Ensis Leei M

The University of Maine DigitalCommons@UMaine

Electronic Theses and Dissertations Fogler Library

Fall 12-20-2019

The Effects of Sediment Acidification and emperT ature on the Immune Capacity of the Atlantic Jackknife (Razor) Clam (Ensis leei M. Huber, 2015)

Brian Preziosi University of Maine, [email protected]

Follow this and additional works at: https://digitalcommons.library.umaine.edu/etd

Part of the Aquaculture and Fisheries Commons, Biogeochemistry Commons, and the Marine Biology Commons

Recommended Citation Preziosi, Brian, "The Effects of Sediment Acidification and emperT ature on the Immune Capacity of the Atlantic Jackknife (Razor) Clam (Ensis leei M. Huber, 2015)" (2019). Electronic Theses and Dissertations. 3133. https://digitalcommons.library.umaine.edu/etd/3133

This Open-Access Thesis is brought to you for free and open access by DigitalCommons@UMaine. It has been accepted for inclusion in Electronic Theses and Dissertations by an authorized administrator of DigitalCommons@UMaine. For more information, please contact [email protected]. THE EFFECTS OF SEDIMENT ACIDIFICATION AND TEMPERATURE ON THE IMMUNE CAPACITY OF THE ATLANTIC JACKKNIFE (RAZOR) CLAM (ENSIS LEEI M. Huber, 2015) By

Brian Preziosi

B.A. Wheaton College, 2010

M.S. University of Maine, 2012

A DISSERTATION

Submitted in Partial Fulfillment of the

Requirements for the Degree of

Doctor of Philosophy

(in Aquaculture and Aquatic Resources)

The Graduate School

The University of Maine

December 2019

Advisory Committee:

Timothy Bowden, Associate Professor of Aquaculture, Advisor

Ian Bricknell, Professor of Aquaculture Biology

Paul Rawson, Associate Professor of Marine Science

Aria Amirbahman, Professor of Civil and Environmental Engineering

Brian Beal, Professor of Marine Ecology

THE EFFECTS OF SEDIMENT ACIDIFICATION AND TEMPERATURE ON THE IMMUNE CAPACITY OF THE ATLANTIC JACKKNIFE (RAZOR) CLAM (ENSIS LEEI M. Huber, 2015) By Brian Preziosi

Dissertation Advisor: Dr. Timothy Bowden

An Abstract of the Dissertation Presented in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy (in Aquaculture and Aquatic Resources) December 2019

Sediment acidification has been shown to negatively impact clams of economic importance such as the soft-shell clam, Mya arenaria Linnaeus, 1758, and hard clam,

Mercenaria mercenaria (Linnaeus, 1758). Effects of sediment acidification on razor clams, including the Atlantic jackknife clam Ensis leei M. Huber, 2015, are unknown. E. leei has been identified as a species with potential for aquaculture operations on the New England coast. E. leei may be resilient to acidification and thus persist in acidified sediments where other clams cannot. To this end, the impact of acidified surface sediment on the internal immune capacity of adult (mean shell length 17.6 ± 0.9 cm) E. leei at 10°C and 20°C was tested. Surface sediment was acidified from pH ~7.2 to ~6.0 using sediment bacteria. Crushed E. leei shell was used to keep the surface sediment of the ambient pH groups ~6.8. Hemocytes (circulating blood cells) are responsible for carrying out immune functions in bivalves, and thus were the target for internal immune capacity assays. Hemocyte parameters measured included cell viability, total hemocyte counts, differential hemocyte counts, and phagocytic capacity. A condition index was calculated using the wet flesh weight and dry shell weight to assess the physiological condition of the clams. In addition, these health parameters were measured in E. leei collected over the

course of a year no more than 30 hours after they were taken from the field site for seasonal trends in baseline immune capacity. The presence or absence of ripe gonads was also noted in these clams. Only total hemocyte count and percent of clams with ripe gonads correlated positively with the temperature of the collection site. A 30-day exposure period to the acidified sediment kept in recirculating aquaria had no significant impact on any of these parameters. In addition, all the clams buried into the sediments of their respective treatments within an hour of being placed there. These results indicate that the internal immune capacity of E. leei adults is not impacted by acidified surface sediments. Additional work will be needed to determine if juvenile E. leei are resilient to acidification as well.

ACKNOWLEDGMENTS

There are many people that deserve a mention here for taking the time out of their schedule to help me with this project. Firstly, I would like to thank my committee members Dr.

Timothy Bowden, Professor Ian Bricknell, Dr. Paul Rawson, Dr. Aria Amirbahman, and Dr.

Brian Beal for all the support and guidance during my research. I would also like to thank Dr.

Brian Beal for not only showing me where razor clams could be collected on Beals Island but also digging the first few batches for me. My thanks to Dana Morse, Mick Devin, and David

Cheney for assistance with acquiring razor clams around the Darling Marine Center. I would like to thank Deborah Bouchard for providing me with the fluorescent bacteria, which ended up being essential to the phagocytosis assay. Many thanks to the staff at the electronic microscopy lab as well for assisting me with the transmission electron microscopy work and the SEANET staff for providing me with some of the skills needed to write a doctoral thesis. Lastly, I would like to thank Dr. Timothy Bowden for going above and beyond the call of duty as my advisor.

iii

TABLE OF CONTENTS

ACKNOWLEDGMENTS ...... iii

LIST OF TABLES ...... ix

LIST OF FIGURES ...... x

CHAPTER 1: GENERAL INTRODUCTION ...... 1

1.1. Biology of Ensis leei ...... 1

1.1.1. Mobility Mechanisms ...... 1

1.1.2. Lifecycle...... 2

1.1.3. Feeding and Growth...... 3

1.2. Current Market Operations for Razor Clams ...... 4

1.3. Aquaculture Potential for Ensis leei...... 6

1.4. Bivalve Immunity...... 7

1.4.1. Organization of the Bivalve Immune System ...... 7

1.4.2. Mucosal Immunity ...... 7

1.4.3. Cellular Immunity...... 8

1.4.4. Humoral Immunity...... 9

1.4.5. Assessing Immune Capacity...... 9

1.5. Environmental Impacts...... 10

1.5.1. Temperature Increases...... 10

1.5.2. Sediment Acidification...... 12

1.6. Thesis Objectives...... 16

CHAPTER 2: CHARACTERIZATION OF ENSIS LEEI HEMOCYTES ...... 17

2.1. Introduction ...... 17

2.2. Methods ...... 20

2.2.1. Clam and Hemolymph Collection ...... 20

2.2.2. Total Hemocyte Counts ...... 21

2.2.3. Hemacolor and Pappenheim’s Panoptical Staining ...... 21

2.2.4. Neutral Red Staining for Lysosomes ...... 22

2.2.5. Differential Counts and Hemocyte Measurements ...... 23

2.2.6. Transmission Electron Microscopy...... 23

2.3. Results ...... 24

2.3.1. Light Microscopy ...... 24

2.3.1.1. Hyalinocytes...... 24

2.3.1.2. Granulocytes ...... 25

2.3.1.3. Vesicular Cells ...... 26

2.3.1.4. Aggregates...... 29

2.3.2. Transmission Electron Microscopy...... 31

2.3.2.1. Hyalinocytes...... 31

2.3.2.2. Granulocytes ...... 33

2.3.2.3. Vesicular Cells ...... 36

2.3.3. Total and Differential Hemocyte Counts ...... 36

2.4. Discussion ...... 37

2.5. Conclusion...... 41

CHAPTER 3: IMMUNE CAPACITY OVER THE COURSE OF A YEAR ...... 42

3.1. Introduction ...... 42

3.2. Methods...... 44

3.2.1. Sediment and Clam Collection ...... 44

3.2.2. Experimental Setup...... 45

3.2.3. Bacteria Preparation...... 46

3.2.4. Hemocyte Collection and Total Counts...... 46

3.2.5. Hemocyte Distribution and Incubation for Immune Assays...... 47

3.2.6. Differential Hemocyte Count Analysis...... 47

3.2.7. Phagocytosis Analysis...... 48

3.2.8. Condition Index Calculation and Check for Gonads...... 48

3.2.9. Statistical Analysis...... 49

3.3. Results...... 49

3.3.1. Environmental Conditions...... 49

3.3.2. Hemocyte Counts...... 52

3.3.3. Phagocytosis...... 54

3.3.4. Condition Index and Percent Ripe...... 55

3.3.5. Power Analysis...... 57

3.4. Discussion...... 57

3.5. Conclusion...... 60

CHAPTER 4: METHOD FOR MANIPULATING SEDIMENT ACIDIFICATION ...... 61

4.1. Introduction ...... 61

4.2. Methods...... 63

4.2.1. Sediment and Shell Collection...... 63

4.2.2. Water Content, Grain Size, and Percent Organic Matter of Sediment...... 65

4.2.3. Experimental Setup...... 65

4.2.4. System Monitoring and Data Analysis...... 66

4.3. Results...... 67

4.3.1. Collection Site Sediment Chemistry and Composition...... 67

4.3.2. Solution Chemistry of Aquarium Water...... 70

4.3.3. Chemistry of Sediment with No Crushed Shell...... 72

4.3.4. Chemistry of Sediment with Crushed Shell...... 77

4.4. Discussion...... 81

4.5. Conclusion...... 85

CHAPTER 5: SEDIMENT PH AND TEMPERATURE EFFECTS ON

ENSIS LEEI IMMUNE CAPACITY ...... 86

5.1. Introduction...... 86

5.2. Methods...... 88

5.2.1. Sediment and Clam Collection...... 88

5.2.2. Experimental Setup...... 89

5.2.3. System Monitoring...... 90

5.2.4. Burrowing Rates and Survival...... 91

5.2.5. Bacteria Preparation...... 93

vii

5.2.6. Hemocyte Collection and Total Counts...... 93

5.2.7. Hemocyte Distribution and Incubation for Immune Assays...... 94

5.2.8. Differential Hemocyte Count Analysis...... 94

5.2.9. Phagocytosis Analysis...... 95

5.2.10. Cell Viability Analysis...... 95

5.2.11. Condition Index...... 95

5.2.12. Statistical Analysis...... 96

5.3. Results...... 97

5.3.1. Total Hemocyte Counts...... 97

5.3.2. Differential Hemocyte Counts...... 99

5.3.3. Phagocytosis...... 101

5.3.4. Cell Viability...... 103

5.3.5. Condition Index...... 105

5.3.6. Laboratory Holding Effects on Measured Parameters...... 106

5.3.7. Power Analysis...... 108

5.4. Discussion...... 109

5.5. Conclusion...... 114

CHAPTER 6: GENERAL DISCUSSION AND CONCLUSION ...... 117

REFERENCES ...... 121

BIOGRAPHY OF THE AUTHOR ...... 135

viii

LIST OF TABLES

Table 2.1. Diameters and nucleus:cell ratio of the hemocyte types found in E. leei .... 28

Table 5.1. Summary of environmental data from collection site and

laboratory aquaria over the course of the experiment ...... 92

LIST OF FIGURES

Figure 2.1. Light micrographs of fixed and stained E. leei hemocytes...... 27

Figure 2.2. Light micrographs of live E. leei hemocytes stained with neutral red...... 28

Figure 2.3. Light micrographs of stained E. leei hemocyte aggregates...... 30

Figure 2.4. Transmission electron micrographs of E. leei hyalinocytes and

vesicular cell ...... 32

Figure 2.5. Transmission electron micrographs of E. leei granulocytes and granules. ... 34

Figure 2.6. Transmission electron micrographs of E. leei granulocytes...... 35

Figure 2.7. Population distribution of hemocyte types in E. leei hemolymph ...... 37

Figure 3.1. Temperature and salinity of the water at the collection site ...... 50

Figure 3.2. The pH and aragonite saturation state of the collection site ...... 51

Figure 3.3. Hemocyte counts of E. leei...... 53

Figure 3.4. Phagocytic activity of E. leei hemocytes...... 54

Figure 3.5. Overall condition of E. leei...... 56

Figure 4.1. Sediment and overlying water pH at collection site ...... 69

Figure 4.2. Sediment and overlying water aragonite saturation state

at collection site ...... 69

Figure 4.3. Overlying tank water pH, aragonite saturation state, and total alkalinity ..... 71

Figure 4.4. Sediment pH and aragonite saturation state (no crushed shell) ...... 73

Figure 4.5. Sediment total alkalinity (no crushed shell) ...... 75

Figure 4.6. Sediment pH and aragonite saturation state (with crushed shell) ...... 78

Figure 4.7. Mean sediment total alkalinity (TA) with crushed shell ...... 79

Figure 5.1. Total hemocyte count results ...... 98

Figure 5.2. Differential hemocyte count results ...... 100

Figure 5.3. Phagocytosis results ...... 102

Figure 5.4. Trypan blue exclusion viability test results ...... 104

Figure 5.5. Condition index results ...... 105

Figure 5.6. Health parameters of E. leei kept in ambient

conditions compared to those in acidification trial...... 107

CHAPTER 1

GENERAL INTRODUCTION

Razor clams are elongate bivalves inhabiting the subtidal and intertidal zones along open coastal areas, lagoons, and estuarine habitats (1). They are highly mobile, have thin shells, and cannot completely seal off their soft tissue from the outside environment (1). A commonality between the many species of razor clams is the growing interest in the bivalve’s potential to further its economic value via aquaculture. Commercial razor clam fisheries have shown significant fluctuations in landings each year as a function of natural stock fecundity, which are inconsistent and unpredictable (2). The inconsistent yield from natural clam beds affects the value in shellfish aquaculture around the globe. Grow out methods and technologies in rearing razor clams for the worldwide market require a greater understanding of the optimal species- specific conditions for spawning, growth, and development. Aquaculture development of New

England’s native razor clam, Ensis leei M. Huber, 2015, is showing steady progress (3) and will soon require information on how environmental conditions impact the species. The purpose of this chapter is to introduce the reader to E. leei, the major components of the bivalve immune system, and the processes driving acidification in Maine’s coastal surface sediments.

1.1. Biology of Ensis leei

1.1.1. Mobility Mechanisms

Perhaps the most unique aspect of the biology and ecology of razor clams is their mobility. E. leei is capable of burrowing, swimming, and even hopping across the surface of exposed sediment. In the Gulf of Maine, the species is most commonly found in shallow subtidal areas in sediments that are either sandy or a mix of sand and mud. Razor clams are incredibly sensitive to tactile sensations and are fast burrowers when disturbed (1). In an undisturbed state,

1 razor clams reside near the sediment surface at the top of their burrows because they have short siphons. The ability to burrow down into the sediment quickly then is crucial to the razor clam’s survival. To accomplish this task, the individual pushes its foot down into the sediment and flares the tip by filling it with hemolymph (circulatory fluid) (4). The resulting club shape at the tip of the foot is used as an anchor. When the clam starts to pull itself towards the anchor, a jet of water is shot out from between the mantle and foot into the sediment around the foot. This action makes the sediment below the foot softer and easier to navigate. The clam uses this repeated motion (rapidly retracting the foot) to fluidize the surrounding sediment, reducing drag during burrowing (1). Recently settled juvenile razor clams can migrate rapidly in the current by using their byssal thread as a sail. Both juveniles and adults can jump across tidal flats by wrapping their foot in circular fashion and quickly straightening it. The shell is relatively light that allows for this added mobility at the cost of the added protection most other clams enjoy (1).

1.1.2. Lifecycle

The reproductive cycle of the razor clam can be broken down into four general phases.

The sexual resting phase (S0) is followed by: gonad development (S1), maturation (S2), spawning (S3), and post spawning gonad exhaustion (S4) at which point spawning ceases and gonadal tissue atrophies or is resorbed (5). An understanding of razor clam reproduction is crucial for successful farming initiatives. The literature indicates that temperature, food availability, and food quality are the most influential parameters controlling reproduction (5).

Razor clams are gonochoristic, meaning an individual has a distinct sex (1). Gametogensis, the process of gonad ripening and spawning, is energetically expensive for bivalves (6). Temperature is indicated as the major parameter to induce spawning since it influences metabolic function, gonad development, and food availability (i.e. the spring diatom bloom is positively correlated

2 with reproduction) (6). Spawning is a large and synchronized event (1). E. leei usually spawns in late spring (7), but timing varies between locations. For instance, spawning occurs between

March-April in the North Sea, while it takes place during June in Canada. In general, spawning occurs later in the year in populations closer to the poles (8).

Reproduction and developmental stages are similar across bivalves that are broadcast spawners. Generally, fertilization is possible after the germinal vesicle breaks and eggs are released through the siphons. The egg continues to divide, entering into the “trochophore” stage

12-16 hours after release (9). Twenty-four hours after release, individuals reach the straight hinge “D” stage, or the early “veliger” stage, characterized by a ciliated, free-swimming, larvae enclosed by two valves composed of CaCO3 (9). At this time the larvae have a complete digestive system with a mouth, foregut, digestive gland, intestine, and anus (2). Furthermore, observations show the gradual disappearance of the “velum” (a large, ciliated membrane utilized until the D-stage for swimming, feeding, and respiration) (9). A few days later the larva reaches metamorphosis as a “pediveliger”. The pediveliger is the last larval stage as the larva then settles to the benthos and moves on to the juvenile stages as “spat” (9). Depending on temperature and species, settlement can occur anywhere between 10 days-3 weeks (1).

1.1.3. Feeding and Growth

Like most bivalves, razor clams are filter feeders. The clam draws in water through the incurrent siphon. The ingested food is controlled by the side palps that send particles to the gills for sorting, and subsequent passage to the mouth. Particles that are not suitable for digestion are bound together with mucous and sent out the excurrent siphon as pseudofeces. Ciliated cells lead digestible food from the mouth to the esophagus. In the esophagus, cells secrete a mucus that covers the food before transport to the stomach. Digestive glands, composed of counter-current

3 ducts absorb food particles and send the remaining fecal matter to the pericardium layer of the mantle for passage to the anus at the end of the posterior adductor muscle (5).

In general, razor clam development is characterized by slow (or no) growth during winter months. During the off-season, the shell grows with a dark coloration while the warmer summer months allow faster growth and create a lighter shell color. Due to differences in shell color per season, age can be counted by “growth rings,” measured using the acetate peel technique. The author notes that existence of “false annuli,” which are added disturbance lines seen on the shell.

The false annuli are caused by a variety of factors that disturb regular growth such as storm events, shell damage from predators, and extreme low tides (8,10). Growth is variable and dependent on species and location, with quicker growth recorded in areas with higher temperatures. Slower growth rates however generally lead to longer life spans and greater maximum shell length (11). The difference in growth rates may be due to food availability and warmer temperatures that occur earlier in the year and last longer in southern regions. E. leei can reach a maximum length of 20 cm and maximum age of 7 years though it only reaches this age in the colder portions of its geographic range, and even then it is rare (1,7,12). Most individuals only live to be 2-4 years (12). E. leei’s native range includes Labrador to South Carolina of

North America’s Atlantic seaboard and the introduced range includes the English Channel to

Sweden of the North Sea coastline (it was introduced through ballast water to the German Bight in 1978) (1).

1.2. Current Market Operations for Razor Clams

The literature suggests other edible shellfish species are still more widely desired, which continues to hinder razor clam farming efforts and developments. For instance, the commercial demand for oysters, mussels, crabs, and gastropods such as whelks and Triton spp. still occupy

4 much larger markets as the razor clam abundance and the wild catch is unreliable, and very variable (7).

Approximately 30% of the market for razor clams centers around the fresh product, whereas the “industrial” (or processed/canned) market averages 70% (7). This depreciates the total value of the razor clam industry since the processed market is worth less, but processed meat is in higher demand. The majority of research on razor clam species is conducted in Europe where, historically, there is a higher demand than in the U.S. Currently, the most commercially valued razor species in Europe are E. silique (Linnaeus, 1758), E. minor (Chenu, 1843), E. magnus

Schumacher 1817, and Solen capensis P. Fischer, 1881 (13). The overall European market for razor clams is worth around $60-$80 million per year (7). The primary European consumers include Spain, Italy, France, Portugal, and the Netherlands (14).

The current value of razor clams in the European market is 4-15 Euros per kg for wild caught clams. The market potential for cultivated clams is estimated to be 30-35 Euros/kg assuming the quality will be higher under controlled conditions (7). In the United States the razor clam industry (E. leei and Siliqua patula (Dixon, 1789)) is still a “work in progress.” E. leei is becoming increasingly more valued. The market price has recently exceeded $5.00 per pound for fresh caught razor clams (1). Their preferred minimum size is 3-6 inches. The value of the catch is measured by multiple factors. The least expensive clam meat is tough in texture and fished from Ireland and Scotland (7). The most expensive and best quality meat is fished from Spain, mainly Galicia. The catch originating from Portugal falls right in the middle as the meat is described as fine quality, but often the shells still contain sand. As described above, the quality of catch is determined by the region it was caught, the nature of the species including their size, and the efficiency of the purification process (7).

One of the challenges facing the razor clam industry is their rapid expiration and sensitivity to handling. One method used to help overcome this challenge is the banding of individuals together. The pressure from the banding alleviates added muscle energy the clam expends to hold its valves together, and extends shelf life up to a week (1). Other factors still limiting the razor clam industry and potential farming efforts include the difficulty in rearing and current inefficient yield rates (7). By investigating the variables limiting the razor clam industry and aquaculture developments, new methodologies may aid in establishing a greater consumer interest worldwide.

1.3. Aquaculture Potential for Ensis leei

E. leei has already been identified as a potential candidate for aquaculture since it grows quickly and local markets have the capacity for rapid expansion given a consistent supply of product (7). Development of the culturing techniques has begun at the University of Maine’s

Darling Marine Center in Walpole, ME, and at the University of Maine at Machias’ Marine

Science Field Station at the Downeast Institute. Operations start in the hatchery with wild broodstock, which are subjected to a temperature shock to induce spawning. The larvae are reared similarly to other cold-water bivalves. Once settled, they can be grown in upwellers without sediment until they reach the juvenile stage (7-10 mm SL). At this point, juveniles must be planted in the estuary since it is not cost-efficient to grow all the algae necessary for their growth in the hatchery. The grow-out method for this species is still under development as the razor clam’s mobility creates a unique challenge. A range of methods have been tried including tents, cages, nets over the seeded area, boarded raceways, and floating sand trays. Most methods were ineffective as they either allowed the clams to leave the site when no netting was used or slowed growth too much when netting was used. Burying boarded raceways into the sediment and placing

6 netting over the top of them was the most effective method; however, this only provides partial protection from subterranean predators (such as the ribbon worm, Cerebratulus lacteus (Leidy,

1851)) (3). Research is still needed to find a method that will provide adequate protection from predators, keep the clams contained, and allow for moderate water flow during grow-out.

1.4. Bivalve immunity

1.4.1. Organization of the Bivalve Immune System

The bivalve immune system consists of external physical barriers, dissolved compounds, and circulating cells (hemocytes). If microbes breach the physical barriers (periostracum, shell, mucous, and epithelium) their pathogen-associated molecular patterns (PAMPs) are detected by hemocyte-bound or dissolved pattern recognition receptors (PRRs). The PRRs trigger the release of cytokines, which recruit additional hemocytes, activate phagocytosis, and/or release of a variety of different microbial compounds (15). Note that invertebrates do not possess antibodies and therefore are not capable of the adaptive immune response found in vertebrates (16).

Invertebrates have some “immune memory,” which is thought to be achieved with the help of specific types of PRRs like thioester-containing proteins (17), C-type lectins, and Down syndrome cell adhesion molecules (18). All these parts work together to neutralize the millions of microbes bivalves encounter on a daily basis; however, each part must be looked at individually to understand why the bivalve immune system is as effective as it is.

1.4.2. Mucosal Immunity

The mucosal immunity of bivalves is a potent yet poorly understudied part of the bivalve immune system (19). Mucous is more than just a physical barrier as it also contains compounds that can kill or assist in the removal of microbes such as hydrolytic enzymes (including lysozyme and proteases) and lectins (a type of PRR) (20). The concentration of these compounds in the

7 mucous has been shown to increase after challenge via bacterial bath exposure, but not by injection (21). This indicates bivalves are capable of altering immune parameters of the internal and exterior immune portion of their immune system individually to localize immune response to infection sites. The external immune system is able to neutralize a great number of microbes on a daily basis however the internal immune system has the tools needed to remove quickly the relatively small number of pathogens that breach the external defenses (22).

1.4.3. Cellular Immunity

Cellular immunity in the bivalve immune system is carried out by hemocytes. These cells are present both in the hemolymph (circulatory fluid) and extrapallial cavity (fluid-filled space between mantle and shell), and can migrate freely between these two spaces as well as into tissue

(23). They destroy foreign or dead cells using either phagocytosis or encapsulation as well as perform basic homeostasis functions including wound healing, food digestion, transport of nutrients, reproduction, excretion, and shell formation. There are two major types of hemocytes: hyalinocytes and granulocytes. The granuloctyes are responsible for executing phagocytosis, a process in which non-self molecules are ingested and destroyed by reactive oxygen species (such as super oxide) and reactive nitrogen species (such as nitric oxide and peroxynitrite).

Hyalinocytes are also capable of phagocytosing nonself molecules, but do so with much lower frequency than granulocytes. Encapsulation is a defense method used for foreign organisms that are too large to be ingested (i.e. multicellular parasites) by phagocytosis. In this process, the invading organism is enclosed by a multitude of hemocytes. Once enclosed, the hemocytes release degradative enzymes and free radicals to kill the foreign organism (24). To prevent the damage of host tissue by reactive oxygen species release, hemocytes produce enzymes (such as superoxide dismutase and catalase) to convert reactive oxygen species to harmless end products

8 such as H2O (25). Hemocytes can also wall of parasites and other pathogens by melanization, in which hemocytes seal off pathogens through the formation of a melanin rich pigmented barrier.

This process is controlled by a diverse group of enzymes known as phenoloxidases (26).

Melanized invaders and/or abiotic irritants may then be enclosed within a layer of calcium carbonate via the process of nacrezation (27).

1.4.4. Humoral Immunity

The numerous types of dissolved immune molecules in the hemolymph, extrapallial fluid, and mucous of bivalves are all part of the humoral portion of the immune system. These immune molecules include many types of PRRs (lectins, peptidoglycan proteins, thioester bearing proteins, lipopolysaccharide and b-1,3-glucan-binding protein, fibrinogen-related proteins), antimicrobial peptides, antimicrobial proteins, hydrolytic enzymes (lysoszyme and proteases), and protease inhibitors. Each group of PRRs specializes in binding to a different type of PAMP.

When a pathogen invades the bivalve, the appropriate PRR triggers the bivalve’s signaling pathway. In turn, the signaling pathway induces the immune response and the production of effectors, which are responsible for the incapacitation and eliminations of pathogens. The antimicrobial peptides, antimicrobial proteins, and hydrolytic enzymes are all used to directly neutralize invading pathogens. The function of protease inhibitors is still debated as they may block microbe protease activity (usually used for food digestion) or block the host’s protease activity when it becomes overactive (24).

1.4.5. Assessing Immune Capacity

The ability of the immune system to neutralize pathogens is known as immune capacity. This can be measured using a variety of biomarkers including but not limited to lysozyme activity, clearance of bacteria from the hemolymph, nitric oxide concentration, phagocytosis frequency,

9 differential cell counts, total cell counts, phenoloxidase activity, superoxide dismutase activity, and super oxide concentrations. The response of any of these parameters can be difficult to measure in-vivo since they are dependent upon a variety of factors. These include the species tested, the time after exposure to the pathogen, the species of pathogen used, the challenge dosage used, season the challenge is done in, location the animals are collected from, sex of the animal, and a variety of environmental factors.

The immune response is also temporally dynamic, which means that the peak activity of each immune parameter may occur at a different number of hours post infection. Analysis at a single time point after infection gives a false impression of immune capacity if prior work has not been done to assess how immune parameters change in the short-term after infection. These complicating factors currently make in-vivo work for razor clams difficult as there are few studies available on their immune system (28,29). In-vitro infection experiments, however, can simplify many these complications since control and experimental groups can be run using hemolymph from the same animal, thus drastically reducing variability between replicates.

This is particularly true of adult razor clams since sufficient hemolymph can be drawn from them to provide multiple time points for assaying multiple immune parameters from a single animal. Once the immune capacity of the species has been established in each season, the effect of environmental parameters of immune capacity can be tested for.

1.5. Environmental Impacts

1.5.1. Temperature Increases

The average temperature of the global-ocean’s surface-waters has been increasing due to the direct and indirect effects of CO2 emissions (30). The present atmospheric CO2 level is over 380 parts per million (ppm), approximately 80 ppm more than any of the maximum CO2 values

10 measured (from ice cores) in at least the last 740,000 years (31). E. leei is native to the Gulf of

Maine (GoM), where surface waters have been warming at a rate of 0.01˚C per year from 1873 to 2001(32). An accelerated warming period then occurred from 2001 to 2011, during which temperatures at all depths in the GoM rose by at least 1˚C (33). The cause of this recent warming period is not well understood. If warming continues at the elevated rate, then E. leei will need to adapt as the average temperature in GoM in the top 20m is likely to rise by at least 5˚C by the year 2070 (ref?).

Temperature is a parameter that is both easy to measure and has a high impact on most biological processes so it should come as no surprise that there is a substantial body of literature available on its effects on bivalve immune function (34). This includes studies with (35–37) and without pathogens (38–40). Simple temperature trials show immune parameters are generally reduced in response to the stress of increasing temperatures in the absence of pathogen challenge

(34), although there are some exceptions where immune parameters are increased (35,36). A pathogen challenge however is required to determine if the selected bivalve can still produce an adequate immune response under any given environmental stressor (35).

The chosen temperature can also have an impact of the pathogen itself. This can be observed readily in a quahog parasite unknown (QPX) infection study on Mercenaria mercenaria

(Linnaeus, 1758) carried out in a variety of temperatures (13, 21, 27˚C). Challenged clams were able to fight off the infection at both 21˚C and 27˚C, but not at 13˚C. Temperature had an impact on both clam immunocompetence and QPX vitality. In this case, M. mercenaria hemocytes were found to have the highest phagocytosis capacity at 21˚C while QPX was noted as having low survival/virulence at higher temperatures (41). Clearly, temperature can have a major impact on both the immune system of bivalves as well as the microbes they encounter.

1.5.2. Sediment Acidification

The study of sediment pH on biological functions such as shell production is a relatively new field of study; however, the evidence gathered so far indicates that there are intertidal locations along the Maine coast where sediment pH is much lower than others (42,43). These areas have been described as “dead mud” by clammers because they find relatively few clams in these sediments. The cause of the low sediment pH is likely the excessive amount of organic matter that settles onto the sediment surface as a result of excess algae growth. Eutrophication

(triggered by runoff of nutrients from land) contributes to excess nitrogen, which is in turn metabolized by bacteria. CO2 is released as an end product from respiration and reacts with

+ seawater to form carbonic acid (H2CO3) (44). The hydrogen ions (H ) can then dissociate from

- 2- this compound to form either bicarbonate (HCO3 ) or carbonate (CO3 ) ions. This increases the number of hydrogen ions in the water and thus lowers pH (45). Atmospheric CO2 also reacts with seawater to form carbonic acid in the seawater that decreases pH in the water column as shown below:

Organic matter (phytoplankton) + O2 → CO2

→ CO2 + H2O  H2CO3

→ - + H2CO3  HCO3 + H

- → 2- + HCO3  CO3 + H

The acidification of the water column from atmospheric CO2 predicted to occur within the next

100 years likely will have no impact on the pore water of the sediments (46).

The break-down of organic matter at the sediment surface, although rapid, is relatively inefficient and allows for excess organic matter to be buried deeper in the sediment. Organic matter that is buried below the sediment surface gets broken down by specialized bacteria that

12 use various forms of anaerobic metabolism. Oxygen usually cannot penetrate below the top centimeter of sediment, making it unavailable to all the microbes found below the sediment surface. Alternative oxidants must be used in the deeper portions of the sediment by specialized

- bacteria. The hierarchy of oxidants used by the bacteria are as follows: O2, NO 3, MnO2, FeOOH,

- SO4 and CO2. The oxidants are used in this specific sequence due to the amount of energy each yields in its respective oxidation reaction. The most reactive substrates (one that will yield the most energy) are consumed first before the next best oxidant is used (47). The possible metabolic pathways the microbial community can make use of in marine sediments are listed below. They are listed from top to bottom in order from those that will produce the most energy to those that will produce the least. The formula “CH20” will be used to generically represent organic matter

(44).

1. Aerobic respiration

1/4“CH20”+1/4O2(g) = 1/4CO2(g)+1/4H2O

2. Denitrification

- + 1/4“CH20” + 1/5NO3 +1/5H = 1/4CO2(g) + 1/10N2(g)+7/20H2O

3. Manganese respiration

+ 1/4“CH20”+1/2MnO2(s)+H = 1/4CO2(g) + 1/2Mn2+ + 3/4H20

4. Iron respiration

+ 2+ 1/4“CH20”+Fe(OH)3(s) + 2H = 1/4CO2(g) Fe +11/4H2O

5. Sulfate reduction

2- + - 1/4“CH20” + 1/8SO 4+1/8H =1/4CO2(g)+1/8HS +1/4H2O

6. Methane fermentation

1/4“CH20” = 1/8CO2(g)+1/8CH4(g)

7. Hydrogen fermentation

1/4“CH20”+1/4H2O = 1/4CO2(g)+1/2H2(g)

These pathways ultimately lead to a pH gradient to being set up in the sediments since reducing capacity decreases with sediment depth (48). Note that not all bacterial respiration is fueled by

-2 the initial addition of organic matter. Sulfate (SO4 ) is constantly supplied from the overlying seawater and supplies anoxic bacteria with an electron receptor for their respiration since the

-2 SO4 penetrates a few meters down into the sediment. The resulting hydrogen sulfide (H2S) is eventually oxidized into intermediate sulfur species (elemental sulfur, polysulfides, thiosulfate,

-2 and sulfite) as well as back into SO4 after being used as an electron receptor by a variety of anoxic bacteria (49).

Marine sediments generally are of a much lower pH (between 0.5 and 1.5 pH units) than the overlying water in a coastal environment. pH within the sediment is stratified into zones, and tends to increase with sediment depth (48). This is because the ample supply of oxygen and organic matter at the sediment-water interface allows for more CO2 to be produced than in the zones below, which only contain the organic matter that is not metabolized at the sediment surface. The depth of each of the redox zones varies seasonally due to changes in organic matter deposition as well as temperature (43,50). In turn, the pH gradient in sediments changes with season. Sediment pH also varies on small spatial scales due to bioturbation as well as differences in organic matter deposition. A simple burrow from a worm can increase the surface area of the sediment-water surface by up to 16-fold and thus drastically increase the oxygen penetration depth to a specific patch of sediment (44). Sediment pH must be measured separately at each site of interest, since variability from one site to the next is high (relative to differences in water column pH). Measuring pH alone however is usually not sufficient to assess how much pressure an acidic environment is placing upon a given bivalve.

Saturation state with respect to the form of calcium carbonate the bivalve of interest builds its shell from (aragonite or calcite) will also need to be calculated for each site if the researcher is interested in finding out if the shell is under dissolution pressure. Aragonite is not as stable as calcite and thus starts to dissolve at a higher pH than calcite. This means that bivalves that use aragonite to form their shells, such as razor clams, are particularly vulnerable to dissolution pressure. The majority of bivalves use aragonite to form their shells however because it comes at a lower metabolic cost than does calcite. Aragonite saturation state (Ωar) can be calculated using the following formula:

2+ 2- Ωar = [Ca ]*[CO 3]/K*sp

2+ 2- where [Ca ] and [CO 3] are the concentrations of dissolved calcium and carbonate ions respectively. K*sp is the stoichiometric solubility product and can be found in the literature if the temperature and salinity of the sample are known (51). When Ωar or ca is 1, there is no net dissolution or precipitation of the indicated from of calcium carbonate. Values above 1 allow for precipitation while values below 1 lead to dissolution (52). Various invertebrates have shown the capacity to counteract this dissolution pressure at a metabolic cost (53,54) (Ries et al., 2008;

Wood et al., 2008). Whether razor clams can cope when Ωar < 1 (and the cost incurred for doing so) has yet to be determined.

A relatively large literature exists on effects of water column acidity on bivalves (55–59); however, little work has been done examining effects of increased sediment acidity on the fate and growth of infaunal bivalves such as E. leei. To date, studies show the valves of juvenile clams will dissolve (60), or that they refuse to burrow (61), or move away from sediments (46) that have been acidified directly with CO2 gas. These studies illustrate the need to obtain data first on effects of sediment pH on clam burrowing behavior before the immune system can be

15 studied, as clams may simply choose to refuse to bury in the sediment if they do not have the option to relocate (as would be the case in an aquaculture scenario). The dissolution resistance of clams has been noted to increase with size (60) as adult clams are more likely to dig in low-pH sediment than juveniles. Even if adult clams burrow and display normal behavior, there is the possibility their immune system will be impacted. To this end, immune parameters were measured in adult E. leei after exposure to elevated water temperatures and decreased sediment pH.

1.6. Thesis Objectives

E. leei has great potential for use as an aquaculture species as literature on handling methods, habitat preference, and starting a broodstock are all available (7). However, very little is known about the pathogens of E. leei (1). Understanding the immune system of E. leei will not only guard against any health issues that may arise in the razor clam industry, but also provide tools to assess immune capacity in the face of environmental impacts. By examining the immune response under potential future environmental conditions, this effort was undertaken to insure that the developing razor clam industry can continue to function under the expected chronic environmental stress from increased acidity and temperature. The objectives for this study were to characterize the hemocytes for this species, establish a seasonal baseline for the unstimulated internal immune parameters of this species, investigate the seasonality of the internal immune capacity of adult E. leei, determine the impacts of increased temperature on the internal immune capacity of this species, and determine the impacts of increased sediment acidity on the internal immune capacity of this species.

CHAPTER 2

CHARACTERIZATION OF ENSIS LEEI HEMOCYTES

2.1. Introduction

The Atlantic jackknife clam, E. leei, is native to the East Coast of North America as well as invasive to the Wadden Sea of Europe. The species currently supports a small fishery along the North Shore, MA as well as in southern Maine, and is in the development phases for large scale aquaculture (1, 3). A small, live and processed market exists for E. leei in the Northeast; yet, an inconsistent supply of product prevents its expansion (1). Sizable markets for razor clams already exist in Netherlands, United Kingdom, Ireland, Spain and Portugal, which will provide an outlet for any product that exceeds the local market’s demand (7). Literature on handling methods, habitat preference, and broodstock are all available; however, very little is known about vulnerability of E. leei to potential pathogens (1,7). Information on E. leei health will be needed to assess the health of stocks as the razor clam aquaculture industry develops in the Northeast.

The circulating cells in bivalves, hemoyctes, have been shown to be indicative of an animal’s immune status (62). This is because hemocytes serve a variety of immune functions as well as wound healing, nutrient digestion, transport, and excretion (62). Establishing the morphology of healthy E. leei hemocytes will make comparisons of hemocytes possible between healthy and diseased bivalves (63). Bivalve hemocytes also can be used as bioindicators for abiotic stressors such as pollutants (64–67), salinity shifts (68), warming (68,69), and acidification (69). Hemocyte morphology, total hemocyte counts (THCs), and differential counts of the hemocytes are some of the basic parameters that can be used to help assess the immune capacity of bivalves (29,70,71). Transmission electron microscopy (TEM) can be used to take a more detailed look at hemocyte morphology and can thus be a useful tool in assessing the

17 immune status of bivalves (66,72,73). These studies have shown that number of granules, integrity of the cytoskeleton, presence of pseudopodia, and size of lysosomes can be used to differentiate between diseased and healthy hemocytes

The two general types of hemocytes (haemocytes) present in most bivalves are granulocytes and hyalinocytes (agranulocytes). Granulocytes are defined by the presence of granules regardless of their size or density. Most granules are considered to be a type of lysosome, as they contain acid hydrolases used to break down invading pathogens and cellular debris (74). Granulocytes can be classified further as large and small. Most bivalves have at least one type of granulocyte, although the lions-paw scallop, Nodipecten subnodosus, has no granulocytes at all (70). Hyalinocytes are distinguished from granulocytes by their lack of granules. Some bivalves also have small hyalinocytes with a large nucleus to cytoplasm ratio sometimes referred to as blast-like cells or hemoblasts. These cells have already been observed in many bivalve species including Ruditapes philippinarum (Adams & Reeve, 1850) (75),

Saccostrea glomerata (Gould, 1850) (76), Crassostrea ariakensis (Fujita, 1913) (77), Saccostrea kegaki Torigoe & Inaba 1981, Crassostrea nippona (Seki, 1934), and Hyotissa hyotis (Linnaeus,

1758) (78). Their presence in a variety of bivalve species, low immune function capacity (77), positive reaction to the mammalian stem cell molecular marker anti-CD34 (in the case of R. philippinarum) (75), and the ability to divide (as found in R. philippinarum) (79) all provide evidence to suggest blast-like cells may actually be stem cells that mature into the other cell types.

In addition to the more common hemocytes types listed above, there are two rare cell types found in only some bivalves: the vesicular cell (vacuolated granulocyte) and the serous cell

(pigmented cell, rhogocyte, pore-cell, brown cell, morula cell). The vesicular cell is

18 characterized by an abundance of clear vesicles and a lack or scarcity of granules. The function of this cell type in bivalves is unknown, although its morphology may be the result of a granulocyte going through degranulation. This cell type has been reported in the mussel

Anodonta cygnea (Linnaeus, 1758) (80) oysters S. kegaki, C. nippona, H. hyotis, Crassostrea gigas (Thunberg, 1793) (78,81), the cockle Cerastoderma edule (Linnaeus, 1758) (82), and the clam Scapharca inaequivalvis (Bruguière, 1789) (83). The serous cell is rounded and has a light brown to almost black coloration when viewed unstained that resembles the appearance of melanin (84,85). It is believed to perform functions such as excretion (74), detoxification of heavy metals (85), removal of moribund parasites (86), and cell aggregation initiation (87).

Bivalves possessing this cell type include the clam R. philippinarum (75), the giant clam

Tridacna crocea Lamarck, 1819 (87), and the cockle C. edule (82).

Classifying the morphology of individual species is necessary as not all bivalves have the same types and frequency of hemocytes (74). For this reason, the number of bivalve species covered in the hemocyte literature has continued to increase over the years. Bivalve species recently added to this literature base include Callista chione (Linnaeus, 1758) (88),

Bathymodiolus japonicas Hashimoto & Okutani, 1994, B. platifrons Hashimoto & Okutani, 1994

B. septemdierum Hashimotot & Okutani 1994 (89), S. kegaki, C. nippona, H. hyotis (78), and A. cygnea (80). The current study contributes to this effort by classifying the hemocytes of E. leei and providing a detailed morphology of them using both transmission electron microscopy

(TEM) and light microscopy. Currently, no information is available on the hemoctyes of E. leei, although work has been conducted on the hemocytes of a related razor clam species, E. siliqua

(29). Total hemocyte counts and differential cell counts will also be presented for wild E. leei.

These counts can be obtained without lethally sampling the animal and can be a valuable

19 indicator of health status. Classifying and describing E. leei hemocytes will lay the foundation for future immune function studies in the species.

2.2. Methods

2.2.1. Clam and Hemolymph Collection

E. leei were collected from the shore of the Damariscotta River Estuary in Walpole, ME

(43°56'05.8"N 69°34'49.1"W) using a clam hoe on 4 December 2014 and 19 January 2015. All were individually banded and transported in a cooler filled with seawater from the Damariscotta

River Estuary. All clams used in the study had no external damage and retracted all soft parts immediately upon removal from water. Clams ranged between 108-145mm SL and 20-25mm in width. Clams were kept in a 150L recirculating tank at the University of Maine, Orono filled with water from the sampling site to keep them in conditions similar to the collection site.

Hemocyte counts and light microscopy were conducted within 48 hours after collection, while

TEM samples occurred a week after collection date. A glass tank kept seawater at 6.5 ±1°C, and had a polypropylene tub placed in it filled with paver’s sand purchased from Home Depot to a depth of 100mm for the clams to burrow. All sand had larger rocks removed from it by sieving it through an onion bag followed by rinsing with artificial seawater (salinity = 30ppt; Crystal Sea

Marinemix) before being placed in the tub. An aquarium circulation pump (Hydor Koralia, 240

GPH) was used to provide constant water movement over the sand tub. A 38 L water change was carried out twice weekly using artificial seawater to maintain water quality when clams were kept more than 48 hours. Hemolymph was withdrawn from the anterior abductor muscle using a

1ml syringe fitted with a 23-gauge needle. A slight, but constant, pressure was maintained until hemolymph started to flow in the syringe. After removing the needle, hemolymph was expelled

20 into a 1.5ml centrifuge tube. There was no need to pool hemolymph from multiple clams as at least 1ml of hemolymph could be obtained from each clam.

An additional batch of five clams (105- 170 mm SL) was collected on 21 January 2016 near Dobbin’s Island, in Beals, ME (44˚30’24.6”N 67˚36’15.1”W). These animals were used exclusively for the Pappenheim’s stain preparations. Clams were banded and transported back to the Bowden lab wrapped in paper towels saturated with natural sea water. These animals were kept in the same tank set-up as described above with the exceptions of a larger aquaria (76 L), and the use of sand from the collection site. A stainless steel sieve with 2.54 mm mesh was used to remove any large animals or pebbles from the sand prior to placing the sand in the tubs.

Salinity was maintained between 32 and 33ppt while temperature was kept at 4.5˚C ± 0.5˚C (±

SD). These clams were fed ad libitum with instant algae (shellfish diet 1800, Reed Mariculture) on a daily basis, and kept at least a week in the lab before hemolymph was drawn. Hemolymph was added to a chilled tube containing an equal amount of 0.38% sodium citrate (Fisher

Scientific) in 0.2µm-filtered artificial seawater (FASW) to reduce clotting (88).

2.2.2. Total Hemocyte Counts

All cell counts were taken immediately by loading freshly drawn hemolymph into a

Neubauer improved hemocytometer and observing the cells at 250x magnification using a Leitz

Wetzlar light microscope. Hemolymph also was checked for gamete contamination at this time.

Small hemocyte aggregates were present, but cells within them were still distinguishable from each other and were thus included in the counts.

2.2.3. Hemacolor and Pappenheim’s Panoptical Staining

Slides to be stained in Hemacolor were prepared by placing 50µl of hemolymph on glass slides within a moisture chamber placed in a refrigerator. Hemolymph was allowed to incubate

21 for 30 min at 6˚C. Excess hemolymph was then removed by tilting the slide to one side and drying the droplet with a paper towel. Slides were immediately stained using a Harleco

Hemacolor® stain kit (65044) consisting of methanol, eosin, and methylene blue. Staining was accomplished according to the manufacturer’s directions. After air drying, the slide was fitted with a cover slip using Triangle Biomedical Sciences SHUR/Mount™ toluene-based liquid mounting media.

Slides to be stained in Pappenheim’s mixture were prepared by placing 50µl of hemolymph on slides within a moisture chamber at room temperature. Hemolymph was allowed to incubate for 30 min before excess hemolymph was removed by pipette. The attached hemocytes were then covered with 60µl of a 1% glutaraldehyde solution made in 32ppt FASW containing 1% sucrose. Slides were fixed for 30 min at 6˚C within their moisture chambers, and then rinsed with filtered artificial seawater before being dipped in May-Grunwald stain (Electron

Microscopy Sciences) for 3 min. Slides were then dipped in distilled water before being dipped in a 5% Giesma (Sigma) bath for 5 min (88). Slides were rinsed once more with distilled water before being mounted with a coverslip as described above.

2.2.4. Neutral Red Staining for Lysosomes

Cells were stained using a Neutral Red staining procedure previously used by Lowe et al.

(90) with a few modifications. The neutral red stock solution was made by dissolving 20 mg neutral red in 1 ml of DMSO. The stock solution was then frozen at -20˚C until a working solution was needed. The working solution was made by mixing 10µl of stock solution with 5 ml of FASW, and then filtered again through a 0.22 µm syringe-driven filter to remove any remaining clumps of neutral red dye. Slides were prepared for staining with neutral red using the same procedure as previously listed for Hemacolor staining preparations. After disposing of the

22 excess hemolymph, 50µl of freshly prepared neutral red working solution was placed onto the attached hemocytes. A cover slip was placed on and hemocytes were allowed to incubate a further 15 min in the moisture chamber kept in a fridge at 6˚C. The sides of the cover slip were then blotted with a tissue before being viewed in an Olympus BX60 light microscope at various magnifications.

2.2.5. Differential Counts and Hemocyte Measurements

An Olympus BX60 light microscope was used to view stained slides at 40x and 100x objective lens. Immersion oil was used with the 100x objective lens. Photographs were taken with an Olympus DP71 microscope camera. Differential counts were taken under the 40x objective lens. For each animal, hemocytes were counted and categorized by cell type until 100 cells were counted. Extremely rare cell types were excluded from the counts. Hemocyte diameters, nucleus diameters, and nucleus to hemocyte ratios were obtained from the Hemacolor micrographs using Image J photo software.

2.2.6. Transmission Electron Microscopy

Freshly drawn hemolymph was used to fill iced 1.5ml centrifuge tubes halfway. As much as 3ml of hemolymph could be extracted from a single clam. Artificial seawater (30ppt) containing 4% glutaraldehyde solution was made by mixing a 0.22 µm filtered artificial seawater

(Crystal Sea Marinemix, salinity of 60ppt) with a stock 8% glutaraldehyde solution (made in DI water) in a 1:1 ratio. This 4% glutaraldehyde fixative solution was then added to the 1.5ml centrifuge tube in a 1:1 ratio. The hemolymph was then gently mixed by inversion and kept in the fridge at 6˚C for 14 hours before being centrifuged at 700xg for 10min. Supernatant was removed and cell pellets were re-suspended in 30ppt salt solution to remove the glutaraldehyde.

This rinse step was repeated two more times before the pellets were postfixed with 1% osmium

23 tetraoxide solution (salinity 30ppt) for 1h at 6°C. The 1% osmium solution was then switched out for chilled 30ppt salt solution and the tube was allowed to sit in the fridge overnight. The

30ppt salt solution was removed by pipette and pellets were dehydrated in an acetone series. The acetone was then replaced with Araldite 502/Embed 812 resin (Electron Microscopy Sciences) and allowed to incubate for 15 minutes while under a vacuum to remove any air bubbles in the resin. This resin rinse was repeated twice more with fresh resin before the samples were cured for 48 hours under vacuum at 55°C. Ultrathin sections were then cut using a diamond knife on a microtome. Sections were dried onto copper grids and stained with uranyl acetate and lead citrate

(81). Sections were then viewed in a Phillips CM10 transmission electron microscope at 100KV and photographs taken with an Orius SC200 digital CCD camera, model 830.

2.3. Results

2.3.1. Light Microscopy

Observation of live hemocytes stained with neutral red and fixed hemocytes stained with

Hemacolor or Pappenheim’s mixture allowed for identification of both hyalinocytes (agranular hemocytes) and granular hemocytes based on the presence or scarcity of granules (Figure 2.1;

Figure 2.2.). Both hemocytes types were then divided into large and small categories based on their nucleus to cytoplasm ratio. Neutral red was used only as a vital stain to check for the presence of lysosomes since it is difficult to discern granulocytes from hyalinocytes when using this stain.

2.3.1.1. Hyalinocytes

Hyalinocytes were classified as two sub-types: large hyalinocytes and small hyalinocytes.

Large hyalinocyte nuclei were oval in shape and stained a dark shade of purple in Hemacolor

(Figure. 2.1.A). These cells were usually spread out onto the glass slide and thus had a high cell

24 diameter. Their average nucleus: cell ratio was mid-range even though their average nucleus size was greater than that found in any of the other hemocyte types (Table 2.1.). Their cytoplasm took up very little stain and had either scattered granules or none at all. Extended pseudopodia were always absent in the hyalinocytes that had no granules (Figure 2.1.B). An abundance of small round vesicles was present in the cytoplasm (Figure 2.1.B) although many of them were difficult to discern due to the lightly-stained cytoplasm. Both round and elongated cell morphologies were observed (Figure. 2.1.A and B).

Small hyalinocytes also were present, and characterized by their rounded, dark blue nucleus. They had both a very high nucleus:cell diameter ratio and small cell diameter (Table

2.1; Figure 2.1.A and E). This made it impossible to discern if granules were present in their scant cytoplasm using light microscopy. A few of these cells had extended pseudopodia (Figure.

2.1.E).

2.3.1.2. Granulocytes

Granulocytes could be identified by the presence of granules in the cytoplasm and were further classified as large and small granulocytes. All the granulocytes had granules that were not well defined in Hemacolor and often blurred together to form one mass (Figure. 2.1.A). In

Pappenheim’s stain preparations, some of the individual granules could be distinguished (Figure

2.1.D). Scattered vesicles were also apparent in some the large granulocytes stained in

Pappenheim’s mixture (Figure 2.1.C). Neutral red preparations showed an abundance of tightly packed red granules filling most of the cytoplasm however the nucleus remained unstained

(Figure 2.2.). Large granulocytes were almost always spread out onto the glass slide and thus had a high average cell diameter (Table 2.1.). They appeared in oval, elongated, and spindle morphologies with extended pseudopodia (Figure. 2.1.A, C, D, and E). The nucleus to cell

25 diameter ratio was the smallest of the five cell types (Table 2.1.). The granules stained either dark pink (Figure 2.1.C) or purple (Figure 2.1.D) and cytoplasm took up very little stain around the edges of the cells (Figure 2.1.C and D). Their nucleus was either round or oval in shape and always stained a dark blue (Figure. 2.1.C and D).

Small granulocytes had a larger nucleus to cell ratio and a smaller cell diameter than large granulocytes (Table 2.1.). In both Hemacolor and Pappenheim’s stain preparations the cytoplasm looked much like that of the large granulocytes (Figure 2.1.A, D, and F). In neutral red preparations the tightly packed masses of granules in the cytoplasm of the small granulocytes stained a dark red, (Figure 2.2.D). The granules tended to blur together in both Hemacolor and neutral red preparations although individual granules were sometimes visible at the edges of theses masses. The nucleus stained a dark blue/black color in Pappenheim’s stain preparations and was round or oval in shape (Figure. 2.1.D and F).

2.3.1.3. Vesicular Cells

Vesicular cells were present in only some of the samples and characterized by an abundance of vesicles surrounding the nucleus. Both round and elongate nucleus morphologies were observed. The cytoplasm stained a light blue color and thus made the large vesicles prominent in this cell type. Numerous extended pseudopodia were observed and the cell type possessed an elongate cell morphology. (Figure 2.1.F).

A B

C D

E F

Figure. 2.1. Light micrographs of fixed and stained E. leei hemocytes. (A) Hemocytes collected without an anti-clotting solution fixed in methanol and stained with Hemacolor. (B-F) Hemocytes collected in a 0.38% sodium citrate solution and fixed in glutaraldehyde before being stained in Papenheim’s mixture. LH- large hyalinocyte, LG- large granulocyte, SG- small granulocyte, SH- small hemocyte, VC- vesicular cell.

A B

C D E

Figure. 2.2. Light micrographs of live E. leei hemocytes stained with neutral red. The stained portions are granules and the clear circle within each cell is the nucleus. (A) Large hemocyte with many extended pseudopodia and some enlarged granules. (B) Large hemocyte filled with enlarged granules. (C) A large hemocyte where the granules have not yet become enlarged. (D) A small rounded hemocyte packed tightly with granules. (E) A small rounded hemocyte with enlarged granules.

Table 2.1. Diameters and nucleus:cell ratio of the hemocyte types found in E. leei. Vesicular cells are excluded from this table because they were far too scarce. Measurements were taken from methanol-fixed hemocytes that were stained using Hemacolor. SH; small hyalinocyte; SG; small granulocyte; LH: large hyalinocyte; LG: large granulocyte; Min: minimum; Max:maximum. Type Nucleus Cell (µm) N/C ratio (µm) Mean ± SD Min Max Mean ± SD Min Max Mean ± SD Min Max SH 3.25±0.56 2.05 4.12 5.41±1.04 3.68 8.07 0.61±0.11 0.40 0.89 SG 3.65±1.83 2.66 5.26 9.43±1.83 5.57 13.76 0.39±0.07 0.28 0.58 LH 4.79±1.01 2.98 6.65 12.51±2.16 8.68 16.71 0.36±0.10 0.21 0.59 LG 3.88±0.79 2.84 5.75 15.32±3.19 10.53 22.03 0.26±0.06 0.16 0.38 n=30 for each hemocyte type.

2.3.1.4. Aggregates

Hemocyte aggregates of various sizes (ranging from 20-320 µm in length) were observed in each of the Hemacolor preparations. Most aggregates formed a round or oval shape however there were the occasional odd shapes such as elongated chains and rings. These aggregates could get quite large when no anti-clotting solution was used (Figure 2.3.A). The use of the sodium citrate anticoagulant solution prevented the hemocytes from forming these large aggregates, but still allowed for the binding of the hemocytes to the glass slides and the formation of small aggregates (Figure 2.3.B and C). The core of the aggregates was much darker in the large aggregates than it was in the small ones. The hemocytes blended in with the dark blue/black color of the core once they joined the aggregate and thus became impossible to identify. Both hyalinocytes and granulocytes showed the capacity to form aggregates as they can be seen extending pseudopodia towards existing aggregates (Figure 2.3.B and C). In neutral red preparations the stain could not penetrate past the outer perimeter of the aggregates so most of the aggregate appeared clear.

B C

Figure. 2.3. Light micrographs of stained E. leei hemocyte aggregates. (A) Low magnification view showing the various possible shapes and sizes of methanol-fixed hemocyte aggregates stained in Hemacolor without anti-clotting solution. (B) Glutaraldehyde-fixed hyalinocytes collected in a 0.38% sodium citrate solution and stained using Pappenheim’s mixture. (C) Glutaraldehyde-fixed granulocytes collected in a 0.38% sodium citrate solution and stained using Pappenheim’s mixture.

2.3.2. Transmission Electron Microscopy

Confirmation of the five morphologically distinct cell types that were identified using light microscopy was made using TEM. This also revealed many details in the hemocytes not visible in the light micrographs. At least two distinct granule types were observed, which were characterized as electron lucent and electron dense. The term electron dense was reserved for the darkest colored granules while the medium to light gray granules were deemed electron lucent granules.

2.3.2.1. Hyalinocytes

Large hyalinocytes were amoeboid in shape and had a low nucleus:cell diameter ratio.

They lacked dense granule concentrations but had some scattered granules and mitochondria.

The nucleus was round and located near the center of the cell. A few large clumps of chromatin could be seen within the nucleus. All large hyalinocytes observed had relatively few granules compared to granulocytes. Granule size was variable (0.16-0.63 µm) and most granules were electron lucent. All granules were either round or oval in shape. Rough endoplasmic reticulum and multiple mitochondria were always present as well as large vesicles. Pseudopodia of various lengths were observed extending from the hyalinocyte’s perimeter (Figure 2.4.A and B).

Small hyalinocytes were round or oval in shape and had a very large nucleus:cytoplasm ratio. The nucleus was always centric and oval in shape. Clumps of chromatin of various shapes were visible in the nucleus. The scarce cytoplasm usually contained at least a few mitochondria, but granules were either relatively scarce compared to granulocytes or completely absent (Fig.

2.4.C and D). When granules were present both electron lucent and electron dense granules were of variable size (1.00-0.18 µm). Long pseudopodia were also observed which extended as much as 4um out from the edge of the cell (Figure 2.4.D).

A B

C D E

Figure. 2.4. Transmission electron micrographs of E. leei hyalinocytes and vesicular cell. (A) Large hyalinocyte with an abundance of empty space in the cytoplasm. (B) Close up of the large hyalinocyte shown in micrograph A. (C) Small hyalinocyte with both extended pseudopodia and some granules. (D) Small hyalinocyte with a large particularly long pseudopodium extended. (E) Vesicular cell with many vesicles and no granules. EDG- electron dense granule, ELG- electron lucent granule, N-nucleus, PS- pseudopodium, M- mitochondria, V-vesicle, rER- rough endoplasmic reticulum.

2.3.2.2. Granulocytes

Large granulocytes had a low nucleus:cytoplasm ratio, and the nucleus contained large clumps chromatin. The cell possessed an elongate morphology with the nucleus at one of the far ends. Only a few vesicles were present as most space was packed densely with various organelles including mitochondria, Golgi apparatus and rough endoplasmic reticulum. Usually, the rough endoplasmic reticulum was found near the nucleus. The most prominent feature however was the abundance of granules. Like those found in the hyalinocytes, these granules occurred in various sizes (0.20- 0.81 µm), and came in both round and oval morphologies. Both electron lucent and electron dense granules were present. Multiple short pseudopodia also were observed extending off the perimeter of each large granulocyte (Figure. 2.6.D).

Small granulocytes had a higher nucleus:cytoplasm ratio than large granulocytes and possessed round or oval morphologies (Figure 2.6.A-C). The nucleus had large clumps of chromatin, and was either centric (in round cell morphologies) or eccentric (in oval cell morphologies) (Figure

2.5.A and D; Figure 2.6.B). Both electron dense and electron lucent granules were found in them although electron dense granules were always more abundant than electron lucent granules within any given small granulocyte. These granules came in various sizes (0.20- 0.87 µm) and were observed to be packed particularly tight in small granulocyte sections that lacked a nucleus

(Figure 2.6A and C). Upon closer inspection of the granules, we noticed some of the electron lucent granules had spots of more intense granular material within them (Figure. 2.5.E and F).

This could not be discerned in the darkest of the electron dense granules, which had a homogenous black coloration (Figure. 2.5.B). The same variety of organelles found in large granulocytes was present and all cells possessed multiple pseudopodia.

A B

D E

Figure. 2.5. Transmission electron micrographs of E. leei granulocytes and granules. (A) Small granulocyte located just outside a hemocyte pellet. (B) Large electron-dense granule with homogeneous coloration within. (C) A trio of granules, each with a different electron density. (D) Small granulocyte with numerous large granules. (E) Large electron lucent granule with only a few small patches of darker areas within. (F) Electron lucent granules with many patches of much darker granular material within. EDG- electron dense granule, ELG- electron lucent granule, N-nucleus, PS- pseudopodium, M- mitochondria, V-vesicle.

A B

C D

Figure. 2.6. Transmission electron micrographs of E. leei granulocytes. (A) Small granulocyte packed tightly with electron dense granules. (B) Small granulocyte with electron dense granules extending multiple pseudopodia. (C) Small granulocyte with a very prominent rER as well as a mixture of electron light and electron dense granules. (D) Large granulocyte with electron dense and electron lucent granules as well as an elongate morphology. EDG- electron dense granule, ELG- electron lucent granule, N-nucleus, PS- pseudopodium, M- mitochondria, V-vesicle, rER- rough endoplasmic reticulum, GA- golgi apparatus.

2.3.2.3. Vesicular Cells

A single vesicular cell was found using TEM (Figure 2.4.E). This hemocyte is characterized by numerous vesicles and a lack of organelles (with the exception of the nucleus).

The vesicular cell has many rounded electron lucid vesicles. The nucleus is rounded and one of the few electron dense areas of the cell. The rest of the cell is entirely electron lucent with the exception a thin network of membranes that form the vesicles and cell perimeter.

2.3.3. Total and Differential Hemocyte Counts

Total cell counts were taken from ten different razor clams, with six sampled in

December and four sampled in January. We pooled results from both sample points since they were sampled about a month apart from the same site, and animals were within the same size and temperature range. The average total hemocyte count was 1.96 x 106 ± 6.78 x 105 cells mL-1.

A total of 1,000 Hemacolor-stained cells (100 per sample) were counted under light microscopy for differential hemocyte counts and were comprised of 11% small hyalinocytes, 12% large hyalinocytes, 59% small granulocytes, and 18% large granulocytes (Figure 2.7.). At least 72% of the hemocytes counted were granulocytes in each of the samples used. The 27 cells that were counted from the electron micrographs were comprised of 26% small hyalinocytes, 15% large hyalinocytes, 48% small granulocytes, and 11% large granulocytes. Vesicular cells were not included in these counts as they were only discernable in Pappenheim’s stain and TEM preparations. Vesicular cells were only present within three of these five Pappenheim’s stain preparations and one of the TEM samples. None of these preparations contained more than two vesicular cells.

Figure 2.7. Population distribution of hemocyte types in E. leei hemolymph. Each bar stands for the mean of the given cell type counted for a total of 1,000 cells across 10 different animals. Error bars represent ± 1 standard error of the mean.

2.4. Discussion

Hemocyte classification in bivalves often varies from author to author as there are both multiple terms used to indicate the same hemocyte type and many different characteristics to classify by. In this work, an effort was made to keep the classification scheme simple by using only granularity and cell sizes as classification criteria. This means affinity for dyes was not used to classify cell types although observed colors are given for each cell type. In both light and electron microscopy, we observed small hyalinocytes (agranular hemocytes), large hyalinocytes, small granulocytes, large granulocytes, and vesicular cells. Serous cells (as described in the

Introduction) seemed to be absent from E. leei hemolymph. Hyalinocytes and granulocytes are

37 present in most bivalves (except some scallops) and their terminology is consistent throughout the literature (74). The classification scheme for the subtypes of these two general cell types however often vary between authors.

The small hyalinocytes identified in the current work resembled the blast-like cells

(hemoblasts) found in many other bivalve species. They had the characteristic high nucleus:cell diameter ratio (74) and stained a dark blue in Pappenheim’s stain (75,76). Some of the small hyalinocytes in the current study possessed a few granules as well as extended pseudopodia

(Figure 2.4.C), both features not usually found in blast-like cells. It is possible, then, that the small hyalinocytes with a few granules are granublasts (Figure. 2.4.C) while the small hyalinocytes without granules are hemoblasts (Figure 2.4.D). Granublasts have been hypothesized to be the stem cell for granulocytes and hemoblasts the stem cells for hyalinocytes

(84). The observations made on the hemocytes of the freshwater gastropod Planorbarius corneus

(Linnaeus, 1758) support this hypothesis (91). We chose to label these cells with a more general term (small hyalinocytes) since functional studies will be needed to determine whether they have the relatively low immune function capability that is characteristic of blast-like cells (78,92).

A rare type of hemocyte, the vesicular cell, was observed under TEM. This cell is unique because it has many vesicles and appeared to lack all organelles except the nucleus. It was only identifiable in TEM and Pappenheim’s stain preparations and was very rare in all samples it was observed in. Their origin are unknown although it could possibly be a recently degranulated granulocyte. Granulocytes have granules comparable in abundance to the vesicles in the vesicular cell so the vesicles could be phagosomes or degranulated granules. If this is the case then a low frequency of vesicular cells would be expected in healthy E. leei specimens. The frequency of this cell type was expected to increase when the animals are exposed to a stressor.

Such a frequency increase has been observed in tunicates exposed to tributyltin (93). Vesicular cell function is largely unknown although they have been observed to be involved with phagocytosis in the sea squirt Phallusia mammillata (Cuvier, 1815) (94) and the inflammatory response in the sea squirt Ciona intestinalis (Linnaeus, 1767) (91,95–97).

The findings in the present study can be compared directly to morphology studies done on other bivalve species. From a broad morphological standpoint, E. leei seems to have a hemocyte population similar to other most other bivalves, as they possess both hyalinocytes and granulocytes of various sizes. This includes the closest relative to E. leei with hemocyte data available, E. siliqua (29). Differential hemocyte counts revealed that E. leei has hemolymph dominated largely by granulocytes (77%) (Figure 2.7.), which is not unusual among bivalves.

Other bivalves with hemocyte populations comprised of over 75% granulocytes include Mytilus edulis Linnaeus, 1758 (29) E. siliqua (29), C. chione (88) and Mya arenaria Linnaeus, 1758

(98).

The average total hemocyte count of E. leei was 1.96 x 106 cells mL-1, which seems to fall into the low range of bivalve average THCs. The bivalves Argopecten irriadians (37.5 x 106 cells mL-1) (99), E. siliqua (6.48 x 106 cells mL-1) (29) M. edulis (5.68 x 106 cells mL-1) (29), T. philippinarum (5.02 x 106 cells mL-1) (79) and Pinctada fucata (Gould, 1850) (4.9 x 106 cells mL-1) (69) all have higher THC than E. leei while C. gigas (1.6 x 106 cells mL-1) (100) and C. chione (1.2 x 106 cells mL-1) (88) both had lower THCs. It should be kept in mind that comparing total and differential hemocyte counts across many studies often does not account for possible seasonal variation, as has been observed in C. gigas (100). This includes the current work for which counts should only be considered representative of the winter season. Still, these

39 results indicate substantial differences are present in the total hemocyte counts among various bivalve species and therefore must be examined on a species by species basis.

The TEM work revealed at least two morphologically distinct types of granules are present in E. leei granulocytes. Multiple granule types have also been identified in a variety of other bivalves such as P. imbricata Röding, 1798 (92), C. gigas (81), and S. glomerata (76). The function of each granule type is unclear at the moment however the presence of dense patches of red granules in neutral red preparations indicates at least one of the granule types has low pH contents. Neutral red staining has been used to confirm the presence of lysosomes in P. imbricata (92); hence, the conclusion that that at least one of the granule types found in E. leei is a lysosome. The small-medium sized electron dense granules are likely to be lysosomes as these were found in high enough densities to be comparable to the densities found in neutral red- stained granulocytes. Release of electron dense granules in response to phagocytosis of latex beads has been recorded in Crassostrea virginica (Gmelin, 1791) (101), which suggests that some of the electron dense granules in E. leei may have hydrolytic enzymes. The differences in electron density of the granules could thus be in part due to variations of hydrolytic enzyme contents. Various hydrolytic enzyme contents have already been detected in other bivalves such as M. arenaria (98) and M. edulis (29).

Evidence from work on the oyster C. rhizophorae (Guilding, 1828) (102) suggests that all hemocyte types in that oyster are actually different developmental stages of a single cell line.

This could be the case for E. leei; however, it is also possible there are two different stem cells as discussed above. Whichever the case, it is likely that granulocytes would degranulate if pathogens or environmental stressors are present (102). After degranulation, the granulocyte would likely have an abundance of empty vesicles and become a vesicular cell. The information

40 from the present study neither confirms nor denies that all the E. leei hemocyte types described here are part of one or more cell lines as functional studies are still needed. The morphology of each hemocyte type established in the current work provides the foundation for these functional studies.

2.5. Conclusion

This study represents the first published record of E. leei hemocytes, as well as the first

TEM work on any razor clam species. The data presented here is both the start of the investigation into E. leei hemocytes, and the foundation for future functional characterization studies. These studies should focus on assessing the capacity of E. leei hemocytes to perform immune functions known to occur in most other bivalve species such as phagocytosis, production of reactive oxygen species, and melanization (15).

This data is also valuable to growers as it gives them an idea of what kinds of total and differential hemocyte counts to look for to identify healthy broodstock clams. Having the established morphologies of the hemocytes also allows growers to identify if there are any irregularly shaped cells in the hemolymph, which can serve as warning signs of neoplasia

(cancer) (103). Cell counts and morphology can be looked at quickly, which enables growers to make on-farm decisions about which animals to remove and which ones to use as broodstock.

The cell counts of E. leei however may change with the season. In the next chapter, the cell counts of healthy E. leei will be examined over the course of year to address this.

CHAPTER 3

IMMUNE CAPACITY OVER THE COURSE OF A YEAR

3.1. Introduction

Internal immune capacity is the ability of an organism to mount an immune response against a pathogen that has penetrated past the host’s exterior barriers. Studying the internal immune capacity of razor clams, Ensis leei, can help determine how these animals are impacted by current or future environmental conditions. Internal immune capacity of invertebrates is usually determined through assays on the circulating blood cells responsible for carrying the immune response, hemocytes. Internal immune capacity can be assessed by studying the hemocytes directly (THC, differential hemocyte counts, phagocytosis activity) as well as the antimicrobial substances they produce (lysozyme, reactive oxygen species, nitric oxide, phenoloxidase) (29,70,71). Assessing many of these parameters concurrently in conjunction with a pathogen challenge can give researchers a more complete understanding of an animal’s immune capacity than measuring a single parameter alone would. Using this method, previous studies have shown bivalve immunity can be impacted by a variety of environmental factors such as pollutant concentrations, food availability, temperature, salinity, oxygen levels, and pH (104).

The immune parameters of bivalves often follow an annual cycle. This is driven by environmental factors that change with the season such as temperature, hours of daylight, and reproductive status (100,105–108). Knowing whether the immune capacity of a given species follows a seasonal pattern is essential to a successful aquaculture operation. Information on immune capacity cycles can help growers further refine their broodstock collection period by notifying them when bivalves have the highest immune capacity. Data on seasonal cycles is also

42 helpful to researchers since it tells them whether the results from an experiment will vary depending on the season it is performed.

When assessing clam health, the acidity of the sediments they live in are particularly important to monitor. Studies that monitor marine sediment acidity are still few in number, and more will be required as sediment acidification intensifies and spreads. Sediment acidification is driven by organic matter. The pH drop often leads to the reduction of carbonate saturation state

(Ω). Calcium carbonate (CaCO3) is formed and dissolved according to the following equation:

2+ 2- CaCO3 ↔ Ca + CO3

2+ 2- where Ca is the concentration of calcium ions and CO3 is the concentration of carbonate ions.

As pH decreases, excess hydrogen ions (H+) bind to carbonate ions to form bicarbonate ions

- (HCO3 ) and thus decrease the concentration of carbonate ions available. This can reduce the Ω, which can be calculated using the following equation:

2- 2+ Ω = [CO3 ] [Ca ]/K*sp

where K*sp is the stoichiometric solubility product, dependent on the carbonate mineral phase

(aragonite, calcite, low Mg calcite, high Mg calcite), temperature, salinity, and pressure. Usually, calcium ions are readily available in marine environments so the concentration of carbonate ions usually drives the Ω (109). Precipitation of the calcium carbonate (CaCO3) shells that invertebrates need to survive is thermodynamically favored when Ω > 1 while dissolution is thermodynamically favored when Ω < 1 (110). Invertebrates are still capable of producing calcium carbonate when Ω < 1, however they must pay a higher metabolic cost to do so (111–

113).

The goal of this study was to determine which, if any, components of E. leei’s internal immune capacity are seasonal. This was done by collecting clams regularly over the course the year and measuring total number of hemocytes, types of hemocytes (differential hemocyte count), and percent of hemocytes capable of phagocytosing bacteria. The wet flesh weight and dry shell weight were also recorded in order to calculate an overall health condition index as done in Vázquez et al. 2013 (114).

3.2. Methods

3.2.1. Sediment and Clam Collection

Sediment and clams (shell length 17.6 ± 0.9 cm) were collected from a site near Dobbins

Island in Beals, ME (44°30'25.0"N 67°36'15.1"W) on 10 sample dates (spread across December

2016 to December 2017) from the lowest portion of the intertidal zone, which was exposed only during astronomically low tides. Samples could not be collected in January, August, and

September due to adverse weather conditions and tides that did not go out far enough. On each sampling trip, the pH values of the overlying water, as well as the sediment at 0-3 cm (surface sediment), 4-7 cm, and 12-15 cm depth were recorded from 10 mL water samples taken via a 20 mL syringe (n = 3 for each depth) and expelled into 15 mL tubes. The contents of the tubes were then immediately analyzed with an Accumet portable pH meter (model AP115) equipped with an

Accumet pH/ATC probe. The sphere of influence for a 10 mL draw in the sediment was determined to be 3 cm or less based on another study that used the syringe method (115). Pore water samples were then gently poured into 15 mL tubes, each containing 10 µL of saturated mercuric chloride solution to preserve the sample until it could be analyzed for TA. The tubes were then inverted five times to insure the mercuric chloride spread throughout the sample.

Overlying water temperature and salinity also were recorded using the pH probe and an ATC

44 portable refractometer. The top ~20 cm of sediment was collected using a shovel and transferred to the laboratory in 5-gallon buckets. All rocks and large invertebrates were removed from the shoveled sediment using a sieve with 2.54 mm mesh size within 10 hr of sediment collection.

The titrations for TA were done back at the University of Maine using the open cell titration method (116), but was modified for smaller sample volumes. A 5 mL sample was placed into a 15 mL tube and titrated with 0.01N hydrochloric acid prepared using a 33 ppt NaCl background solution down to a pH of 3.00. The pH and temperature were recorded. Mixing between acid additions was achieved by moving the probe up and down. The Ωar was calculated from total alkalinity, pH, and salinity values using CO2SYS (ver. 2.1 http://cdiac.ess- dive.lbl.gov/ftp/co2sys/) and the K1 and K2 values recommended for estuarine waters (117).

3.2.2. Experimental Setup

The sediment collected on each trip was distributed among 6 rectangular polystyrene containers measuring 18  12.5  12.5 cm (height, length and width). Each container was filled to the 15 cm depth with sieved sediment and placed within an aquarium (75 L capacity) that had been filled with ̴ 60 L of 33 ppt artificial seawater made with Tropic Marin Pro Reef mix. Each container then had a single clam (shell length 17.6 ± 0.9 cm) added to each of them. All clams buried into their respective sediments within an hour. The aquarium was constantly aerated with aquarium air pumps and water at the surface level of the sediment was kept in constant motion using an aquarium circulation pump (Hydor Koralia, 240 GPH). Aquarium water pH, temperature, salinity, and dissolved oxygen were recorded using the Accumet® portable pH meter, a refractometer, and a YSI EcoSense® dissolved oxygen meter (model DO200A). All clams were sampled and removed from the aquarium within 30 hours of being taken from the collection site. 45

3.2.3. Bacteria Preparation

The Serotype 02β V. anguillarum strain NV15812 was selected for use with the phagocytosis assay in this study since it had been modified to carry the broad-host-range plasmid p67T1 that encodes ampicillin resistance and constitutive production of the red fluorescent protein d-Tomato (118). Vibrio anguillarum carrying p67T1 was grown at 22°C on a 1.5% agar made with 2.5% Difco Luria-Bertani Broth powder, 1.5% NaCl, and 600 µg ml-1 of ampicillin.

This selective medium insured growth of a pure culture of V. anguillarum with d-Tomato. The bacteria was then re-suspended in a 33 ppt solution of filtered artificial seawater (made with

Crystal Sea Marine Mix) at a concentration of about 2.0 x 108th cells ml-1 (0.2 absorbance at 620 nm).

3.2.4. Hemocyte Collection and Total Counts

Within 30 hours of clam collection, hemolymph was withdrawn from the mid-section of the foot using a 1 mL syringe fitted with a 23-gauge needle. A slight but constant pressure was maintained until hemolymph started to flow in the syringe. After removing the needle, the hemolymph (1 ml per draw) was expelled into a 30 mL tube (topped with a Fisherbrand 100 µm cell strainer) kept on ice containing 1.5 mL of 2.5% sodium citrate (Fisher Scientific). This was repeated so a total of 2 mL of hemolymph could be collected from each clam. A separate 30 mL tube was used for each clam. Cell counts were carried out on the tube contents using a hemocytometer and the cell count was adjusted down to 1.0 x 106 cells/ml if needed using filtered artificial seawater (FASW).

3.2.5. Hemocyte Distribution and Incubation for Immune Assays

The hemolymph from each tube was distributed via pipette across three wells (300 µl per well) of a 24 well culture plate for differential hemocyte counts and trypan blue exclusion cell viability assay (120). May-Grünwald stain (80 µl) was added to one well for the differential hemocyte counts. The remaining hemolymph was distributed to two cover slips (300 µl each) for later use with the phagocytosis assay. The plates and cover slips were then incubated at 20°C for

45 min to allow hemocytes to attach. During this incubation period, the wells with May-

Grünwald stain in them were then photographed on an inverted microscope so differential hemocyte counts could be obtained. Unattached hemocytes were then washed away from the rest of the wells and cover slips using two 500 µl rinses of FASW. After the overlying solutions were removed, both of the cover slips and 1 of the wells had 200 µl of the V. anguillarum solution placed on them. The remaining well had FASW placed on it to serve as a control. The cover slips were incubated for 3.5 hours and then photographed under a combination of fluorescent (wave length) and white light. The 24 well plate was incubated for five hours at 20°C before 50 µl of a

0.22 µm filtered 0.3% trypan blue solution was added. The hemocytes in these wells were then photographed using a camera mounted on an inverted microscope. All cell counting for subsequent analysis was done manually using the cell counter tool in ImageJ photo software

(ver. 1.8.0 for windows https://imagej.nih.gov/ij/download.html).

3.2.6. Differential Hemocyte Count Analysis

A total of at least 150 hemocytes over 6 fields of view were counted for each clam from micrographs. Each hemocyte was identified as either a granulocyte or hyalinocyte based on the presence or absence of their granules (appearing as purple circles when stained) respectively.

The percent granulocytes present in a given animal was calculated by dividing the number of granulocytes by the total number of hemocytes counted per animal and then multiplying by 100.

3.2.7. Phagocytosis Analysis

For each animal, at least 100 hemocytes were counted across two cover slips. At least six fields of view were used per cover slip. Hemocytes that had one or more V. anguillarum cells within them (visible due to red fluorescence) were counted as phagocytic hemocytes while those with no V. anguillarum cells were counted as non-phagocytic hemocytes. The number of phagocytic hemocytes was then divided by the total number of hemocytes and multiplied by 100 to obtain the percent phagocytic hemocytes for a single animal. All the percent phagocytic hemocyte values within a given treatment group was used to calculate the average for that group.

3.2.8. Condition Index Calculation and Check for Gonads

After hemolymph extraction, the clam meat was shucked and immediately weighed. The mass of the hemolymph drawn from the meat was added to the weight. In addition, shell mass was recorded to the nearest 0.1g after being allowed to dry for at least 24 hours on bench top at

21-23°C. The condition index was then calculated by dividing the wet meat weight by the dry shell weight and multiplying by 100 (114). Values for this condition index usually range from 75 to 250 and are unitless. Values get smaller as flesh weight decreases and shell mass stays the same. Insufficient food sources as well as some diseases can lead to this. Values get larger when flesh mass increases faster than shell growth, which is the case when clams build up their gonads. The presence or absence of ripe gonads was also noted for each clam through visual inspection for a white or beige mass on the anterior abductor muscle. If a one of these masses was present, it was viewed under the microscope to insure it was gonad.

3.2.9. Statistical Analysis

One-way ANOVA was used to check for the presence of significant differences between any of the data points taken over the course of the year for each clam health parameter. The

Pearson correlation test was used to test if the parameters that had a difference in their data set were correlated to temperature, which was used as a proxy for the seasonality. GraphPad Prism

(ver. 8.0.1 https://www.graphpad.com/scientific-software/prism/) was used to generate figures and perform statistical tests. All data analyzed passed normality and equal variance tests. All tests were performed at the 0.05 significance level.

A power analysis was carried out with the goal of determining if the statistical tests carried out had sufficient power and minimum detectable difference to detect the mean differences observed in the data. G*Power (ver. 3.1.9.2 https://g-power.apponic.com/) was used to determine the power and effect size of each statistical test used. Effect size was expressed as the number of standard deviations from the mean. The minimum detectable difference was then calculated from the effect size by multiplying the average standard deviation of all groups for a given health parameter by the effect size. The minimum detectable difference was then compared to the differences between means.

3.3. Results

3.3.1. Environmental Conditions

The temperature and salinity of the water at the collection site changed throughout the year. Average overlying water temperature and salinity ranges were 0.0 to 19.1°C and 30 to 35 ppt respectively over the course of the sampling period. Average pH ranges of the overlying water, sediment surface, 4-7 cm depth and 12-15 cm sediment depth were 7.67 to 8.04, 6.51 to

7.30, 6.95 to 7.31, and 7.10 to 7.50 respectively. Average total alkalinity ranges of the overlying water, 0-3 cm sediment depth, 4-7 cm sediment depth and 12-15 cm sediment depth were 1490

-1 to 2633, 1147 to 3760, 1417 to 3760, and 777 to 4187 µmol kg respectively. The Ωar ranges of the overlying water, sediment surface, 4-7 cm depth and 12-15 cm sediment depth were 0.64 to

1.09, 0.07 to 0.24, 0.06 to 0.28, and 0.10 to 0.48 respectively.

2 0 3 6 S a lin ity

)

C T e m p e ra tu re ( C )

 1 5 3 4

(

S e

a r

l u

i t

1 0 3 2 n a

t r

y e p

m 5 3 0

e T

0 2 8

6 7 7 7 7 7 7 7 7 7 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 3 7 9 9 9 7 6 -6 -4 -2 -1 -2 -2 -2 -2 -2 -2 0 1 2 2 2 3 4 5 6 7 1 1 1 1

Figure 3.1. Temperature and salinity of the water at the collection site. Error bars = ± S.D, n = 3 for each point.

8 .0 O v e rly in g W a te r

) S e d im e n t S u rfa c e e

l 7 .5 a 1 2 -1 5 c m D e e p

S 7 .0

N (

H 6 .5 p

6 .0

6 7 7 7 7 7 7 7 7 7 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 3 7 9 9 9 7 6 -6 -4 -2 -1 -2 -2 -2 -2 -2 -2 0 1 2 2 2 3 4 5 6 7 1 1 1 1 B

e 1 .2 5 t

a O v e rly in g W a te r t

1 .0 0 S e d im e n t S u rfa c e

n o

i 1 2 -1 5 c m D e e p

t a

r 0 .7 5 u

a S

0 .5 0

i n

o 0 .2 5

a r

A 0 .0 0

6 7 7 7 7 7 7 7 7 7 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 3 7 9 9 9 7 6 -6 -4 -2 -1 -2 -2 -2 -2 -2 -2 0 1 1 2 2 3 4 5 6 7 1 1 1 1 Figure 3.2. The pH and aragonite saturation state of the collection site. (A) The pH of the sediments and overlying water at the collection site over the course of a year. (B) The aragonite saturation state of the sediments and overlying water at the Beals Island collection site over the course of a year. Error bars = ± S.D, n = 3 for each point.

3.3.2. Hemocyte Counts

Average THCs and granulocyte proportion ranges were 1.80 x 106 to 6.07 x 106 cells mL-

1 and 63% to 77% granulocytes respectively over the course of the sampling period. The THCs did not stay constant across all sample dates (ANOVA, p = 0.0131), and positively correlated with the temperature of the collection site water (Pearson correlation, p = 0.001) (Figure 3.3.).

THC increased as waters got warmer and decreased as they got colder. Differential hemocyte counts also varied across sample dates, but only a few of the sample dates were different from each other (ANOVA, p = 0.0405). Differential hemocyte counts did not correlate with temperature (Pearson, p = 0.7908). The granules of the granulocytes stained dark purple while hyalinocytes appeared clear or lightly stained.

A 1 .0  1 0 7 2 0 T o ta l H e m o c y te C o u n t

8 .0  1 0 6 T e m p e ra tu re ( C ) l 1 5

m /

e s 6 m

e 6 .0  1 0 t

p y 1 0

( c

 o

4 .0  1 0 6 C m

)

H 5 2 .0  1 0 6

0 0

6 7 7 7 7 7 7 7 7 7 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 3 7 9 9 9 7 6 -6 -4 -2 -1 -2 -2 -2 -2 -2 -2 0 1 2 2 2 3 4 5 6 7 1 1 1 1 B 1 0 0 2 0 % G ra n u lo c y te s

s 8 0 T e m p e ra tu re ( C )

e 1 5

T t

e y

c 6 0 o

p l

u 1 0

(

 n

a 4 0 r

)

G 5

% 2 0

0 0

6 7 7 7 7 7 7 7 7 7 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 3 7 9 9 9 7 6 -6 -4 -2 -1 -2 -2 -2 -2 -2 -2 0 1 2 2 2 3 4 5 6 7 1 1 1 1

Figure 3.3. Hemocyte counts of E. leei. All E. leei were taken from the collection site and sampled within 30 hours of collection. (A) Total hemocyte counts of E. leei taken from the collection site over the course of the year. p = 0.001 for correlation test between hemocyte counts and temperature (Pearson). (B) Percent granulocytes of E. leei hemocytes taken from the collection site over the course of the year. p = 0.791 for correlation between hemocyte counts and temperature (Pearson). Error bars = ± S.D, n = 3-4 clams per point.

3.3.3. Phagocytosis

Average phagocytic activity ranged between 49 to 88% over the course of the sampling period. The percent phagocytic cells did not stay constant across all sample dates (ANOVA, p <

0.001), and was not correlated with the temperature of the collection site water (Pearson correlation, p = 0.324) (Figure 3.4.). Vibrio anguillarum were clearly visible since they remained fluorescent even after they were internalized.

1 0 0 2 0

% P h a g o c y tic C e lls

s l l 8 0 T e m p e ra tu re ( C ) e 1 5

e c

m i

t 6 0

p y

c 1 0 (

 o

g 4 0

)

a h

P 5

2 0 %

0 0 6 7 7 7 7 7 7 7 7 7 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 3 7 9 9 9 7 6 -6 -4 -2 -1 -2 -2 -2 -2 -2 -2 0 1 2 2 2 3 4 5 6 7 1 1 1 1

Figure 3.4. Phagocytic activity of E. leei hemocytes. All animals were sampled within 30 hours of collection. p = 0.324 for correlation between percent phagocytic hemocytes and temperature (Pearson). Error bars = ± S.D, n = 3-4 clams per point.

3.3.4. Condition Index and Percent Ripe

Average condition index ranged from 129-189 over the course of the sampling period.

The condition index did not stay constant across all sample dates (ANOVA, p < 0.001), but did not correlate with the temperature of the collection site water (Pearson correlation, p = 0.905)

(Figure 3.5.A). The condition index did not stay constant across all sample dates (ANOVA, p <

0.001), and did not correlate with temperature of the collection site water (Pearson correlation, p

= 0.905) (Figure 3.5.A). Gonads only were present in the sample points at warmest time of the year (5-29-17, 6-27-17, and 7-26-17) and their presence was correlated with temperature

(Pearson, p = 0.025).

0 2 2 5 2 0

0 C o n d itio n In d e x 1

* 2 0 0 ) s T e m p e ra tu re ( C )

s 1 5

a 1 7 5

e m

l 1 5 0

p e 1 0

( h



s 1 2 5

C /

) h

s 1 0 0

e 5

t 7 5 e

w ( 5 0 0

6 7 7 7 7 7 7 7 7 7 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 3 7 9 9 9 7 6 -6 -4 -2 -1 -2 -2 -2 -2 -2 -2 0 1 2 2 2 3 4 5 6 7 1 1 1 1

B s

d 1 0 0 2 0 a

n % R ip e

G T e m p e ra tu re ( C )

7 5 1 5 e

T p

e i

m R

p h

t 5 0 1 0

( i



) s

m 2 5 5

0 0 % 6 7 7 7 7 7 7 7 7 7 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 3 7 9 9 9 7 6 -6 -4 -2 -1 -2 -2 -2 -2 -2 -2 0 1 2 2 2 3 4 5 6 7 1 1 1 1

Figure 3.5. Overall condition of E. leei. (A) Condition index and (B) percent of E. leei individuals with ripe gonads taken from the collection site over the course of the year and sampled within 30 hours. p = 0.905 and 0.0205 for condition index and percent ripe clams with temperature respectively (Pearson). Error bars = ± S.D, n = 3-4 clams per point.

3.3.5. Power Analysis

The power of the one-way ANOVAs (n of all groups combined = 37) was determined to be 0.86 at an effect size of 0.8. This means there was a 86% chance to correctly identify a difference among the means if that difference was 0.8 average standard deviations or larger. The value that 0.8 average standard deviations equates to for a given measured parameter is called the minimum detectable difference. The mean detectable differences for each health parameter were smaller than all but the smallest mean differences among the groups in their respective one-way

ANOVA. The differences too small for the test to detect were only 4% or less of the average mean. This indicates the sample size was sufficiently large to detect most of the observed mean differences. The power of the Pearson correlation tests (n of temperature and health parameter data combined = 67) was determined to be 0.93 at an effect size of 0.4. This effect size was small enough to detect correlations where present.

3.4. Discussion Results of the clam health survey show the THCs and proportion of clams with ripe gonads were correlated to collection site temperature while phagocytic activity, condition index, and granulocyte proportion are not (Figs 3.3-3.5). THCs and percent of clams with ripe gonads were lower in the colder months and higher in the warmer ones. These parameters also had similar values at the December collection points in 2016 and 2017, which suggests these parameters follow an annual cycle (Figs. 3.3., 3.5.B). It is likely then that THC and gonad ripeness are dependent upon the season since temperature was used as a proxy for seasonality.

The THCs likely increased in response to the higher temperatures as has been reported in many other bivalve field studies (105,108,121). The increase in heart rate of bivalves with warmer temperatures has been suggested as the primary reason for the THC increase (122).

Increasing THC count also better prepares bivalves for the greater number of pathogens present in the warmer months (123,124). The THC count increase in the current study may have resulted from hemocyte proliferation and/or hemocyte migration from tissues into the hemolymph (125).

The percent of clams with ripe gonads likely correlated with temperature because E. leei follows a seasonal spawning cycle with spawning occurring in the warmer months (1).

Unfortunately, the exact month(s) the E. leei spawn(s) in the current study could not be pinpointed since the sample point at the end of August could not be obtained. Field studies on other bivalves have shown most bivalves have at least one simultaneous spawning event in the warmer portion of the year (105,121). The link between the bivalve reproductive cycle and temperature also can be seen in hatcheries, where an increase in temperature is often used to induce spawning (7). The reproductive state of bivalves can also affect THC since additional hemocytes are produced after spawning to resorb left over sex cells (104). This may have contributed to the THC increase in the current study since the month with the highest average

THC (give the name of the month) came just after a potential spawning month for E. leei (Figure

3.3.A).

Phagocytic activity, condition index, and granulocyte proportion did not correlate to temperature and thus are not considered seasonal in E. leei. It is not unusual for some health parameters among bivalves to follow a seasonal cycle while others do not, even within the same species (104). Results from laboratory tests can also differ from field tests. For example, phagocytosis rates of bivalve hemocytes usually follow temperature trends in laboratory experiments however this is usually not always the case with field experiments (126,127). This is likely because of the variety of changing environmental parameters (temperature, hour of air exposure, hours of daylight, salinity, sediment chemistry, etc) to which the animals are subjected

58 in the wild. All of these environmental parameters also can have synergetic effects on clam health parameters since the animals can potentially be exposed to changes in all of them at the same time. In addition, each health parameter may respond differently to the environmental conditions. Even within the same size range, differences can exist between individual animals as well. Energy reserves, metabolic rates, and parasite loads are just a few of the biological factors that can differ among individuals at any given sample point and impact immunity parameters

(77,128,129).

3.5. Conclusion

The lack of correlation between phagocytic activity, condition index, and granulocyte proportion and seasonality highlight how unpredictable the effects of environment factors in the field can be in E. leei. The variability of the hemocyte system is likely due to a combination of both internal and environmental factors. For this reason, any researchers looking to carry out lab experiments should sample a portion of their animals close to collection time so they can check if the animals in their control group have been impacted by holding conditions. Laboratory mesocosms can never replicate the natural environment exactly but checking the health parameters of interest against animals freshly acquired from the environment will make it more likely the results will hold true for animals in the field.

The fact that proportion of ripe clams and total hemocyte counts follow a seasonal cycle makes it easier for both E. leei growers and stock regulators to monitor the health of E. leei stocks. Growers can check the health of their broodstock during any season they need to knowing what the approximate cell count and of a healthy clam is for the given season. The time to collect that broodstock in the Beals, ME area seems to be in late May, when clams with ripe gonads start appearing. Stock regulators now also have a seasonal baseline for total hemocyte counts they can compare results to when monitoring E. leei stocks for any changes in health. In these ways, the seasonal E. leei data can serve as a tool to aid growers and stock regulators alike.

CHAPTER 4

METHOD FOR MANIPULATING SEDIMENT ACIDIFICATION

4.1. Introduction

Sediment acidification is caused by rapid decay of organic matter and often leads to decreases in sediment pore water pH, alkalinity, and carbonate saturation state. Most acidification occurs at the sediment-water interface because oxygen is required for rapid organic matter decomposition, which is generally only available in the top few centimeters (50,130). In coastal environments, there is an abundance of organic matter settling from the water column above to fuel this process. Organic matter, however, is not evenly distributed on the sediment surface. Large concentrations of organic matter can be deposited on the sediment surface through die-offs of algal blooms that result from nutrient run-off from the land (131). This organic matter can then be relocated through the burrowing action of benthic macrofauna, which also affects sediment exposure to oxygen (130). As a result, the sediment surface pH can vary significantly from 6.0 to 8.2 across different sites and times of the year (130,132). Values on the lower end of this range have been observed at intertidal locations in both the Bay of Fundy (132) and the coast of Maine (60,133).

Sediment acidification has already been identified as a major environmental stressor for a variety of clam species, including, the soft-shell clam M. arenaria (60,61,109), the hard shell clam M. mercenaria (60,134), and the Manila clam R. philippinarum (135). Young clams are particularly vulnerable to this stressor. M. arenaria in the ~0.2 mm size class begin to suffer increased mortality rates at sediment pH < 7.3. Clams become more resilient to the stressor as they grow larger (~0.4 mm M. arenaria survival is not impacted until pH 7.0; (60), but even R.

61 philippinarum adults have 100% mortality after being exposed to pH 6.1 sediments for 10 days

(135). The need to cope with acidic sediments becomes especially important for razor clams, such as the Atlantic jackknife clam E. leei. This species must spend the majority of its life at the surface of the sediment due to its short siphons, and has no way to seal its soft tissues off from the external environment (1). It is essential to protect these clam species as they have ecologic and economic value (1,136).

Laboratory studies can be used to assess how much acidification each clam species can tolerate, and these results would be valuable to the growers. Techniques for manipulating and monitoring seawater chemistry by changing the carbonate concentration in the laboratory setting are well defined (116,137). Techniques for manipulating marine sediment chemistry by changing carbonate concentration are still in development. Furthermore, research on the impacts of sediment acidification on infaunal marine bivalves is limited (136). The majority of laboratory studies conducted involving sediment acidification impacts on clams make use of quick bursts of

CO2 gas to acidify the sediment while air from an aquarium pump is bubbled in a control group.

This method produces a stable sediment pH; however, it does not mimic the natural acidification process, which is driven by the microbial decay of organic matter (136).

Adding crushed bivalve shells has been shown to increase sediment pH, and thus, might serve as a natural method of restricting the sediment pH in the laboratory to collection site values. An increase of ~0.3 pH units was achieved using crushed M. arenaria shells (1 mm grain size) during a 30 day field study (138). Field sediments also have been successfully buffered using a mix of 70% C. gigas shells and 30% mixed-species clam shells crushed to a grain size of

5 cm or less. Sediments treated with this mixture were ~0.2 pH units higher than untreated sediments after 55 days (139). The controlled conditions of laboratory aquaria should allow for

62 these field methods to be modified so that the crushed shell counteracts the acidification caused by the sediment bacteria without raising the pH higher than collection site values.

The current work explores how coastal sediments can be acidified in the laboratory using the natural metabolic processes of sediment bacteria. Spreading crushed E. leei shell is proposed as a method to counteract acidification and keep the sediment chemistry close to collection site parameters in the control group. Acidifying sediment in the laboratory using the decay of the pre- existing organic matter would allow the creation of mesocosms that more closely mimic the natural environment. The primary concern with this method, however, is the pH stability that is largely dependent on the metabolic processes of the sediment bacteria with respect to the production of carbonate species (130). The carbonate chemistry of coastal sediments housed in recirculating aquaria were monitored over 74 days at 24°C and 6.5°C. In addition, stability of the solution chemistry was assessed once acidification ceased. The reported solution chemistry parameters included sediment pH, total alkalinity, and aragonite saturation state (Ωar), because aragonite (see Chapter 1 for further details on aragonite) is the calcium carbonate mineral phase found in most clam shells (140,141). Successful manipulation of sediment chemistry allows for testing the acidification tolerance of clams as well as other burrowing invertebrate species in the laboratory.

4.2. Methods

4.2.1. Sediment and Shell Collection

Sediment was collected from a site near Dobbins Island in Beals, Maine (44°30'25.0"N

67°36'15.1"W) during the summer 2017 from a 10 square meter region in the lowest portion of the intertidal zone. This area was exposed only during astronomically large tides and well

63 populated with E. leei. The pH values of the overlying water, as well as the sediment at 0-3 cm

(surface sediment), 4-7 cm, and 12-15 cm depth were recorded from 10 mL water samples taken via a 20 mL syringe (n = 3 for each depth) and expelled into 15 mL tubes. The contents of the tubes were then immediately analyzed with an Accumet portable pH meter (model AP115) equipped with an Accumet pH/ATC probe. The sphere of influence for a 10 mL draw in the sediment was determined to be 3 cm or less based on another study that used the syringe method

(115). Pore water samples were then gently poured into 15 mL tubes, each containing 10 µL of saturated mercuric chloride solution to preserve the sample until it could be analyzed for total alkalinity. The tubes were then inverted five times to insure the mercuric chloride spread throughout the sample. Overlying water temperature and salinity were also recorded using the pH probe and an ATC portable refractometer (n = 3 for each). The top ~20 cm of sediment (~ 18

L) was collected using a shovel and transferred to the laboratory in 5-gallon buckets. Sediment cores (n = 2) also were taken from the collection site by pushing 50 ml polypropylene centrifuge tubes straight down into the sediment to a depth of ~6 cm. These cores, in addition to 3 samples of sieved (2.54 mm mesh) sediment from the top ~20 cm, were analyzed later for water content, grain size distribution, and percent organic matter (see section below for full methods). All rocks and large invertebrates were removed from the shoveled sediment using a sieve with 2.54 mm mesh size within 10 hr of sediment collection. Dead valves of E. leei also were collected at the

Beals Island site. The shells were later rinsed with distilled water and heated (~100 °C for 2 hr) before they were ground up to a grain size of < 3 mm using a mortar and pestle. E. leei shells were chosen because they are made primarily of aragonite, which dissolves faster than other forms of calcium carbonate (141,142).

4.2.2. Water Content, Grain Size, and Percent Organic Matter of Sediment

Mean percent organic matter and sediment grain size distribution was determined for both 6 cm sediment cores (n = 2) and sieved (2.54 mm mesh) sediment from the top ~20 cm (n =

3). Organic matter analysis on cores was performed for the 0-2, 2-4, and 4-6 cm sections of each core. Grain size analysis was carried out on the sediment from the entirety of the 2 cores combined as well as 3 samples (~100 g each) from the sediment collected by shovel from the top

~20 cm of the sediment. All sediment was dried at 120°C for 24 hours to determine both water content and prepare it for analysis. Percent water content was determined by dividing the mass of the dried sediment by the wet sediment, multiplying it by 100, and then subtracting it from 100.

Grain size analysis was carried out by putting dried sediment through a series of sieves stacked on top of each other (143). The sieve numbers and mesh sizes used going from the top to the bottom of the stack were #20 (840 µm), #40 (425 µm), #60 (250 µm), #20 (75 µm). The mass of the sediment remaining on each sieve was divided by the initial mass of the dry sample before being sieved and multiplied by 100 to determine the percent composition for each grain size range. The sediment was then characterized according to 4 grain size categories: gravel or very coarse sand (greater than 840 µm), coarse to medium sand (840-425 µm), fine to very fine sand

(425-75 µm), and silt or mud (less than 75 µm) (144). Percent organic matter was determined using the loss on ignition method. The weight of the sediment after combustion at 550°C was subtracted from the initial dry weight of the sediment. This was then divided by the initial dry weight and multiplied by 100 to obtain the percent organic matter (145).

4.2.3. Experimental Setup

To determine how temperature affects sediment acidification rates, the sediment was distributed among 12 rectangular polystyrene containers measuring 18  12.5  12.5 cm (height,

65 length and width). Each container (henceforth referred to as sediment beds) was filled to 15 cm with sieved sediment and placed within 6  75 L aquaria (2 beds per aquarium) filled with 30 ppt artificial seawater made with Tropic Marin Pro Reef mix for 74 days. Three aquaria were kept at a mean (± 1 SE) of 24 ± 0.04°C (using aquarium heaters, while the other three were kept at 6.5 ±

0.03°C using chillers. All aquaria were constantly aerated although one of the air pumps used for a 24°C aquarium failed during the experiment (this aquarium was removed from the data set). On day 33, one sediment bed was taken out from each aquarium to have 150 g of crushed E. leei shell (1.4 g cm-2) added to its sediment surface to determine if crushed shell can bring the pH back up to the levels found at the collection site. The crushed shell was then lightly pressed into the top one centimeter of sediment by hand before the sediment beds were slowly placed back into respective aquaria. All aquaria were covered with black polythene for the first 10 days of the experiment to reduce the light levels during the initial acidification of the sediment that limited potential photosynthetic activity from diatoms present on the sediment surface.

4.2.4. System Monitoring and Data Analysis

Monitoring of pH and total alkalinity of these 4 different groups of sediment beds over a

74-day period allowed for observation of both short and long term trends of sediment chemistry.

Pore water samples were taken via syringe at the top of the sediment column (0-3 cm) at every 2-

3 days, and from the middle (4-7 cm) and bottom (12-15 cm) depths at every 10-12 days. pH of a

10 mL pore water sample (extracted via syringe) was measured immediately after being expelled into a 15 mL tube which was then poisoned with 10 µL of saturated mercuric chloride to preserve the sample until it could be analyzed for total alkalinity. Titrations for total alkalinity were carried out using the open cell titration method (116), but was modified for smaller sample volumes. A 5 mL sample was placed into a 15 mL tube and titrated with 0.01N HCl prepared

66 using a 30 ppt NaCl solution down to a pH of 3.00. The pH and temperature were recorded.

Mixing between acid additions was achieved by moving the probe up and down. The Ωar was calculated from total alkalinity, pH, and salinity values using CO2SYS (ver. 2.1 http://cdiac.ess- dive.lbl.gov/ftp/co2sys/) and the K1 and K2 values recommended for estuarine waters (117). pH, total alkalinity, and Ωar of the three measured depths and overlying water were then compared to each other using one-way ANOVA and Tukey’s multiple comparisons tests. Linear regression analysis was used to determine if significant changes in pH, alkalinity, or Ωar occurred in any of the treatments being kept in the aquaria by testing if the line of best fit had a slope significantly different from zero. Repeated measures one-way ANOVA followed by Tukey’s multiple comparisons was used instead of linear regression for the 4-7 and 12-15 cm depths of the treatments with crushed shell since only 3 data points were acquired for these depths. All data analyzed met normality and equal variance assumptions. GraphPad Prism (ver. 8.0.1 https://www.graphpad.com/scientific-software/prism/) was used to generate figures and perform statistical tests (all at the 0.05 significance level).

4.3. Results

4.3.1. Collection Site Sediment Chemistry and Composition

The pH of the collection site sediment was lowest at the surface and increased with depth.

Mean pH of the surface sediment (6.75 ± 0.06) was significantly lower than the sediment at the

12-15 cm depth (7.14 ± 0.07) (Tukey’s multiple comparison test, p = 0.009), but none of the other depths were significantly different from each other (Tukey’s multiple comparison test, p ≥

0.183). Mean pH of the water overlying the sediments (7.79 ± 0.01) was significantly higher than at any of the measured sediment depths (Tukey’s multiple comparison test, p < 0.001 for all comparisons), which were 6.75 ± 0.06 (0-3 cm), 6.95 ± 0.08 (4-7 cm), and 7.14 ± 0.07 (12-15

67 cm) (Figure 4.1.). No significant differences in total alkalinity occurred between the overlying water (1947 ± 18 µmol kg-1) and any of the sediment depths, which were 3760 ± 540 (0-3 cm),

3843 ± 328 (4-7 cm), and 4187 ± 1426 (12-15 cm) µmol kg-1 (Tukey’s multiple comparison test, p ≥ 0.254). In addition, no significant differences in total alkalinity were detected among any of those sediment depths (Tukey’s multiple comparison test, p ≥ 0.979). Mean Ωar followed the same trends as mean pH. That is, the overlying water was saturated with respect to aragonite

(just over 1.0), and all measured sediment depths were undersaturated with respect to aragonite with means ranging from ~0.1 to ~0.7. Both the sediment surface (0.22 ± 0.05) and 4-7 cm depth

(0.30 ± 0.04) had amean Ωar significantly lower than that of the overlying water (1.05 ± 0.03)

(Tukey’s multiple comparison test, p ≤ 0.005). Surface sediment had a lower mean Ωar (0.22 ±

0.05) than the 4-7 cm (0.30 ± 0.04) and 12-15 cm depths (0.67 ± 0.2) did, but there were no significant differences between any of these sediment depths (Tukey’s multiple comparison test, p ≥ 0.066) (Figure 4.2.). The average salinity and temperature of the overlying water was 30 ppt and 19.1°C respectively.

8 .0

) 7 .5

a c

s 7 .0

S B

N 6 .5 (

p 6 .0

5 .5 w a te r c o l 0 -3 4 -7 1 2 -1 5

S e d im e n t D e p th (c m )

Figure 4.1. Sediment and overlying water pH at collection site. Samples were taken on 6/27/17. Each bar represents an average. The “water col” label indicates pH of the overlying water. n = 3 per bar. Error bars = standard error of the mean.

e 1 .2 5

a t

1 .0 0

o i

t a

r 0 .7 5

u t

a S

0 .5 0

e t

i n

o 0 .2 5 g

a r

A 0 .0 0 w a te r c o l 0 -3 4 -7 1 2 -1 5 S e d im e n t D e p th (c m )

Figure 4.2. Sediment and overlying water aragonite saturation state at collection site. Samples were taken 6/27/17. Each bar represents an average. The “water col” label indicates pH of the overlying water. n = 3 per bar. Error bars = standard error of the mean.

The 6 cm sediment cores from the site had a water content of 31%. Sediment cores were comprised of 17.3% gravel or very course sand (greater than 840 µm), 5.0% coarse to medium sand (840-425 µm), 76.0% fine to very fine sand (425-75 µm), and 1.7% silt or mud (less than

75 µm). The sieved sediment from the top ~20 cm had a water content of 30.4%. This sediment was comprised of 9.5% gravel or very coarse sand, 4.1% coarse to medium sand, 84.6% fine to very fine sand, and 1.8% of silt or clay. The organic matter of all the sediment was between 4.0 and 4.3%.

4.3.2. Solution Chemistry of Aquarium Water

Mean pH of the overlying water in all aquaria remained between 7.83 and 8.11 for the duration of the experiment (Figure 4.3.A), although the gradual increase from the low to high end of this range was significant at 24°C (linear regression for 24°C, p < 0.001; linear regression for 6.5°C, p = 0.114). Changes in total alkalinity varied with temperature. The total alkalinity significantly increased in the 24℃ group from 2420 ± 40 to 3360 ± 470 µmol kg-1 during the experiment (linear regression, p < 0.001). In the 6.5℃ group, the total alkalinity fluctuated between 2467 ± 587 and 3140 ± 68 µmol kg-1 (linear regression, p = 0.034) (Figure 4.3.B). The

Ωar at both temperatures followed the trends observed in total alkalinity (Figure 4.3.A). In the

24°C group, mean Ωar increased significantly from approximately 1.6 to 3.4 (linear regression, p

< 0.001) while the 6.5°C group fluctuated between 0.9 and 1.5 (linear regression, p = 0.006)

(Figure 4.3.A).

e 4 .0 8 .5 t

a 2 4  C p H

t 3 .5 S

p 6 .5  C p H

n 3 .0 o

( t 8 .0 N 2 4  C s a t. s ta te

a 2 .5

B r

u 6 .5  C s a t. s ta te t

2 .0 a

S S

1 .5 a e 7 .5

l t

e i

)

n 1 .0 o

g 0 .5 a r

A 0 .0 7 .0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 D a y

) 4 0 0 0 1 - 2 4  C T A

l 3 5 0 0 6 .5  C T A

o m

µ (

3 0 0 0

n i

l 2 5 0 0

l A

2 0 0 0

l a

t o

T 1 5 0 0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 D a y

Figure 4.3. Means of overlying tank water pH, aragonite saturation state, and total alkalinity. Graphs showing the pH, aragonite saturation state (A), and total alkalinity (TA) (B) of the overlying water in both 6.5℃ and 24℃ tanks over time. Blue points represent the 6.5℃ tanks (n = 3 per point) while red points represent the 24 ℃ tanks (n = 2 per point). Error bars are ± standard error of the mean.

4.3.3. Chemistry of Sediment with No Crushed Shell

Significant acidification of the surface sediment from initial mean pH values (~6.9) to a final mean pH ~6.0 was achieved in all sediment beds without crushed shell (linear regression, p

< 0.001; Figure 4.4.A). The time that it took to reach this pH and time the pH remained around this value were temperature dependent. For the 24°C treatment, mean pH dropped below 6.0 after 8 days, while for the 6.5 °C treatment, it took 20 days. Mean pH for the 24 °C and 6.5 °C treatment was ≤ 6.0 for 18 and 28 days, respectively (Figure 4.4.A). Mean pH of the sediment at

4-7 cm followed a similar trend but never went below 6.4 (Figure 4.4.B). Mean pH values at 12-

15 cm sediment depth remained between 6.6 and 7.0 (Figure 4.4.C).

e 1 .0 0 7 .5

t 2 4  C p H

t S

p 7 .0 6 .5  C p H

n 0 .7 5 o

( t

N 2 4  C s a t. s ta te

r 6 .5

u 6 .5  C s a t. s ta te t

0 .5 0 a

S S

6 .0

a e

l t

e i

)

n 0 .2 5

o 5 .5 g

a r

A 0 .0 0 5 .0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 B D a y

e 2 .0 0 7 .5

t a

t 1 .7 5 S

7 .0

n 1 .5 0 o

( t N

a 1 .2 5

r 6 .5

u t

1 .0 0

S S

6 .0

0 .7 5 a e

l t

e i

)

n 0 .5 0

o 5 .5

g 0 .2 5

a r

A 0 .0 0 5 .0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 C D a y

e 2 .0 0 7 .5

t a

t 1 .7 5 S

7 .0

n 1 .5 0 o

( t N

a 1 .2 5

r 6 .5

u t

1 .0 0 a

S S

6 .0

0 .7 5 a e

l t

)

n 0 .5 0

o 5 .5

g 0 .2 5

a r

A 0 .0 0 5 .0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 D a y Figure 4.4. Mean sediment pH and aragonite saturation state (no crushed shell). Graphs showing the pH and aragonite saturation state of the sediment beds that had no crushed shell added. Each graph represents a different sediment depth: (A) 0-3 cm; (B) 4-7 cm; and, (C) 12-15 cm. Blue points represent the 6.5 ℃ tanks (n = 3) while red points represent the 24 ℃ tanks (n = 2). Error bars represent ± standard error of the mean.

Total alkalinity (TA) of the surface sediment followed the same trends observed in the pH. Mean TA of the 24°C and 6.5°C sediment surface groups significantly decreased from their respective initial values of 4095 ± 145 and 3063 ± 554 µmol kg-1 down to 410 ± 310 and 560 ±

285 µmol kg-1 by day 39 (linear regression, p = 0.006 for 24°C and p < 0.001 for 6.5°C) before starting to increase (Figure 4.5.A). At the 4-7 cm depth, an increase mean TA in the 24°C group was detected from 3820 ± 200 to 9935 ± 2695 µmol kg-1. In contrast, mean TA in the 6.5°C treatment decreased from 3023 ± 168 to 2433 ± 508 µmol kg-1 (Figure 4.5.B). Neither of these changes in TA at the 4-7 cm depth were significant (linear regression, p = 0.066 for 24°C and p =

0.352 for 6.5°C). At the 12-15 cm depth, mean TA increased in the 24°C treatment over the duration of the experiment from 3885 ± 575 to 14505 ± 6965 µmol kg-1, but high variation between replicates prevented this increase from being statistically significant (linear regression; p = 0.172). In the 6.5°C treatment, mean TA fluctuated between 3513 ± 941 – 5907 ± 490 µmol kg-1 (linear regression; p = 0.951) (Figure 4.5.C).

) 5 0 0 0 1

- 2 4  C T A

l 4 0 0 0 6 .5  C T A

m µ

(

3 0 0 0

i n

l 2 0 0 0

k l

1 0 0 0

a t

o T 0 B 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 D a y

) 1 5 0 0 0

g k

1 2 5 0 0

l o

m 1 0 0 0 0

µ (

y t

i 7 5 0 0

n i

a 5 0 0 0

l A

l 2 5 0 0

t o

T 0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 C D a y

) 2 2 5 0 0

1 -

g 2 0 0 0 0

l 1 7 5 0 0 o

m 1 5 0 0 0 µ

(

y 1 2 5 0 0

t i

n 1 0 0 0 0

i l

a 7 5 0 0

k l

A 5 0 0 0

l a

t 2 5 0 0

o T 0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 D a y

Figure 4.5. Mean sediment total alkalinity (no crushed shell). Graphs showing the total alkalinity (TA) of the sediment beds that had no crushed shell added. Each graph represents a different sediment depth: (A) 0-3 cm; (B) 4-7 cm; and, (C) 12-15 cm. Blue points represent the 6.5℃ tanks (n = 3) while red points represent the 24℃ tanks (n = 2). Error bars represent ± standard error of the mean.

The Ωar for surface sediment followed the same pattern observed for the pH. Mean Ωar was the lowest at the sediment surface in both temperature treatments since it dropped below

0.04 by day 20 and remained this low until day 53 (Figure 4.4.A). By the end of the experiment, mean Ωar at the 24°C had risen back to ~0.15 while mean Ωar at 6.5°C stayed below 0.04 (linear regression, p = 0.840 for 24°C and p = < 0.001 for 6.5°C). For sediments at the 4-7 cm and 12-15 cm depths in the 6.5°C treatment, initial mean Ωar was ~0.25, which decreased significantly to

~0.10 by the end of the experiment (linear regression, p ≤ 0.013). In the 24°C treatment, the Ωar of the 4-7 cm and 12-15 cm depths increased over the course of the experiment until they exceeded 0.5 and 1.0 respectively; however, the changes were not statistically significant (linear regression, p = 0.052 for 4-7 cm and p = 0.096 for 12-15 cm) (Figure 4.4.B; Figure 4.4.C).

4.3.4. Chemistry of Sediment with Crushed Shell

When crushed shell was added on day 33, mean pH of the sediment surface increased from 6.24 ± 0.01 to 6.80 ± 0.02 for the 24°C treatment, and from 6.16 ± 0.03 to 6.55 ± 0.10 for the 6.5°C treatment within 6 days (Figure 4.6.A). During this time, mean TA also increased from

790 ± 690 to 4585 ± 255 µmol kg-1 for the 24 °C treatment and from 443 ± 99 to 3816 ± 474

µmol kg-1 for the 6.5°C treatment. The changes in sediment surface mean pH and mean TA for

24°C and 6.5°C crushed shell groups over the course of the experiment were both statistically significant (linear regression, p ≤ 0.012). It is possible that mean pH and TA increases occurred in less than 3 days after the crushed shell addition since sediment chemistry was not checked sooner than this. The sediment surface mean Ωar (starting at ~0.01) significantly increased after the crushed shell was applied to 0.62 ± 0.01 and 0.22 ± 0.01 for the 24°C and 6.5°C treatments, respectively, by the end of the experiment (linear regression, p ≤ 0.001) (Figure 4.7.A). Mean pH, TA, and Ωar all remained above the values they reached after 6 days of exposure to crushed shell for the rest of the experiment. (Figure 4.6.A; Figure 4.7.A). A full layer of crushed shell was still visible on the sediment surface at the end of the experiment.

e 1 .0 0 7 .5

t 2 4  C p H a

t S

p 7 .0 6 .5  C p H

n 0 .7 5 o

( t

N 2 4  C s a t. s ta te a

r 6 .5

u 6 .5  C s a t. s ta te t

0 .5 0 a

S S

c 6 .0

a e

l t

e i

) n 0 .2 5

o 5 .5

g a r

A 0 .0 0 5 .0 B 3 0 4 0 5 0 6 0 7 0 D a y

e 2 .0 0 7 .5

t a

t 1 .7 5 S

p 7 .0

n 1 .5 0 o

( t

a 1 .2 5

r 6 .5

u t

1 .0 0 a

S S

6 .0

0 .7 5 a e

l t

e i

)

n 0 .5 0

o 5 .5

g 0 .2 5

a r

A 0 .0 0 5 .0 3 0 4 0 5 0 6 0 7 0 C D a y

e 2 .0 0 7 .5 t

t 1 .7 5 S

p 7 .0

n 1 .5 0 o

( t

a 1 .2 5

B r 6 .5

u t

1 .0 0 a

S S

6 .0

0 .7 5 a e

l t

e i

) n 0 .5 0

o 5 .5

g 0 .2 5 a

A 0 .0 0 5 .0 3 0 3 5 4 0 4 5 5 0 5 5 6 0 6 5 7 0 7 5 D a y

Figure 4.6. Mean sediment pH and aragonite saturation state (with crushed shell). Graphs showing the pH and aragonite saturation state of the sediment beds after the addition of crushed shell 33 days after storage in the 20 gallon tanks. Each graph represents a different sediment depth: (A) 0-3 cm; (B) 4-7 cm; and, (C) 12-15 cm. Blue points represent the 6.5℃ tanks (n = 3) while red points represent the 24°C tanks (n = 2). Error bars represent ± standard error of the mean.

) 8 0 0 0 1 - 2 4  C T A

7 0 0 0

l 6 .5 C T A

o 6 0 0 0 

m µ

( 5 0 0 0

y t

i 4 0 0 0

n i

l 3 0 0 0

a k

l 2 0 0 0 A

l a

t 1 0 0 0

o T 0 B 3 0 3 5 4 0 4 5 5 0 5 5 6 0 6 5 7 0 7 5 D a y

) 2 5 0 0 0

1 -

l 2 0 0 0 0

µ (

1 5 0 0 0

n i

l 1 0 0 0 0

l A

5 0 0 0

a t

o T 0 C 3 0 3 5 4 0 4 5 5 0 5 5 6 0 6 5 7 0 7 5 D a y

) 2 0 0 0 0

1 -

o 1 5 0 0 0

(

y t

i 1 0 0 0 0

n i

a k

l 5 0 0 0 A

a t

o T 0 3 0 3 5 4 0 4 5 5 0 5 5 6 0 6 5 7 0 7 5 D a y Figure 4.7. Mean sediment total alkalinity (TA) with crushed shell. Graphs showing TA of the sediment beds that had crushed shell added to them 33 days after being placed in the 20 gallon tanks. Each graph represents a different sediment depth: (A) 0-3 cm; (B) 4-7 cm; and, (C) 12-15 cm. Blue points represent the 6.5°C tanks (n = 3 ) while red points represent the 24 °C tanks (n = 2). Error bars represent ± standard error of the mean.

Some of the sediment chemistry parameters at the 4-7 cm depth of the crushed shell treatments also changed during the experiment. Mean pH remained between ~6.4 and ~6.8 in both temperature treatments (Figure 4.6.B), however, mean TA in the 24°C and 6.5°C treatments

(initially 2915 ± 1425 and 3720 ± 702 µmol kg-1, respectively) increased to 15830 ± 7480 and

6397 ± 610 µmol kg-1 within 20 days of crushed shell application (Figure 4.7.B). Mean aragonite

Ωar at 24°C increased from an initial value of 0.12 ± 0.07 to 0.65 ± 0.05 after being exposed to crushed shell for 20 days, but in the 6.5°C treatment, mean Ωar remained around 0.10 (Figure

4.6.B). The change in mean pH from day 53 to 74 at 6.5°C was statistically significant (Repeated measures ANOVA, p = 0.024; Tukey’s multiple comparison test, p = 0.024), but all other changes were not (Repeated measures ANOVA, p ≥ 0.107).

Sediment chemistry at the 12-15 cm depth of the groups treated with crushed shell were relatively stable for the duration of the experiment. Mean pH for both temperature groups varied between 6.75 and 7.00 for the duration of experiment (Figure 4.6.C). Mean TA for the 24°C treatment increased from 10975 ± 1455 µmol kg-1 on day 33 to 15780 ± 1090 µmol kg-1 on day

53, but the change was not significant (Repeated measures ANOVA, p = 0.431). The increase in the 6.5°C treatment from 5637 ± 43 to 7247 ± 123 µmol kg-1 during this period however was significant (Repeated measures ANOVA, p = 0.008; Tukey’s multiple comparison test, p =

0.006) (Figure 4.7.C). Mean Ωar for the 24°C treatment increased from 0.72 ± 0.14 to 1.25 ± 0.06 by day 53, but this change was not statistically significant (Repeated measures ANOVA, p =

0.308). No significant changes were detected in the Ωar of the 6.5°C treatment, which remained around 0.25 (Repeated measures ANOVA, p = 0.138) (Figure 4.6.C).

4.4. Discussion

Sediment acidification rate is temperature dependent because it is fueled by the metabolism of the resident bacteria. Consequently, incubation temperature will dictate the duration of a stable surface sediment pH. Two different temperatures (6.5°C and 24°C) were used to demonstrate the effect temperature has on acidification rate, and the stability of pH and total alkalinity. At both temperatures, surface sediment pH stabilized at ~6.0 before starting to increase, possibly because of the limitation on the bioavailable organic matter (130). In the 24°C group, mean pH began to increase ~30 days sooner than the pH of sediments in the 6.5°C group because bacteria consume their substrates faster at higher temperatures (Figure 4.4.A) (146). The surface sediment at both temperatures reached a pH of 6.2–6.5 by the end of the experiment

(Figure 4.4.A). The majority of acidification occurs in the top few centimeters of the sediment in natural settings due to the availability of oxygen. This allows for the rapid production of CO2

(which decreases pH) during aerobic respiration and oxidation of H2S from depth, which lowers

+ 2- pH when it is converted into H and SO 4 (147). The ammonia and phosphate produced during organic matter remineralization increase pH but many times more CO2 is produced than ammonia or phosphate on a molar basis (130,148,149). Mean pH changed in the mid depths of the sediment during the trial, but only about half as much as occurred at the surface sediment

(Figure 4.4.B). The sediments at the bottom of the beds and the overlying water however both had a stable pH for the duration of the experiment (Figure 4.3.A; Figure 4.4.C). This may be because some of the CO2 produced from the various types of anaerobic respiration at lower depths of the sediment column were used to fuel methanogenesis, the type of anaerobic respiration used when CO2 is the only electron acceptor available (130).

Mean total alkalinity levels of the overlying water and surface sediment followed the pH trends. The total alkalinity levels of the 4-7 and 12-15 cm sediment depths in the 24°C treatment; however, did not follow pH trends. Instead, they increased over time to values that were at least double the collection site averages (Figure 4.5.B, Figure 4.5.C, Figure 4.7.B, Figure 4.7.C). Total alkalinity increased with depth in the sediment beds regardless whether shell was added to the surface (Figure 4.5.; Figure 4.7.). The total alkalinity of the 24℃ overlying water also increased over time but not in the 6.5°C overlying water (Figure 4.3.B). This suggests that the increase in mean total alkalinity may have been caused by sulfate reduction occurring at depth in the sediment (150). Sulfate reduction produces alkalinity, but the added alkalinity can be removed by regular water exchange between the sediment and water column (150). The absence of water changes in the current study likely allowed alkalinity to build up in the sediments and thus act as an alkalinity source to the overlying waters. The rate of microbial metabolism increases with temperature, which is likely why the large total alkalinity increases were seen only in the 24°C group (151). Changing the overlying water regularly should prevent total alkalinity increases from occurring in the overlying water and sediment depths.

Mean pH, total alkalinity, and Ωar of the 24°C and 6.5°C acidified surface sediments all increased from the time of crushed shell application such that the Ωar reached values equal or greater to those found at the collection site surface sediment (0.22 ± 0.05) by the end of the experiment (Figure 4.6.A; Figure 4.7.A). The crushed shell might also be applied to the sediment before it acidifies as a preventive measure as other researchers have done (60,138). A treatment with a similarly sized material not made of calcium carbonate (such as gravel) was not used in the current study so it is difficult to pinpoint how much of the change in chemistry is due to possible mixing of overlying water with sediment pore water and how much of it was due to

CaCO3 dissolving. Either way, the crushed shell seems to be an effective of way of counteracting natural surface sediment acidification in the lab. A single treatment of 150 grams crushed shell per sediment bed (spread evenly across the sediment surface to obtain 1.4 g cm-2) kept sediment pH at ~7.0 for at least a month (Figure 4.6.A). All treated sediment beds had a full layer of crushed shell at the end of the experiment, which indicates there was likely enough crushed shell remaining to buffer the surface sediments beyond a month. Researchers may be able to obtain the same effect with less crushed shell than 150 grams if they crush the shell down to a size of 1 mm instead of the 3 mm size used in the current study, since smaller shell pieces dissolve faster

(138).

The method described here can be used to acidify surface sediments for experiments with benthic organisms lasting about around 30 days if a sediment surface pH of around 6.0 is desired.

The possibility exists for experiments to be carried out with sediment surface pH values around

6.5 after 55 days of incubation at 24°C, but the stability of this pH past 13 days is unknown. This would allow researchers to investigate effects of short-term or medium-term sediment acidification impacts on benthic organisms. The organisms to be used in an experiment should not be added to the sediment until after the initial sediment incubation period of 10 to 20 days

(depending on researcher’s choice of temperature) in the aquaria to insure the chemistry has stabilized within the desired range. Crushed shell should be added to the sediment beds intended to be used as controls at the start of the sediment incubation period to keep chemistry similar to that of the collection site. Regardless of the desired pH or temperature, it should not be assumed that this method works the same for all types of sediments because the sediment pH can change with organic matter content and grain size (132). Researchers should run a preliminary experiment with the sediment they intend to use in their experiments to insure acidification

83 intensity and duration are sufficient for the planned experiment. Application of crushed shell as a way to counteract (buffer) acidification in control groups should be included in the preliminary experiment. The duration and temperature of the sediment incubation as well as the crushed shell amounts can all be modified to meet the researcher’s needs.

The natural acidification method may not be suitable for all acidification experiments since it has some limitations. Temperature affects a variety of biological and physiological processes (such as behavioral responses of clams to acidification), which can cause confounding effects when trying to measure the impacts of both acidification and temperature together (152).

A control and acidified group with replicate tanks should be run at each temperature so the effects of temperature alone can be isolated. If only one temperature is being used then experiments can be run for as long as the sediment chemistry will stay within the desired range.

Sediment pH can only be constrained to values of ~6.0 or ~6.5 rather than the wide range of values that can be obtained by bubbling CO2. The sediment must also be collected at least 10 days before the animals are placed in the sediment to allow for acidification to take place. Spatial and temporal variation in sediment bacterial populations between different samples of sediment may also affect the acidification rates. Multiple sediment beds should be used within each group of the experiment to help account for this. If other environmental parameters or pollutants are being tested, in addition to sediment chemistry, the researchers must insure during a preliminary trial to determine if the additional variable(s) impact the acidification rate of the sediment.

Other methods exist to acidify sediments, although they each have their own set of limitations as well (136). The method used most often is to bubble CO2 gas directly into the sediment. This allows for precise control of the sediment pH for any duration, but does not mimic natural acidification. Gelatin can be placed under the acidified sediment to keep the pH

84 stable in both lab and field sediments, but the stability remains for a short duration (e.g., 24-48 hours) (109). Another method is to find sediments already at the desired pH in the field and collect them using a core sampler. This method mimics the natural environment but can only be used for short-term experiments since the native sediment bacteria will begin to change the sediment chemistry (109).

4.5. Conclusion

The primary purpose of the current work was to show that natural acidification is possible in the laboratory. The method described here can be used to acidify naturally the surface of sediments that consist of at least 70% fine to very fine sand kept within recirculating aquaria down to pH ~6.0 and an Ωar ~0 for 20 to 50 days, depending on temperature (Figure 4.4.A). This pH corresponds to some of the more acidic values recorded in surface sediments (132). Adding crushed shells of E. leei to the sediment surface increased the mean pH to values similar to those at the collection site (~6.8) within three days of application (Figure 4.6.A). Crushed shells can thus be used to maintain a control group with a pH close to that of the field sediment. Additional studies will be needed to test if the natural acidification method works with other sediment types as well as flow-through systems. There is no single method that can be used for all possible sediment acidification experiments; however, the current study should help expand the tool kit for researchers studying the effect of sediment acidification on benthic organisms.

CHAPTER 5

SEDIMENT PH AND TEMPERATURE EFFECTS ON

ENSIS LEEI IMMUNE CAPACITY

5.1. Introduction

The pH values of surface sediments can range from 6.0-8.2 across various marine environments including temperate and tropical coastal seas, and deep-sea sediments. This means the sediment surface is often more acidic than the seawater above it (130,132). The sediment acidification is caused by the rapid decay of organic matter at the sediment surface often leads to the release of CO2 as well as decreases in sediment pore water pH, alkalinity, and Ωar. The most intense acidification occurs at the sediment surface because oxygen is required for rapid organic matter decomposition, which is usually only available in the top few centimeters (50,130).

Organic matter settling from the water column above fuels this process. Organic matter, however, is not evenly distributed on the sediment surface. In coastal areas, large concentrations of organic matter can be deposited on the sediment surface through die-offs of algal blooms that result from nutrient run-off from the land (131). The organic matter can then be redistributed through the burrowing action of benthic macrofauna, which also affects sediment exposure to oxygen (130). Values on the lower end of the 6.0-8.2 range have been observed at individual sites in both the Bay of Fundy (132) and the coast of Maine (60,133), making sediment acidification a concern for these areas. Sediment acidification is expected to intensify in the future although the site-specific nature of sediment acidification and limited data available make it difficult to predict how quickly acidified areas will expand (134,153).

One of the primary concerns with sediment acidification is its impact on clams. For example, soft-shell clams have great economic value, with commercial landings in Maine worth over $16,000,000 in 2018 alone (154). The developing clam aquaculture industry has the potential to create even more revenue by supplying clams to offset the decreasing yields of wild harvests (155). Research on the impacts of sediment acidification on clams is still in its infancy, but studies done to date indicate changes in sediment pH (lowering) can affect shell formation, burrowing behavior, rate of heavy metal uptake, and mortality rate (109,135,138). In some cases, pH levels in the field have been low enough to dissolve clam shells (60). Impacts of sediment pH on clam immunity are currently unknown.

Coping with corrosive sediments is particularly important for razor clams, such as the

Atlantic jackknife clam E. leei. This species must spend the majority of its life at the surface of the sediment due to its short siphons and has no way to seal its soft tissues off from the external environment (1). Although very little is known about the potential pathogens of E. leei (1), the potential remains for sediment acidification to impact the health of E. leei. Information on the health of the E. leei will be needed to assess the health of stocks as the razor clam aquaculture industry develops in the Northeast.

The goal of the current study was to assess if the immune capacity of E. leei is impacted by temperature and acidified sediment, as both individual and combined stressors. Surface sediment acidification was achieved by incubating the sediment beds in recirculating aquaria, allowing the sediment bacteria to produce CO2 as they metabolize the organic matter in the sediment. A series of hemocyte parameters were measured to determine immune capacity after

30 days of exposure to experimental treatments, including total hemocyte counts, differential hemocyte counts, cell viability, and proportion of phagocytic cells. In addition, a condition index

87 was calculated using the wet flesh and dry shell weight as an overall health indicator. These parameters were measured in clams less than 30 hours after they were removed from their natural habitat to determine if holding the clams in aquaria altered any of the parameters.

5.2. Methods

5.2.1. Sediment and Clam Collection

Sediment and clams (shell length 17.6 ± 0.9 cm) were collected from Dobbins Island in

Beals, ME (44°30'25.0"N 67°36'15.1"W) in November 2017 from the lowest portion of the intertidal zone, which was exposed only during astronomically low tides. A separate trip was made for just sediment in October 2017. The pH of the overlying water, as well as the sediment at 0-3 cm (surface sediment), 4-7 cm, and 12-15 cm depth were recorded from 10 mL pore water samples taken via a 20 mL syringe (n = 3 for each depth). The contents of the tubes were immediately analyzed with an Accumet portable pH meter (model AP115) equipped with an

Accumet pH/ATC probe. The sphere of influence for a 10 mL draw in the sediment was determined to be 3 cm or less based on a separate study that used the syringe method (115). Pore water samples were gently poured into 15 mL tubes, each containing 10 µL of saturated mercuric chloride solution to preserve the sample until it could be analysed for total alkalinity. The tubes were inverted five times to insure the mercuric chloride spread throughout the sample. Overlying water temperature and salinity were also recorded using a pH probe and an ATC portable refractometer.

The top ~20 cm of sediment was collected using a shovel and transferred to the laboratory in 5-gallon buckets. All rocks and large invertebrates were removed from the sediment using a sieve with 2.54 mm mesh within 10 hr of sediment collection. Ensis leei shells

88 also were collected at the Dobbins Island site for later application as a buffering agent on the sediment to be used in the experiment. Upon return to laboratory, shells were rinsed with distilled water and heated (~100 °C for 2 hr) before they were ground up to a chip size of < 3 mm using a mortar and pestle. Ensis leei shells were chosen because they are made primarily of aragonite, which dissolves faster than other forms of calcium carbonate (141,142).

5.2.2. Experimental Setup

The sediment collected in October was distributed among 16 rectangular polystyrene containers measuring 18  12.5  12.5 cm (height, length and width). Each container was filled to 15 cm with sieved sediment and placed within eight aquaria (75 L capacity) filled with about

60 L of 33 ppt artificial seawater made with Tropic Marin Pro Reef mix. Each aquarium had two of the sediment containers placed in them. In each aquarium, one sediment container was briefly removed to add 50 g of crushed E. leei shell (0.5 g cm-2) sprinkled on its surface to keep pH levels similar to collection values. The sediment containers were then incubated for 14 days at 10

± 0.2°C (mean ± standard deviation) to allow acidification to occur in the containers that had no crushed shell added. Another four sediment containers were prepared in November (on the same day all clams were collected) and distributed among four of the aquaria. At this point, half of the aquaria had three sediment containers in them and half of the aquaria had two sediment containers in each of them. A single clam was placed on the surface of the sediment in each of these sediment containers. All clams buried into their respective sediments within an hour. Less than 30 hours after being placed on the sediment, the aquaria with three clams each had a clam sampled and removed from the November sediment container. All of the November sediment containers were then removed from the aquaria so that each one had two sediment containers containing a clam remaining (one treated with crushed shell and one without). The seawater

89 temperature in half of the tanks was then gradually raised over a 10-day period to 20 ± 0.5°C using aquarium heaters, while the other four remained at 10 ± 0.2°C. Tanks were maintained at these temperatures for the 30-day experimental exposure period that began after this acclimation period. The four treatment groups were as follows: 6.8 pH (Ωar 0.19 ± 0.01) sediment at 10°C

(control group), 6.8 pH (Ωar 0.22 ± 0.13) sediment at 20°C, 6.0 pH (Ωar 0.01 ± 0.01) at 10°C, and

6.1 pH (Ωar 0.04 ± 0.05) sediment at 20°C. Each of these groups was replicated four times with one clam in each of the sediment containers. All tanks were kept in the same laboratory and constantly aerated with aquarium air pumps and water at the surface level of the sediment was kept in constant motion using an aquarium circulation pump (Hydor Koralia, 240 GPH). The clams were fed ad libitum with instant algae (shellfish diet 1800, Reed Mariculture) on a daily basis during the both the acclimation and exposure periods.

5.2.3. System Monitoring

Tank water pH, temperature, salinity and dissolved oxygen were monitored daily using the Accumet® portable pH meter, a refractometer, and a YSI EcoSense® dissolved oxygen meter

(model DO200A) respectively. The pH, salinity, and TA of the sediment was monitored via pore water samples taken by syringe at the top of the sediment column (0-3 cm) every 2-3 days, and from the middle (4-7 cm) and bottom (12-15 cm) depths at the start and end of the exposure period. The pH and salinity of pore water was measured immediately after 10 mL of it was expelled into a 15 mL tube that was then poisoned with 10 µL of saturated mercuric chloride to preserve the sample until it could be analyzed for TA. The titrations for total alkalinity were performed using the open cell titration method (116), which was modified for the smaller sample volumes used in this analysis. A 5 mL sample was placed into a 15 mL tube and titrated with

0.01N hydrochloric acid prepared using a 33 g/L NaCl solution down to a pH of 3.00. Mixing

90 between acid additions was achieved by moving the probe up and down. The Ωar was calculated from TA, pH, and salinity values using CO2SYS (ver. 2.1 http://cdiac.ess- dive.lbl.gov/ftp/co2sys/) and the K1 and K2 values recommended for estuarine waters (117). A

90% water change was carried out on all tanks on a weekly basis to maintain water quality.

Water quality was monitored using API® ammonia, nitrite, and nitrate test kits.

5.2.4. Burrowing Rates and Survival

Each of the four treatments had four replicates comprised of a single clam in a single sediment container. All of these clams burrowed into their respective sediment beds within an hour of being placed there. All clams in the 10°C 6.0 pH group survived the 30-day exposure period while three clams survived in the 10°C 6.8 pH group and 20°C 6.1 pH group. Only one clam in the 20°C 6.8 pH group survived.

Table 5.1. Summary of environmental data from collection site and laboratory aquaria over the course of the experiment. The dissolved oxygen of all tanks was always kept above 7.50 mg/L through constant aeration. pCO2 and Ωar were calculated for each tank using CO2SYS from the temperature, salinity, pH and total alkalinity values. All values are means ± SD.

Sample Description Salinity Temp pH Total pCO2 Ωar (ppt) (˚C) (NBS Alkalinity µatm Scale) (µmol kg-1) Collection site overlying 33 7.2 ± 0.1 7.74 ± 2263 ± 40 762 1.03 ± water 0.02 0.03 Collection site sediment, 33 7.1 ± 0.1 7.18 ± 2083 ± 263 3298 0.31 ± 0-3 cm deep 0.30 0.18 Collection site sediment, 33 6.8 ± 0.1 7.29 ± 2053 ± 407 2220 0.32 ± 4-7 cm deep 0.02 0.06 Collection site sediment, 33 6.7 ± 0.1 7.24 ± 1653 ± 625 2047 0.23 ± 12-15 cm deep 0.06 0.09 20°C tank water 33 20.0 ± 7.95 ± 3099 ± 369 718 3.55 ± 0.5 0.06 0.64 20°C tank sediment, 33 20.0 ± 6.05 ± 1915 ± 2253 44052 0.04 ± 0-3 cm deep, no shell 0.5 0.24 0.05 20°C tank sediment, 33 20.0 ± 6.82 ± 2526 ± 1494 8910 0.25 ± 4-7 cm deep, no shell 0.5 0.18 0.20 20°C tank sediment, 33 20.0 ± 6.98 ± 2466 ± 1431 6723 0.34 ± 12-15 cm deep, no shell 0.5 0.20 0.24 20°C tank sediment, 33 20.0 ± 6.78 ± 2527 ± 1525 11859 0.22 ± 0-3 cm deep, w/shell 0.5 0.17 0.13 20°C tank sediment, 33 20.0 ± 6.90 ± 2493 ± 794 7669 0.26 ± 4-7 cm deep, w/shell 0.5 0.08 0.09 20°C tank sediment, 33 20.0 ± 6.96 ± 2743 ± 566 7203 0.34 ± 12-15 cm, w/shell 0.5 0.09 0.11 10°C tank water 33 10.0 ± 7.90 ± 2896 ± 275 751 1.87 ± 0.2 0.05 0.30 10°C tank sediment, 33 10.0 ± 6.00 ± 804 ± 648 16896 0.01 ± 0-3 cm deep, no shell 0.2 0.24 0.01 10°C tank sediment, 33 10.0 ± 6.66 ± 1565 ± 1032 8292 0.07 ± 4-7 cm deep, no shell 0.2 0.18 0.05 10°C tank sediment, 33 10.0 ± 6.92 ± 1706 ± 886 4254 0.15 ± 12-15 cm deep, no shell 0.2 0.17 0.11 10°C tank sediment, 33 10.0 ± 6.82 ± 2894 ± 1320 10219 0.19 ± 0-3 cm deep, w/shell 0.2 0.18 0.10 10°C tank sediment, 33 10.0 ± 6.96 ± 2569 ± 728 6247 0.23 ± 4-7 cm deep, w/shell 0.2 0.17 0.13 10°C tank sediment, 33 10.0 ± 7.05 ± 2614 ± 482 5347 0.28 ± 12-15 cm deep, w/shell 0.2 0.13 0.14

5.2.5. Bacteria Preparation

This selective medium insured growth of a pure culture of V. anguillarum with d-Tomato. The bacteria was then re-suspended in a 33 ppt solution of filtered artificial seawater (made with

Crystal Sea Marine Mix) at a concentration of about 2.0 x 108th cells ml-1 (0.2 absorbance at 620 nm).

5.2.6. Hemocyte Collection and Total Counts

At the end of the 30 day exposure period, hemolymph was withdrawn from the mid- section of the foot using a 1 mL syringe fitted with a 23-gauge needle. A slight, but constant, pressure was maintained until hemolymph started to flow in the syringe. After removing the needle, the hemolymph (1 ml per draw) was expelled into a 30 mL tube (through a Fisherbrand

100 µm cell strainer), kept on ice containing 1.5 mL of 2.5% sodium citrate (Fisher Scientific) to inhibit hemocyte aggregation, while still allowing for attachment after removal from ice. This was repeated so a total of 2 mL of hemolymph could be collected from each clam. A separate 30 mL tube was used for each clam. Cell counts were carried out using a hemocytometer and the cell count was adjusted down to 1.0 x 106 cells/ml if needed using filtered artificial seawater

(FASW).

5.2.7. Hemocyte Distribution and Incubation for Immune Assays

The hemolymph from each tube was distributed via pipette into three wells (300 µl per well) of a 24 well culture plate for differential hemocyte counts and trypan blue exclusion cell viability assays. May-Grünwald stain (80 µl) was added to one well for the differential hemocyte counts. The remaining hemolymph was distributed to two cover slips (300 µl each) for later use with the phagocytosis assay. The plates and cover slips were then incubated at 20°C for 45 min to allow hemocytes to attach. During this incubation period, the wells containing the May-

Grünwald stain were photographed on an inverted microscope so differential hemocyte counts could be obtained. Unattached hemocytes were then washed away from the rest of the wells and cover slips using two 500 µl rinses of FASW. After the overlying solutions were removed, both of the cover slips and one of the wells received 200 µl of the V. anguillarum solution. The remaining well received FASW to serve as a control. The cover slips were incubated for 3.5 hours and then photographed under a combination of fluorescent and white light. The 24 well plate was incubated for 5 hours at 20°C before 50 µl of a 0.22 µm filtered 0.3% trypan blue solution was added. The hemocytes in these wells were then photographed using a camera mounted on an inverted microscope. This procedure was carried out for all 11 (of an initial 16) of the surviving clams. All cell counting for subsequent analysis was done manually using the cell counter tool in ImageJ photo software (ver. 1.8.0 for windows https://imagej.nih.gov/ij/download.html).

5.2.8. Differential Hemocyte Count Analysis

A total of at least 150 hemocytes over six fields of view were counted for each clam from micrographs. Each hemocyte was identified as either a granulocyte or hyalinocyte based on the presence or absence of their granules respectively. The percent granulocytes present in a given

94 sample was calculated by dividing the number of granulocytes by the total number of hemocytes counted per animal and then multiplying by 100 (120).

5.2.9. Phagocytosis Analysis

For each clam, at least 100 hemocytes were counted across two cover slips. At least six fields of view were used per cover slip. Hemocytes that had one or more V. anguillarum cells within them (visible due to red fluorescence) were counted as phagocytic hemocytes while those with no V. anguillarum cells were counted as non-phagocytic hemocytes. The number of phagocytic hemocytes was then divided by the total number of hemocytes and multiplied by 100 to obtain the percent phagocytic hemocytes for a single clam. Averages for each treatment group were calculated by taking the average of all clams within that group.

5.2.10. Cell Viability Analysis

A total of at least 50 hemocytes over four fields of view were counted for each well (two wells per clam) from micrographs. Each of these hemocytes was identified as non-viable or viable based on the presence or absence of trypan blue in the cell, respectively. The percent viable cells present in a given animal was calculated by dividing the number of cells that did not take up the trypan blue by the total number of hemocytes counted per animal and then multiplying by 100.

5.2.11. Condition Index

After hemolymph extraction, the clam meat was shucked and immediately weighed. The mass of the hemolymph drawn from the meat was added to the weight. The mass of the shell was recorded to the nearest 0.1 g after being allowed to dry for at least 24 hours on bench top at 21-

23°C. The condition index was then calculated by diving the wet meat weight by the dry shell weight and multiplying by 100 (114).

5.2.12. Statistical Analysis

Two-way ANOVA was used to check for the presence of significant differences caused by temperature (a = 2; 10oC vs. 20oC), pH (b = 2; 6.0 vs. 6.8), and the interaction between these variables in total hemocyte count, differential hemocyte count, phagocytosis, and condition index data sets. Tukey’s multiple comparisons test was used to determine which groups differed from each other if any significant effects were found in any two-way ANOVA test. A one-way

ANOVA test was used to compare the hemocyte viability among all treatments since the purpose of the trypan exclusion test was to insure the hemocytes were still viable after being incubated at

20°C for up to 5 hours. T-tests were used to compare the health parameters of clams held in recirculating tanks for 30 hours to those held for 41 days to determine if holding the clams in the aquaria impacted any of the health parameters. GraphPad Prism (ver. 8.0.1 https://www.graphpad.com/scientific-software/prism/) was used to generate figures and perform statistical tests. All tests were run at the 0.05 significance level (alpha of 0.05). All data analyzed passed normality and equal variance tests.

A power analysis was run to determine if the statistical tests used had sufficient power and minimum detectable difference to detect the mean differences observed in the data. G*Power

(ver. 3.1.9.2 https://g-power.apponic.com/) was used to determine the power and effect size of each statistical test used. Effect size was expressed as a number of standard deviations. For example, an effect size of 1.0 indicates the effect size that can be detected in that test is equal to

1 standard deviation from the mean. If 1 standard deviation is equal to say 5% for a given health parameter then the minimum detectable difference for that test is 5%. Any differences between

96 means smaller than 5% would be undetectable by that test and the experiment would need to be run again with a greater sample size to detect differences smaller than 5%. For this reason, the minimum detectable difference was determined for each health parameter and compared to the differences between means. The number of clams needed in a re-run of the experiment was determined for each health parameter that had a minimum detectable difference larger than the differences observed between means. The total number of clams needed for re-runs of the experiment ranged from 32 to 72 for the two-way ANOVAs used to check for pH and temperature effects, and 34 to 500 for the t-tests used to determine if holding time affects health parameters.

5.3. Results

5.3.1. Total Hemocyte Counts

Neither pH nor temperature had a significant effect on THC (p = 0.790 and 0.868 respectively). The two-way ANOVA interaction result indicated there was no significant interaction between pH and temperature (p = 0.515). The average THC for the 20°C 6.1 pH,

20°C 6.8 pH, 10°C 6.0 pH, and 10°C 6.8 pH groups were 5.83 ± 3.01 x 106, 3.84 x 106, 3.08 ±

2.29 x 106, and 4.52 ± 1.20 x 106 cells mL-1 respectively (Figure 5.1.A). The 2.5% sodium citrate solution and chilling with ice was an effective method of keep hemocytes from aggregating within the first few minutes of being drawn out of the clam (156) (Figure 5.1.B). The 2.5% sodium citrate, however, did not prevent the hemocytes from attaching to the substrate below them during the 20°C incubation period (Figure 5.1.C).

A 8 .0  1 0 6 2 0  C

l 6 .0  1 0 6 1 0  C

s e

y 4 .0  1 0 6

m e 6

H 2 .0  1 0

H H H H p p p p .1 .8 .0 .8 6 6 6 6  C  C  C  C 0 0 0 0 2 2 1 1

T r e a tm e n t

B C

Figure. 5.1. Total hemocyte count results. (A) Graph showing average total hemocyte count. n = 1 for the 20°C 6.8 pH group and n = 3-4 for all other groups. Errors bars represent ± SD. p = 0.790, 0.868, and 0.515 for mean pH, temp, and interaction respectively (2-way ANOVA). (B) Micrograph shows E. leei hemocytes under phase contrast just after being drawn from clam foot. None of the hemocytes were attached to the well bottom at this point. (C) Micrograph of hemocytes attached to bottom of well after a 45 min incubation at 20°C.

5.3.2. Differential Hemocyte Counts

Both pH and temperature had a significant effect on differential hemocyte counts (p =

0.031 and 0.021 respectively). Tukey’s multiple comparisons tests for each possible group comparison, however, all returned a non-significant result (p ≥ 0.59). The two-way ANOVA interaction result indicated there was no significant interaction between pH and temperature (p =

0.125). The mean percent granulocytes for the 20°C 6.1 pH, 20°C 6.8 pH, 10°C 6.0 pH, and

10°C 6.8 pH groups were 58 ± 8%, 33%, 65 ± 7%, 59 ± 9% respectively (Figure 5.2.A). The granulocytes stained dark purple while hyalinocytes appeared clear or lightly stained. All cells appeared round since they were stained before they had a chance to attach to a solid substrate

(Figure 5.2.B).

1 0 0 2 0  C

s 8 0 1 0  C

t y

c 6 0

u n

a 4 0

% 2 0

H H H H p p p p .1 .8 .0 .8 6 6 6 6  C  C  C  C 0 0 0 0 2 2 1 1

T r e a tm e n t

Figure 5.2. Differential hemocyte count results. (A) Graph showing average percent of hemocytes counted in differential stain micrographs that were granulocytes. n = 1 for the 20°C 6.8 pH group and n = 3-4 for all other groups. Errors bars represent ± SD. p = 0.031, 0.021, and 0.125 for mean pH, temp, and interaction respectively (2-way ANOVA). p ≥ 0.59 for multiple comparisons between all groups (Tukey’s multiple comparisons) (B) Micrograph shows E. leei hemocytes stained with May-Grünwald stain, which stains the granules purple.

100

5.3.3. Phagocytosis

Neither pH nor temperature had a significant effect on the percent phagocytic cells (p =

0.282 and 0.166 respectively). The average percent granulocytes for the 20°C 6.1 pH, 20°C 6.8 pH, 10°C 6.0 pH, and 10°C 6.8 pH groups were 88.0 ± 7.0%, 82%, 91.5 ± 4.4%, 89.3 ± 4.2% respectively. The two-way ANOVA interaction result indicated there was no significant interaction between pH and temperature (p = 0.602) (Figure 5.3.A). Vibrio anguillarum were clearly visible since they remained fluorescent even after they were internalized (Figure 5.3.B).

101

A 1 0 0

s 2 0  C

l l

e 1 0  C

C 7 5

t c

o 5 0

a h

f 2 5

% 0

H H H H p p p p .1 .8 .0 .8 6 6 6 6  C  C  C  C 0 0 0 0 2 2 1 1

T r e a tm e n t

Figure 5.3. Phagocytosis results. (A) Graph showing average percent of phagocytic cells. n = 1 for the 20°C 6.8 pH group and n = 3-4 for all other groups. Errors bars represent ± SD. p = 0.282, 0.166, and 0.602 for mean pH, temp, and interaction respectively (2-way ANOVA). (B) Micrograph shows hemocytes after being exposed to 2.0 x 108th cells ml-1 V. anguillarum for 3.5 hours.

102

5.3.4. Cell Viability

There were no significant differences in cell viability among both treatment groups whether or not V. anguillarum was added (p = 0.687). Average cell viability was always at least

91% after the cells had been incubated in the wells for 5 hours at 20°C. Standard deviation for cell viability never exceeded 5%. This indicated that the hemocytes were viable for the immune capacity assays since they were carried out within 5 hours of hemocyte extraction (Figure 5.4.A).

Cells that were non-viable stained a dark blue while the viable ones remained clear (Figure

5.4.B, C).

103

A 1 0 0 2 0  C

8 0 1 0  C

l e

C 6 0

b a

i 4 0 V

% 2 0

g g g g g g g g n n n n n n n n a a a a a a a a ...... V /V V /V V /V V /V o w o w o w o w n n n n H H H H H p H p H p H p p p p p 1 8 0 8 1 . 8 . 0 . 8 . . 6 . 6 . 6 . 6 6 6 6 6 C C C C C  C  C  C   0  0  0  0 0 2 0 2 0 1 0 1 2 2 1 1

T r e a tm e n t

B C

Figure. 5.4. Trypan blue exclusion viability test results. Viable cells remain unstained and the non-viable ones are stained blue. (A) Graph showing average percent of viable cells as determined by the trypan blue exclusion test. n = 1 for the 20°C 6.8 pH groups and n = 3-4 for all other groups. Errors bars represent ± SD. p = 0.687 (one-way ANOVA). (B) Micrograph of hemocytes (no V. anguillarum added) exposed to trypan blue stain. (C) Micrograph of hemocytes (with 2.0 x 108th cells ml-1 V. anguillarum added) exposed to trypan blue stain.

104

5.3.5. Condition Index

Neither pH nor temperature had a significant effect on the condition index of the clams (p

= 0.129 and 0.632 respectively). The two-way ANOVA interaction result indicated there was no significant interaction between pH and temperature on clam condition index (p = 0.544). The average condition index for the 20°C 6.1 pH, 20°C 6.8 pH, 10°C 6.0 pH, and 10°C 6.8 pH groups were 178 ± 11%, 131%, 175 ± 19%, 153 ± 48% respectively (Figure 4.5.).

2 2 5 0

0 2 0  C 1

* 2 0 0 )

s 1 0  C

s 1 7 5

l 1 5 0

e h

s 1 2 5 /

s 1 0 0

t 7 5 e

w ( 5 0

H H H H p p p p .1 .8 .0 .8 6 6 6 6  C  C  C  C 0 0 0 0 2 2 1 1

T r e a tm e n t Figure 5.5. Condition index results. n = 1 for the 20°C 6.8 pH group and n = 3-4 for all other groups. Errors bars represent ± SD. p = 0.129, 0.632, and 0.544 for pH, temp, and interaction respectively (2-way ANOVA).

105

5.3.6. Laboratory Holding Effects on Measured Parameters

The t-tests for the THC, differential hemocyte count, phagocytic activity, and condition index of the clams held in recirculating aquaria for 41 days (11 acclimation days and 30 exposure days) returned a result which suggests the parameters do not differ significantly from the clams held in the aquaria for only 30 hours (p ≥ 0.09 for all comparisons; Figure 5.6.). Power analysis results, however, indicated the sample sizes were not sufficient for these tests to detect a significant difference in the observed differences. The holding effect experiment should be run again with greater numbers to confirm holding clams in aquaria for 41 days has no impact on health parameters (see section 5.3.7. for details). The data in the 41-day groups have been taken from the 10°C, sediment surface pH of 6.8 of the previous figures (Figure 5.1.A; Figure 5.2.A;

Figure 5.3.A; Figure 5.4.A). The clams in the 30-hour group were held at 10°C in sediment taken from the collection site at the same time as the clams (Table 5.1.).

106

A B

T o ta l H e m o c y te C o u n t D iffe re n tia l H e m o c y te C o u n ts

8 .0  1 0 1 0 0 l

6 .0  1 0 6 s 8 0

e m

s c

e 6 0

o l

y 4 .0  1 0 6

n o

a 4 0

r e

6 G

H 2 .0  1 0

% 2 0

0 0 3 0 h o u r s 4 1 d a ys 3 0 h o u r s 4 1 d a ys T im e K e p t in T a n k s T im e K e p t in T a n k s

C D P h a g o c y to s is o f V . a n g u illa r u m C o n d itio n In d e x

1 0 0

0 2 2 5

1 2 0 0

l )

l 8 0 s

e 1 7 5

C a

1 5 0

m i

t 6 0

l l

y 1 2 5

c h

o 1 0 0 s

g 4 0

/ a

h 7 5 h

l 5 0

2 0 f

% t

e 2 5 w 0 ( 0 3 0 h o u r s 4 1 d a ys 3 0 h o u r s 4 1 d a ys T im e K e p t in T a n k s T im e K e p t in T a n k s

Figure 5.6. Health parameters of E. leei kept in ambient conditions compared to those in acidification trial. Treatment from acidification trial being used for comparison was the 10°C tank sediment, 0-3 cm deep, w/shell treatment shown in Table 5.1. Ambient condition clams were sampled within 30 hours of being removed from the collection site. n = 4 for the 30 hours group and n = 3 for the 41 day group. Errors bars represent ± SD. p ≥ 0.09 for the series of t-test run on the 2 groups in each graph.

107

5.3.7. Power Analysis

The power of the two-way ANOVAs (total n of all groups = 11) was determined to be

0.78 at an effect size (also known as the minimum detectable difference) of 1.3. This means there was a 78% chance to correctly identify a difference among the means if that difference was 1.3 average standard deviations or larger. The minimum detectable differences for the differential hemocyte count, phagocytosis, and condition index data were all within the range of mean differences found although they were only lower than the differences of the higher side of the range for their respective data set. Sample size would need to be increased to 8 clams per group

(total of 32 clams) to bring the effect size down to 0.7 while increasing power to 0.89. At an effect size of 0.7, the minimum detectable differences for differential hemocyte count, phagocytosis, and condition index data are lower than all the mean differences except the smallest mean difference for their respective data set. For total hemocyte data, the experiment would need to run again with 18 clams per group (total of 72 clams) to bring the effect size down to 0.4 while maintaining power at 0.80. At an effect size of 0.4, the minimum detectable difference for total hemocyte count data is lower than all but the smallest mean difference in the data set.

The power of the one-way ANOVA for hemocyte viability was 0.78 with an effect size of

1.0 (minimum detectable difference of 3%). Sample size would not need to be increased for this assay if the experiment was ran again. The power for the t-tests used to compare 30 hour and 41 day clams was 0.80 with an effect size of 2.7. The minimum detectable difference for all the health parameters was greater than any of the mean differences seen their respective data set. For differential hemocyte count and phagocytosis data, the experiment would need to be run again with 17 clams in each group (34 clams total) in order to bring effect size down to 1.0 and keep

108 power around 0.80. This would bring the minimum detectable differences below the mean differences for differential hemocyte count and phagocytosis data. For total hemocyte count and condition index data, the experiment would need to be run again with 250 clams in each group

(500 clams total) in order to bring effect size down to 0.25 and keep power around 0.80. At this effect size, minimum detectable differences for total hemocyte count and condition index data would be below the mean differences.

5.4. Discussion All clams burrowed within an hour of being placed on the sediment. The acidified groups

(10°C 6.0 pH and 20°C 6.1 pH group) had mean survival rates of 100% and 75%, respectively, but these were not statistically significant. Results showed that adult E. leei still find sediment habitable when the surface layer has been acidified to a pH 6.0 (more than a whole pH unit lower than the surface sediment at their collection site). A study on another razor clam species,

Sinonovacula constricta (Lamarck, 1818), demonstrated that adult clams (5.3 ± 1.1 cm) can burrow in sediment within a 24-hour timeframe when exposed to acidified overlying water (pH

7.4) although they reduce their burrowing depth by about 0.5 cm (157). These results are in contrast to a study by Clements et al. on M. arenaria juveniles in which only 10% of clams burrowed into sediment cores from their collection site when placed on surface sediments with a pH of 6.84 for 5 hours (109). Similarly, Rodriguez et al. found that less than 5% of R. philippinarum (3.92 ± 0.16 cm long) burrowed into sediments with a pH of 6.1 after 48 hours.

Survival for clams in this treatment after 10 days was 0% (135). This suggests that wild adult E. leei can live in sediments that are too acidic for other clam species. The reason may be related to the large size of the clams (17.6 ± 0.9 cm long) because a previous study showed mortality by acidified sediment decreases as size increases with 0.2 mm, 0.4 mm, and 0.6 mm long M.

109 mercenaria (60). This size dependence of mortality in acidified sediment may be because the larger animals are more resilient to shell dissolution. This is likely because the type of calcium carbonate M. mercenaria uses for its shell switches from amorphous calcium carbonate to the less soluble aragonite when it reaches the later stages of larval development (134,158). The greater shell thickness of adult clams may also make them more resilient to sediment acidification than juveniles since more shell must be dissolved before the clam’s tissues are exposed to the surrounding environment (60). If the same applies for E. leei, then the shells of adults would be expected to be more resilient to dissolution than those of larvae or juveniles even though more shell surface area is exposed to surrounding sediment.

The total hemocyte counts of E. leei in the current study were not significantly impacted by temperature or pH (Figure 5.1.). There are currently no other studies available on how sediment acidification impacts bivalve immunity to allow for direct comparisons. However, studies are available on the impacts of overlying water acidification (usually referred to as ocean acidification) on bivalve immunity. The impact of ocean acidification on bivalve total and differential hemocyte counts appears to vary by species and exposure time. Short-term exposure experiments (3-7 days) to pH 7.3-7.5 seawater have resulted in decreased, unchanged, and increased THC in the Mediterranean mussel M. galloprovincialis Lamarck, 1819, Venus clam

Chamelea gallina (Linnaeus, 1758), and pearl oyster P. fucata respectively (69,159). In medium- term experiments, total hemocyte counts of the blood clam Tegillarca granosa (Linnaeus, 1758) decreased after 14-day exposure to acidified overlying seawater (pH 7.3) while total hemocyte counts of M. mercenaria and M. edulis were unaffected by a month-long exposure to seawater with pH values of 7.59 and 6.50 respectively (59,160–162). Results of the current study are in agreement with these latter 2 month-long studies while the short-term studies suggest the

110 possibility of E. leei THC being impacted by acidification at an earlier point in the 30-day study.

If the THC of E. leei did change early in the study then it must have reverted by the time sampling occurred on day 30. Warming generally leads to higher total hemocyte counts in bivalves (41,104,126,163). However, there are exceptions to this generalization. Total hemocyte counts can decrease if the exposure to elevated temperature continues past four months (41) or the changes may become undetectable due to high variations of metabolism levels between individuals (129). The THC of the E. leei in the current work was not impacted by temperature within 30 days but may have increased if the experiment was run for longer.

The two-way ANOVA for the differential hemocyte counts of E. leei in the current study returned a significant result for temperature and pH impacts, however the comparisons between each group returned a non-significant result (Figure 5.2.). This suggests there is no temperature or pH effect on differential hemocyte counts. Studies over a similar timeframe (32 days) found that the granulocyte proportions in the blue mussel (M. edulis) (59,164) also were unaffected by acidified seawater exposure (pH 6.5). Short term exposures to acidified water, however, seem to impact the differential hemocyte counts. The granulocyte proportion in M. unguiculatus

Valenciennes, 1858 decreased after 72 hour exposure to acidified seawater (pH 7.3) (165). In P. fucata, a 72 hour exposure to pH 7.5 seawater caused an increase in the proportion of both large granulocytes and large hyalinocytes (69). As with the total hemocyte counts, this means it is possible the differential hemocyte counts of the E. leei in the current study changed early in the exposure period. Bibby et al. reported M. edulis can change their differential hemocyte counts 16 days after being placed in aquaria set to ambient conditions (59). The clams in the current study may have changed their differential hemocyte counts at a similar time point to prepare themselves for any potential pathogens in their new environment. An increase in the percent

111 granulocytes would indicate an increase in immune capacity since E. leei granulocytes contain many more granules (filled with low-pH contents for eliminating microbes) than hyalinocytes do

(156). The lack of a significant difference between the differential hemocyte counts of clams sampled at the 30-hour and 41-day time point suggest however that any change in granulocyte proportion earlier on the study would have only been temporary.

Hemocyte viability remained above 90% for all of the E. leei in the current study (Figure

5.4.). The viability of M. edulis hemocytes (164) also was unaffected by acidified seawater exposure although the mortality rate of hemocytes in M. unguiculatus increased after 72 hour exposure to 7.3 pH seawater. It is still possible the viable hemocytes had damaged organelles (as found in M. edulis) since electron microscopy was not used (164). However, the damage was likely minimal since over 80% of the hemocytes in each of treatment groups phagocytosed V. anguillarum.

The proportion of phagocytic cells of E. leei found in the current study was unaffected by temperature or sediment acidification (Figure 5.3.). This is in contrast to the bivalve species that experience phagocytosis inhibition in response to short-term acidified seawater exposure (pH 7.1 to 7.4) such as M. edulis (164), M. unguiculatus (165), P. fucata (69), and T. granosa (162).

Phagocytic capacity in M. edulis did not, however, decrease when exposed for six months to pH

7.65 seawater (166). Similarly, a 28-day exposure to acidified seawater (pH 7.55) did not impact phagocytic activity in the Pacific oyster C. gigas (167). Phagocytic activity of E. leei in the current study was measured only after 30 days of exposure to acidified sediments so it is possible their phagocytic activity decreased during the early days of the experimental period and then reverted back by the end of the experiment. Another possibility is that the hemocytes remain unaffected by the surface sediment acidification because they circulate through the entire body of

112 the clam, and thus spend little time in the acidified portion of the sediment column (top 3 cm).

Temperature (10°C and 20°C) also had no impact on the E. leei hemocytes, even though the phagocytic activity of bivalves generally increases with temperature within their physiological range (41,104,126,127). The lack of a phagocytic activity increase due to warming in E. leei may be due to the high percent of phagocytic cells in the 10°C groups (at least 85%). This leaves little potential for increases in phagocytic rate.

The condition index of the E. leei in the current work was not significantly affected by temperature or pH (Figure 5.5.). The condition index of adult R. philippinarum, juvenile grooved carpet clams (R. decussatus (Linnaeus, 1758)), and adult common cockles (C. edule) were also found to be unaffected by medium term exposure to acidification (28-75 days in 7.30, 7.46, and

7.75 pH overlying seawater respectively) although C. edule condition index decreased when the acidification was combined with a 3°C temperature increase (168–170). Even when food is limited, M. edulis condition index remain unaffected by a 6 month exposure period to pH 7.65 seawater. An increase in temperature from 12°C to 16°C at ambient pH, however, lowered M. edulis condition index significantly (171). Another long-term study (105 days) showed condition index decreases in a linear fashion as temperature increases from 4- 25°C and 4-16°C for M. edulis and the ocean quahog (Arctica islandica (Linnaeus, 1767)) respectively (172). Results of the current study agree with most of the above condition index studies since acidification alone seems to have no effect on adult bivalve condition index regardless of whether the exposure period is medium or long term. The lack of change in condition index suggests that E. leei did not experience any major shell dissolution since this would have decreased the shell mass and thus changed the condition index. Temperature appears to decrease bivalve condition index as it increases within the physiological range of the species, but only when the exposure period is at

113 least 42 days (172). The exposure period for current study (30 days) may have been too short to reveal any condition index changes due to temperature.

5.5. Conclusion

The present study was the first to examine the effects of sediment acidification on the immune capacity of a bivalve species. During the 30-day exposure period, both pH and temperature had no effect on any of the parameters measured. There was also no interaction between pH and temperature for any of the measured parameters. Measuring the total number, type, and phagocytic activity of hemocytes from each clam insured that the lack of change in one of these parameters is not due to a difference in another. The lack of a difference in these parameters between clams sampled after being kept in the tanks for 41 days and clams sampled within 30 hours of being collected (all kept at ambient conditions) suggests the clams were still representative of the wild population at the end of the acidification trial (Figure 5.6.). These holding effect experiments however should be re-run to confirm this result since the t-tests all had minimum detectable differences that are higher than the observed mean differences. In the 2- way ANOVAs used to test for acidification and temperature effects, only the total hemocyte data had a minimum detectable difference that was higher than any of the observed mean differences.

Researchers interested in determining if there were treatment effects that went undetected for total hemocyte count data should repeat the experiment with 18 clams per group. If interest is restricted to the smaller mean differences in the differential hemocyte count, phagocytosis, and condition index data then only 8 clams are required per group.

Overall, these results indicate the internal immune capacity of adult E. leei are unaffected by 30-day exposure to acidified surface sediments (~6.0 pH). The results also suggest this is the case at both 20°C and 10°C, but the experiment should be run again with at least 8

114 clams per group to verify this since only one clam survived in the 20°C 6.8 pH group. Immune effectors responsible for stopping bacteria before they can enter the clam hemolymph (mucous on soft tissue and hemocytes in the extrapallial fluid) were not examined in this study and could therefore still be affected by acidification (19). It should also be noted the clams in this study were not food-limited so it is possible E. leei can only tolerate acidified conditions when food is abundant, which is the case for juvenile M. edulis (173).

Most sediment acidification studies performed to date show juvenile clams actively avoid acidified sediments, likely because it can have negative impacts on their survival. In addition, acidification of the overlying water has been shown to impact negatively the immune system of a variety of bivalve species. E. leei has the potential to be a species that is more resilient to sediment acidification than the other clams tested so far. That said, the vulnerability of clams to sediment acidification is likely not dependent on species alone. Temperature has already been shown to influence burrowing responses to sediment acidification (152). Even field tests on the same species have yielded very different results. Green et. al showed that M. arenaria are susceptible to dissolution in acidified sediments (aragonite values of 0.40) while Beal and Otto showed a survival rate of at least 56% is possible for juvenile M. arenaria (< 12 mm shell length) in sediments with pH 7.11 ± 0.21 (aragonite value of 0.33 ± 0.12) when the area is protected with anti-predator netting (138,174). This may be because clam calcification response to acidification can vary between populations (134).

115

Regardless, Beal and Otto have demonstrated that juvenile M. arenaria are capable of surviving in corrosive sediments and will choose to burrow in them if predators are present (174). The current study provides the first example of a clam that is resilient to extreme sediment surface acidification (pH 6.0, Ωar of 0.04). Survival and growth studies with juveniles are still needed however to insure this resilience is not limited to the adults.

116

CHAPTER 6 GENERAL DISCUSSION AND CONCLUSIONS Concerns in Maine about nearshore sediment acidification began in 2009 when Green et al. (60) made clammers aware of the hazard with his study entitled “death by dissolution”. Since then, the issue has been brought to the attention of a wider audience through both scientific papers and news articles, all of which point out the decreasing number of productive mud flats

(133). The problem is primarily being highlighted in Maine since it is home to a large clam fishery. Note that not all unproductive clam flats behave similarly due to acidified sediments since predation pressure can also cause a once productive clam flat to become barren (174,175).

The increasing populations of the predatory green crab Carcinus maenas (Linnaeus, 1758) on the

Maine coast over the last decade is particularly concerning. C. maenas populations are predicted to continue to increase as Maine waters warm and it becomes less likely the winters will be cold enough to reduce crab population sizes (176,177). Both predation and sediment acidification are likely to continue to reduce the number of productive clam flats, but this discussion will be focused on ways to address the acidification. Sediment acidification is likely to intensify and spread as the amount of nitrogen run off from land increases although there are no estimates for the rate of this intensification since it is site dependent (134).

Sediment acidification only recently started to be researched, although a variety of possible mitigation methods have already been suggested. Crushed shells can be deposited onto the acidified sediment, which may help bring the saturation state up as the shells dissolve.

Crushed shells do not increase clam abundance however if predation pressure is high at the site

(174). However care should be taken such that crushed shell addition to a mud flat does not contaminate sediment with pathogens. Another mitigation method is to reduce the outflow excess nitrogen into the waterways. This can be done at point sources (sewage treatment plants, 117 large finfish aquaculture operations) and non-point sources (private residence). The excess nitrogen from these sources allows for quick algae growth, which means the algae will all die off at relatively the same time. The CO2 that bacteria produce as they decompose the algae drives the pH of the surface sediment down (178). Planting kelp beds in these areas can help reduce the nitrogen levels since they take up nitrogen to grow and live much longer than algae does

(179,180). The run off containing fertilizers adds to nitrogen excess, and can be reduced by educating the public on how fertilizer impacts the aquatic environment. The size at which species become resilient enough to plant in acidified sediments can be determined so growers know how long they must keep them in the hatchery before planting them on their lease site (134). This should be determined for multiple clam species since some species, such as the razor clam E. leei, may be more resilient to sediment acidification than others.

Climate change should also be considered in addition to sediment acidification.

Interactive effects of increased temperature and acidified sediment on bivalves have yet to be studied since studies to date have focused on sediment acidification or temperature effects alone.

Chapter 5 included both 10°C and 20°C treatments to examine the effects of temperature on E. leei health, which tested them in a climate change scenario since they do not currently experience temperatures around 20°C for a 30 day duration at the Beals collection site. Clams were given 10 days to acclimate to the temperature as it was increased from 10 to 20°C (by 1°C per day), but future experiments should allow for about double this time to increase survival in the 20°C groups. Factoring in the effects of temperature will be essential to understanding how clams will tolerate sediment acidification in the field at temperatures predicted to occur in the Gulf of

Maine by the end of the century.

118

In addition to sediment acidification, clams must also deal with ocean acidification. This acidification of the overlying water usually has no impact on the adults since they tolerate even more acidic water within the sediments (55). The larvae of clams though seem to have their shell growth reduced by pH and carbon dioxide levels expected in the Gulf of Maine by end of the century. The hard clam M. mercenaria and Pacific razor clam S. patula both have their shell growth reduced (181,182). Hatcheries however can mitigate this problem by buffering the water in their tanks with sodium carbonate. This has been shown to be successful practice for oyster hatcheries on the west and east coast of the US (183).

The works in this dissertation should be helpful in mitigating sediment acidification impacts on clams. Studies done in this research area so far have primarily focused on clam survival and growth. While these certainly are the first parameters that should be measured, sub- lethal effects will also need to be examined for size class of each species that can survive being planted out in acidified sediments. Chapter 4 gives researchers a way to naturally acidify sediments to aid researchers in designing experiments to investigate sub-lethal effects in addition to survival. The variability that comes with using bacteria to acidify the sediment, however means researchers will need to run a preliminary trial with sediment from their collection to test if acidification duration and intensity meets their needs for the experiment. Chapters 2 and 3 provide the hemocyte classification and annual cycle in E. leei respectively, which can be used to plan further research immune capacity studies with E. leei. The January and August sample points in the annual cycle were attempted, but could not be obtained due to heavy snow and wind respectively. The sample points could not be reattempted at a different time in the month since the clams can only be collected when the tides are out at an extreme. Having these points would have made for a more complete view of the E. leei’s annual cycle, although the trends are still

119 discernible in the data that was collected. Lastly, chapter 5 shows E. leei has the potential to be planted out in the estuary earlier than other species if the juvenile forms are anywhere near resilient to acidification as the adults, which can tolerate 30 days of 6.0 pH surface sediment (as seen in Chapter 5).

The natural sediment acidification method and E. leei immune capacity assays were successfully developed and used to test the immune capacity of E. leei. The seasonality of these parameters was determined and the resilience of adult E. leei to sediment surface acidification was established through a laboratory study. These all represent novel research accomplishments since the immune system of E. leei has never been studied, a natural acidification method has never been published, and the effects of sediment acidification on bivalve immunity have never been studied. The next step in this line of work would be to test if juvenile E. leei survival rates, growth rates, and immune capacity are impacted by sediment acidification. Additional or modified health assays would have to be developed since only a limited amount of hemolymph can be drawn out of each juvenile clam. If juvenile E. leei are shown to be tolerant of acidified surface sediment (pH ~6.0), this would make it possible for growers to raise clams in acidified sediment flats.

120

REFERENCES

1. Leavitt D. Biology of the Atlantic Jackknife (Razor) Clam (Ensis directus Conrad, 1843) [Internet]. Northeastern Regional Aquaculture Center; 2010. Available from: https://agresearch.umd.edu/sites/agresearch.umd.edu/files/_images/uploaded/NRAC%202 17-2010.pdf

2. da Costa FD, Martinez-Patiño D, Ojea J, S. Nóvoa. Larval rearing and spat production of the razor clam Ensis siliqua (Bivalvia: Pharidae). J Mar Sci. 2010;29(2):347–51.

3. Burt W, Leavitt D, Rivara G, Flimlin G. An industry directed feasibility study of the razor clam (Ensis directus) as a candidate for intertidal and shallow subtidal culture in the northeastern U.S. 2005. (Final Report to NRAC).

4. Winter AG, Hosoi AE. Identification and Evaluation of the Atlantic Razor Clam (Ensis directus) for Biologically Inspired Subsea Burrowing Systems. Integr Comp Biol. 2011 Jul 1;51(1):151–7.

5. Guerra A; L C; Gaspar, M; da Costa, F. Razor clams: biology, aquaculture, and fisheries. Xunta Galicia Santiago Compost Cons Do. 2011 Mar;

6. da Costa F, Martinez-Patino D. Culture potential of the razor clam Solen marginatus (Pennant, 1777). Aquaculture. 2009;288(1–2):57–64.

7. Freudendahl A, Nielsen M. Cultivation of American razor clam-screening the potential for commercial cultivation. Univ Aarhus Dan Shellfish Cent. 2005;

8. Lassuy DR & Simons, D 1989. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (Pacific Northwest). Pacific razor clam. TR-EL. 1989;82(4):16.

9. Loosanoff VL, Davis HC. Rearing of bivalve mollusks. Adv Mar Biol. 1963;1:1–136.

10. McMullen JC. Some aspects of the life history of razor clams, Siliqua patula (Dixon) in Cook Inlet, Alaska. 1967;110–8.

11. Weymouth F, McMillin H, Rich WH. Latitude and relative growth in the razor clam, Siliqua patula. J Exp Biol. 1931;8(3):228–49.

12. Cardoso JF, Nieuwland G, Wijsman J, Witbaard R, van der Veer HW. Validation of a method for age determination in the razor clam Ensis directus. 2011;

13. Constantino R, Gaspar MB, Pereira F, Carvalho S, Cúrdia J, Matias D, et al. Environmental impact of razor clam harvesting using salt in Ria Formosa lagoon (Southern Portugal) and subsequent recovery of associated benthic communities. Aquat Conserv Mar Freshw Ecosyst. 2009;19(5):542–53.

121

14. da Costa F, Darriba S, Martinez-Patino D, Guerra A. Culture possibilities of the razor clam Ensis arcuatus (Pharidae: Bivalvia). Aquac Res. 2011;42(10):1549–57.

15. Allam B, Raftos D. Immune responses to infectious diseases in bivalves. J Invert Pathol. 2015;131:121–36.

16. Matozzo V. Aspects of eco-immunology in molluscs. ISJ. 2016;13:116–21.

17. Rodrigues J, Brayner FA, Alves LC, Dixit R, Barillas-Mury C. Hemocyte Differentiation Mediates Innate Immune Memory in Anopheles gambiae Mosquitoes. Science. 2010 Sep 10;329(5997):1353–5.

18. Ng TH, Chiang Y-A, Yeh Y-C, Wang H-C. Review of Dscam-mediated immunity in shrimp and other arthropods. Dev Comp Immunol. 10;46(2):129–38.

19. Allam B, Pales Espinosa E. Bivalve immunity and response to infections: Are we looking at the right place? Fish Shellfish Immunol. 2016;53:4–12.

20. Russell MW, Mestecky J, Strober W, Lambrecht BN, Kelsall BL, Cheroutre H. — Overview: the Mucosal Immune System. In: Mucosal Immunology. Academic Press Boston; 2015. p. 3–8.

21. Jing X, Espinosa EP, Perrigault M, Allam B. Identification, molecular characterization and expression analysis of a mucosal C-type lectin in the eastern oyster, Crassostrea virginica. Fish Shellfish Immunol. 2011;30(3):851–8.

22. Hartland BJ, Timoney JF. Vivo clearance of enteric bacteria from the hemolymph of the hard clam and the American oyster. Appl Environ Microbiol. 1979;37(3):517–20.

23. Lau J, Sussman L, Katalai S, Pales Espinosa E, Allam B. Characterization of hemocytes from different body fluids of the eastern oyster Crassostrea virginica. In 2013.

24. Söderhäll K. Invertebrate immunity: developmental and comparative immunology. 2010.

25. Manduzio H, Rocher B, Durand F, Galap C, Leboulenger F. The point about oxidative stress in molluscs. ISJ. 2005;2(2):91–104.

26. Cerenius L, Söderhäll K. The prophenoloxidase‐activating system in invertebrates. Immunol Rev. 2004;198(1):116–26.

27. Jameson HL, Nicoll W. On some Parasites of the Scoter Duck (ædemia nigra), and their Relation to the Pearl‐inducing Trematode in the Edible Mussel (Mytilus edulis). In Wiley Online Library; 1913. p. 53–63.

28. Peng M, Niu D, Wang F, Chen Z, Li J. Complement C3 gene: Expression characterization and innate immune response in razor clam Sinonovacula constricta. Fish Shellfish Immunol. 2016;55:223–32.

122

29. Wootton E, Dyrynda E, Ratcliffe N. Bivalve immunity: comparisons between the marine mussel (Mytilus edulis), the edible cockle (Cerastoderma edule) and the razor-shell (Ensis siliqua). Fish Shellfish Immunol. 2003;15(3):195–210.

30. Pachauri RK, Allen MR, Barros V, Broome J, Cramer W, Christ R, et al. Climate change 2014: synthesis Report. Contribution of working groups I, II and III to the fifth assessment report of the intergovernmental panel on climate change. IPCC; 2014.

31. Hoegh-Guldberg O, Mumby PJ, Hooten AJ, Steneck RS, Greenfield P, Gomez E, et al. Coral reefs under rapid climate change and ocean acidification. science. 2007;318(5857):1737–42.

32. Shearman RK, Lentz SJ. Long-term sea surface temperature variability along the US East Coast. J Phys Oceanogr. 2010;40(5):1004–17.

33. Morrison J, Pettigrew N, O’Donnell J, Runge J. Rapid detection of climate scale environmental variability in the Gulf of Maine. In IEEE; 2012. p. 1–5.

34. Matozzo V, Marin M. Bivalve immune responses and climate changes: is there a relationship. ISJ. 2011;8:70–7.

35. Ellis RP, Widdicombe S, Parry H, Hutchinson TH, Spicer JI. Pathogenic challenge reveals immune trade-off in mussels exposed to reduced seawater pH and increased temperature. J Exp Mar Biol Ecol. 2015;462:83–9.

36. Richard G, Le Bris C, Guérard F, Lambert C, Paillard C. Immune responses of phenoloxidase and superoxide dismutase in the manila clam Venerupis philippinarum challenged with Vibrio tapetis – Part II: Combined effect of temperature and two V. tapetis strains. Fish Shellfish Immunol. 2015;44(1):79–87.

37. Wang K, Pales Espinosa E, Tanguy A, Allam B. Alterations of the immune transcriptome in resistant and susceptible hard clams (Mercenaria mercenaria) in response to Quahog Parasite Unknown (QPX) and temperature. Fish Shellfish Immunol. 2016;49:163–76.

38. Monari M, Matozzo V, Foschi J, Cattani O, Serrazanetti GP, Marin MG. Effects of high temperatures on functional responses of haemocytes in the clam Chamelea gallina. Fish Shellfish Immunol. 2007;22(1–2):98–114.

39. Munari M, Matozzo V, Marin MG. Combined effects of temperature and salinity on functional responses of haemocytes and survival in air of the clam Ruditapes philippinarum. Fish Shellfish Immunol. 2011;30(4–5):1024–30.

40. Yu JH, Song JH, Choi MC, Park SW. Effects of water temperature change on immune function in surf clams, Mactra veneriformis (Bivalvia: Mactridae). J Invertebr Pathol. 2009;102(1):30–5.

123

41. Perrigault M, Dahl SF, Espinosa EP, Gambino L, Allam B. Effects of temperature on hard clam ( Mercenaria mercenaria) immunity and QPX (Quahog Parasite Unknown) disease development: II. Defense parameters. J Invertebr Pathol. 2011;106(2):322–32.

42. Green MA, Aller RC. Early diagenesis of calcium carbonate in Long Island Sound sediments: Benthic fluxes of Ca2+ and minor elements during seasonal periods of net dissolution. J Mar Res. 2001;59(5):769–94.

43. Green MA, Aller RC, Aller JY. Carbonate dissolution and temporal abundances of foraminifera in Long Island Sound sediments. Limnol Oceanogr. 1993;38(2):331–45.

44. Libes SM. Introduction to marine biogeochemistry [Internet]. Vol. 2nd. Amsterdam;Boston; Elsevier/Academic Press; 2009. Available from: http://umaine.summon.serialssolutions.com

45. Doney SC, Fabry VJ, Feely RA, Kleypas JA. Ocean acidification: the other CO₂ problem. Ann Rev Mar Sci. 2009;1:169–92.

46. Clements JC, Woodard KD, Hunt HL. Porewater acidification alters the burrowing behavior and post-settlement dispersal of juvenile soft-shell clams (Mya arenaria). J Exp Mar Biol Ecol. 2016 Apr;477:103–11.

47. Mackin JE, Swider KT. Organic matter decomposition pathways and oxygen consumption in coastal marine sediments. J Mar Res. 1989;47(3):681–716.

48. ZoBell CE. Studies on redox potential of marine sediments. AAPG Bull. 1946;30(4):477– 513.

49. Jørgensen BB, Findlay AJ, Pellerin A. The Biogeochemical Sulfur Cycle of Marine Sediments. Front Microbiol [Internet]. 2019 Apr 24 [cited 2019 Nov 3];10. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6492693/

50. Green MA, Gulnick JD, Dowse N, Chapman P. Spatiotemporal patterns of carbon remineralization and bio-irrigation in sediments of Casco Bay Estuary, Gulf of Maine. Limnol Oceanogr. 2004;49(2):396–407.

51. Mucci A. The solubility of calcite and aragonite in seawater at various salinities, temperatures, and one atmosphere total pressure. Am J Sci. 1983;283(7):780–99.

52. Kuchinke M, Tilbrook B, Lenton A. Seasonal variability of aragonite saturation state in the Western Pacific. Mar Chem. 2014;161:1–13.

53. Ries J, Cohen A, McCorkle D. Marine biocalcifiers exhibit mixed responses to CO2- induced ocean acidification. 11th Int. In 2008.

54. Wood HL, Spicer JI, Widdicombe S. Ocean acidification may increase calcification rates, but at a cost. Proc R Soc Lond B Biol Sci. 2008;275(1644):1767–73.

124

55. Gazeau F, Parker LM, Comeau S, Gattuso J-P, O’Connor WA, Martin S, et al. Impacts of ocean acidification on marine shelled molluscs. Mar Biol. 2013;160(8):2207–45.

56. Amaral V, Cabral HN, Bishop MJ. Moderate acidification affects growth but not survival of 6-month-old oysters. Aquat Ecol. 2012 Mar 1;46(1):119–27.

57. Bamber RN. The effects of acidic sea water on young carpet-shell clams Venerupis decussata (L.)(Mollusca: Veneracea). J Exp Mar Biol Ecol. 1987;108(3):241–260.

58. Barton A, Hales B, Waldbusser GG, Langdon C, Feely RA. The Pacific oyster, Crassostrea gigas, shows negative correlation to naturally elevated carbon dioxide levels: Implications for near-term ocean acidification effects. Limnol Oceanogr. 2012;57(3):698– 710.

59. Bibby R, Widdicombe S, Parry H, Spicer J, Pipe R. Effects of ocean acidification on the immune response of the blue mussel Mytilus edulis. Aquat Biol. 2008;2(1):67–74.

60. Green MA, Waldbusser GG, Reilly SL, Emerson K, O’Donnell S. Death by dissolution: sediment saturation state as a mortality factor for juvenile bivalves. Limnol Oceanogr. 2009;54(4):1037–47.

61. Clements JC, Hunt HL. Influence of sediment acidification and water flow on sediment acceptance and dispersal of juvenile soft-shell clams (Mya arenaria L.). J Exp Mar Biol Ecol. 2014;453(0):62–9.

62. Cheng TC. A classification of molluscan hemocytes based on functional evidences. In: Cheng TC, editor. Invertebrate Blood. New York: Springer; 1984. p. 111–46.

63. Mateo DR, Siah A, Araya MT, Berthe FC, Johnson GR, Greenwood SJ. Differential in vivo response of soft-shell clam hemocytes against two strains of Vibrio splendidus: changes in cell structure, numbers and adherence. J Invertebr Pathol. 2009;102(1):50–6.

64. Cheng TC, Sullivan JT. Effects of heavy metals on phagocytosis by molluscan hemocytes. Mar Env Res. 1984;14(1):305–15.

65. Gagnaire B, Thomas-Guyon H, Burgeot T, Renault T. Pollutant effects on Pacific oyster, Crassostrea gigas (Thunberg), hemocytes: screening of 23 molecules using flow cytometry. Cell Biol Toxicol. 2006;22(1):1–14.

66. Seiler GR, Morse MP. Kidney and hemocytes of Mya arenaria (Bivalvia): Normal and pollution-related ultrastructural morphologies. J Invert Pathol. 1988;52(2):201–14.

67. Winston GW, Moore MN, Kirchin MA, Soverchia C. Production of reactive oxygen species by hemocytes from the marine mussel, Mytilus edulis: lysosomal localization and effect of xenobiotics. Comp Biochem Physiol C Pharmacol Toxicol Endocrinol. 1996;113(2):221–9.

125

68. Fisher WS, Auffret M, Balouet G. Response of European flat oyster (Ostrea edulis) hemocytes to acute salinity and temperature changes. Aquaculture. 1987;67(1):179–90.

69. Li S, Liu Y, Liu C, Huang J, Zheng G, Xie L, et al. Morphology and classification of hemocytes in Pinctada fucata and their responses to ocean acidification and warming. Fish Shellfish Immunol. 2015;45(1):194–202.

70. Estrada N, Velázquez E, Rodríguez-Jaramillo C, Ascencio F. Morphofunctional study of hemocytes from lions-paw scallop Nodipecten subnodosus. Immunobiology. 2013;218(8):1093–103.

71. Le Foll F, Rioult D, Boussa S, Pasquier J, Dagher Z, Leboulenger F. Characterisation of Mytilus edulis hemocyte subpopulations by single cell time-lapse motility imaging. Fish Shellfish Immunol. 2010;28(2):372–86.

72. Cajaraville M, Abascal I, Etxeberria M, Marigómez I. Lysosomes as cellular markers of environmental pollution: Time‐and dose‐dependent responses of the digestive lysosomal system of mussels after petroleum hydrocarbon exposure. Env Toxicol Water Qual. 1995;10(1):1–8.

73. Moore CA, Beckmann N, Patricia Morse M. Cytoskeletal structure of diseased and normal hemocytes of Mya arenaria. J Invertebr Pathol. 1992;60(2):141–7.

74. Hine PM. The inter-relationships of bivalve haemocytes. Fish Shellfish Immunol. 1999;9(5):367–85.

75. Cima F, Matozzo V, Marin MG, Ballarin L. Haemocytes of the clam Tapes philippinarum (Adams & Reeve, 1850): morphofunctional characterisation. Fish Shellfish Immunol. 2000;10(8):677–93.

76. Aladaileh S, Nair SV, Birch D, Raftos DA. Sydney rock oyster (Saccostrea glomerata) hemocytes: morphology and function. J Invert Pathol. 2007;96(1):48–63.

77. Donaghy L, Lambert C, Choi K-S, Soudant P. Hemocytes of the carpet shell clam (Ruditapes decussatus) and the Manila clam (Ruditapes philippinarum): current knowledge and future prospects. Aquaculture. 2009;297(1):10–24.

78. Hong H-K, Kang H-S, Le TC, Choi K-S. Comparative study on the hemocytes of subtropical oysters Saccostrea kegaki (Torigoe & Inaba, 1981), Ostrea circumpicta (Pilsbry, 1904), and Hyotissa hyotis (Linnaeus, 1758) in Jeju Island, Korea: morphology and functional aspects. Fish Shellfish Immunol. 2013;35(6):2020–5.

79. Matozzo V, Marin MG, Cima F, Ballarin L. First evidence of cell division in circulating haemocytes from the Manila clam Tapes philippinarum. Cell Biol Int. 2008;32(7):865–8.

80. Salimi L, Jamili S, Motalebi A, Eghtesadi-Araghi P, Rabbani M, Rostami-Beshman M. Morphological characterization and size of hemocytes in Anodonta cygnea. J Invert Pathol. 2009;101(2):81–5.

126

81. Chang S-J, Tseng S-M, Chou H-Y. Morphological characterization via light and electron microscopy of the hemocytes of two cultured bivalves: a comparison study between the hard clam (Meretrix lusoria) and Pacific oyster (Crassostrea gigas). Zool Stud. 2005;44(1):144–53.

82. Russell-Pinto F, Reimão R, de Sousa M. Haemocytes in Cerastoderma edule (Mollusca, Bivalvia): distinct cell types engage in different responses to sheep erythrocytes. Fish Shellfish Immunol. 1994;4(5):383–97.

83. Holden JA, Pipe RK, Quaglia A, Ciani G. Blood cells of the arcid clam, Scapharca inaequivalvis. J Mar Biol Assoc UK. 1994;74(02):287–99.

84. Cheng T. Invertebrate Blood Cells. Ratcliffe NA Rowley AF Acad Press Inc N Y. 1981;233.

85. Haszprunar G. The molluscan rhogocyte (pore-cell, Blasenzelle, cellule nucale), and its significance for ideas on nephridial evolution. J Mollus Stud. 1996;62(2):185–211.

86. Mackin J. Histopathology of infection of Crassostrea virginica (Gmelin) by Dermocystidium marinum Mackin, Owen, and Collier. Bull Mar Sci. 1951;1(1):72–87.

87. Nakayama K, Nomoto AM, Nishijima M, Maruyama T. Morphological and Functional Characterization of Hemocytes in the Giant Clam Tridacna crocea. J Invert Pathol. 1997;69(2):105–11.

88. Matozzo V, Bailo L. A first insight into haemocytes of the smooth venus clam Callista chione. Fish Shellfish Immunol. 2015;42(2):494–502.

89. Tame A, Yoshida T, Ohishi K, Maruyama T. Phagocytic activities of hemocytes from the deep-sea symbiotic mussels Bathymodiolus japonicus, B. platifrons, and B. septemdierum. Fish Shellfish Immunol. 2015;45(1):146–56.

90. Lowe D, Fossato V, Depledge M. Contaminant-induced lysosomal membrane damage in blood cells of mussels Mytilus galloprovincialis from the Venice Lagoon: an in vitro study. Mar Ecol Prog Ser. 1995;7(43):734.

91. Franceschi C, Cossarizza A, Monti D, Ottaviani E. Cytotoxicity and immunocyte markers in cells from the freshwater snail Planorbarius corneus (L.)(Gastropoda pulmonata): implications for the evolution of natural killer cells. Eur J Immunol. 1991;21(2):489–93.

92. Kuchel RP, Raftos DA, Birch D, Vella N. Haemocyte morphology and function in the Akoya Pearl Oyster, Pinctada imbricata. J Invert Pathol. 2010;105(1):36–48.

93. Radford JL, Hutchinson AE, Burandt M, Raftos DA. Effects of Metal-Based Environmental Pollutants on Tunicate Hemocytes. J Invert Pathol. 11;76(4):242–8.

94. Scippa S, Iazzetti G, De Vincentiis M. Observations on the ascidian Phallusia mammillata blood cells during development. Acta Embryol Morphol Exp. 1987;8:341–6.

127

95. Parrinello N, De Leo G, Di Bella MA. Fine structural observations of the granulocytes involved in the tunic inflammatory-like reaction of Ciona intestinalis (Tunicata). J Invert Pathol. 1990 Sep 1;56(2):181–9.

96. Parrinello N, xf, Patricolo E, Canicatt, xec, Calogero. Inflammatory-Like Reaction in the Tunic of Ciona intestinalis (Tunicata). I. Encapsulation and Tissue Injury. Biol Bull. 1984;167(1):229–37.

97. Parrinello N, xf, Patricolo E. Inflammatory-Like Reaction in the Tunic of Ciona intestinalis (Tunicata). II. Capsule Components. Biol Bull. 1984;167(1):238–50.

98. Huffman JE, Tripp M. Cell types and hydrolytic enzymes of soft shell clam (Mya arenaria) hemocytes. J Invert Pathol. 1982;40(1):68–74.

99. Zhang W, Wu X, Wang M. Morphological, structural, and functional characterization of the haemocytes of the scallop, Argopecten irradians. Aquaculture. 2006;251(1):19–32.

100. Duchemin MB, Fournier M, Auffret M. Seasonal variations of immune parameters in diploid and triploid Pacific oysters, Crassostrea gigas (Thunberg). Aquaculture. 2007;264(1):73–81.

101. Hinsch GW, Hunte M. Ultrastructure of phagocytosis by hemocytes of the American oyster. California: Academic Press, Inc.; 2012. 479 p. (Pathology in Marine Science).

102. de Freitas Rebelo M, de Souza Figueiredo E, Mariante RM, Nóbrega A, de Barros CM, Allodi S. New insights from the oyster Crassostrea rhizophorae on bivalve circulating hemocytes. PloS One. 2013;8(2).

103. Delaporte M, Synard S, Pariseau J, McKenna P, Tremblay R, Davidson J, et al. Assessment of haemic neoplasia in different soft shell clam Mya arenaria populations from eastern Canada by flow cytometry. J Invertebr Pathol. 2008 Jun 1;98(2):190–7.

104. Anisimova AA. Morphofunctional parameters of hemocytes in the assessment of the physiological status of bivalves. Russ J Mar Biol. 2013 Nov 1;39(6):381–91.

105. Flye-Sainte-Marie J, Soudant P, Lambert C, Le Goïc N, Goncalvez M, Travers M-A, et al. Variability of the hemocyte parameters of Ruditapes philippinarum in the field during an annual cycle. J Exp Mar Biol Ecol. 2009 Aug;377(1):1–11.

106. Lemaire N, Pellerin J, Fournier M, Girault L, Tamigneaux E, Cartier S, et al. Seasonal variations of physiological parameters in the blue mussel mytilus spp. from farm sites of eastern Quebec. Aquaculture. 2006 Nov 24;261(2):729–51.

107. Matozzo V, Ros LD, Ballarin L, Meneghetti F, Marin MG. Functional responses of haemocytes in the clam Tapes philippinarum from the Lagoon of Venice: fishing impact and seasonal variations. Can J Fish Aquat Sci. 2003 Aug;60(8):949–58.

128

108. Mosca F, Lanni L, Cargini D, Narcisi V, Bianco I, Tiscar PG. Variability of the hemocyte parameters of cultivated mussel Mytilus galloprovincialis (Lmk 1819) in Sabaudia (Latina, Italy) coastal lagoon. Mar Environ Res. 2013 Dec;92:215–23.

109. Clements JC, Woodard KD, Hunt HL. Porewater acidification alters the burrowing behavior and post-settlement dispersal of juvenile soft-shell clams (Mya arenaria). J Exp Mar Biol Ecol. 2016;477:103–11.

110. Waldbusser GG, Hales B, Haley BA. Calcium carbonate saturation state: on myths and this or that stories. ICES J Mar Sci. 2016 Mar 1;73(3):563–8.

111. Cohen AL, Holcomb M. Why Corals Care About Ocean Acidification: Uncovering the Mechanism. Oceanography. 2009;22(4):118–27.

112. Palmer AR. Calcification in marine molluscs: how costly is it? Proc Natl Acad Sci. 1992 Feb 15;89(4):1379–82.

113. Pan T-CF, Applebaum SL, Manahan DT. Experimental ocean acidification alters the allocation of metabolic energy. Proc Natl Acad Sci. 2015 Mar 25;201416967.

114. Vazquez N, Bruno EP, Marquez F, Van der Molen S, Gilardoni C, Cremonte F. A histopathological survey of the razor clam Ensis macha (Pharidae) along the Patagonian Argentina coast. J Invertebr Pathol. 2013;112(3):253–9.

115. Drupp PS, De Carlo EH, Mackenzie FT. Porewater CO2–carbonic acid system chemistry in permeable carbonate reef sands. Mar Chem. 2016 Oct;185:48–64.

116. Dickson AG, Sabine CL, Christian JR. Guide to Best Practices for Ocean CO2 Measurements. [Internet]. North Pacific Marine Science Organization; 2007 [cited 2018 Sep 18]. Available from: https://www.oceanbestpractices.net/handle/11329/249

117. Millero FJ. Carbonate constants for estuarine waters. Mar Freshw Res. 2010 Mar 18;61(2):139–42.

118. Singer JT, Phennicie RT, Sullivan MJ, Porter LA, Shaffer VJ, Kim CH. Broad-Host- Range Plasmids for Red Fluorescent Protein Labeling of Gram-Negative Bacteria for Use in the Zebrafish Model System. Appl Environ Microbiol. 2010 Jun 1;76(11):3467–74.

119. Pietrak MR, Molloy SD, Bouchard DA, Singer JT, Bricknell I. Potential role of Mytilus edulis in modulating the infectious pressure of Vibrio anguillarum 02β on an integrated multi-trophic aquaculture farm. Aquaculture. 2012;326:36–39.

120. Strober W. Trypan blue exclusion test of cell viability. Curr Protoc Immunol. 1997;21(1):A–3B.

121. Soudant P, Paillard C, Choquet G, Lambert C, Reid HI, Marhic A, et al. Impact of season and rearing site on the physiological and immunological parameters of the Manila clam Venerupis philippinarum. Aquaculture. 2004;229:401–18.

129

122. Feng SY. Heart rate and leucocyte circulation in Crassostrea virginica (gmelin), ,. Biol Bull. 1965 Apr 1;128(2):198–210.

123. Travers M-A, Le Goïc N, Huchette S, Koken M, Paillard C. Summer immune depression associated with increased susceptibility of the European abalone, Haliotis tuberculata to Vibrio harveyi infection. Fish Shellfish Immunol. 2008 Dec 1;25(6):800–8.

124. Dang VT, Speck P, Benkendorff K. Influence of elevated temperatures on the immune response of abalone, Haliotis rubra. Fish Shellfish Immunol. 2012 May 1;32(5):732–40.

125. Pipe RK, Coles JA. Environmental contaminants influencing immune function in marine bivalve molluscs. Fish Shellfish Immunol. 1995 Nov 1;5(8):581–95.

126. Chu F, La Peyre J. Perkinsus marinus susceptibility and defense-related activities in eastern oysters Crassostrea virginica. temperature effects. Dis Aquat Organ. 1993;16:223–34.

127. Carballal MJ, López C, Azevedo C, Villalba A. In vitrostudy of phagocytic ability of Mytilus galloprovincialis Lmk. haemocytes. Fish Shellfish Immunol. 1997 Aug 1;7(6):403–16.

128. Donaghy L, Kim B-K, Hong H-K, Park H-S, Choi K-S. Flow cytometry studies on the populations and immune parameters of the hemocytes of the Suminoe oyster, Crassostrea ariakensis. Fish Shellfish Immunol. 2009;27(2):296–301.

129. Ashton-Alcox KA, Ford SE. Variability in molluscan hemocytes: a flow cytometric study. Tissue Cell. 1998 Apr 1;30(2):195–204.

130. Widdicombe S, Spicer JI, Kitidis V. Effects of ocean acidification on sediment fauna. Ocean Acidif. 2011;16.

131. Pimenta A, Grear J. Guidelines for Measuring Changes in Seawater pH and Associated Carbonate Chemistry in Coastal Environments of the Eastern United States. 2018;59.

132. Clements JC, Hunt HL. Testing for Sediment Acidification Effects on Within-Season Variability in Juvenile Soft-Shell Clam (Mya arenaria) Abundance on the Northern Shore of the Bay of Fundy. Estuaries Coasts. 2018 Mar;41(2):471–83.

133. Friends of Casco Bay. 2013 Casco Bay clam flat pH study [Internet]. 2014. Available from: https://www.cascobayestuary.org/wp-content/uploads/2018/01/Casco-Bay-Clamflat- pH-Study-FOCB-2013.pdf

134. Waldbusser G, Bergschneider H, Green M. Size-dependent pH effect on calcification in post-larval hard clam Mercenaria spp. Mar Ecol Prog Ser. 2010 Nov 4;417:171–82.

135. Rodríguez-Romero A, Jiménez-Tenorio N, Basallote MD, Orte MRD, Blasco J, Riba I. Predicting the Impacts of CO2 Leakage from Subseabed Storage: Effects of Metal

130

Accumulation and Toxicity on the Model Benthic Organism Ruditapes philippinarum. Environ Sci Technol. 2014 Oct 21;48(20):12292–301.

136. Clements JC, Hunt HL. Effects of CO2 -driven sediment acidification on infaunal marine bivalves: A synthesis. Mar Pollut Bull. 2017 Apr;117(1–2):6–16.

137. Riebesell U, Fabry VJ, Hansson L, Gattuso J-P. Guide to best practices for ocean acidification research and data reporting [Internet]. Riebesell U, Fabry VJ, Hansson L, Gattuso J-P, editors. Luxembourg: Office for Official Publications of the European Communities; 2011 [cited 2018 Sep 18]. 258 p. Available from: https://oceanrep.geomar.de/8471/

138. Green MA, Waldbusser GG, Hubazc L, Cathcart E, Hall J. Carbonate Mineral Saturation State as the Recruitment Cue for Settling Bivalves in Marine Muds. Estuaries Coasts. 2013 Jan;36(1):18–27.

139. Greiner CM, Klinger T, Ruesink JL, Barber JS, Horwith M. Habitat effects of macrophytes and shell on carbonate chemistry and juvenile clam recruitment, survival, and growth. J Exp Mar Biol Ecol. 2018 Dec 1;509:8–15.

140. Cubillas P, Köhler S, Prieto M, Chaïrat C, Oelkers EH. Experimental determination of the dissolution rates of calcite, aragonite, and bivalves. Chem Geol. 2005 Mar 15;216(1):59– 77.

141. Areibat LEM, Kamari A. Razor clam (Ensis directus) shell as a low-cost adsorbent for the removal of Congo red and Rhodamine B dyes from aqueous solution. AIP Conf Proc. 2017 May 19;1847(1):040004.

142. Ries JB. Skeletal mineralogy in a high-CO2 world. J Exp Mar Biol Ecol. 2011 Jul 15;403(1):54–64.

143. López GI. Grain Size Analysis. In: Gilbert AS, editor. Encyclopedia of Geoarchaeology [Internet]. Dordrecht: Springer Netherlands; 2017 [cited 2018 Dec 28]. p. 341–8. Available from: https://doi.org/10.1007/978-1-4020-4409-0_18

144. Wentworth CK. A Scale of Grade and Class Terms for Clastic Sediments. J Geol. 1922 Jul;30(5):377–92.

145. Heiri O, Lotter AF, Lemcke G. Loss on ignition as a method for estimating organic and carbonate content in sediments: reproducibility and comparability of results. J Paleolimnol. 2001;25(1):10.

146. Pomeroy L, Wiebe W. Temperature and substrates as interactive limiting factors for marine heterotrophic bacteria. Aquat Microb Ecol. 2001;23:187–204.

147. Cai W-J, Huang W-J, Luther GW, Pierrot D, Li M, Testa J, et al. Redox reactions and weak buffering capacity lead to acidification in the Chesapeake Bay. Nat Commun. 2017 Aug 28;8(1):1–12.

131

148. Soetaert K, Hofmann AF, Middelburg JJ, Meysman FJR, Greenwood J. The effect of biogeochemical processes on pH. Mar Chem. 2007 Apr 25;105(1):30–51.

149. Widdicombe S, Dashfield SL, McNeill CL, Needham HR, Beesley A, McEvoy A, et al. Effects of CO2 induced seawater acidification on infaunal diversity and sediment nutrient fluxes. Mar Ecol Prog Ser. 2009 Mar 30;379:59–75.

150. Whitworth KL, Silvester E, Baldwin DS. Alkalinity capture during microbial sulfate reduction and implications for the acidification of inland aquatic ecosystems. Geochim Cosmochim Acta. 2014 Apr 1;130:113–25.

151. Arnosti C, Jørgensen B, Sagemann J, Thamdrup B. Temperature dependence of microbial degradation of organic matter in marine sediments:polysaccharide hydrolysis, oxygen consumption, and sulfate reduction. Mar Ecol Prog Ser. 1998;165:59–70.

152. Clements JC, Bishop MM, Hunt HL. Elevated temperature has adverse effects on GABA- mediated avoidance behaviour to sediment acidification in a wide-ranging marine bivalve. Mar Biol [Internet]. 2017 Mar [cited 2018 Sep 18];164(3). Available from: http://link.springer.com/10.1007/s00227-017-3085-1

153. Andersson AJ, Mackenzie FT, Lerman A. Coastal ocean CO2–carbonic acid–carbonate sediment system of the Anthropocene. Glob Biogeochem Cycles [Internet]. 2006 [cited 2019 Aug 15]; Available from: https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2005GB002506%4010.1002/%2 8ISSN%291944-9224.GNEWS1

154. DMR. Commercial Fishing Historical Landings Data: Maine Department of Marine Resources [Internet]. 2019 [cited 2019 Jul 10]. Available from: https://www.maine.gov/dmr/commercial-fishing/landings/historical-data.html

155. Hagan JM, Wilkerson E. How to Install a Soft-shell Clam Farm. 2018;23.

156. Preziosi BM, Bowden TJ. Morphological characterization via light and electron microscopy of Atlantic jackknife clam (Ensis directus) hemocytes. Micron. 2016;84:96– 106.

157. Peng C, Zhao X, Liu S, Shi W, Han Y, Guo C, et al. Ocean acidification alters the burrowing behaviour, Ca2+/Mg2+-ATPase activity, metabolism, and gene expression of a bivalve species, Sinonovacula constricta. Mar Ecol Prog Ser. 2017 Jul 20;575:107–17.

158. Weiss IM, Tuross N, Addadi L, Weiner S. Mollusc larval shell formation: amorphous calcium carbonate is a precursor phase for aragonite. J Exp Zool. 2002;293(5):478–91.

159. Matozzo V, Chinellato A, Munari M, Finos L, Bressan M, Marin MG. First Evidence of Immunomodulation in Bivalves under Seawater Acidification and Increased Temperature. PLOS ONE. 2012 Mar 27;7(3):e33820.

132

160. Ivanina AV, Hawkins C, Sokolova IM. Interactive effects of copper exposure and environmental hypercapnia on immune functions of marine bivalves Crassostrea virginica and Mercenaria mercenaria. Fish Shellfish Immunol. 2016 Feb 1;49:54–65.

161. Ivanina AV, Hawkins C, Sokolova IM. Immunomodulation by the interactive effects of cadmium and hypercapnia in marine bivalves Crassostrea virginica and Mercenaria mercenaria. Fish Shellfish Immunol. 2014 Apr;37(2):299–312.

162. Liu S, Shi W, Guo C, Zhao X, Han Y, Peng C, et al. Ocean acidification weakens the immune response of blood clam through hampering the NF-kappa β and toll-like receptor pathways. Fish Shellfish Immunol. 2016 Jul 1;54:322–7.

163. Paillard C, Allam B, Oubella R. Effect of temperature on defense parameters in Manila clam Ruditapes philippinarum challenged with Vibrio tapetis. Dis Aquat Organ. 2004;59(3):249–62.

164. Sun T, Tang X, Jiang Y, Wang Y. Seawater acidification induced immune function changes of haemocytes in Mytilus edulis: a comparative study of CO2 and HCl enrichment. Sci Rep. 2017 Feb 6;7:41488.

165. Sui Y, Kong H, Shang Y, Huang X, Wu F, Hu M, et al. Effects of short-term hypoxia and seawater acidification on hemocyte responses of the mussel Mytilus coruscus. Mar Pollut Bull. 2016 Jul 15;108(1):46–52.

166. Mackenzie CL, Lynch SA, Culloty SC, Malham SK. Future Oceanic Warming and Acidification Alter Immune Response and Disease Status in a Commercial Shellfish Species, Mytilus edulis L. PLOS ONE. 2014 Jun 13;9(6):e99712.

167. Wang Q, Cao R, Ning X, You L, Mu C, Wang C, et al. Effects of ocean acidification on immune responses of the Pacific oyster Crassostrea gigas. Fish Shellfish Immunol. 2016 Feb 1;49:24–33.

168. Range P, Chícharo MA, Ben-Hamadou R, Piló D, Matias D, Joaquim S, et al. Calcification, growth and mortality of juvenile clams Ruditapes decussatus under increased pCO2 and reduced pH: Variable responses to ocean acidification at local scales? J Exp Mar Biol Ecol. 2011 Jan 1;396(2):177–84.

169. Velez C, Figueira E, Soares AMVM, Freitas R. Combined effects of seawater acidification and salinity changes in Ruditapes philippinarum. Aquat Toxicol. 2016 Jul 1;176:141–50.

170. Ong EZ, Briffa M, Moens T, Van Colen C. Physiological responses to ocean acidification and warming synergistically reduce condition of the common cockle Cerastoderma edule. Mar Environ Res. 2017 Sep 1;130:38–47.

171. Mackenzie CL, Ormondroyd GA, Curling SF, Ball RJ, Whiteley NM, Malham SK. Ocean Warming, More than Acidification, Reduces Shell Strength in a Commercial Shellfish Species during Food Limitation. PLOS ONE. 2014 Jan 28;9(1):e86764.

133

172. Hiebenthal C, Philipp EER, Eisenhauer A, Wahl M. Interactive effects of temperature and salinity on shell formation and general condition in Baltic Sea Mytilus edulis and Arctica islandica. Aquat Biol. 2012 Mar 6;14(3):289–98.

173. Thomsen J, Casties I, Pansch C, Körtzinger A, Melzner F. Food availability outweighs ocean acidification effects in juvenile Mytilus edulis: laboratory and field experiments. Glob Change Biol. 2013;19(4):1017–27.

174. Beal BF, Otto W. How acidic sediments and seawater affect interactive effects of predation on survival, growth, and recruitment of wild and cultured soft-shell clams, Mya arenaria L., along a tidal gradient at two intertidal sites in eastern Maine. 2019;50.

175. Beal BF, Kraus MG. Interactive effects of initial size, stocking density, and type of predator deterrent netting on survival and growth of cultured juveniles of the soft-shell clam, Mya arenariaL., in eastern Maine. Aquaculture. 2002;208(1):81–111.

176. Tepolt CK, Somero GN. Master of all trades: thermal acclimation and adaptation of cardiac function in a broadly distributed marine invasive species, the European green crab, Carcinus maenas. J Exp Biol. 2014;217(7):1129–1138.

177. Beal BF, Nault D-M, Annis H, Thayer P, Leighton H, Ellis B. Comparative, Large-Scale Field Trials Along the Maine Coast to Assess Management Options to Enhance Populations of the Commercially Important Softshell Clam, Mya arenaria L. J Shellfish Res. 2016 Dec;35(4):711–27.

178. Wallace RB, Baumann H, Grear JS, Aller RC, Gobler CJ. Coastal ocean acidification: The other eutrophication problem. Estuar Coast Shelf Sci. 2014 Jul;148:1–13.

179. Varela DA, Hernríquez LA, Fernández PA, Leal P, Hernández-González MC, Figueroa FL, et al. Photosynthesis and nitrogen uptake of the giant kelp Macrocystis pyrifera (Ochrophyta) grown close to salmon farms. Mar Environ Res. 2018 Apr;135:93–102.

180. Kim JK, Kraemer GP, Yarish C. Use of sugar kelp aquaculture in Long Island Sound and the Bronx River Estuary for nutrient extraction. Mar Ecol Prog Ser. 2015 Jul 2;531:155– 66.

181. Talmage SC, Gobler CJ. Effects of past, present, and future ocean carbon dioxide concentrations on the growth and survival of larval shellfish. Proc Natl Acad Sci. 2010 Oct 5;107(40):17246–51.

182. Talmage SC, Gobler CJ. Effects of Elevated Temperature and Carbon Dioxide on the Growth and Survival of Larvae and Juveniles of Three Species of Northwest Atlantic Bivalves. PLOS ONE. 2011 Oct 31;6(10):e26941.

183. Lockwood D. Can Shellfish Adapt to Ocean Acidification? ACS Cent Sci. 2017 Apr 26;3(4):263–5.

134

BIOGRAPHY OF THE AUTHOR

Brian Matthew Preziosi was born on June 10, 1988 in Wilmington, Massachusetts. He attended Wilmington Public High School. He received his bachelor’s degree from Wheaton

College of Norton Massachusetts in May, 2010. He majored in biology and minored in environmental studies. During his college studies, he designed an effective template for collecting green crabs at Salem State’s Cat Cove, monitored the growth of soft-shell clams in

Boston Harbor in collaboration with the Massachusetts Division of Marine Fisheries, and completed an animal husbandry summer internship at Woods Hole Oceanographic. He graduated from the Marine Biology program at University of Maine in 2012 with a Master of Science degree. His Master’s thesis was titled “The effects of ocean acidification and climate change on the reproductive processes of the marine copepod Calanus finmarchicus” Brian is currently a student member of the National Shellfisheries Association. Brian is a candidate for the Doctor of

Philosophy degree in Aquaculture and Aquatic Resources from the University of Maine in

December 2019.

135