have otherwise inhabited the sponges were displaced because the shelter-forming sponges were no longer available. Consequently, an increasing number of lobsters sought shelter in other structures (e.g., solution holes, artificial structures) (Butler et al. 2005). In the present study, areas with a higher abundance of sponges contained fewer infected lobsters, which may be explained by an increase in the number of dens with solitary occupants. In a sponge-limited habitat, density-dependent transmission of PaV1 may be driven by the cohabitation of healthy and infected lobsters in solution holes. The next step in identifying drivers of PaV1 would be to compare the number of infected lobsters with abiotic factors (e.g., temperature, directionality of currents, etc.).

Table 5-1. Summary of lobster surveys. PaV1 prevalence includes all clinically- and subclinically-infected lobsters.

Site # of lobsters PaV1 prevalence (%) Lobster density (collected/min) Tavernier Key 23 61 0.38 Don 11* 48 25 0.80 Don 14* 31 90 0.52 Don 8* 26 58 0.43 Long Key Lookout 30 57 0.50 Long Key State Park 25 68 0.42 Don 1* 12 42 0.20 Don 7* 12 42 0.20 Don 2* 24 79 0.40 Don 3* 19 100 0.55 Don 5* 33 55 0.55 Litho5 23 87 0.38 Porpoise Key 21 57 0.35 Venture Key 92 98 1.52 Gopher Key 44 48 0.73 Cudjoe Bay 10 0 0.17 Don 20* 28 82 0.47 Don 21* 9 89 0.15 Little Swash Key 19 68 0.32 Lakes Passage 41 58 0.68 *denotes sites from annual monitoring program

Table 5-2. Summary of community characteristics.

Site Total sponge Large sponge Gorgonian Hard coral Seagrass Algal density (/m2) density (/m2) density density cover cover (/m2) (/m2) (%) (%) Tavernier Key 1.96 0.32 0.08 0 2.78 9.72 Don 11* 1.45 0.06 0.06 0.04 10.8 46.1 Don 14* 0.39 0.01 0.11 0.02 3.61 51.7 Don 8* 0.1 0.01 1.35 0 25 27.8 Long Key Lookout 0.17 0.01 0 0.01 3.61 25 Long Key State 0.4 0.14 0.23 0 34.2 8.61 Park Don 1* 1.36 0.07 0.96 0.01 4.72 56.7 Don 7* 0.86 0.08 0.02 0 7.5 26.7 Don 2* 0.9 0.04 0.02 0.01 0 33.3 Don 3* 1 0.14 3.55 0 15 27.2 Don 5* 1.9 0.39 0.22 0 4.17 6.39 Litho5 0.66 0.32 0.07 0.01 18.3 53.3 Porpoise Key 0.64 0.12 0.04 0 6.39 5 Venture Key 1 0.2 0 0.02 4.72 20.8 Gopher Key 0.32 0.16 0 0 30.6 39.2 Cudjoe Bay 0.98 0.05 0 0 4.72 15.6 Don 20* 0.92 0.16 0 0 16.1 58.9 Don 21* 0.43 0.13 0 0 31.4 88.3 Little Swash Key 0.66 0.3 0.02 0 0 96.1 Lakes Passage 0.95 0.38 0.01 0.02 36.7 0 *denotes sites from annual monitoring program

Table 5-3. Results of a generalized logistic mixed model. All variables were fixed, with the exception of site. Source Estimate Standard Error z value P Lobster Density 3.0334 1.0606 2.860 0.0042* Lobster Size -0.0383 0.0097 -3.964 <0.001* Number of Blocks 0.0047 0.0184 0.255 0.7991 PC Score 1 0.2632 0.2251 1.169 0.2425 PC Score 2 -0.022 0.0222 -0.994 0.3204 *represents significant p values

Figure 5-1. Linear regression of lobsters infected with PaV1 and the lobster density at each sampling site (n = 20).

Figure 5-2. Distribution of sampling sites (n = 20) and the corresponding prevalence of PaV1 as determined by qPCR. The lower, middle, and upper regions of the Florida Keys are shown in panels A, B, and C, respectively.

Figure 5-3. Sponge density and species diversity of sampling sites (n = 7) across the lower region of the Florida Keys. The diameter of each marker reflects the sponge density based on the total abundance of all sponge species. The pie charts only reflect the diversity of large sponges (≥ 20 cm). S. vesparium (shown in blue) has the highest abundance of large sponges.

Figure 5-4. Sponge density and species diversity of sampling site (n = 7) across the middle region of the Florida Keys. The diameter of each marker reflects the sponge density based on the total abundance of all sponge species. The pie charts only reflect the diversity of large sponges (≥ 20 cm). S. vesparium and I. campana account for a high proportion of large sponges in this region.

Figure 5-5. Sponge density and species diversity of sampling site (n = 6) across the upper region of the Florida Keys. The diameter of each marker reflects the sponge density based on the total abundance of all sponge species. The pie charts only reflect the diversity of large sponges (≥ 20 cm). Ircinia sp. accounts for many of the large sponges, especially in the westernmost locations of this region. The green area represents the uppermost region of the Florida Keys.

Figure 5-6. Scree plot generated from the PCA. Variance, or eigenvalues, are displayed on the y- axis and the principal components are on the x-axis. The arrow represents the “elbow” of the scree, which corresponds two axes, or components, ultimately used to calculate PC scores.

Figure 5-7. Biplot of the first two principal components. The red arrows depict factor loadings, calculated previously using eigenvalues and eigenvectors. The plot is separated into four quadrats by the blue intersecting lines.

CHAPTER 6 CONCLUSIONS

The objective of this project was to investigate interactions between PaV1, its lobster host, and the environment. To determine and quantity these interactions, a TaqMan quantitative real-time (qPCR) assay was developed for PaV1 using representative haplotypes from Florida

(USA) and from across the Caribbean (Chapter 2). Our qPCR assay had a diagnostic sensitivity and specificity of 100% and 84%, respectively. This highly sensitive and specific assay has extensive applications given that it is also highly efficient and cost-effective. Not only can the assay be used to diagnose PaV1 in lobsters, but the tool can also be applied to comparative studies. Similar to previous work done on White Spot Syndrome Virus (WSSV) in giant tiger prawns (Jeswin et al. 2015), our assay can be used to compare the viral load of PaV1 in different tissues of infected lobsters. Another future application of our assay would be to measure levels of

PaV1 in lobsters co-infected with other pathogens (e.g., Vibrio spp.; Porter et al. 2001). In the present study, qPCR was used to help determine the following: the viability of PaV1, the relationship between PaV1 progression and lobster size, as well as the role that habitat may play in the spatial epidemiology of PaV1.

The goal of Chapter 3 was to determine at which point PaV1 loses its infectivity. Because

PaV1 is a non-enveloped virus, it would likely remain viable longer than enveloped viruses. The enveloped WSSV persists in seawater for up to 12 d (Satheesh Kumar et al. 2013), so it was hypothesized that PaV1 would be infectious for ± 14 d. To test this hypothesis, we inoculated artificial seawater one time with purified PaV1. Next, we added PaV1-negative lobsters to the water at regular intervals before removing the lobsters following a 14-day exposure period. After retesting every lobster for PaV1 using qPCR, we discovered that PaV1 was still detectable 21 d post-inoculation of the water. Therefore, it is possible that PaV1 remains viable for 16 d longer

than originally thought. Previous studies concluded that, based on the assumption that PaV1 is infectious for 5 d, waterborne transmission does not drive the Caribbean-wide distribution of

PaV1. Instead, postlarval lobsters contract and are capable of spreading the PaV1 virus across the greater Caribbean (Kough et al. 2015). However, since our findings reveal that PaV1 is potentially infectious beyond 5 d, it is possible that waterborne transmission plays a greater role in the dispersal of PaV1 throughout the Caribbean region. When considering the viability and movement of PaV1, temperature should also be taken into account. Since temperature and PaV1 progression are positively linked, one possible future direction for this study would be to determine how increasing sea surface temperatures affects the viability of PaV1.

In Chapter 4, the objective was to determine the relationship between infectious PaV1 and lobster size. Using qPCR to measure the response of juvenile lobsters to various viral loads of PaV1, it was hypothesized that both small and large juveniles would have a high viral count when inoculated with high viral load. Moreover, the large lobsters would collectively have a lower viral count. Two size classes of juvenile lobsters (small and large) were screened using qPCR and then injected with diluted or undiluted hemolymph collected from clinically-infected lobsters. All lobsters, excluding those among the control groups, tested positive for PaV1. The small lobsters had a higher overall net increase in viral count as compared to that of the large lobsters thus supporting our hypothesis. The small lobsters also suffered higher mortality relative to their larger conspecifics. The high viral count associated with small lobsters may explain why

PaV1 infections are size-dependent and are, in particular, fatal to small juveniles. In the future, we would recommend repeating this study with other modes of PaV1 transmission (e.g., waterborne, contact) to simulate natural infections.

The goal of Chapter 5 was to identify the factors that drive geographical variability in the prevalence of PaV1. Preliminary data showed that hard-bottom habitats had a higher rate of transmission than seagrass-dominated habitats. It was hypothesized that sponges, characteristic organisms of hard-bottom substrate, would explain the associated levels of PaV1. Twenty locations across the Florida Keys were surveyed to measure PaV1 prevalence and the following parameters: sponges, corals, algae, seagrass, and lobster density. Lobsters were collected by divers and subsequently tested using qPCR to determine the prevalence of PaV1. Belt transect surveys were conducted to determine the size and abundance of sponges. Using principal component analysis (PCA) and a generalized logistic mixed model, lobster size and density explained the observed variation in PaV1 prevalence across the twenty survey sites. The density of sponges and hard corals did not have an effect on PaV1 prevalence because healthy lobsters avoid cohabitation with diseased conspecifics, even when the availability of shelter-forming structures is limited. However, habitat covered predominantly by seagrass, algae, and gorgonians may have had a marginal effect on the prevalence of PaV1. A next step for this study would be to record the environmental variables at these survey sites since fluctuations in temperature, for example, may cause variability in PaV1 prevalence.

By developing a qPCR assay for PaV1, there is now a greater understanding of the virus and how it interacts with its host and environment. The results of this comprehensive study may be used, in the case of PaV1 viability, to inform models that predict the movement patterns of

PaV1. To manage the PaV1 virus and to mitigate its effect on lobster populations, it is necessary to continue studying the “epidemiological triangle” of PaV1.

LIST OF REFERENCES

Agaisse H (2007) An adaptive immune response in Drosophila? Cell Host Microbe 1:91–93

Albrecht GT, Valentin AE, Hundertmark KJ, Hardy SM (2014) Panmixia in Alaskan populations of the snow crab Chionoecetes opilio (Malacostraca: Decapoda) in the Bering, Chukchi, and Beaufort Seas. J Crustac Biol 34:31–39

Alderman DJ (1996) Geographical spread of bacterial and fungal diseases of crustaceans. Rev Sci Tech Off Int Epiz 15:603–632

Altshuler B (1981) Modeling of dose-response relationships. Environ Health Perspect 42:23– 27

Anderson JR, Behringer DC (2013) Spatial dynamics in the social lobster Panulirus argus in response to diseased conspecifics. Mar Ecol Prog Ser 474:191–200

Arala-Chaves M, Sequeira T (2000) Is there any kind of adaptive immunity in invertebrates? Aquaculture 191:247–258

Barber I, Arnott SA, Braithwaite VA, Andrew J, Huntingford FA (2001) Indirect fitness consequences of mate choice in sticklebacks: offspring of brighter males grow slowly but resist parasitic infections. Proc R Soc Lond B 268:71–76

Basset C, Holton J, O’Mahony R, Roitt I (2003) Innate immunity and pathogen-host interaction. Vaccine 21:S2/12–S2/23

Bateman KS, Munro J, Uglow B, Small HJ, Stentiford GD (2012) Susceptibility of juvenile European lobster Homarus gammarus to shrimp products infected with high and low doses of white spot syndrome virus. Dis Aquat Org 100:169–184

Behringer DC (2003) The ecological ramifications of density and disease in the Caribbean spiny lobster Panulirus argus. PhD thesis, Old Dominion University, Norfolk, VA

Behringer DC, Butler MJ IV (2006) Density-dependent population dynamics in juvenile Panulirus argus (Latreille): The impact of artificial density enhancement. J Exp Mar Biol Ecol 334:84–95

Behringer DC, Butler MJ IV (2010) Disease avoidance influences shelter use and predation in Caribbean spiny lobster. Behav Ecol Sociobiol 64:747–755

Behringer DC, Butler MJ IV, Moss J, Shields JD (2012) PaV1 infection in the Florida spiny lobster (Panulirus argus) fishery and its effects on trap function and disease transmission. Can J Fish Aquat Sci 69:136–144

Behringer DC, Butler MJ, Shields JD (2006) Avoidance of disease in social lobsters. Nature 441:421

Behringer DC, Butler MJ IV, Shields JD (2008) Ecological and physiological effects of PaV1 infection on the Caribbean spiny lobster (Panulirus argus Latreille). J Exp Mar Biol Ecol 359:26–33

Behringer DC, Butler MJ IV, Shields JD (2010) A review of the lethal spiny lobster virus PaV1 – ten years after its discovery. Proceedings of the 62nd Gulf and Caribbean Fisheries Institute

Behringer DC, Butler MJ IV, Shields JD, Moss J (2011) Review of Panulirus argus Virus 1– a decade after its discovery. Dis Aquat Org 94:153–160

Bertelsen RD, Butler MJ IV, Herrnkind WF, Hunt JH (2009) Regional characterisation of hard-bottom nursery habitat for juvenile Caribbean spiny lobster (Panulirus argus) using rapid assessment techniques. N Z J Mar Freshw Res 43:299–312.

Bly JE, Lawson LA, Szalai AJ, Clem LW (1993) Environmental factors affecting outbreaks of winter saprolegniosis in channel catfish, Ictalurus punctatus (Rafinesque). J Fish Dis 16:541–549

Bowden TJ, Thompson KD, Morgan AL, Gratacap RML, Nikoskelainen S (2007) Seasonal variation and the immune response: a fish perspective. Fish Shellfish Immunol 22:695– 706

Breyta R, Black A, Kaufman J, Kurath G (2016) Spatial and temporal heterogeneity of infectious hematopoietic necrosis virus in Pacific Northwest salmonids. Infect Genet Evol 45:347–358

Briones-Fourzán P, Candia-Zulbarán RI, Negrete-Soto F, Barradas-Ortiz C, Huchin-Mian JP, Lozano-Álvarez E (2012) Influence of local habitat features on disease avoidance by Caribbean spiny lobsters in a casita-enhanced bay. Dis Aquat Org 100:135–148

Brisbin MM, McElroy AE. Espinosa EP, Allam B (2015) Antimicrobial activity in the cuticle of the American lobster, Homarus americanus. Fish Shellfish Immunol 44:542–546.

Bruneaux M, Visse M, Gross R, Pukk L, Saks L, Vasemägi A (2017) Parasite infection and decreased thermal tolerance: impact of proliferative kidney disease on a wild salmonid fish in the context of climate change. Funct Ecol 31:216–226

Brunner JL, Richards K, Collins JP (2005) Dose and host characteristics influence virulence of ranavirus infections. Oecologia 144:399–406

Bruno JF, Selig ER, Casey KS, Page CA, Willis BL, Harvell CD, Sweatman H, Melendy AM (2007) Thermal stress and coral cover as drivers of coral disease outbreaks. PLOS Biol 5:1220–127

Burge CA, Eakin CM, Friedman CS, Froelich B, Hershberger PK, Hofmann EE, Petes LE, Prager KC, Weil E, Willis BL, Ford SE, Harvell CD (2014) Climate change influences on marine infectious diseases: implications for management and society. Ann Rev Mar Sci 6:249–277

Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, Mueller R, Nolan T, Pfaffl MW, Shipley GL, Vandesompele J, Wittwer CT (2009) The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem 55:611-622

Butler MJ IV, Behringer DC, Dolan TW III, Moss J, Shields JD (2015) Behavioral immunity suppresses an epizootic in Caribbean spiny lobsters. PLOS ONE DOI: 10.1371/journal.pone.0126374

Butler MJ, Behringer DC, Shields JD (2008) Transmission of Panulirus argus virus 1 (PaV1) and its effect on the survival of juvenile Caribbean spiny lobster. Dis Aquat Org 79:173– 182

Butler MJ IV, Dolan TW III, Hunt JH, Rose KA, Herrnkind WF (2005) Recruitment in degraded marine habitats: a spatially explicit, individual-based model for spiny lobster. Ecol App 15:902–918

Butler MJ IV, Herrnkind WF (1991) Effect of benthic microhabitat cues on the metamorphosis of pueruli of the spiny lobster Panulirus argus. J Crustac Biol 11:23–28

Butler MJ IV, Hunt JH, Herrnkind WF, Childress MJ, Bertelsen R, Sharp W, Matthews T, Field JM, Marshall HG (1995) Cascading disturbances in Florida Bay, USA: cyanobacteria blooms, sponge mortality, and implications for juvenile spiny lobsters Panulirus argus. Mar Ecol Prog Ser 129:119–125

Candia-Zulbarán RI, Briones-Fourzán P, Lozano-Álvarez E, Barradas-Ortiz C, Negrete-Soto F (2015) Caribbean spiny lobsters equally avoid dead and clinically PaV1-infected conspecifics. ICES J Mar Sci 72:i164–i169

Castro KM, Cobb JS, Gomez-Chiarri M, Tlusty M (2012) Epizootic shell disease in American lobsters Homarus americanus in southern New England: past, present and future. Dis Aquat Org 100:149–158

Cattell RB (1966) The scree test for the number of factors. Multivariate Behav Res 1:245– 276

Centurioni LR, Niiler PP (2003) On the surface currents of the Caribbean Sea. Geophys Res

100

Lett 30:1279

Chaikeeratisak V, Tassanakajon A, Armstrong PB (2014) Interaction of pathogenic Vibrio bacteria with the blood clot of the Pacific white shrimp, Litopenaeus vannamei. Biol Bull 226:102–110

Chambers G, Millan D, Cuschieri K, Cubie HA, Graham SV (2014) Assessing the detection of human papillomavirus late mRNA in liquid base cytology samples for risk stratification of cervical disease. J Med Virol 86:627–633

Chang PS, Chen HC, Wang YC (1998) Detection of white spot syndrome associated baculovirus in experimentally infected wild shrimp, crab and lobsters by in situ hybridization. Aquaculture 164:233–242

Chiappone M, Sullivan KM (1994) Ecological structure and dynamics of nearshore hard- bottom communities in the Florida Keys. Bull Mar Sci 54:747–756

Childress MJ (1995) The ontogeny and evolution of gregarious behavior in juvenile Caribbean spiny lobster, Panulirus argus. PhD thesis, Florida State University, Tallahassee, FL

Chinchar VG, Hyatt A, Miyazaki T, Williams T (2009) Family Iridoviridae: poor viral relations no longer. In: James L. Van Etten (ed) Lesser known large dsDNA viruses (current topics in microbiology and immunology). Springer Publishing Company, New York, NY, p 123–170

Chistoserdov AY, Quinn RA, Gubbala SL, Smolowitz R (2012) Bacterial communities associated with lesions of shell disease in the American Lobster, Homarus americanus Milne-Edwards. J Shellfish Res 31:449–462

Chou PH, Chang HS, Chen IT, Lin HY, Chen YM, Yang HL, Wang HC (2009) The putative invertebrate adaptive immune protein Litopenaeus vannamei Dscam (LvDscam) is the first reported Dscam to lack a transmembrane domain and cytoplasmic tail. Dev Comp Immunol 33:1258–1267

Clark KF, Greenwood SJ, Acorn AR, Byrne PJ (2013) Molecular immune response of the American lobster (Homarus americanus) to the White Spot Syndrome Virus. J Invertebr Pathol 114:298–308

Claverie JM, Abergel C, Ogata H (2009) Mimivirus. In: James L. Van Etten (ed) Lesser known large dsDNA viruses (current topics in microbiology and immunology). Springer Publishing Company, New York, NY, p 89–121

Colling G, Matthies D (2004) The effects of plant population size on the interactions between the endangered plant Scorzonera humilis, a specialized herbivore, and a phytopathogenic fungus. Oikos 105:71–78

101

Cooke BD, Fenner F (2002) Rabbit haemorrhagic disease and the biological control of wild rabbits, Oryctolagus cuniculus, in Australia and New Zealand. Wildl Res 29:689–706

Cross PC, Caillaud D, Heisey DM (2013) Underestimating the effects of spatial heterogeneity due to individual movement and spatial scale: infectious disease as an example. Landscale Ecol 28:247–257

Crowden AE, Broom DM (1980) Effects of the eyefluke, Diplostomum spathaceum, on the behavior of dace (Leuciscus leuciscus). Anim Behav 28:287–294

Dolan TW III, Butler MJ IV, Shields JD (2014) Host behavior alters spiny lobster – viral disease dynamics: a simulation study. Ecology 95:2346–2361

Duan H, Jin S, Zhang Y, Li F, Xiang J (2014) Granulocytes of the red claw crayfish Cherax quadricarinatus can endocytose beads, E. coli and WSSV, but in different ways. Dev Comp Immunol 46:186–193

Dwyer G, Elkinton JS, Buonaccorsi JP (1997) Host heterogeneity in susceptibility and disease dynamics: tests of a mathematical model. Am Nat 150:685-707

Easson CG, Slattery M, Momm HG, Olson JB, Thacker RW, Gochfeld DJ (2013) Exploring individual- to population-level impacts of disease on coral reef sponges: using spatial analysis to assess the fate, dynamics, and transmission of Aplysina Red Band Syndrome (ARBS). PLOS ONE 8:e79976

Ericson L, Burdon JJ, Müller WJ (1999) Spatial and temporal dynamics of epidemics of the rust fungus Uromyces valerianae on populations of its host Valeriana salina. J Ecol 87:649–658

Felício MTD, Hald T, Liebana E, Allende A, Hugas M, Nguyen-The C, Johannessen GS, Niskanen T, Uyttendaele M, McLauchlin J (2015) Risk ranking of pathogens in read-to- eat unprocessed foods of non-animal origin (FoNAO) in the EU: initial evaluation using outbreak data (2007-2011). Int J Food Microbiol 195:9–19

Field RH, Chapman CJ, Taylor AC, Neil DM, Vickerman K (1992) Infection of the Norway lobster Nephrops norvegicus by a Hematodinium-like species of dinoflagellate on the west coast of Scotland. Dis Aquat Org 13:1–15

Firquet S, Beaujard S, Lobert PE, Sané F, Caloone D, Izard D, Hober D (2015) Survival of enveloped and non-enveloped viruses on inanimate surfaces. Microbes Environ 30:140– 144

Flannery J, Rajko-Nenow P, Keaveney S, O’Flaherty V, Doré W (2013) Simulated sunlight inactivation of norovirus and FRNA bacteriophage in seawater. J Appl Microbiol 115:915–922

102

Foley K, Fazio G, Jensen AB, Hughes WHO (2014) The distribution of Aspergillus spp. opportunistic parasites in hives and their pathogenicity to honey bees. Vet Microbiol 169:203–210

Food and Agriculture Organization FishStat (2014) Retrieved April 3 2017 from www.fao.org/fishery/

Hall TA (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl Acids Symp Ser 41:95–98

Hanson LA, Rudis MR, Vasquez-Lee M, Montgomery RD (2006) A broadly applicable method to characterize large DNA viruses and adenoviruses based on the DNA polymerase gene. Virol J 3:28

Hart BL (1990) Behavioral adaptations to pathogens and parasites: five strategies. Neurosci Biobehav Rev 14:273–294

Harvell D, Aronson R, Baron N, Connell J, Dobson A, Ellner S, Gerber L, Kim K, Kuris A, McCallum H, Lafferty K, McKay B, Porter J, Pascual M, Smith G, Sutherland K, Ward J (2004) The rising tide of ocean diseases: unsolved problems and research priorities. Front Ecol Environ 2:375–382

Harvell CD, Kim K, Burkholder JM, Colwell RR, Epstein PR, Grimes DJ, Hofmann EE, Lipp EK, Osterhaus ADME, Overstreet RM, Porter JW, Smith GW, Vasta GR (1999) Emerging marine diseases – climate links and anthropogenic factors. Science 285:1505– 1510

Hauton C (2012) The scope of the crustacean immune system for disease control. J Invertebr Pathol 110:251–260.

Hellberg ME, Burton RS, Neigel JE, Palumbi SR. 2002. Genetic assessment of connectivity among marine populations. Bull Mar Sci 70:273–290

Hernroth BB, Sköld H, Wiklander K, Jutfelt F, Baden S. 2012. Simulated climate change causes immune suppression and protein damage in the crustacean Nephrops norvegicus. Fish Shellfish Immunol 33:1095–1101

Herrnkind WF, Butler MJ IV, Hunt JH, Childress M (1997) Role of physical refugia: implications from a mass sponge die-off in a lobster nursery in Florida. Mar Freshw Res 48:759–769

Hokanson RM, TerWee J, Choi I, Coates J, Dean H, Reddy DN, Wolf AM, Collisson EW (2000) Dose response studies of acute feline immunodeficiency virus PPR strain infection in cats. Vet Microbiol 76:311–327

103

Howie R, Alfa MJ, Coombs K (2008) Survival of enveloped and non-enveloped viruses on surfaces compared with other micro-organisms and impact of suboptimal disinfectant exposure. J Hosp Infect 69:368–376

Huchin-Mian JP, Briones-Fourzán P, Simá-Álvarez R, Cruz-Quintana Y, Pérez-Vega JA, Lozano-Álvarez E, Pascual-Jiménez C, Rodríguez-Canul R (2009). Detection of Panulirus argus Virus 1 (PaV1) in exported frozen tails of subadult-adult Caribbean spiny lobsters Panulirus argus. Dis Aquat Org 159–162

Huchin-Mian JP, Rodríguez-Canul R, Briones-Fourzán P, Lozano-Álvarez E (2013) Panulirus argus virus 1 (PaV1) infection prevalence and risk factors in a Mexican lobster fishery employing casitas. Dis Aquat Org 107:87–97

Janeway CA Jr, Medzhitov R (2002) Innate immune recognition. Annu Rev Immunol 20:197–216

Jeswin J, Anju A, Thomas PC, Paulton MP, Vijayan KK (2015) Analysis of viral load between different tissues and rate of progression of white spot syndrome virus (WSSV) in Penaeus monodon. Aquacult Res 46:2003–2012

Jones EO, Webb SD, Ruiz-Fons FJ, Albon S, Gilbert L (2011) The effect of landscape heterogeneity and host movement on a tick-borne pathogen. Theor Ecol 4:435–448

Katoh K, Toh H (2008) Recent developments in the MAFFT multiple sequence alignment program. Brief Bioinform 9:286–298

Kough AS, Paris CB, Behringer DC, Butler MJ IV (2015) Modelling the spread and connectivity of waterborne marine pathogens: the case of PaV1 in the Caribbean. ICES J Mar Sci 72:i139–i146

Kuehl DW, Haebler R, Potter C (1991) Chemical residues in dolphins from U.S. Atlantic coast including Atlantic bottlenose obtained during the 1987/1988 mass mortality. Chemosphere 22:1071–1084

Kurtz J, Armitage SAO (2004) Alternative adaptive immunity in invertebrates. Trends Immunol 27:493–496

Kurtz J, Franz K (2003) Innate defence: evidence for memory in invertebrate immunity. Nature 425:37–38

Laramore SE, Scarpa J, Laramore CR, Lin J (2009) Virulence variation of white spot syndrome virus in pacific white shrimp Litopenaeus vannamei. J Aquat Anim Health 21:82–90

Lavallée J, Hammell KL, Spangler ES, Cawthorn RJ (2001) Estimated prevalence of Aerococcus viridans and Anophryoides haemophila in American lobster Homarus

104

americanus freshly captured in the waters of Prince Edward Island, Canada. Dis Aquat Org 46:231–236

Le Morvan C, Troutaud D, Deschaux P (1998) Differential effects of temperature on specific and nonspecific immune defences in fish. J Exp Biol 201:165–168

Lessios HA (1988) Mass mortality of Diadema antillarum in the Caribbean: what have we learned? Ann Rev Ecol Syst 19:371–393

Lessios HA, Robertson DR, Cubit JD (1984) Spread of Diadema mass mortality through the Caribbean. Science 226:335–337

Li C, Shields JD (2007) Primary culture of hemocytes from the Caribbean spiny lobster, Panulirus argus, and their susceptibility to Panulirus argus Virus 1 (PaV1). J Invertebr Pathol 94:48–55

Li C, Shields JD, Small HJ, Reece KS, Hartwig CL, Cooper RA, Ratzlaff RE (2006) Detection of Panulirus argus Virus 1 (PaV1) in the Caribbean spiny lobster using fluorescence in situ hybridization (FISH). Dis Aquat Org 72:185–192

Li YY, Xia XA, Wu QY, Liu WH, Lin YS (2008) Infection with Hematodinium sp. in mud crabs Scylla serrata cultured in low salinity water in southern China. Dis Aquat Org 82:145–150

Liang T, Wu T, Du J, Ji H, Li Y, Gu W, Wang W (2011) Characterization of a tailless white spot syndrome virus from diseased Penaeus vannamei and Procambarus clarkii in China. Afr J Biotechnol 10:13936–13942

Loehle C (1995) Social barriers to pathogen transmission in wild animal populations. Ecology 76:326–335

Lozano-Álvarez E, Briones-Fourzán P, Ramírez-Estévez A, Placencia-Sánchez D, Huchin- Milan JP, Rodríguez-Canul R (2008) Prevalence of Panulirus argus Virus 1 (PaV1) and habitation patterns of healthy and diseased Caribbean spiny lobsters in shelter-limited habitats. Dis Aquat Org 80:95–104

Marancik DP, Wiens GD (2013) A real-time polymerase chain reaction assay for identification and quantification of Flavobacterium psychrophilum and application to disease resistance studies in selectively bred rainbow trout Oncorhynchus mykiss. FEMS Microbiol Lett 339:122–129

Martin LB, Hopkins WA. Mydlarz LD, Rohr JR (2010) The effects of anthropogenic global changes on immune functions and disease resistance. Ann N Y Acad Sci 1195:129–148

McCallum H, Harvell D, Dobson A (2003) Rates of spread of marine pathogens. Ecol Lett 6:1062–1067

105

McCallum HI, Kuris A, Harvell CD, Lafferty KD, Smith GW, Porter J (2004) Does terrestrial epidemiology apply to marine systems? Trends Ecol Evol 19:585–591

Messick GA, Jordan SJ, Van Heukelem WF (1999) Salinity and temperature effects on Hematodinium sp. in the blue crab Callinectes sapidus. J Shellfish Res 18:657–662

Meyers TR, Botelho C, Koeneman TM, Short S, Imamura K (1990) Distribution of bitter crab dinoflagellate syndrome in southeast Alaskan Tanner crabs Chionoecetes bairdi. Dis Aquat Org 9:37–43

Meyers TR, Koeneman TM, Botelho C, Short S (1987) Bitter crab disease: a fatal dinoflagellate infection and marketing problem for Alaskan Tanner crabs Chionoecetes bairdi. Dis Aquat Org 3:195–216

Mogensen TH (2009) Pathogen recognition and inflammatory signaling in innate immune defenses. Clin Microbiol Rev 22:240–273

Montgomery-Fullerton MM, Cooper RA, Kauffman KM, Shields JD, Ratzlaff RE (2007) Detection of Panulirus argus virus 1 in Caribbean spiny lobsters. Dis Aquat Org 76:1–6

Mori K, Stewart JE (2006) Immunogen-dependent quantitative and qualitative differences in phagocytic responses of the circulating hemocytes of the lobster Homarus americanus. Dis Aquat Org 69:197–203

Moss J, Behringer D, Shields JD, Baeza A, Aguilar-Perera A, Bush PG, Dromer C, Herrera-Moreno A, Gittens L, Matthews TR, McCord MR, Schärer MT, Reynal L, Truelove N, Butler MJ (2013) Distribution, prevalence, and genetic analysis of Panulirus argus virus 1 (PaV1) from the Caribbean Sea. Dis Aquat Org 104:129–140

Moss J, Butler MJ IV, Behringer DC, Shields JD (2012) Genetic diversity of the Caribbean spiny lobster virus, Panulirus argus virus 1 (PaV1), and the discovery of PaV1 in lobster postlarvae. Aquat Biol 14:223–232

Murcia C (1995) Edge effects in fragmented forests: implications for conservation. Trends Ecol Evol 10:58–62

Navarrete CD, Fessler DMT (2006) Disease avoidance and ethnocentrism: the effects of disease vulnerability and disgust sensitivity on intergroup attitudes. Evol Hum Behav 27:270–282

Newman MW, Johnson CA (1975) A disease of blue crabs (Callinectes sapidus) caused by a parasitic dinoflagellate, Hematodinium sp. J Parasitol 61:554–557

106

Nolan T, Hands RE, Bustin SA (2006) Quantification of mRNA using real-time RT-PCR. Nat Protoc 1:1559–1582

OIE (2016) Chapter 1.1.2 Principles and methods of validation of diagnostic assays for infectious diseases (World Organisation for Animal Health) http://www.oie.int/international-standard-setting/aquatic-manual/access-online/

Oliver ECJ, Benthuysen JA, Bindoff NL, Hobday AJ, Holbrook NJ, Mundy CN, Perkins- Kirkpatrick SE (2017) The unprecedented 2015/16 Tasman Sea marine heatwave. Nat Commun 8:16101

Park AW (2012) Infectious disease in animal metapopulations: the importance of environmental transmission. Ecol Evol 2:1398–1407

Perkins TE, Matlack GR (2002) Human-generated pattern in commercial forests of southern Mississippi and consequences for the spread of pests and pathogens. For Ecol Manage 157:143–154

Plantegenest M, Le May C, Fabre F (2007) Landscape epidemiology of plant diseases. J R Soc Interface 4:963–972

Polymenakou PN, Mandalakis M, Stephanou EG, Tselepides A (2008) Particle size distribution of airborne microorganisms and pathogens during an intense African dust event in the Eastern Mediterranean. Environ Health Perspect 116:292–296

Porter L, Butler MJ, Reeves RH (2001) Normal bacterial flora of the spiny lobster Panulirus argus and its possible role in shell disease. Mar Freshw Res 52:1401–1405

Powell EN, Gauthier JD, Wilson EA, Nelson A, Fay RR, Brooks JM (1992) Oyster disease and climate change. Are yearly changes in Perkinsus marinus parasitism in oysters (Crassostrea virginica) controlled by climatic cycles in the Gulf of Mexico? Mar Ecol 13:243–270

Prior S, Browdy CL, Shepard EF, Laramore R, Parnell PG (2003) Controlled bioassay systems for determination of lethal infective doses of tissue homogenates containing Taura syndrome or white spot syndrome virus. Dis Aquat Org 54:89–96

Quinn RA, Metzler A, Smolowitz RM, Tlusty M, Chistoserdov AY (2012) Exposures of Homarus americanus shell to three bacteria isolated from naturally occurring epizootic shell disease lesions. J Shellfish Res 31:485-493

Raoult D, Audic S, Robert C, Abergel C, Renesto P, Ogata H, La Scola B, Suzan M, Claverie JM (2004) The 1.2-Megabase Genome Sequence of Mimivirus. Science 306:1344–1350

Real LA, Biek R (2007) Spatial dynamics and genetics of infectious diseases on heterogeneous landscapes. J R Soc Interface 4:935–948

107

Robblee MB, Barber TR, Carlson PR Jr, Durako MJ, Fourqurean JW, Muehlstein LK, Porter D, Yarbro LA, Zieman RT, Zieman JC (1991) Mass mortality of the tropical seagrass Thalassia testudinum in Florida Bay (USA). Mar Eco Prog Ser 71:297–299

Rodríguez-Rey GT, Solé-Cava AM, Lazoski C (2014) Genetic homogeneity and historical expansions of the slipper lobster, Scyllarides brasiliensis, in the south-west Atlantic. Mar Freshwat Res 65:59–69

Satheesh Kumar S, Ananda Bharathi R, Rajan JJS, Alavandi SV, Poornima M, Balasubramanian CP, Ponniah AG (2013) Viability of white spot syndrome virus (WSSV) in sediment during sun-drying (drainable pond) and under non-drainable pond conditions indicated by infectivity to shrimp. Aquaculture 402–403:119–126

Scholthof KBG (2007) The disease triangle: pathogens, the environment and society. Nat Rev Microbiol 5:152–156

Shields JD, Behringer DC (2004) A new pathogenic virus in the Caribbean spiny lobster Panulirus argus from the Florida Keys. Dis Aquat Org 59:109–118

Shields JD, Squyars CM (2000) Mortality and hematology of blue crabs, Callinectes sapidus, experimentally infected with the parasitic dinoflagellate Hematodinium perezi. Fish Bull 98:139–152

Shridhar S, Hassan K, Sullivan DJ, Vasta GR, Fernandez Robledo JA (2013) Quantitative assessment of the proliferation of the protozoan parasite Perkinsus marinus using a bioluminescence assay for ATP content. Int J Parasitol 3:85–92

Smith BP, Habasha F, Reina-Guerra M, Hardy AJ (1979) Bovine salmonellosis: experimental production and characterization of the disease in calves, using oral challenge with Salmonella typhimurium. Am J Vet Res 40:1510–1513

Smolowitz R, Chistoserdov AY, Hsu A (2005) A description of the pathology of epizootic shell disease in the American lobster, Homarus americanus. J Shellfish Res 24:749–756

Smolowitz R, Quinn RA, Cawthorn RJ, Summerfield RL, Chistoserdov AY (2014) Pathology of two forms of shell disease of the American lobster Homarus americanus Milne Edwards in Atlantic Canada. J Fish Dis 37:577–581

Snieszko SF, Taylor CC (1947) A bacterial disease of the lobster (Homarus americanus). Science 105:500

Staggemeier R, Heck TMS, Demoliner M, Ritzel RGF, Röhnelt NMS, Girardi V, Venker CA, Spilki FR (2017) Enteric viruses and adenovirus diversity in waters from 2016 Olympic venues. Sci Total Environ 586:304–312

108

Stebbing PD, Pond MJ, Peeler E, Small HJ, Greenwood SJ, Verner-Jeffreys D (2012) Limited prevalence of gaffkaemia (Aerococcus viridans var. homaria) isolated from wild-caught European lobsters Homarus gammarus in England and Wales. Dis Aquat Org 100:159–167

Stentiford GD, Ramilo A, Abollo E, Kerr R, Bateman KS, Feist SW, Bass D, Villalba A (2017) Hyperspora aquatica n.gn., n.sp. (Microsporidia), hyperparasitic in Marteilia cochillia (Paramyxida), is closely related to crustacean-infecting microsporidian taxa. Parasitology 144:186–199

Stewart JE, Arie B (1973) Depletion of glycogen and adenosine triphosphate as major factors in the death of lobsters (Homarus americanus) infected with Gaffkya homari. Can J Microbiol 19:1103–1110

Stewart JE, Cornick JW, Spears DI, McLeese DW (1966) Incidence of Gaffkya homari in natural lobster (Homarus americanus) populations of the Atlantic region of Canada. J Fish Res Board Can 23:1325–1330

Subramaniam K, Gotesman M, Smith CE, Steckler NK, Kelley KL, Groff JM, Waltzek TB (2016) Megalocytivirus infection in cultured Nile tilapia Oreochromis niloticus. Dis Aquat Org 119:253–258

Sussman M, Loya Y, Fine M, Rosenberg E (2003) The marine fireworm Hermodice carunculata is a winter reservoir and spring-summer vector for the coral-bleaching pathogen Vibrio shiloi. Environ Microbiol 5:250–255

Syed Musthaq S, Sudhakaran R, Balasubramanian G, Sahul Hameed AS (2006) Experimental transmission and tissue tropism of white spot syndrome virus (WSSV) in two species of lobsters, Panulirus homarus and Panulirus ornatus. J Invertebr Pathol 93:75–80

Takagishi N, Yoshinaga T, Ogawa K (2009) Effect of hyposalinity on the infection and pathogenicity of Miamiensis avidus causing scuticociliatosis in olive flounder Paralichthys olivaceus. Dis Aquat Org 86:175–179

Takahashi H, Takahashi M, Ohshima C, Izawa Y, Uema M, Kuda T, Kimura B, Noda M (2016) Differences in the viability of murine norovirus in different aquatic locations. Mar Pollut Bull 112:313–317

Tan WGH, Barkman TJ, Chinchar VG, Essani K (2004) Comparative genomic analyses of frog virus 3, type species of the genus Ranavirus (family Iridoviridae). Virology 323:70- 84

Taylor DM, Khan RA (1995) Observations on the occurrence of Hematodinium sp. (Dinoflagellata: Syndinidae), the causative agent of bitter crab disease in Newfoundland snow crab (Chionoectes opilio). J Invertebr Pathol 65:283–288

109

Tellier MS, Bertelsen R, Hunt J, Butler M, Matthews T (2008) Monitoring the flora and fauna of the nearshore hardbottom habitats of the Florida Keys. Florida’s Wildlife Legacy Initiative (FWRI) File Code: F2196-05-08-F

Tlusty MF, Metzler A (2012) Relationship between temperature and shell disease in laboratory populations of juvenile American lobsters (Homarus americanus). J Shellfish Res 31:533–541

Van Abel N, Schoen ME, Kissel JC, Meschke JS (2017) Comparison of risk predicted by multiple norovirus dose-response models and implications for quantitative microbial risk assessment. Risk Anal 37:245–264

Vander Wal E, Garant D, Calmé S, Chapman CA, Festa-Bianchet M, Millien V, Rioux- Paquette S, Pelletier F (2014) Applying evolutionary concepts to wildlife disease ecology and management. Evol Appl 7:856–868

Vazquez L, Alpuche J, Maldonado G, Agundis C, Pereyra-Morales A, Zenteno E (2009) Immunity mechanisms in crustaceans. Innate Immun 15:179–188

Waltzek TB, Miller DL, Gray MJ, Drecktrah B, Briggler JT, MacConnell B, Hudson C, Hopper L, Friary J, Yun SC, Malm KV, Weber ES, Hedrick RP (2014) New disease records for hatchery-reared sturgeon. I. Expansion of frog virus 3 host range into Scaphirhynchus albus. Dis Aquat Org 111:219–227

Ward JR, Lafferty KD (2004) The elusive baseline of marine diseases: are diseases in ocean ecosystems increasing? PLOS Biol 2:542–547

Watson LR, Milani A, Hedrick RP (1998) Effects of water temperature on experimentally- induced infections of juvenile white sturgeon (Acipenser transmontanus) with the white sturgeon iridovirus (WSIV). Aquaculture 166:213–228

Watts M, Munday BL, Burke CM (2001) Immune responses of teleost fish. Aust Vet J 79:570–574

Whitten MMA, Davies CE, Kim A, Tlusty M, Wootton EC, Chistoserdov A, Rowley AF (2014) Cuticles of European and American lobsters harbor diverse bacterial species and differ in disease susceptibility. MicrobiologyOpen 3:395–409

Wilberg MJ, Livings ME, Barkman JS, Morris BT, Robinson JM (2011) Overfishing, disease, habitat loss, and potential extirpation of oysters in upper Chesapeake Bay. Mar Ecol Prog Ser 436:131–144

Wilson EA, Powell EN, Wade TL, Taylor RJ, Presley BJ, Brooks JM (1992) Spatial and temporal distributions of contaminant body burden and disease in Gulf of Mexico oyster

110

populations: the role of local and large-scale climatic controls. Helgol Meeresunters 46:201–235

Xia X, Wang J, Ji J, Zhang J, Chen L, Zhang R (2015) Bacterial communities in marine aerosols revealed by 454 pyrosequencing of the 16S rRNA gene. J Atmos Sci 72:2997– 3008

Yanjiao Z, Kangsen M, Hongming MA, Qinghui AI, Wenbing Z, Wei X (2011) Rearing in intermediate salinity enhances immunity and disease-resistance of turbot (Scophthalmus maximus L.). Acta Oceanol Sin 30:122–128

111

BIOGRAPHICAL SKETCH

Abigail was born in Boston, MA. She spent most of her childhood in New Hampshire before returning to Boston where she received a Bachelor of Science in biology at Emmanuel

College. Abigail graduated with distinction in the field of biology after having completed a research internship at the New England Aquarium (Boston). This introduction to American lobster, Homarus americanus, research inspired her to pursue graduate studies at the University of New Hampshire (Durham, NH). Under the supervision of Dr. Winsor Watson, Abigail continued her research of the American lobster eventually awarding her a Master of Science in zoology. Her master’s thesis research determined how well catch by standard, vented traps and ventless traps correspond to the relative abundance of lobster populations. She, simultaneously, characterized trap saturation of said trap types using an underwater video surveillance system.

Next, Abigail advanced her lobster research at the University of Florida (Gainesville, FL). Under the direction of Dr. Donald Behringer, she studied the Caribbean spiny lobster, Panulirus argus, and its associated virus, Panulirus argus Virus 1 (PaV1). She earned a Doctorate of Philosophy in Fisheries and Aquatic Sciences through her dissertation work on the disease ecology of PaV1.

112