Heterosigma Akashiwo As a Potential Source of Prey For

HETEROSIGMA AKASHIWO AS A POTENTIAL SOURCE OF PREY FOR

THE TOXIC DINOFLAGELLATE, DINOPHYSIS ACUMINATA

Amanda K. Williams

A thesis submitted to the Faculty of the University of Delaware in partial fulfillment of the requirements for the degree of Master of Science in Marine Studies

Summer 2020

HETEROSIGMA AKASHIWO AS A POTENTIAL SOURCE OF PREY FOR

THE TOXIC DINOFLAGELLATE, DINOPHYSIS ACUMINATA

Amanda K. Williams

Approved: ______Kathryn J. Coyne, Ph.D. Professor in charge of thesis on behalf of the Advisory Committee

Approved: ______Mark A. Moline, Ph.D. Director of the School of Marine Science and Policy

Approved: ______Estella Atekwana, Ph.D. Dean of the College of Earth, Ocean, and Environment

Approved: ______Douglas J. Doren, Ph.D. Interim Vice Provost for Graduate and Professional Education and Dean of the Graduate College ACKNOWLEDGMENTS

Thank you to all members of my committee- Dr. Kathryn Coye, Dr. Jon Cohen, and Dr. Tye Pettay. Special thanks to Amanda Pappas who supported me both inside and outside of the lab. Thank you also to Yanfei Wang, Emmie Healey, Dr. Mark Warner, Dr. Sylvain Le Marchand, Dr. Matthew Johnson and Dr. Juliette Smith for helping me in various ways. Thanks to Dr. Ed Whereat and the University of Delaware Citizens Monitoring Program for insightful conversations and sharing data. Thank you to Kathy for taking me on as her student and welcoming me with support and guidance.

I also want to acknowledge funding for this project, which was provided by the Carolyn Thompson Thoroughgood Graduate Student Research Award and Delaware Sea Grant R/HCE-32 to DTP and KJC. In addition, thank you to Delaware Sea Grant for sponsoring me to attend the U.S. Harmful Algae ID Course which took place at Bigelow Laboratories in Maine, it was a valuable experience that I will never forget.

I would also like to acknowledge those who played a significant role in me getting to this point in my life. Thank you to my parents for their love and many sacrifices to ensure that I get the best education. Thank you to my late Nana Mary and Popop for supporting me and teaching me the honor and value of living a life of love towards others and all of life. Thank you to Nana Alice for the many conversations, notes, funny stories, and bouquets of flowers. Thank you to my brother, Dustin for always offering a smile and a shoulder to lean on. Thank you to Minnie, who has always been there for me to talk to and comfort me when I need it the most. And last but certainly not least, thank you to my dear love Amanda who has been here for me in every way and brought a smile to my face every day since I met her 2 years ago.

In addition, thank you to the Indian River, which fostered in me a love of marine biology and nature. Thank you to the algae who have guided me the past 2 years, what glorious and immaculate beings! Thank you to the birds that sing every morning and the crickets that follow at night. Thanks to the many trees and rocks that share my love and the tiny tardigrades that inspire my dreams of outer space. Thanks to the stars, the sun and moon for being a constant- a rock to stand on in the midst of so much change. And finally, thank you to the plants, which have been my life-long teacher, it is to them that I commit my next steps, in hopes to unlock the healing and freedom I experience every day in the hearts of others.

iii DEDICATION

I dedicate the sum of this work and my efforts: To the waters of Delaware, where the algae dance. And to the algae who worship the sun, in eternal bliss.

iv TABLE OF CONTENTS

LIST OF TABLES ...... vii LIST OF FIGURES ...... viii ABSTRACT ...... x

Chapter

1 LITERATURE REVIEW ...... 1

1.1 Dinophysis acuminata and its presence in the Delaware Inland Bays .. 1 1.2 Dinophysis acuminata ecology ...... 2 1.3 Analysis of pigments in plastids of Dinophysis acuminata ...... 3 1.4 Research Objectives ...... 4

2 HETEROSIGMA AKASHIWO AS A POTENTIAL SOURCE OF PREY FOR THE TOXIC DINOFLAGELLATE DINOPHYSIS ACUMINATA ... 6

2.1 Abstract ...... 6 2.2 Introduction ...... 7 2.3 Methods...... 9

2.3.1 Algal and ciliate cultures ...... 9 2.3.2 Growth Curves ...... 10 2.3.3 Predation Experiments ...... 10 2.3.4 Microscopic analysis ...... 11 2.3.5 Analysis of accessory pigments ...... 11 2.3.6 Occurrence of Heterosigma akashiwo and Dinophysis acuminata at Torquay Canal ...... 12 2.3.7 Molecular analysis of bloom samples- looking for Heterosigma akashiwo ...... 13 2.3.8 Statistical Analysis ...... 14

2.4 Results ...... 14

2.4.1 Growth Rates ...... 14 2.4.2 Ingestion Rates ...... 15 2.4.3 Analysis of nuclear inclusions ...... 15 2.4.4 A ratiometric approach using chlorophyll a emission to evaluate plastid origin ...... 15 2.4.5 Relationships between D. acuminata and H. akashiwo abundances in Torquay Canal ...... 16

2.5 Discussion ...... 17

v 2.6 Conclusions ...... 23

REFERENCES ...... 39

Appendix

DATA……………………………………………………………………………..... 43

vi LIST OF TABLES

Table 2.1 Environmental metadata, cell density of D. acuminata (Pappas in prep), and presence of H. akashiwo during the winter 2019-2020 bloom of D. acuminata at site TQB in Torquay Canal, Rehoboth Beach, DE. .. 25

Table A1 Cell counts from the growth rate experiment where Dinophysis acuminata was offered either Heterosigma akashiwo, Heterosigma akashiwo and Mesodinium rubrum, or Mesodinium rubrum as prey. All values are representative of the mean of three biological replicates...... 43

Table A2 Growth rates of Heterosigma akashiwo and Mesodinium rubrum in control treatment, used to offset ingestion rate of Dinophysis acuminata when offered these as prey...... 44

Table A3 Ingestion rates of Dinophysis acuminata when offered Heterosigma akashiwo or Mesodinium rubrum as prey...... 45

Table A4 Growth rates of Heterosigma akashiwo and Teleaulax amphioxeia in control treatment, used to offset ingestion rate of Dinophysis acuminata when offered these as prey...... 48

Table A5 Ingestion rates of Mesodinium rubrum when offered Heterosigma akashiwo or Teleaulax amphioxeia as prey...... 49

Table A6 Emission intensity of chlorophyll a in Dinophysis acuminata offered Heterosigma akashiwo, both Heterosigma akashiwo and Mesodinium rubrum, or Mesodinium rubrum as prey...... 50

Table A7 Ratio of intensity of emission of chlorophyll a when excited at 488 nm and 633 nm in plastids of Heterosigma akashiwo, Teleaulax amphioxeia, and Dinophysis acuminata that was given Mesodinium rubrum as prey...... 51

Table A8 Ratio of intensity of emission of chlorophyll a when excited at 488 nm and 561 nm in plastids of Dinophysis acuminata when offered Heterosigma akashiwo, Heterosigma akashiwo and Mesodinium rubrum, or Mesodinium rubrum as prey...... 52

vii LIST OF FIGURES

Figure 2.1 Log Heterosigma akashiwo cell abundance vs. Log Dinophysis acuminata cell abundances in Torquay Canal between 2014 and 2017 for samples in which both species were present. A Pearson’s correlation revealed that Heterosigma akashiwo and Dinophysis acuminata cell abundance are not significantly correlated (p = 0.16)...... 27

Figure 2.2 Map of Torquay Canal, Rehoboth Bay, Delaware (38.699031, - 75.112409) sampling sites and transect route...... 28

Figure 2.3 Cell abundances of D. acuminata over a period of 10 days, when offered only Heterosigma akashiwo, both H. akashiwo and Mesodinium rubrum, or only M. rubrum as prey. Error bars represent standard error (n=3 for each treatment) and asterisk indicates a statistically significant difference between cell density on day 10 of D. acuminata offered M. rubrum as prey and D. acuminata offered either H. akashiwo or H. akashiwo and M. rubrum as prey (One-way ANOVA and post-hoc Tukey test, p = 0.0028)...... 29

Figure 2.4 Cell abundances of Heterosigma akashiwo and Mesodinium rubrum over a period of 10 days, when offered to Dinophysis acuminata as prey. Error bars represent standard error (n=3 for each treatment). There was a statistically significant difference between cell abundance of H. akashiwo and Mesodinium rubrum on days 1, 3, 6, 8, & 10 (One-way ANOVA and post-hoc Tukey test, p = 0.041, 0.0016, 0.0038, 9.64 E-7, and 3.0 E-8, respectively)...... 30

Figure 2.5 Ingestion rates of Mesodinium rubrum when offered Heterosigma akashiwo and Teleaulax amphioxeia as prey. Error bars represent standard error (n=6 for each treatment) and asterisk indicates a statistically significant difference between ingestion rates (One-way ANOVA, p = 2.0 E-6)...... 31

Figure 2.6 Ingestion rates of Dinophysis acuminata when offered Heterosigma akashiwo and Mesodinium rubrum as prey. Error bars represent standard error (n=6 for each treatment) and asterisk indicates a statistically significant difference between ingestion rates (One-way ANOVA, p = 2.0 E-5)...... 32

viii Figure 2.7 Dinophysis acuminata stained with Hoechst 33342, with no nuclear inclusions when offered Heterosigma akashiwo as prey (A), with nuclear inclusions when offered Mesodinium rubrum as prey (B), and in brightfield (C)...... 33

Figure 2.8 Fluorescence intensity of chlorophyll a in Dinophysis acuminata offered only Heterosigma akashiwo, both Heterosigma akashiwo and Mesodinium rubrum, and only Mesodinium rubrum as prey. Error bars represent standard error (n=30 for each treatment) and asterisk indicates a statistically significant difference between fluorescence intensity of chlorophyll a (One-way ANOVA and post-hoc Tukey test, p = 1.0 E-13)...... 34

Figure 2.9 Ratio of intensity of emission of chlorophyll a between 649 and 698 nm when using 488 nm and 633 nm excitation wavelengths, in plastids of Heterosigma akashiwo and Teleaulax amphioxeia, and in Dinophysis acuminata that was given Mesodinium rubrum as prey. Error bars represent standard error (n=30 for each treatment) and asterisk indicates a statistically significant difference (One-way ANOVA and post-hoc Tukey test, p = 9.37 E-6)...... 35

Figure 2.10 Ratio of intensity of emission of chlorophyll a when excited using 488 nm vs. 561 nm, in plastids of Dinophysis acuminata when offered only Heterosigma akashiwo, Heterosigma akashiwo and Mesodinium rubrum, or only Mesodinium rubrum as prey. 20 cells of each treatment were observed. Error bars represent standard error (n=30 for each treatment). There was no significant difference between the ratios of intensity of emission of chlorophyll a in any of the treatments (One-way ANOVA, p = 0.752)...... 36

Figure 2.11 Nitrate/nitrite, ammonia, and orthophosphate levels at site TQB during the bloom of Dinophysis acuminata which occurred between December of 2019 and February of 2020 in Torquay Canal, Rehoboth Bay, Delaware...... 37

Figure 2.12 Cell abundance relative to temperature and salinity at site TQB during co-occurring blooms between 2014 and 2017. Bubble size indicates relative cell abundance based on microscopic observation. Panel (A) Dinophysis acuminata (1.0 x 103 - 1.42 x 106 cells L-1) and (B) Heterosigma akashiwo (1.0 x 104 – 3.30 x 106). This data was collected and by UDCMP...... 38

ix ABSTRACT

Dinophysis acuminata is a toxic dinoflagellate that is found worldwide in coastal regions, including the Delaware Inland Bays (DIBs). It produces the toxin, okadaic acid, and has the potential to cause diarrhetic shellfish poisoning. Previous studies demonstrated that lab cultures of D. acuminata acquire plastids from its prey, the ciliate, Mesodinium rubrum, which, in turn, acquired plastids from the cryptophyte,

Teleaulax amphioxeia. Reports of D. acuminata from field samples, however, have found plastids of the raphidophyte, Heterosigma akashiwo within D. acuminata cells, which suggests a broader range of prey than originally thought. This project seeks to better understand prey selection by D. acuminata in laboratory culture and wild populations found in the DIBs. There are three goals for this project: (1) evaluate the relationship between co-occurring blooms of D. acuminata and H. akashiwo in the

DIBs, (2) determine if H. akashiwo is a source of prey for either D. acuminata or M. rubrum, and (3) design a dual excitation ratiometric approach using fluorescence microscopy to identify species-specific pigments in chloroplasts retained by D. acuminata. Our results show that in Torquay Canal, blooms of D. acuminata do not occur in response to the presence of H. akashiwo D. acuminata does not prey on or retain plastids from H. akashiwo. A better understanding of biotic and abiotic factors that contribute to blooms of D. acuminata, such as nutrients, temperature, and prey availability, will help in forecasting blooms and monitoring for toxicity.

x Chapter 1

LITERATURE REVIEW

1.1 Dinophysis acuminata and its presence in the Delaware Inland Bays

Dinophysis is a genus of dinoflagellates common in marine coastal waters. Its widespread distribution and ability to produce toxins harmful to human health have prompted research into understanding its life-cycle, growth mechanisms, distribution, and toxic nature (reviewed by Reguera et al., 2014). Dinophysis acuminata produces the toxins okadaic acid (OA), dinophysistoxins, and pectonotoxins (Yasumoto et al., 1985). Human exposure to OA produced by D. acuminata may cause diarrhetic shellfish poisoning (DSP). Additionally, OA is a tumor promoter in animal carcinogenesis experiments, and has the potential to alter DNA and cellular components, negatively affect the immune and nervous systems, and impact embryonic development (Cohen et al., 1990; Valdiglesias et al., 2013). Dinophysis acuminata releases OA directly into the seawater while also retaining the toxin intracellularly (reviewed by Reguera et al., 2014). Due to its lipophilic nature, OA accumulates in the fat tissue of filter feeders, such as shellfish. The guidance level of

OA, set by the United States Food and Drug Administration (FDA) is 16 µg OA equivalents per 100 g of shellfish tissue (0.16 ppm) (Visciano et al. 2016). When OA equivalents surpass this guidance level, shellfish harvesting closures go into effect. The first U.S. report of shellfish harvesting closures due to OA occurred in 2010 at the Texas Gulf Coast (Deeds et. al, 2010). Since then, DSP closures have spread throughout the U.S. from the state of Washington to the Chesapeake Bay. D.

1 acuminata was first observed in the Delaware Inland Bays (DIBs) in a tributary, Love Creek, Angola, DE, in 2003 by the University of Delaware Citizens Monitoring Program (UDCMP) (Whereat, 2016). Since then, it has been found in various locations throughout the DIBs, including in Torquay Canal, a tributary of the northern Rehoboth Bay. In Torquay Canal, D. acuminata has been detected at cell densities greater than 1 million cells L-1 with toxin levels surpassing the guidance level of 0.16 ppm (Wolny et al., 2020). In 2013, the Delaware Legislature authorized leasing of subaqueous bottom in the DIBs for shellfish aquaculture operations (Regulatons.delaware.gov/AdminCode/title7/3000/3800/3801.shtml). The designated aquaculture areas in Rehoboth Bay, one of the bays in the DIBs are within 3 miles of Torquay Canal, and the potential risk D. acuminata blooms pose to aquaculture operations and human health is unknown. Understanding the bloom dynamics and ecological drivers of blooms of D. acuminata within Torquay Canal is essential to development of monitoring strategies for toxins at aquaculture sites to protect human health.

1.2 Dinophysis acuminata ecology To prevent or mitigate blooms of D. acuminata, an understanding of their ecology is required. Studies of D. acuminata in Northport Harbor, New York, an area with a history of blooms similar to Torquay Canal, revealed the abundance of D. acuminata is significantly correlated with water temperature in Northport Harbor, chlorophyll a concentration, and ciliate density, specifically of the Mesodinium genus (Hattenrath-Lehmann et al., 2013). Nutrients have also been found to directly and indirectly promote blooms of D. acuminata. Ammonia and urea directly promote blooms, while other nutrients such as nitrate and phosphate indirectly promote blooms

2 by stimulating growth of other phytoplankton species that may serve as prey for D. acuminata (Burkholder et al., 2008; Hattenrath-Lehmann et al., 2013; Hattenrath- Lehmann et al., 2015; Garcia-Portela et al., 2020). Although DSP events have been recorded as far back as 1976, the only successful method for long-term culturing of D. acuminata in the laboratory was not established until 2006 (Park et al., 2006, Yasumoto et al., 1985). This method entails feeding D. acuminata the ciliate Mesodinium rubrum, which has itself ingested a cryptophyte, usually Teleaulax sp., (although other cryptophytes have been used; Park et al, 2006). Dinophysis acuminata retains the plastids of M. rubrum for photosynthesis (kleptoplasty), which M. rubrum originally retained from its cryptophyte prey (Johnson et al., 2006, Mafra Jr. et al., 2016, Park et al., 2006). While this method is the only way known to maintain laboratory cultures, wild populations of D. acuminata have been found to contain plastids from other prey sources. For example, wild D. acuminata isolated from Korea contained plastids originating from Pyramimonas sp., T. amphioxeia, Teleaulax acuta, Chroomonas sp., and the raphidophyte Heterosigma akashiwo (Kim et al., 2012). Identifying alternate sources of prey for D. acuminata may help predict blooms, as D. acuminata has been shown to divide more readily in the presence of its prey (Burkholder et al., 2008, Hattenrath- Lehmann et al., 2013).

1.3 Analysis of pigments in plastids of Dinophysis acuminata Accessory pigments are used to widen the spectrum of light being transferred to chlorophyll a, which in turn generates more energy through photosynthesis. These pigments may be unique to different classes of algae. For example, the carotenoid fucoxanthin is abundant in the raphidophyte H. akashiwo but not T. amphioxeia, while

3 phycoerythrin is an accessory pigment in the cryptophyte, T. amphioxeia, and plastids retained by the ciliate, M. rubrum (Hilton et al. 2012; Glazer, A. 1994), but not found in H. akashiwo. Confocal microscopy uses lasers at defined wavelengths that can be used to excite specific accessory pigments within a single cell. For example, the accessory pigments fucoxanthin and phycoerythrin can be excited in algal plastids using 488 nm and 561 nm as excitation wavelengths, respectively, and energy transferred from the accessory pigments to chlorophyll can be measured as emission intensity at 648-698 nm wavelengths. However, the process is not entirely clear cut because other pigments such as chlorophylls c1 & c2, zeaxanthin, violaxanthin, and alloxanthin, which have similar excitation wavelengths as fucoxanthin, may cause interference.

1.4 Research Objectives In Torquay Canal, blooms of D. acuminata and H. akashiwo have historically co-occurred (UDCMP). This project seeks to determine the potential for D. acuminata to prey on and retain plastids from H. akashiwo, and the role of H. akashiwo in promoting blooms of D. acuminata. The specific objectives of this project are to:

• Measure growth rates for D. acuminata when offered H. akashiwo as prey either alone or with the control prey species M. rubrum.

• Measure ingestion rates for M. rubrum and for D. acuminata when offered H. akashiwo as prey.

• Examine cells of D. acuminata for nuclei of prey (nuclear inclusions) when cultured in the presence of H. akashiwo.

4 • Measure the intensity of chlorophyll a in whole cells of D. acuminata when offered H. akashiwo as prey either alone or with M. rubrum.

• Evaluate a dual ratiometric method to identify the contribution of accessory pigments to chlorophyll fluorescence in chloroplasts of recently fed D. acuminata cells.

• Identify potential correlations between cell densities of H. akashiwo and D. acuminata during blooms of D. acuminata in Torquay Canal. The set of controlled laboratory experiments described here will investigate the potential role of H. akashiwo in promoting blooms of D. acuminata in the DIBs, while the field observations will…. In a broader context, results of this project will increase our understanding of predator-prey interactions among HAB dinoflagellates and the contribution of this interaction specifically to D. acuminata bloom dynamics.

5 Chapter 2

HETEROSIGMA AKASHIWO AS A POTENTIAL SOURCE OF PREY FOR THE TOXIC DINOFLAGELLATE DINOPHYSIS ACUMINATA

2.1 Abstract

Dinophysis acuminata blooms co-occurred with the raphidophyte, Heterosigma akashiwo between April and June of 2014-2017 in Torquay Canal, Delaware, USA. In the study presented here, predation on H. akashiwo by D. acuminata and the potential for D. acuminata to retain plastids from H. akashiwo was investigated. Growth rates of D. acuminata were measured when cultured with H. akashiwo either alone or with its known source of prey, Mesodinium rubrum. Ingestion rates were also measured and the presence of nuclear inclusions evaluated in cells of D. acuminata when presented with H. akashiwo as prey. Retention of plastids from H. akashiwo was investigated by measuring chlorophyll a fluorescence intensities in cells presented with H. akashiwo as prey compared to chlorophyll a fluorescence intensities in cells presented with control species M. rubrum. Additionally, a method was developed to identify the presence of the accessory pigment fucoxanthin from plastids originating from H. akashiwo in cells of D. acuminata using confocal microscopy. Results of this investigation do not support the hypothesis that D. acuminata preys on H. akashiwo. Dinophysis acuminata had significantly lower growth and ingestion rates when offered H. akashiwo as prey and exhibited positive growth and ingestion rates only when fed control species M. rubrum. In addition, nuclear inclusions were observed when fed M. rubrum, but not H. akashiwo. Intensity of chlorophyll a fluorescence was lower in D. acuminata when H. akashiwo was offered as prey compared to predation on M. rubrum and the fucoxanthin was not detected in any of the cells examined. A bloom of D. acuminata

6 was monitored between 23 December, 2019 and 11 February, 2020. During this time, H. akashiwo was detected in only two samples out of 37 samples collected during the bloom. These results indicate that the co-occurrence of blooms of D. acuminata and H. akashiwo previously observed was likely due to factors other than predation on H. akashiwo by D. acuminata, which stimulated the growth of each species.

2.2 Introduction Dinophysis acuminata (Dinophysis genus) is a predatory dinoflagellate that produces okadaic acid (OA), a toxin associated with diarrhetic shellfish poisoning (DSP) (Yasumoto et al., 1985). In laboratory culture, D. acuminata is known to prey on the ciliate, Mesodinium rubrum, harboring plastids that originated from one of several cryptophytes, such as Teleaulax amphioxeia, which are then retained by D. acuminata in the process of kleptoplasty (Park et al., 2006; Takishita et al., 2002). Due to its obligate kleptoplastic nature, recently established protocols for culturing D. acuminata strictly use M. rubrum as prey (Park et al., 2006, Burkholder et al., 2008). Dinophysis acuminata has been shown to divide prolifically in the presence of M. rubrum, suggesting that blooms of this species are driven in part by the presence of prey (Burkholder et al., 2008, Hattenrath-Lehmann et al., 2013). However, wild D. acuminata cells have been found to contain plastids of multiple algal origin, including those originating not only from Teleaulax amphioxeia, but also Teleaulax acuta, Chroomonas sp., Pyramimonas sp., and the raphidophyte Heterosigma akashiwo (Kim et al., 2012). The finding of plastids originating from H. akashiwo in D. acuminata cells raises the question of whether H. akashiwo is a source of prey for D. acuminata, thus playing a role in initiating D. acuminata blooms.

7 H. akashiwo is of interest to the Delaware region where blooms of H. akashiwo have co-occurred with blooms of D. acuminata (Figure 2.1). Both H. akashiwo and D. acuminata form dense blooms in Torquay Canal, a tributary of Rehoboth Bay, Delaware (Wolny et al., 2020). In addition, both H. akashiwo and D. acuminata cell densities have been recorded in the millions of cells per liter and Dinophysis toxin levels from extracted shellfish tissue have exceeded the US FDA recommended exposure levels (Wolny et al., 2020). Understanding what initiates blooms of the toxic D. acuminata in Torquay Canal is important in preserving human health, as shellfish aquaculture farming occurs nearby in Rehoboth Bay, DE. This study aims to determine if H. akashiwo is a source of prey for D. acuminata and if D. acuminata retains plastids of H. akashiwo in laboratory culture. To do so, ingestion and growth rates of D. acuminata when offered H. akashiwo as prey were measured. Individual D. acuminata cells that were cultured with H. akashiwo were also examined using a confocal microscope to determine if they contained nuclei of prey, also known as nuclear inclusions. Nuclear inclusions would indicate recent predation, as D. acuminata is known to feed with a peduncle and retain the organelles of its prey, which includes chloroplasts and nuclei. To investigate the origin of plastids in D. acuminata, I designed a dual excitation ratiometric approach. This technique measures the relative contribution of the accessory pigments fucoxanthin (Ex: 488 nm) and phycoerythrin (Ex: 561 nm) to chlorophyll fluorescence. Since fucoxanthin is present in H. akashiwo but not M. rubrum, and phycoerythrin is present in M. rubrum but not H. akashiwo, the ratio of chlorophyll fluorescence intensity when excited at 488 vs. 561 nm in D. acuminata might be used to indicate the presence of H. akashiwo plastids in D. acuminata cells.

8 Finally, I examined the hypothesis that blooms of D. acuminata co-occur with H. akashiwo. I used historical data to determine the correlation between H. akashiwo and D. acuminata cell abundances in Torquay Canal and analyzed historical trends in temperature and salinity of co-occurring blooms to evaluate environmental factors that may contribute to these blooms. I also monitored a bloom of D. acuminata that occurred in Torquay Canal during the winter of 2019-2020, and the presence/absence of H. akashiwo in samples collected during the bloom.

2.3 Methods

2.3.1 Algal and ciliate cultures

Cultures of D. acuminata originally isolated from Torquay Canal, Rehoboth Beach, DE (Lat. = 38.699031, Long. = -75.112409) were shared by Dr. Juliette Smith (Virginia Institute of Marine Science, Gloucester Point, VA). Cultures of M. rubrum and T. amphioxeia were shared by Dr. Matthew Johnson (Woods Hole Oceanographic Institution, Woods Hole, MA). H. akashiwo (strain CCMP 2393, National Center for Marine Algae and Microbiata, Bigelow, ME) was originally isolated from Torquay Canal. All species were cultured in 20 PSU f/6 media at 18 °C with a 12:12 light: dark regime (Goa et al., 2019). Cultures of D. acuminata were maintained with M. rubrum at a predator: prey ratio of 4:1 and cultures of M. rubrum were fed T. amphioxeia at a predator: prey ratio of 4:1, both on a weekly basis. As needed, cultures of D. acuminata were filtered through a 10 µm mesh sieve to prevent overgrowth of the cryptophyte, T. amphioxeia.

9 2.3.2 Growth Curves Dinophysis acuminata were maintained for two weeks without prey prior to the start of the experiment. Cells were enumerated using a Sedgewick rafter chamber (Goa et al., 2019). D. acuminata was transferred into 150 mL flasks and subjected to the following treatments: only H. akashiwo, both H. akashiwo and M. rubrum, and only M. rubrum were added separately to each of 3 flasks at a predator: prey ratio of 1:4. On days 1, 3, 6, 8, and 10, the cell density of D. acuminata was determined using a Sedgewick rafter chamber, counting at least 100 cells. Cell densities of M. rubrum and

H. akashiwo were measured using a hemocytometer. The following calculation was used to determine growth rate (µ):

ln(%&'' )*+,-),%& . %&'' )*+,-),%& . ) µ = /0123 0104023 number of days

2.3.3 Predation Experiments Dinophysis acuminata, M. rubrum, H. akashiwo, and T. amphioxeia were cultured as above and enumerated using a Sedgewick rafter chamber or hemocytometer. M. rubrum was transferred to separate wells of a 12-well plate (n=6) and cultured with H. akashiwo and T. amphioxeia using a predator: prey ratio of 1: 20. This experiment was repeated with D. acuminata (n=6) with M. rubrum and H. akashiwo as prey using a predator: prey ratio of 1:20. At 24 hours, prey cell density was enumerated using a hemocytometer, counting at least 100 cells. Growth and grazing rates were calculated as in Frost (1972), Heinbokel (1978) and Jeong and Latz (1994), along with the grazing constant, clearance rate, and ingestion rate (see Appendix 1 for equations and examples).

10 2.3.4 Microscopic analysis During the growth curve experiment, a 2-mL aliquot from each flask was removed at 48 hours and 0.2 µL of the nuclear stain Hoechst 33342 was added. Samples were then filtered onto 25 mm 0.2 µm black polycarbonate filters and mounted on a microscope slide. Each filter received one drop of Citifluor AF1 mountant media (Ted Pella, Inc., Redding, CA) and was covered with a coverslip (Martin-Cereceda et al., 2008; Martinez et al., 2014). Cells were imaged on a Zeiss LSM 710 inverted confocal microscope fitted with a 40x C-Apochromat water immersion objective (NA = 1.2) (Carl Zeiss Inc., Thornwood, NY). 405 nm and 633 nm laser lines were used to excite Hoechst 33342 and chlorophyll a, respectively from individual cells. Fluorescence emission from Hoechst 3342 was collected between 415 and 497nm, and fluorescence emission from chlorophyll a was collected at 649-698nm. Z-stack images with a step of 0.3 µm were acquired from bottom to top of the cells. Mean intensity values of chlorophyll a emission were determined using the ImageJ/Fiji (NIH, Bethesda, MD) software. Briefly, maximum intensity projections of the z-stacks were segmented to measure intensities from single cells.

2.3.5 Ratiometric analysis of accessory pigments

This method was evaluated using confocal microscopy to image pure cultures of H. akashiwo, T. amphioxeia, and D. acuminata. Cells were filtered onto 25 mm 0.2 µm black polycarbonate filters and mounted on a microscope slide. Each filter received one drop of Citifluor AF1 mountant media (Ted Pella, Inc., Redding, CA) and was covered with a coverslip (Martin-Cereceda et al., 2008; Martinez et al., 2014). 488 nm and 633 nm laser lines were used to excite fucoxanthin and chlorophyll a,

11 respectively in cells. Fluorescence emission when excited at these wavelengths was collected between 649-698 nm. The intensity of emission when excited at each wavelength was then determined in ImageJ/Fiji (NIH, Bethesda, MD) software. Ratios of emission intensity with excitation at 488 nm compared to emission intensity when excited at 633 nm determined for the pure cultures of H. akashiwo, T. amphioxeia and D. acuminata Dinophysis acuminata was then cultured for 48 hours with H. akashiwo, a mix of H. akashiwo and M. rubrum, or only M. rubrum as prey as described above for growth experiments. Cells were mounted onto microscope slides as above and the 488 nm and 561 nm laser lines were used to excite fucoxanthin and phycoerythrin, respectively in cells of D. acuminata. Fluorescence emission when excited at these wavelengths was collected between 649-698 nm and the ratio of chlorophyll fluorescence at each excitation wavelength (Ex:488 to Ex:561) was calculated as above.

2.3.6 Occurrence of Heterosigma akashiwo and Dinophysis acuminata at Torquay Canal

Between 2014 and 2017, bi-weekly water samples between April and October were collected at Torquay Canal by the University of Delaware Citizens Monitoring program. Salinity, temperature, and cell density of D. acuminata and H. akashiwo were recorded from these samples. In this study, the relationship between salinity, temperature and H. akashiwo cell density and co-occurring blooms of D. acuminata between 2014 and 2017 were analyzed (Data shared by UDCMP). Weekly sampling of site TQB in the Torquay Canal was performed during a winter bloom of D. acuminata between 23 December, 2019 and 11 February, 2020.

12 Transects of Torquay Canal, which included sites TQB, TQ 12, 11, 10, 9, and 8, were completed on January 3, 7, and 14, 2020, and on February 4, 2020 (Figure 2.2). Surface water samples were collected in 1-liter bottles and filtered through 120 um mesh to remove debris and zooplankton. Temperature, salinity, pH, conductivity, and dissolved oxygen were measured using a YSI 5564 probe (Xylem Incorporated, Yellow Springs, OH) and Secchi disk readings were collected. Aliquots of each sample were filtered in duplicate onto GF/F glass microfiber filters for chlorophyll analysis. Chlorophyll was extracted in 90% acetone in the dark at -20 °C for 24 hours and in vitro concentration of chlorophyll was measured using a Turner 10 AU Fluorometer (Turner Designs, Sunnyvale, CA). Water samples were preserved in Lugol’s to determine cell density of D. acuminata using light microscopy. Triplicate water samples were filtered through a 3.0 µm polycarbonate membrane filter, submerged in CTAB extraction buffer (cetyltrimethyl ammonium bromide + 4% 2-

Mercaptoethanol) with pGEM at a concentration of 20 ng µL-1, and stored at -80 °C until extraction of DNA (Coyne et al., 2001). The filtrate from one subsample was then passed through a 0.2 µm polycarbonate membrane filter and the filtrate stored at -

3- + - 20 °C for dissolved nutrient analysis (PO4 , NH4 , NOx ) with a Seal AA3 autoanalyzer (SEAL Analytical Inc., Mequon, WI).

2.3.7 Molecular analysis of bloom samples- looking for Heterosigma akashiwo DNA was extracted as described in Coyne et al. (2001) and resuspended in LoTE (3 mM Tris-HCl, 0.2 mM EDTA at a pH of 7.5). Polymerase chain reaction (PCR) assays were conducted using primers targeting the H. akashiwo 18S rRNA gene in samples as described in Coyne et al. (2005) using primers Hs1350F (5’- CTAAATAGTGTCGGTAATGCTTCT-3’) and Hs1750R (5’-

13 GGCAAGTCACAATAAAGTTCCAT-3’). The PCR cycling parameters consisted of 35 cycles of 30 s at 94°C, 30 s at 56 °C, and 1 min at 72 °C, followed by a 5-min extension at 72 °C (Coyne et al., 2005). PCR products were visualized using gel electrophoresis.

2.3.8 Statistical Analysis Statistical analysis was performed using a one-way ANOVA test in excel on individual data collected during the following experiments: growth curve, ingestion rate, chlorophyll a intensity, and the ratio of intensity of emission of both fucoxanthin: chlorophyll a and fucoxanthin: phycoerythrin. If data were found to be significantly different (p ≤ 0.05), then a Tukey honestly significant difference (HSD) post hoc test was conducted. Pearson’s correlation coefficient was determined between D. acuminata and H. akashiwo cell density in Torquay Canal between 2011 and 2018.

2.4 Results

2.4.1 Growth Rates The growth rate of D. acuminata was determined when offered H. akashiwo or a mix of H. akashiwo and M. rubrum as prey and compared to the growth of D. acuminata control cultures when offered M. rubrum alone (Figure 2.3). The mean growth rate of D. acuminata when offered only H. akashiwo as prey was -0.037 d-1 and when offered both M. rubrum and H. akashiwo was -0.0076 d-1 (Appendix 2). The growth rate of D. acuminata in control cultures with only M. rubrum was 0.050 d-1 and significantly higher (p = 0.0028) than the growth rate when in the presence of both H. akashiwo and M. rubrum or H. akashiwo alone (Appendix 2). During this experiment, the growth rate of H. akashiwo in treatments with D. acuminata was 0.28

14 d-1 and of H. akashiwo when in the presence of M. rubrum and D. acuminata was 0.30 d-1 (Figure 2.4, Appendix 2). M. rubrum cells were not detected after 3 days in flasks containing only D. acuminata and those containing H. akashiwo and D. acuminata (Figure 2.4).

2.4.2 Ingestion Rates In predation experiments, the ingestion rate of M. rubrum when offered H. akashiwo as prey (-1.6 cells d-1) was significantly lower (p = 2.0 E-6) than when offered T. amphioxeia as prey (4.26 cells d-1) (Figure 2.5, Appendix 2). The ingestion rate of D. acuminata when offered H. akashiwo as prey (-2.96 cells d-1) was also significantly lower (p = 2.0 E-5) than when offered M. rubrum as prey (14.7 cells d-1) (Figure 2.6, Appendix 2).

2.4.3 Analysis of nuclear inclusions

Nuclear inclusions were not detected in cells of D. acuminata that were offered H. akashiwo or a combination of M. rubrum and H. akashiwo as prey, and were only detected in cells that were offered M. rubrum as sole prey (Figure 2.7). Chlorophyll a emission intensity was significantly lower (p = 1.03 E-13) in D. acuminata that were presented with only H. akashiwo as prey (140.3 ± 15.5) or a combination of H. akashiwo and M. rubrum as prey (192.8 ± 15.9) than in cells of D. acuminata that were presented with only M. rubrum as prey (361.0 ± 16.2) (Figure 2.8, Appendix 2).

2.4.4 A ratiometric approach using chlorophyll a emission to evaluate plastid origin

The ratiometric approach was first evaluated using pure cultures of H. akashiwo, T. amphioxeia and D. acuminata that were given M. rubrum as prey. The

15 ratio of chlorophyll a emission intensity in H. akashiwo cells when excited at 488 nm

(for fucoxanthin) and 633 nm (for chlorophyll a) was significantly higher (1.31 ± 0.12; p = 9.37 E-6) than the ratio of chlorophyll a emission intensity collected from T. amphioxeia plastids (1.21 ± 0.01) and plastids in D. acuminata that were offered M. rubrum as prey (0.98 ± 0.28) (Figure 2.9, Appendix 2). There was no statistically significant difference (p = 0.752) between the ratio of chlorophyll a emission intensity in cells of D. acuminata that were presented with only H. akashiwo, H. akashiwo and M. rubrum, or only M. rubrum as prey when excited at 488 nm (for fucoxanthin) vs. 561 nm (for phycoerythrin) (Figure 2.10, Appendix 2). However, several individual cells of D. acuminata which had been given either H. akashiwo or both H. akashiwo and M. rubrum had a higher ratio of chlorophyll a emission intensity. Examination of data indicates that this was due to a lower emission intensity when excited at 561 nm (for phycoerythrin).

2.4.5 Relationships between D. acuminata and H. akashiwo abundances in Torquay Canal

Pearson’s correlation between presence of D. acuminata and of H. akashiwo in Torquay Canal between 2011 and 2018 showed no correlation between the presence of D. acuminata and of H. akashiwo (r = 0.22, p = 0.16) (Figure 2.1). During this time, salinity ranged from 18.5 to 27.6 and temperature ranged from 17.7 °C to 25.9 °C. A bloom of D. acuminata occurred in Torquay Canal, Rehoboth Bay, DE, first detected on 23 December, 2019 and was monitored until 11 February, 2020. Cell density ranged from 217,000 cells L-1 to 4,000 cells L-1 at site TQB, the site with the highest cell density. At TQB, mean temperature was 7.41 °C and mean salinity was 28.05. Temperature at TQB ranged from 4.47 °C to 9.42 °C and salinity ranged from

16 26.79 to 28.70 (Table 2.1). Using the H. akashiwo specific PCR and 18S rDNA primers Hs1350F and Hs1750R, H. akashiwo was detected on 23 December, 2019 at sites TQB and TQ 11 and on 3 January, 2020 at site TQB (Table 2.1). Mean ammonium concentration at site TQB was 4.93 ± 0.77, mean orthophosphate concentration was 0.33 ± 0.07, and mean nitrate and nitrite concentration was 16.40 ± 3.10 (Appendix 2). Maximum ammonia concentrations (9.58 uM) occurred at TQB on 9 January, 2020, maximum orthophosphate (0.95 uM) occurred on 14 January, 2020, and maximum nitrate and nitrite (48.64 uM) occurred on 3 February, 2020 as (Figure 2.11).

2.5 Discussion Dinophysis acuminata blooms have been found to co-occur with many phytoplankton species, including raphidophytes such as H. akashiwo (Burkholder et al., 2008; Ma et al., 2006). When D. acuminata is practicing kleptoplasty over autotrophy, there is an increase in its division rate, resulting in a bloom (Burkholder et al., 2008, Hattenrath-Lehmann et al., 2013). Laboratory cultures of D. acuminata have only been observed to feed on and retain plastids from the ciliate M. rubrum, and when offered it as prey, has a growth rate that is three times higher than when no prey is provided, due to the increased trophic efficiency associated with mixotrophy over autotrophy (Figure 2.3; Park et al., 2006, Sanders, 1991). During a bloom of D. acuminata off the coast of Korea, however, plastids of multiple algal origins, including plastids from H. akashiwo were detected in cells using PCR and restriction fragment length polymorphism (RFLP) analysis (Kim et al., 2012). The co-occurrence of blooms of D. acuminata and H. akashiwo in the DIBs, and the finding of D. acuminata cells containing plastids originating from H. akashiwo in other coastal

17 regions, warranted further research to determine if D. acuminata preys on H. akashiwo. The objective of the research presented here was to examine the link between co-occurring blooms of H. akashiwo and D. acuminata in Torquay Canal and the potential for D. acuminata to prey on H. akashiwo and retain plastids from this species. The present findings indicate that D. acuminata does not prey on H. akashiwo in controlled laboratory culture experiments. The growth rate of D. acuminata when provided with H. akashiwo alone (-0.037 d-1) or a combination of M. rubrum and H. akashiwo (-0.0076 d-1) as prey was significantly lower (p = 0.0028) than when provided with M. rubrum as sole prey (0.050 d-1) (Figure 2.3). This suggests that growth of D. acuminata is inhibited by the presence of H. akashiwo, even when its preferred prey (M. rubrum) is present. Tong et al. (2010) calculated growth rates of D. acuminata that were provided with M. rubrum as prey to be 0.11 d-1 at 4 °C and 0.23 d-1 at 10 °C. These growth rates were higher than what was measured in this study. The lower growth rate reported here may be due to a lower initial predator: prey ratio. This experiment used a 4: 1 predator: prey ratio while Tong et al. (2010) used a ratio of 20: 1. Further evidence that D. acuminata selectively feeds on M. rubrum is the disappearance of M. rubrum and increase in H. akashiwo cell abundance in predation experiments (Figure 2.4). As the ingestion rate experiments (discussed below) demonstrate that D. acuminata does not prey on H. akashiwo, the initial decrease in H. akashiwo cell abundance in growth experiments was likely due to growth inhibition after transfer to a 12-well plate from a larger flask. Results of this study show that the ingestion rate of D. acuminata when provided with H. akashiwo as prey (-2.96 cells d-1) was significantly lower (p = 2.0 E-

18 5) than when provided with M. rubrum as prey (14.7 cells d-1) (Figure 2.6, Appendix 2). Higher temperatures have been shown to increase the activity and feeding rate of plankton (Hansen et al., 1997; Kamiyama et al., 2009), with reported average ingestion rates of M. rubrum by D. acuminata ranging from 1.18 cells d-1 at 4 °C and 2.06 cells d-1 at 10 °C (Tong et al. 2010). Predation experiments conducted here were at an even higher temperature (18 °C), which may have contributed to the higher ingestion rates measured in this study. In spite of the higher activity, there was no measurable predation on H. akashiwo. The negative ingestion rate of D. acuminata when offered H. akashiwo as prey shows that H. akashiwo grew during the experiment. This growth may have occurred in response to nutrients present in the media or exudates like dissolved organic matter originating from D. acuminata. To determine if an intermediate may be required to introduce H. akashiwo plastids into D. acuminata, I measured ingestion rated for M. rubrum when offered H. akashiwo as prey. Results demonstrate that M. rubrum does not prey on H. akashiwo. Ingestion rates for M. rubrum when offered H. akashiwo as prey (-1.6 cells d-1) were significantly lower (p = 2.0 E-6) than when offered T. amphioxeia as prey (4.26 cells d-1) (Figure 2.5, Appendix 2). Dinophysis acuminata has been observed to feed with a peduncle, acquiring all contents of its prey (Park et al., 2006). Nuclear inclusions in M. rubrum cells that recently fed on T. amphioxeia have been well documented (Taylor et al., 1969, 1971; Hibberd, 1977; Oakley and Taylor, 1978, Kim et al., 2017). Here, nuclear inclusions were used as a metric to indicate recent feeding by D. acuminata and were only observed in D. acuminata when offered M. rubrum as prey. There was no evidence of nuclear inclusions in D. acuminata when offered H. akashiwo as prey (Figure 2.7).

19 An increase in the photosynthetic rate and chlorophyll a concentration has been documented in D. acuminata fed M. rubrum (Nielsen et al., 2012; Hansen et al., 2016). In this project, chlorophyll a intensity was used as an additional metric for recent predation. Chlorophyll a intensity was significantly lower in D. acuminata cells provided with H. akashiwo (16.3 ± 1.2) or no prey (20.3 ± 1.5) than in cells offered M. rubrum as prey and which contained nuclei of M. rubrum (29.2 ± 1.4) (Figure 2.8, Appendix 2). Despite finding no nuclear inclusions in D. acuminata offered H. akashiwo as prey, it was still possible that plastids were retained. A ratiometric approach to evaluate energy transfer from accessory pigments present in each plastid type was developed. Here, chlorophyll a emission was measured using fluorescence microscopy after excitation at 488 nm to measure energy transfer from fucoxanthin, and 561 nm to measure energy transfer from phycoerythrin. To validate this method, chlorophyll emission was measured in pure cultures of H. akashiwo and T. amphioxeia, as well as D. acuminata that had been given M. rubrum as prey, when cells were excited at 488 nm and normalized to emission when excited at 633 nm (excitation wavelength for chlorophyll a). H. akashiwo was found to have a significantly higher ratio of chlorophyll emission intensity (normalized value=1.31 ± 0.12) than that of T. amphioxeia (normalized value =1.21 ± 0.01) or D. acuminata (normalized value =0.98 ± 0.28) (Figure 2.9, Appendix 2). This suggests that chlorophyll a emission intensity when excited at 488 nm in D. acuminata cells retaining plastids from H. akashiwo would be greater than in cells retaining plastids without fucoxanthin from M. rubrum. For this experiment, D. acuminata were offered only H. akashiwo, both H. akashiwo and M. rubrum, and only M. rubrum as prey. Chlorophyll emission in D. acuminata

20 was measured after excitation at 488 nm and normalized to the emission when excited at 561 nm (for phycoerythrin). There was no significant difference (p = 0.21) in chlorophyll emission intensity at 488 nm excitation wavelength (normalized to emission at 561 nm wavelength) in any of the treatment groups (Figure 2.10). This suggests that plastids from H. akashiwo were not present in D. acuminata and is consistent with results of other experiments in this study. In Torquay Canal, blooms of H. akashiwo and D. acuminata co-occurred between April and June of each year between 2014 and 2017 when water temperatures were between 15.0 °C and 26.4 °C, the optimal temperature range for blooms of each species (Figure 2.12; UDCMP; Zhang et al., 2006; Handy et al., 2008; Hattenrath- Lehmann, 2013). Bloom co-occurrence of D. acuminata and H. akashiwo in Torquay Canal is likely driven by fluxes in seasonal temperature, nutrient inputs, and/or predator-prey dynamics (Connell & Jacobs, 1997; Hattenrath-Lehman et al., 2013; Hattenrath-Lehman et al., 2015). The Pearson’s correlation for samples collected at Torquay Canal between 2011 and 2018 revealed that the cell abundance of D. acuminata is not correlated with H. akashiwo, which suggests that the co-occurrence of each species are independent of one another (p = 0.16) (Figure 2.1). A bloom of D. acuminata in Torquay Canal was detected on 23 December, 2019 and monitored until 4 February, 2020 (Pappas et al., 2020). Highest cell density (217,000 cells L-1) occurred at TQB on January 6th (Pappas et al., in progress) (Table 2.1). M. rubrum was present in live water samples over the course of the entire bloom (Williams and Pappas, personal observations). Using a species-specific PCR assay, H. akashiwo was detected on the first 2 days of sampling, 23 December, 2019 at sites TQ11 and TQ12 and 3 January, 2020 at site TQB, but not observed in live water

21 samples (Table 2.1). For the remainder of the bloom, H. akashiwo was not detected at any sites. It should be noted that H. akashiwo forms cysts, allowing survival through adverse conditions such as high and low temperatures (Portune et al., 2009). The optimal range of temperature for H. akashiwo cyst germination in the DIBs is between

13 °C and 17 °C (Portune et al., 2009), higher than temperatures recorded during the bloom of D. acuminata recorded here (4.47 °C to 9.80 °C). Due to temperature constraints for cyst germination, seasonal blooms of H. akashiwo typically occur in Torquay Canal beginning in April or May after a rise in water temperature (Zhang et al., 2006; Handy et al., 2008; Portune et al., 2009). A cyst stage for D. acuminata has not yet been identified, suggesting that the winter bloom in Dec. 2019 was seeded by D. acuminata cells in the water column. In spite of the low water temperatures during the winter bloom of D. acuminata, the presence of preferred prey and elevated nutrients together likely stimulated the bloom. In fact, M. rubrum was observed in most water samples collected during the bloom (Williams and Pappas, personal observations). In a study by Hattenrath-Lehman et al. (2015), blooms of D. acuminata in Northport Harbor were found to be correlated with nutrient levels, suggesting that nutrients may have played a role in bloom initiation of D. acuminata in Torquay Canal in December 2019

(Figure 2.11). Peak nutrient levels at site TQB during the bloom (NOx = 48.64 uM,

NH3= 9.58 uM, OP= 0.95 uM) were higher than average nutrients levels reported by Hattenrath-Lehman et al. (2015) during blooms of D. acuminata in Northport Harbor over a 4-year period (NOx= 12.10 uM, NH3= 3.42 uM, DOP = 0.80 uM). High concentrations of ammonia, in particular, may have promoted cell division in D. acuminata (Garcia-Portela et al., 2020), while high nitrate and phosphate

22 concentrations may have contributed to the bloom by stimulating growth of M. rubrum and/or it’s cryptophyte prey (Tong et al., 2015; Johnson et al., 2015).

2.6 Conclusions While D. acuminata cells isolated in the wild have been found to contain plastids from H. akashiwo, predation on H. akashiwo by D. acuminata or by an intermediate, M. rubrum, was not observed in the laboratory (Kim et al., 2012). Growth rates of D. acuminata when provided with H. akashiwo as prey were close to zero and ingestion rates were negative. Microscopic analysis showed an absence of nuclear inclusions, suggesting that even low levels of predation were not occurring. Furthermore, chlorophyll a intensity when offered H. akashiwo as prey was indistinguishable from starved cells, and there was no significant difference in the ratio of chlorophyll a emission when excited at 488 nm to chlorophyll emission when excited at 561 when D. acuminata was offered H. akashiwo as prey compared to predation on M. rubrum. While results presented here also do not support predation of H. akashiwo by M. rubrum, it’s possible that other ciliates may provide a conduit for transfer of plastids from H. akashiwo to D. acuminata such as Strombidinopsis acuminatum, Metacylis sp., or Favella sp. (Strom et al., 2013). Dinophysis acuminata has not yet been observed to feed on any of these ciliates, however these ciliates have been observed to retain plastids from H. akashiwo, raising the possibility that they are grazed by D. acuminata in the absence of M. rubrum (Strom et al., 2013). It is possible that the H. akashiwo plastids discovered in cells of D. acuminata by Kim et al., (2012) were not obtained directly by D. acuminata but through one of the above ciliates. To better understand the ecology of D. acuminata and factors that contribute to bloom initiation, further research into the predator-prey dynamics will be necessary.

23 Research into the life history of D. acuminata may also be insightful as it could reveal the ability for D. acuminata to form cysts or a resting stage where cells remain in the sediment until environmental conditions become favorable. In sum, temperature and nutrient availability likely provide an environment that favors blooms of both D. acuminata and H. akashiwo in Torquay Canal. Further analysis of environmental data in this and other mid-Atlantic estuaries with frequent blooms of D. acuminata could help determine conditions that initiate blooms and what causes blooms to re-occur annually in these areas.

24 Table 2.1 Environmental metadata, cell density of D. acuminata (Pappas in prep), and presence of H. akashiwo during the winter 2019-2020 bloom of D. acuminata at site TQB in Torquay Canal, Rehoboth Beach, DE.

Dinophysis acuminata cell Heterosigma Date & Site Temp. C Sal. abundance akashiwo Sampled (cells mL-1) presence/absence (Pappas in prep.)

12/23/2019 TQB 6.7 28.15 29 + TQ12 8.00 28.54 8.00 - TQ11 1.00 28.89 1.00 +

1/3/2020 TQB 8.61 28.35 12.33 + TQ12 8.39 28.48 7.67 - TQ11 8.03 28.71 0.67 - TQ10 8.29 29.18 0.33 - TQ9 8.44 29.77 2.00 - TQ8 8.33 29.78 0.33 - 1/6/2020 7.08 26.79 217.33 - TQB 1/7/2020 TQB 8.25 27.92 27.67 - TQ12 7.97 27.84 30.33 - TQ11 6.98 27.73 10.33 - TQ10 7.40 27.17 4.33 - TQ9 7.08 28.79 2.33 - TQ8 6.69 28.3 0.00 - 1/8/2020 8.63 27.76 46.00 - TQB 1/9/2020 7.73 27.56 7.67 - TQB

25 1/14/2020 TQB 9.42 26.49 4.00 - TQ12 9.55 26.35 15.67 - TQ11 6.98 26.62 11.67 - TQ10 9.74 26.7 0.33 - TQ9 9.75 29.11 1.67 - TQ8 9.78 29.41 1.00 - TQBay 9.81 29.61 1.00 - 1/17/2020 7.5 28.28 29.00 - TQB 1/24/2020 4.47 29.39 16.33 - TQB 1/31/2020 6.31 28.59 41.33 - TQB 2/3/2020 6.6 28.7 6.00 - TQB 2/4/2020 TQB 7.63 28.65 30.67 - TQ12 7.52 28.77 3.00 - TQ11 7.57 28.58 2.00 - TQ10 7.83 28.23 0.33 - TQ9 8.07 28.34 0.00 - TQ8 8.21 28.91 0.33 - TQBay 7.45 29.61 0.33 -

Figure 2.2 Map of Torquay Canal, Rehoboth Bay, Delaware (38.699031, -75.112409) sampling sites and transect route.

Nuclear inclusions A B

C 10 um

Figure 2.7 Dinophysis acuminata stained with Hoechst 33342, with no nuclear inclusions when offered Heterosigma akashiwo as prey (A), with nuclear inclusions when offered Mesodinium rubrum as prey (B), and in brightfield (C).

acuminata

Dinophysis n

Fluorescence intensity of chlorophyll

Figure 2.12 Cell abundance relative to temperature and salinity at site TQB during co- occurring blooms between 2014 and 2017. Bubble size indicates relative cell abundance based on microscopic observation. Panel (A) Dinophysis acuminata (1.0 x 103 - 1.42 x 106 cells L-1) and (B) Heterosigma akashiwo (1.0 x 104 – 3.30 x 106). This data was collected and by UDCMP.

38 REFERENCES

Burkholder, J., Glibert, P., Skelton, H. (2008). Mixotrophy, a major mode of nutrition for harmful algal species in eutrophic waters. Harmful Algae, 8, 77-93

Connell L. B., Jacobs M. A.. Strickland R. M. (1997). Anatomy of a bloom: Heterosigma akashiwo in Puget Sound, Puget Sound Research 98:830- 834

Coyne, K., Handy, S., Demir, E., Whereat, E., Hutchins, D., Portune, K., Doblin, M., Cary, S. (2005). Improved quantitative real-time PCR assays for enumeration of harmful algal species in field samples using an exogenous DNA reference standard. Limnology and Oceanography: Methods, 3:381-391

Coyne, K. J., Hutchins, D.A., Hare, C.E., Cary, S.C. (2001). Assessing temporal and spatial variability in Pfiesteria piscicida distributions using molecular probing techniques. Aquatic Microbial Ecology, 24:275-285

Demir-Hilton, E., Hutchins, D., Czymmek, K., Coyne, K. (2012). Description of Viridilobus marinus (Gen. Et Sp. Nov.), a new raphidophyte from Delaware Inland Bays. Journal of Phycology, 48:1220-1231

Frost, B.W. (1972). Effects of size and concentration of food particles on the feeding behavior of the marine planktonic copepod Calanus pacificus. Limnology and Oceanography,17, 6:805-815

Garcia-Portela, M., Reguera, B., Gago, J., Le Gac, M., Rodriguez, F. (2020). Uptake of inorganic and organic nitrogen sources by D. acuminata and D. acuta. Microorganisms, 8, 187:1-15

Goa, H., Tong, M., An, X., Smith, J. (2019) Prey lysate enhances growth and toxin production in an isolate of Dinophysis acuminata. Toxins, 11, 57:1-16

Handy, S., Coyne, K., Portune K., Demir, E., Doblin, M., Hare, C., Cary, S., Hutchins, D. (2005) Evaluating vertical migration behavior of harmful raphidophytes in the Delaware Inland Bays utilizing quantitative real-time PCR. Aquatic Microbial Ecology, 40, 2:121–132

Hansen P J, Bjørnsen P K, Hansen B W. 1997. Zooplankton grazing and growth: Scaling within the 2-2,000-µm body size range. Limnology and Oceanography, 42(4): 687-704

39 Hansen, P., Ojamae, K., Berge, T., Trampe, E., Nielsen, L., Lips, I., Kuhl, M. (2016). Photoregulation in a kleptochloroplastidic dinoflagellate, Dinophysis acuta. Frontiers in Microbiology: 7:785:1-11

Hattenrath-Lehman, T., Marcoval, M., Mittlesdorf H., Goleski, J., Wang, Z., Haynes, B., Morton, S., Gobler, C. (2015). Nitrogenous nutrients promote the growth and toxicity of Dinophysis acuminata during estuarine bloom events. PloS One, 10, 4: e0124148

Hattenrath-Lehmann, T., Marcoval, M., Berry, D., Fire, S., Wang, Z., Morton, S., Gobler, C. (2013). The emergence of Dinophysis acuminata blooms and DSP toxins in shellfish in New York waters. Harmful Algae, 26:33-44

Heinbokel, J.F. (1978). Studies on the functional role of tintinnids in the Sothern California Bight. I. Grazing and growth rates in laboratory cultures. Marine Biology, 47: 177-189

Jeong, H., Latz, M. (1994). Growth and grazing rates of the heterotrophic dinoflagellates Protoperidinum spp. On red tide dinoflagellates. Marine Ecology Progress Series, 106:173-185

Kamiyama T, Suzuki T. 2009. Production of dinophysistoxin-1 and pectenotoxin-2 by a culture of Dinophysis acuminata (Dinophyceae). Harmful Algae, 8(2): 312- 317

Kim, M., Drumm, K., Daugbjerg, N., Hansen, P. (2017). Dynamics of sequestered cryptophyte nuclei in Mesodinium rubrum during starvation and refeeding. Frontiers in Microbiology, 8, 423

Kim, M., Kim, S., Yih, W., Park, M. (2012). The marine dinoflagellate genus Dinophysis can retain plastids of multiple algal origins at the same time. Harmful Algae, 13:105-111

Lindahl, O., Lundve, B., Johansen, M. (2007). Toxicity of Dinophysis spp. in relation to population density and environmental conditions on the Swedish west coast. Harmful Algae, 6: 218-231

Ma, S., Whereat, E., Luther, G., (2006). Shift of algal community structure in dead end lagoons of the Delaware Inland Bays during seasonal anoxia. Aquatic Microbial Ecology, 44:279-290

MacKenzie, L., Beuzenberg, V., Holland, P., McNabb, P., Selwood, A., (2004). Solid phase adsorption toxin tracking (SPATT): a new monitoring tool that simulates

40 the biotoxin contamination of filter feeding bivalves. Toxicon, 44:901-918

Martin-Cereceda, M., Williams, R., Novarino, G. (2008). Easy visualization of the protist Oxyrrhis marina grazing on a live fluorescently labelled heterotrophic nanoflagellate. Current Microbiology, 57:45-50

Martinez, R., Isari, S., & Calbet, A. (2014). Use of live, fluorescently-labeled algae for measuring microzooplankton grazing in natural communities. Journal of Experimental Marine Biology and Ecology, 457:59-70

Nielsen, L., Krock, B., Hansen, P. (2012). Effects of light and food availability on production, growth, and photosynthetic rate of Dinophysis acuminata. Marine Ecology Progress Series, 471:37-50

Nishitani, G., Nagai, S., Hayakawa, S., Kosaka Y., Sakurada, K. 2012. Multiple plastids collected by the dinoflagellate Dinophysis mitra through kleptoplasty. Applied and Environmental Microbiology, 78:813-821

Pappas, A. In progress. Winter bloom dynamics and molecular analysis of benthic sediments for the toxic dinoflagellate, Dinophysis acuminata, at Torquay Canal, Rehoboth Bay, Delaware, USA. Thesis submitted to Delaware State University

Park, M., Kim, S., Kim, H., Myung, G., Kang, Y., & Yih, W. (2006). First successful culture of the marine dinoflagellate Dinophysis acuminata. Aquatic Microbial Ecology, 45:101-106

Portune, K., Coyne, K., Hutchins, D., Handy, S., Cary, S. (2009). Quantitative real- time PCR for detecting germination on Heterosigma akashiwo and Chattonella subsalsa cysts from Delaware’s Inland Bays, USA, Aquatic Microbial Ecology, 55:229-239

Rial, P., Garrido, J., Jaen, D., Rodriguez, F. (2012). Pigment composition in three Dinophysis species (Dinophyceae) and the associated cultures of Mesodinium rubrum and Teleaulax amphioxeia. Journal of Plankton Research, 35, 2: 433– 437.

Sanders, R.W., 1991. Mixotrophic protists in marine and freshwater ecosystems. Journal of Eukaryotic Microbiology, 31:76–81.

Smith J., Tong, M., Kulis, D., Anderson, D. (2018). Effect of ciliate strain, size, and nutritional content on the growth and toxicity of mixotrophic Dinophysis acuminata. Harmful algae, 78:95-105

Strom, S., Harvey, E., Fredrickson, K., Menden-Deuer, S. (2013). Broad salinity tolerance as a refuge from predation in the harmful raphidophyte alga Heterosigma akashiwo (Raphidophyceae). Journal of phycology, 49: 20-31

Tong, M., Zhou, Q., Kulis, D., Jiang, T., Qi, Y., Anderson, D. (2010). Culture techniques and growth characteristics of Dinophysis acuminata and its prey. Chinese Journal of Oceanology and Limnology, 28, 6:1230-1239

Tong M., Smith, J., Kulis, D., Anderson, D. (2015). Role of dissolved nitrate and phosphate in isolates of Mesodinium rubrum and toxin-producing Dinophysis acuminata. Aquatic Microbial Ecology, 75: 169-185

Whereat, E. (2015). Dinophysis acuminata in Delaware’s inland bays and coastal waters 2001-2015. From: https://www.inlandbays.org/wp- content/uploads/Dinophysis-in-DE-2001-2015-Whereat-STAC-DEC-2015.pdf

Wolny, J., Egerton, T., Handy, S., Stutts, W., Smith, J., Whereat, E., Bachvaroff, T., Henrichs, D., Campbell, L., Deeds, J. (2020). Characterization of Dinophysis spp. (Dinophyceae, Dinphysiales) from the Mid-Atlantic region of the United States. Journal of phycology, DOI: 10.1111/jpy.12966:1-21

Yasumoto T., Murata, M., Oshima, Y., Sano, M., Matsumoto, G.K., Clardy, J. (1985). Diarrhetic shellfish toxins. Tetrahedron, 41, 6:1019-1025

Zhang, Y., Fu, F.-X., Whereat, E., Coyne, K.J., Hutchins, D.A., 2006. Bottom-up controls on a mixed-species HAB assemblage: a comparison of sympatric Chattonella subsalsa and Heterosigma akashiwo (Raphidophyceae) isolates from the Delaware Inland Bays, USA. Harmful Algae 5 (3), 310–320

Map of shellfish harvesting sites: http://www.dnrec.delaware.gov/swc/wa/Pages/GrowingWaters.aspx

APPENDIX

Heterosigma Heterosigma akashiwo & Mesodinium Prey offered to Dinophysis acuminata akashiwo Mesodinium rubrum rubrum

Mean (cells/mL) 1,544 1,424 1,535 0 Standard Deviation 0 0 0

Mean (cells/mL) 1,161 947 1,166 1 Standard Deviation 237 163 409

Mean (cells/mL) 1,003 996 1,565 3 Standard Deviation 85 184 251 Mean (cells/mL) 830 972 1,383

Day Since Fed Day Since 6 Standard Deviation 50 154 113 Mean (cells/mL) 881 1,324 2,328 8 Standard Deviation 230 337 178 Mean (cells/mL) 1,062 1,321 2,520 10 Standard Deviation 285 199 80 Growth rate of -0.0374 -0.0076 0.0495 Dinophysis acuminata

43 Table A2 Growth rates of Heterosigma akashiwo and Mesodinium rubrum in control treatment, used to offset ingestion rate of Dinophysis acuminata when offered these as prey.

Initial cell Final cell Mean Time Growth rate Trial abundance (X , abundance growth 0 (t, hours) (µ) cells/mL) (Xt, cells/mL) rate 1 41,644 222,000 24 0.0697 0.0623

2 41,644 172,000 24 0.0591 3 41,644 164,000 24 0.0571 4 41,644 171,667 24 0.0590 5 41,644 235,000 24 0.0721

Mesodinium rubrumMesodinium 6 41,644 162,000 24 0.0566 1 54,775 123,333 24 0.0338 0.0474 2 54,775 164,000 24 0.0457

3 54,775 178,000 24 0.0491 4 54,775 240,000 24 0.0616 akashiwo

Heterosigma 5 54,775 174,000 24 0.0482 6 54,775 166,000 24 0.0462

Growth rate calculation (Frost 1972, Heinbokel 1978, and Jeong & Latz 1994): (µ, day -1)

ln(%&'' )*+,-),%& . %&'' )*+,-),%& . ) µ = /0123 0104023 number of days

Example using data from Table A2:

C0 = 41,644, Ct = 222,000, t = 24

µ = (ln (222,000/41,644))/24 = 0.0679 day -1

44 Table A3 Ingestion rates of Dinophysis acuminata when offered Heterosigma akashiwo or Mesodinium rubrum as prey.

45 Grazing constant calculation (Frost 1972, Heinbokel 1978, and Jeong & Latz 1994): (g, h-1)

CD∗ 31 ( ) g =µ − CF 4

Where t is time, µ is growth rate, C0 is initial cell abundance of prey in the absence of predator, and Ct is the final cell abundance of prey in the presence of predator

Example using data from Tables A2 and A3:

-1 -1 µ = 0.0679, C0 = 41,644 cells mL , Ct = 132,000 cells mL , t = 24 hours g = 0.0679 * -((ln(132,000/41,644)/24) = 0.0142 h-1

Mean prey concentration calculation (Frost 1972, Heinbokel 1978, and Jeong & Latz 1994): (C̅ , cells)

D(IJK) C̅ = GF(H LM) 4(NLO)

Where C0 is the initial prey cell abundance in the absence of predator, µ is the growth rate of the prey in the absence of predator, g is the grazing constant, and t is time

Example using data from Tables A2 and A3:

-1 C0 = 41,644 cells mL , µ = 0.0679, t = 24 hours, g = 0.0142

C̅ = (41,644*(e24(0.062-0.014))-1))/ 24 * (0.062-0.014) = 78,323 cells mL-1

Clearance rate calculation (Frost 1972, Heinbokel 1978, and Jeong & Latz 1994): (F, volume grazer-1 h-1)

P = QR/ C0p

Where V is flask volume, g is grazing constant, and C0p is initial concentration of predators

Example using data from Tables A2 and A3:

-1 V= 2.29 mL , g = 0.0142, C0p = 4,773 cells mL

F = (2.29 x 0.0142)/4,773 = 6.82353E-06 volume grazer-1 h-1

Ingestion rate calculation (Frost 1972, Heinbokel 1978, and Jeong & Latz 1994): (I, cells d-1):

T = C̅ F x t

Where C̅ is mean prey cell abundance in the presence of predator, F is clearance rate, and t is time

Example using data from Tables A2 and A3:

C̅ = 78,323 cells mL -1, F = 6.82353E-06, t = 24 hours

I = 78,323 x 6.82353E-06 x 24 = 12.8 cells d-1

47 Table A4 Growth rates of Heterosigma akashiwo and Teleaulax amphioxeia in control treatment, used to offset ingestion rate of Dinophysis acuminata when offered these as prey.

48 Table A5 Ingestion rates of Mesodinium rubrum when offered Heterosigma akashiwo or Teleaulax amphioxeia as prey.

49 Table A6 Emission intensity of chlorophyll a in Dinophysis acuminata offered Heterosigma akashiwo, both Heterosigma akashiwo and Mesodinium rubrum, or Mesodinium rubrum as prey.

50 Table A7 Ratio of intensity of emission of chlorophyll a when excited at 488 nm and 633 nm in plastids of Heterosigma akashiwo, Teleaulax amphioxeia, and Dinophysis acuminata that was given Mesodinium rubrum as prey.

51 Table A8 Ratio of intensity of emission of chlorophyll a when excited at 488 nm and 561 nm in plastids of Dinophysis acuminata when offered Heterosigma akashiwo, Heterosigma akashiwo and Mesodinium rubrum, or Mesodinium rubrum as prey.

Heterosigma Dinophysis Heterosigma akashiwo & Mesodinium acuminata in the akashiwo Mesodinium rubrum presence of rubrum Mean Ratio 0.16 0.17 0.17 Standard Error 0.019 0.014 0.0094 p-value = 0.75