ABSTRACT
FLOOD, STACIE LYNN. Ecotoxicology of Estuarine Phytoplankton Growth and Toxicity
in Response to Atrazine Exposures. (Under the direction of committee chair JoAnn M.
Burkholder).
Estuarine phytoplankton play integral roles in ecosystem functioning by providing a
basis for energy fluxes (through their carbon fixation capacity) and nutrient cycling, and by
influencing community structure (through oxygen and food biosynthesis, and through both
competitive interactions and interactions with other trophic levels). Some phytoplankton
populations (strains) proliferate to such an extent that they can form nearly monospecific
blooms, with devastating impacts on ecosystems. Anthropogenic inputs of chemical
environmental contaminants such as toxic substances and nutrient pollution are frequently
associated with developing harmful algal blooms (HABs). A critical knowledge gap in
estuarine ecology concerns how phytoplankton assemblages respond to stressors in chronically
disturbed habitats, and why some populations respond by producing phycotoxins. The
available evidence suggests that herbicides such as atrazine can significantly affect
phytoplankton assemblage composition, but little is known about how estuarine phytoplankton
respond to multiple stressors in chronically disturbed habitats, and why some strains respond
by producing toxins that can kill fish and other aquatic life. The goals of this research were to
establish a robust protocol for testing the effects of a ubiquitous herbicide, atrazine, on
estuarine phytoplankton, and then to use that protocol to compare the effects of atrazine with
versus without nutrient enrichment on selected benign and harmful species. The benign species
used was the cosmopolitan estuarine/marine alga, Dunaliella tertiolecta (Chlorophyta),
whereas the HAB species included the ichthyotoxic haptophyte Prymnesium parvum
(Haptophyta) and the toxigenic raphidophyte Chattonella subsalsa (Heterokontophyta,
Raphidophyceae). The relative toxicity of atrazine to the three species in salinity 20 nutrientreplete media was D. tertiolecta > P. parvum > C. subsalsa (mean 96-hr IC50 = 139.3, 126.5,
and 111.3 µg atrazine L-1, respectively). Response estimates were influenced by the strain, the
geographic origin, the salinity, and the timing of the analysis endpoint. Nutrient limitation
alleviated the growth-inhibiting effects of this herbicide for all three algal taxa to varying
extent. Production of hemolytic substances in the HAB species increased under co-occurring
nutrient stress and herbicide exposure, and the greatest toxicity was observed under combined
imbalanced nitrogen/phosphorus levels and high atrazine exposures. These findings advance
knowledge about how nutrient regimes and herbicides interact to control harmful estuarine
phytoplankton population dynamics. The test platform that was developed here can be
extended for use with other herbicides and other estuarine phytoplankton species.
© Copyright 2017 by Stacie Lynn Flood
All Rights Reserved
Ecotoxicology of Estuarine Phytoplankton Growth and Toxicity in Response to Atrazine
Exposures
by
Stacie Lynn Flood
A dissertation submitted to the Graduate Faculty of
North Carolina State University
in partial fulfillment of the
requirements for the degree of
Doctor of Philosophy
Plant Biology
Raleigh, North Carolina
2017
APPROVED BY:
______________________________
Dr. JoAnn M. Burkholder
Committee Chair
______________________________
Dr. W. Gregory Cope
______________________________
Dr. Stacy A. C. Nelson
______________________________
Dr. Parke A. Rublee
______________________________
Dr. Thomas R. Wentworth
ii
DEDICATION
This work is dedicated to my mother, Cindy Flood, who always supported my efforts to grow
and encouraged my ambitions. This work is also dedicated to the memory of my
grandmothers Pat Walton, who always believed in me and taught me to live with dignity and
grace, and Mary Wilkinson, who kept me grounded, kept me close, and taught me to love
fish, and to my great-grandmother Thelma Seals, who became a Ph.D. chemist under much
more challenging circumstances than I will ever know. You are my heroes, and together you
brought me to this point.
This work is also dedicated to the memory of Charlie, who was by my side throughout it all,
making me laugh and keeping me focused on the beauty of the world around us. The wind has
taken you from my sight, but my heart will always see you.
iii
BIOGRAPHY
Stacie Flood was born and raised in Tulsa, Oklahoma where she attended Edison
High School and developed a long-abiding interest in keeping and breeding aquarium fishes.
She owned and operated a local aquarium supply shop in Tulsa and specialized in the
building and care of planted aquaria and in the husbandry of Symphysodon species (Discus
fish) before moving to the Greater St. Louis area to work at a tropical fish hatchery. While
living in Missouri, Stacie’s primary interests shifted from the captive care of tropical fish to
the ecology of North American native fishes, and she began to devote her attention to
environmental sciences and the factors influencing aquatic ecosystems in United States. After
earning an Associate’s Degree in General Sciences at John A. Logan College in Carbondale,
Illinois, she enrolled in an undergraduate zoology program at Southern Illinois University
where she was fortunate to work, in turns, on projects with or in the labs of Dr.s Matt Whiles,
Frank Wilhelm and Brooks Burr studying aquatic invertebrate ecology, limnology and
ichthyology, respectively. Following graduation, she applied to North Carolina State
University and was accepted into the Center of Applied Aquatic Ecology under the direction
of Dr. JoAnn Burkholder, who has guided her through an advanced level examination of the
ecology of surface and ground waters of North Carolina, the U.S., and throughout globe.
These discoveries and discussions were often challenging and always fascinating, and have
provided Stacie with a greater appreciation for the scale of economic, political, cultural and
civic influences on natural aquatic ecosystems - and never dulled the brilliance of the small
moments of wonder found glancing through snorkel or microscope at the beauty that lies
hidden under the surface of a stream.
iv
TABLE OF CONTENTS
LIST OF TABLES ................................................................................................................. vi
LIST OF FIGURES ............................................................................................................. viii
LIST OF ABBREVIATIONS .................................................................................................x
CHAPTER 1: Assessment of Atrazine Toxicity to the Benign Estuarine Phytoplankter,
Dunaliella tertiolecta (Chlorophyta) Under Varying Nutrient Conditions .........................1
1.1 INTRODUCTION .................................................................................................................... 1
1.2 MATERIALS AND METHODS ........................................................................................... 5
1.2.i Basic Platform Design ................................................................................................. 5
1.2.ii Culture Conditions ...................................................................................................... 7
1.2.iii Optical Density Versus Direct Cell Counts ............................................................ 8
1.2.iv Population Growth Rate Measurements .................................................................. 9
1.2.v Algal Bioassays ............................................................................................................ 9
1.2.vi Statistical Analysis .................................................................................................... 12
1.3 RESULTS ................................................................................................................................ 13
1.4 DISCUSSION ......................................................................................................................... 15
1.5 FIGURES AND TABLES .................................................................................................... 20
1.6 REFERENCES ........................................................................................................................ 35
CHAPTER 2: Persistence Under Sub-Optimal Conditions; Imbalanced Nutrient
Regimes Increase Prymnesium parvum Resilience to Herbicide Exposure.......................46
2.1 INTRODUCTION .................................................................................................................. 46
2.2 MATERIALS AND METHODS ......................................................................................... 49
2.2.i Experimental Organism, Strains, and Culture Conditions .................................... 49
2.2.i.a. Toxic Prymnesium parvum .......................................................................................... 49
2.2.i.b. Strains and Culturing .................................................................................................. 51
2.2.ii Optical Density Versus Direct Cell Counts ........................................................... 52
2.2.iii Population Growth Rate Measurements ............................................................... 53
2.2.iv Algal Bioassays ......................................................................................................... 53
2.2.v Hemolytic Activity Assay Protocols .......................................................................... 55
2.2.v.a Blood Collection and Erythrocyte Preparation ........................................................... 56
v
2.2.v.b Preparation of Algal Extracts ...................................................................................... 56
2.2.v.c Hemolytic Assays ......................................................................................................... 57
2.2.vi Statistical Analysis .................................................................................................... 58
2.3 RESULTS ................................................................................................................................ 59
2.3.i Atrazine x Nutrient Bioassays .................................................................................... 59
2.3.ii Erythrocyte Lysis Assays ........................................................................................... 62
2.4 DISCUSSION ......................................................................................................................... 64
2.5 FIGURES AND TABLES .................................................................................................... 70
2.6 REFERENCES ........................................................................................................................ 94
Chapter 3: Examination of Chattonella subsalsa (Raphidophyceae) Growth and
Hemolytic Activity Responses to Co-Occurring Environmental Stressors ....................106
3.1 INTRODUCTION ................................................................................................................ 106
3.2 MATERIALS AND METHODS ....................................................................................... 108
3.2.i Experimental Organism ............................................................................................ 108
3.2.ii Culture Conditions ................................................................................................... 109
3.2.iii Optical Density Versus Direct Cell Counts ........................................................ 110
3.2.iv Population Growth Rate ........................................................................................ 111
3.2.v Bioassays ................................................................................................................... 112
3.2.vi Protocols for Assays of Hemolytic Activity .......................................................... 115
3.2.vi.a Blood Collection and Erythrocyte Preparation ............................................................. 115
3.2.vi.b Preparation of Algal Extracts ................................................................................... 116
3.2.vi.c Hemolytic Assays ...................................................................................................... 117
3.2.vii Statistical Analysis ................................................................................................ 118
3.3 RESULTS .............................................................................................................................. 119
3.3.i Growth ......................................................................................................................... 119
3.3.ii Atrazine Effects ......................................................................................................... 120
3.3.iii Toxicity as Hemolytic Activity ............................................................................... 121
3.4 DISCUSSION ....................................................................................................................... 123
3.5 FIGURES AND TABLES .................................................................................................. 127
3.6 REFERENCES ...................................................................................................................... 144
APPENDIX ...........................................................................................................................156
vi
LIST OF TABLES
Assessment of Atrazine Toxicity to the Benign Estuarine Phytoplankter, Dunaliella
tertiolecta (Chlorophyta) Under Varying Nutrient Conditions
Table 1.1
Table 1.2
Table 1.3
Table 1.4
Table 1.5
Table 1.6
Table 1.7
Table 1.8
Table 1.9
List of species used for aquatic toxicity testing in the ECOTOX database,
ranked by the number of different studies on all chemicals .................... 20
List of algal groups and species used for aquatic toxicity testing in the
ECOTOX database, ranked by number of different studies on atrazine . 21
Marine algal species recommended for use in guideline testing protocols
by international organizations ................................................................. 22
Media recipes used in D. tertiolecta culture and testing ......................... 23
The 96-hr point estimates based on effects on biomass estimated by
manual cell counts ................................................................................... 25
The 96-hr point estimates based on effects on biomass estimated by OD
................................................................................................................. 26
The atrazine effect estimates over time based on D. tertiolecta growth
rates, as estimated by OD. ....................................................................... 29
Influence of treatment variables on D. tertiolecta growth at 96-hr for
nutrient x atrazine bioassays ................................................................... 33
Influence of treatment variables on D. tertiolecta growth at day 20 ....... 34
Persistence Under Sub-Optimal Conditions; Imbalanced Nutrient Regimes Increase
Prymnesium parvum (Haptophyta) Resilience to Herbicide Exposure
Table 2.1
Table 2.2
Table 2.3
Table 2.4
Table 2.5
Table 2.6
Table 2.7
Table 2.8
Collection locations in the U.S. for the tested Prymnesium parvum
culture isolates ......................................................................................... 71
Media recipes used in Prymnesium testing and cultures ......................... 72
Erythrocyte lysis assay buffer preparation .............................................. 72
Blood storage medium preparation ......................................................... 73
Changes in atrazine effects in nutrient-replete media over time (µg/L)
and maximum growth rate kmax for all strains tested. .............................. 75
Median half-maximal activity estimates (units, µg L-1) over time for all
strains in nutrient-replete media. ............................................................. 76
Heat map of kmax values for the growth response of P. parvum strains
tested in different treatments over time ................................................... 77
Results of 4-way ANOVA testing growth rates of P. parvum (all three
strains) at 96-hr or 10-day endpoints as dependent variables. Fixed
effects are atrazine concentration (Atra), nutrient regime (Nutr),
geographic origin (strain) and salinity .................................................... 83
vii
Table 2.9
Table 2.10
Table 2.11
Table 2.12
Results of a 3-way ANOVAs for each of the three P. parvum strains (TX,
NC, SC) testing growth rates at 96-hr or 10-day endpoints as dependent
variables. Fixed effects are atrazine concentration (Atra), nutrient regime
(Nutr), and salinity .................................................................................. 84
Results from a 3-way ANOVA testing normalized hemolytic activity as
the dependent variable and atrazine concentration (Atra), nutrient regime
(Nutr), and geographic origin (Strain) as the fixed effects...................... 86
Results from a 4-way ANOVA testing hemolytic effects at all salinities
(NC and TX strains only) against atrazine concentration (Atra), nutrient
regime (Nutr), geographic origin (Strain), and salinity ........................... 87
Results from a 3-way ANOVA testing normalized hemolytic activity in
individual strains as the dependent variable and atrazine concentration
(Atra), nutrient regime (Nutr), and salinity as the fixed effects .............. 88
Examination of Chattonella subsalsa (Raphidophyceae) Growth and Hemolytic
Activity Responses to Co-Occurring Environmental Stressors
Table 3.1
Table 3.2
Table 3.3
Table 3.4
Table 3.5
Culture media recipes used for culturing and assays of Chattonella
subsalsa…………. ................................................................................ 127
Growth inhibition and lowest-observed effect concentrations over time in
10-day bioassays with Chattonella subsalsa ......................................... 128
Growth inhibition and lowest-observed effect concentrations over time in
28-day replete media bioassays with Chattonella subsalsa .................. 129
Results of two-way ANOVAs examining atrazine concentration and
nutrient regime effects on growth rates of Chattonella subsalsa in
bioassays at different endpoints. ........................................................... 130
Results of two-way ANOVAs examining atrazine concentration and
nutrient regime effects on hemolytic activity of Chattonella subsalsa in
bioassays at the 10-day and 28-day endpoints. ..................................... 131
viii
LIST OF FIGURES
Assessment of Atrazine Toxicity to the Benign Estuarine Phytoplankter, Dunaliella
tertiolecta (Chlorophyta) Under Varying Nutrient Conditions
Figure 1.1
Figure 1.2
Figure 1.3
Figure 1.4
Figure 1.5
Figure 1.6
Figure 1.7
Light micrograph of Dunaliella tertiolecta, the benign test species used
to design the testing protocol ...................................................................20
Correlation statistics between optical density and manual cell counts
describing procedures for selection of most appropriate wavelength for
use in monitoring test activity. .................................................................24
Results of 96 hr atrazine exposure for Dunaliella tertiolecta .................27
Distribution of biomass-derived effect estimates determined by manual
cell counts and OD ...................................................................................28
Test control performance over time for the 20-day D. tertiolecta atrazine
bioassays ..................................................................................................30
Growth response of Dunaliella tertiolecta to atrazine exposure under
different nutrient regimes.........................................................................31
Atrazine-induced changes in Dunaliella tertiolecta growth under
different nutrient regimes over time ........................................................32
Persistence Under Sub-Optimal Conditions; Imbalanced Nutrient Regimes Increase
Prymnesium parvum (Haptophyta) Resilience to Herbicide Exposure
Figure 2.1
Figure 2.2
Figure 2.3
Figure 2.4
Figure 2.5
Figure 2.6
Figure 2.7
Figure 2.8
Figure 2.9
Light micrograph Prymnesium parvum (Texas strain) ............................70
Example of the relationship between optical density (OD) and
Prymnesium parvum population density (PD) .........................................74
Cell densities of control cultures for the Texas strain of P. parvum at
salinity 20 .................................................................................................78
Growth rates of the NC strain of P. parvum at different salinities in all
treatments at day 10 .................................................................................79
Atrazine inhibition curves in salinity 10 for the South Carolina strain of
P. parvum over time.................................................................................80
Atrazine inhibition curves for all P. parvum assays over time, by nutrient
regime ......................................................................................................81
Atrazine inhibition curves for all P. parvum assays over time, by salinity
level ..........................................................................................................82
Differences in relative production of hemolytic compounds by P. parvum
strains (TX, NC) in control treatments at two salinity levels (10, 20) .....85
Relative production of hemolytic activity plotted against atrazine
concentration for all tested strains, showing line-of-fit for non-linear
relationships and corresponding R2, and confidence intervals as colored
zones corresponding to different nutrient treatments (half maximal
ix
Figure 2.10
Figure 2.11
Figure 2.12
Figure 2.13
growth effect concentrations for atrazine are provided as insets with each
assay display) ...........................................................................................89
Log-normal relationships between hemolytic activity and atrazine
concentration, in salinity 20 f/2 media.....................................................90
Log-normal relationships between hemolytic activity and atrazine
concentration, in salinity 10 f/2 media.....................................................91
Linear relationships between growth rates and hemolytic activity in P.
parvum strains tested in salinity 10 .........................................................92
Linear relationships between growth rates and hemolytic activity in P.
parvum strains tested in salinity 20 .........................................................93
Examination of Chattonella subsalsa (Raphidophyceae) Growth and Hemolytic
Activity Responses to Co-Occurring Environmental Stressors
Figure 3.1
Figure 3.2
Figure 3.3
Figure 3.4
Figure 3.5
Figure 3.6
Figure 3.7
Figure 3.8
Figure 3.9
Figure 3.10
Figure 3.11
Figure 3.12
Chattonella subsalsa cells fixed with 1% PFA + 1% glut or fixed with
acidic Lugol’s solution ………….......................................................... 132
Example of the modeled relationships between optical density (OD) and
C. subsalsa cell density that was used to monitor population growth for
atrazine inhibition assays ....................................................................... 133
Growth of C. subsalsa in control cultures during 10-day and 28-day
bioassays. ............................................................................................... 134
Growth curves for Chattonella subsalsa in 10-day and 28-day bioassays
with atrazine under nutrient-replete conditions ..................................... 135
Variation in atrazine half-maximal effective concentrations for
Chattonella subsalsa in 28-day nutrient-replete bioassays over time ... 136
Dose-response relationships showing increasing potency of atrazine to C.
subsalsa in 10-day, nutrient-replete growth bioassays …………. ........ 137
Inhibition concentration curves for different nutrient regimes tested in 10day C. subsalsa atrazine exposure assay ............................................... 138
Inhibition concentration curves for different nutrient regimes tested in 28day atrazine exposure assays, and measured at 10 and 28 days ............ 139
Normalized hemolytic activity in control cultures, measured at end of 10and 28-day assays .................................................................................. 140
Growth rates (divisions day-1) and normalized hemolytic activity in all
tested atrazine concentrations, nutrient treatments and controls for C.
subsalsa.................................................................................................. 141
Log-normal relationships between hemolytic activity (normalized to cell
density and sample volume) and atrazine concentration or growth rates at
10-day endpoint ..................................................................................... 142
Hemolytic activity at different test endpoints, normalized to cell density
and sample volume, in various nutrient regimes under a range of atrazine
concentrations ........................................................................................ 143
x
LIST OF ABBREVIATIONS
APHA: American Public Health Association
ASTM: American Society for Testing and Materials
CAAE: Center for Applied Aquatic Ecology
CASRN: Chemical Abstracts Service (CAS) Registration Number
ECx: Effect Concentration (x% of treatment controls)
ELA: Erythrocyte Lysis Assay
HA: Hemolytic Activity
ICx: Inhibition Concentration (x% of treatment controls)
ISO: International Organization for Standardization
IU: International Units
LOEC: Lowest Observed Effect Concentration
NOEC: No Observed Effect Concentration
OD: Optical Density
OECD: Organization for Economic Cooperation and Development
ROS: Reactive Oxygen Species
RPMI: Roswell Park Memorial Institute medium
U.S. EPA: United States Environmental Protection Agency
1
CHAPTER 1: Assessment of Atrazine Toxicity to the Benign Estuarine Phytoplankter,
Dunaliella tertiolecta (Chlorophyta) Under Varying Nutrient Conditions
1.1 INTRODUCTION
Located at the land-water interface, estuaries are extremely vulnerable to land-based
pollution (Howarth 2008; Crossett et al. 2014). They have been evaluated as the most
anthropogenically degraded habitat types on Earth (Lytle and Lytle 2001; Scott et al. 2006),
and are especially threatened by nutrient pollution (cultural eutrophication - Bricker et al.
2008) and other chemical environmental contaminants (CECs) such as herbicides
(DeLorenzo et al. 2001, McCarthy et al. 2007). Agricultural nonpoint source pollution is a
leading source of water quality impairment and a major contributor to contamination of
estuaries (Bricker et al. 2008) and according to the United States Environmental Protection
Agency (U.S. EPA 2000, 2002), the leading causes of coastal impairment include pesticides
and nutrients. Most pesticides entering estuaries are herbicides and, among these, atrazine (2chloro-4-ethylamino-6-isopropylamino-s-triazine) is one of the most widely used agricultural
herbicides in the U.S. (U.S. EPA 2003). This triazine herbicide is used to control annual
broadleaf and grassy weeds in crops such as corn, sorghum, and sugarcane, and on golf
courses (Solomon et al. 1996). Atrazine is relatively mobile and persistent (Koc 40-394, Kd
0.20–12.6; Pereira and Rostad 1990; Solomon et al. 1996; Giddings 2005; Jablonowski et al.
2011) and is the most commonly detected pesticide in U.S. surface waters (Gilliom et al.
2006). More than 20.4 million kilograms (kg) (45 million pounds) of atrazine were applied in
the U.S. during 2014 (U.S. Department of Agriculture – National Agricultural Statistics
Service 2017). The Albemarle-Pamlico Estuarine System was evaluated by the National
Oceanic and Atmospheric Administration as historically having the highest amount of
pesticide use of any estuarine watershed in the U.S. (Pate et al. 1992), but pesticide data are
collected at relatively few sites, quarterly at multi-year intervals or less frequently (U.S. EPA
STORage and RETrieval (STORET) database).
2
Atrazine can remain in some soil types for as long as 4-22 years (U.S. EPA 2007a;
Jablonowski et al. 2008, 2009) and has a long aqueous half-life (335 days in light - Rohr and
McCoy 2010; 742 days in darkness - Ryberg et al. 2010). Biodegradation of atrazine is
minimal due to the presence of the s-triazine ring (Cunningham et al. 1984; Kemp et al.
1985; Howard 1991) leading to potential for prolonged exposures to aquatic biota (Glotfelty
et al. 1984; Pinckney et al. 2002). Environmental concentrations as high as 691 µg/L have
been reported in Midwestern streams (Lockert et al. 2006). Atrazine is also a common
estuarine contaminant (Readman et al. 1993; Alegria et al. 2000; Pennington et al. 2001;
Southwick et al. 2003; McCarthy et al. 2007). During significant runoff periods, 2-3% of the
atrazine applied to crops can be transported from the point of application to estuaries (e.g.
Glotfelty et al. 1984). Atrazine is frequently detected in the Chesapeake Bay region, where
stormwater runoff has contained concentrations as high as 480 µg/L (Forney and Davis 1981;
Eisler 1989; Lehotay et al. 1998). Like most pesticides, it is infrequently measured in many
aquatic systems, so the databases for many waterbodies have concerning gaps in available
information (Starr et al. 2017 and references therein).
Atrazine is a photosynthetic inhibitor that targets the QB plastoquinone-binding niche
in the D1 protein of the PSII apparatus (Forney and Davis 1981; Fuerst and Norman 1991)
and interrupts photosynthetic electron transport via competitive binding, resulting in
oxidative stress, photooxidation of chlorophyll and, ultimately, cell necrosis (Moreland and
Hill 1962). Exposure to phototrophs can also inhibit hormones, cell division, pigment
synthesis, lipid synthesis, and cell metabolism (Radosevich et al. 2007). In aquatic
ecosystems, atrazine has depressed primary productivity of phytoplankton assemblages
(Solomon et al. 1996; Graymore et al. 2001). Toxicants affecting phytoplankton populations
can potentially affect higher trophic levels and food web structure (Walsh and Merrill 1984;
Hylland and Vethaak 2011). Adverse effects of atrazine on the photosynthesis of
phytoplankton and other primary producers in aquatic systems have occurred at
concentrations as low as 1-10 µg L-1 (Solomon et al. 1996, Giddings 2005), which commonly
occur in surface waters (Kemp et al. 1985; Lakshminarayana et al. 1992).
3
Although phytoplankton are the foundation of pelagic food webs, toxicity data for
estuarine phytoplankton and herbicides are sparse, especially in estuarine and marine waters
(Starr et al. 2017). In addition, there has been no systematic investigation about how nutrient
availability affects atrazine toxicity in estuarine phytoplankton. Such assessments require
development of novel approaches. Historically in the regulatory decision-making process,
much of the information on aquatic toxicity of chemicals was derived from tests on daphnids
and fish (Mohan and Hosetti 1999). Algae repeatedly have been found to be more sensitive to
many toxic substances, including many herbicides, than animals (Walsh et al. 1980; Blanck
et al. 1984; Hughes et al. 1988; Lewis 1990b, 1995). Yet, there is still a paucity of
toxicological data available for microalgae compared to other groups (Eisentraeger et al.
2003; e.g., Table 1.1). Toxicity determinations for chemical hazard assessments under the
Toxic Substances Control Act have begun to include short-duration (< 14 days) inhibition
tests with microalgae (algal toxicity bottle tests – e.g., Stephan et al. 1985; U.S. EPA 1985).
Only a handful of species, typically “hardy” species from freshwaters, are commonly used in
direct toxicity tests of algae (Table 1.2) (Lewis 1990a; Fairchild et al. 1998). The freshwater
bias applies to indirect effects as well, which are considered as important or more important
than direct effects in natural environments. More than 80% of studies about indirect effects
of contaminants in aquatic systems have been conducted in freshwaters (Fleeger et al. 2003).
Clearly, estuarine and marine microalgae have been too infrequently considered in
developing protective criteria for chemical contaminants (Blanck et al. 1984; Lewis 1995;
Lytle and Lytle 2001; Eklund and Kautsky 2003). The data generated from freshwater
microalgae are inadequate to protect estuarine phytoplankton assemblages from chemical
hazards because the freshwater microalgae were largely selected for being “good laboratory
organisms,” (Nyholm and Källqvist 1989), meaning easy to culture, rather than for sensitivity
or ecological relevance. The “hardy” freshwater species are generally more tolerant of
chemical hazards than sensitive species; they are not good indicators of freshwater
phytoplankton assemblage response, and they are not good surrogates for estuarine or marine
microalgae (Lewis 1995; Graymore et al. 2001).
4
Microalgal taxa are known to vary substantially, often by several orders of
magnitude, in response to the same toxicant (Blanck et al. 1984; Lewis 1995; Rojíčková and
Maršálek 1999; Faust et al. 2003; Janssen and Heijerick 2003). Moreover, bioassay toxicity
estimates are chemical-, test species-, and endpoint-dependent, and also extremely sensitive
to variations in physical conditions such as media, lighting, and culture volume (Blanck et al.
1984; Kasai et al. 1993; Lewis 1995; Stauber 1995; Blaise et al. 1997; Choi et al. 2012).
There typically is high intra-species variation (strain differences) as well; remarkably,
different populations of the same test species had EC50 (concentration causing an effect on
50% of the tested population) estimates that differed among laboratories by many orders of
magnitude, even for simple compounds (Nyholm 1985; Kasai et al. 1993; Behra et al. 1999).
Several researchers have recommended an alternate approach to develop protective criteria –
that is, the use of standardized methods on a number of species in a “test battery”, that is,
tests of several algal species from a broad taxonomic range, selected considering the
community structure of ecosystems potentially affected by the toxicant (Blanck et al. 1984;
Peterson et al. 1994; Rojíčková and Maršálek 1999). Guideline testing protocols have been
developed to examine marine microalgae (Table 1.3), and these data are gradually being
incorporated chemical safety assessments. Where “non-standard” test species have been used
in toxicity testing, the approach has generally been to screen a small number of “data-rich”
compounds against the responses of multiple test species; for example, Swanson et al. (1991)
reported phytotoxicity of various pesticides to 68 algal species. This approach enables
assessment of common test species to a given toxicant for quality assurance purposes, while
also providing novel information about the scope of potential toxicant activity in non-model
organisms – but taxonomic resolution generally has been poor, and restricted to a small
number of freshwater groups. A survey of the ECOTOX database (Table 1.2) revealed that
more than a third (36%) of the studies on atrazine listed under the grouping “Algae, Moss,
Fungi” were of green algae, 18% were of diatoms, 13% were of cyanobacteria (blue-green
algae), and 7% were of other groups - and only ~25% of the 719 studies on atrazine studies
listed under the group “Algae” provided genus and species information.
5
Here, ecologically relevant endpoints were used to assess the responses of selected
estuarine phytoplankton to atrazine exposures, using methods designed (i) to allow
comparisons with toxicity estimates from the published literature, and (ii) to enable screening
of multiple species in future studies. A robust testing platform was developed to examine
ecologically relevant responses to atrazine by a robust model organism. Use of a robust,
benign model organism enabled comparison of the response data with assessments of relative
test performance observed in the same species by other researchers, which provided
confidence in the experimental regime. This platform can then be used to screen multiple
species under similar exposure conditions. By reporting standardized endpoints, the toxicity
estimates developed here can be compared to values previously reported in the literature or
publicly available databases. The testing platform designed here will allow for herbicide
toxicity testing of as-yet-untested, ecosystem-disrupting harmful algal species (Sunda et al.
2006) to ensure that recommended exposure levels to atrazine adequately protect ecosystem
function by including harmful algae in assessments. The experiments also enabled an
examination of how the estuarine microalgal response to atrazine is influenced by unbalanced
nutrient regimes, which now characterize most U.S. estuaries (Bricker 2008; Glibert et al.
2011; Glibert and Burkholder 2017).
1.2 MATERIALS AND METHODS
1.2.i Basic Platform Design
A hardy model organism with an established history of use in toxicity testing,
Dunaliella tertiolecta Butcher (Chlorophyta) (DeLorenzo 2009) (Fig. 1.1), was selected to
assess effects of atrazine, a compound with an extensive toxicity reference library, so that the
testing platform can be easily applied to tests of other estuarine phytoplankton. This species
has a cosmopolitan distribution in estuaries and marine coastal waters (McLachlan 1960). It
is relatively easy to maintain in laboratory cultures (Byrd et al. 2016) and is a standard test
organism (U.S. EPA 1974; ASTM 1990; APHA 1995; DeLorenzo 2009), allowing for interlaboratory comparisons of results.
6
The responses of D. tertiolecta to atrazine were quantified using modifications to
standard 96-hr algal bioassays (ASTM 1994), including benchmarks to enable comparison of
sensitivities with findings from other researchers. Control performance criteria were also
selected from available guideline protocol recommendations (U.S. EPA 1971; Walsh 1988;
ASTM 1996; Blaise et al. 1997; Geis et al. 2000) to establish quality checks on test
performance. These minimum acceptance criteria included control variability less than 20%;
control biomass increased by more than 20% by test end; and growth in solvent controls did
not vary from media-only cultures by more than 20%. Other environmental conditions were
monitored to confirm that assay performance was compatible with standard test methods: pH
did not vary by more than 1.5 units (ISO 1995); consistent lighting was maintained at ~100
µE of photosynthetically active radiation m-2 s-1 (106.3 + 26.2 µE m-2 s-1; cool white
fluorescent tubes); and temperature was maintained at 24 ± 2°C (ambient air temperature). In
addition, the vessel volume, cell inoculation density, and growth phase were stipulated. It
was important to define these environmental factors because within-species variation in
tolerance versus sensitivity to toxicants has been related to pH (Peterson et al. 1984; Fahl et
al. 1995), temperature (Chalifour and Juneau 2011), light regime (Guasch et al. 1997;
Guasch and Sabater 1998; Sjollema et al. 2014), growth phase (Stauber 1995), cell density
(Tang et al. 1998), and container volume (Stratton and Giles 1990).
There has been debate about which endpoint parameter – growth rate or biomass – is
most appropriate for use in deriving effect estimates to describe toxicity (Nyholm 1985,
1990). Therefore, both biomass (ICbx) and growth rate (ICrx) (concentrations which caused
inhibition relative to control cultures) values were presented based on 96-hr IC10 and IC50,
along with the no-observed effect (NOEC) and lowest-observed effect (LOEC)
concentrations. Biomass-based effect estimates can show a higher degree of scatter, and are
more influenced by slight variations in the specific test system used, so growth rate has
previously been shown to be more robust (Nyholm 1985; Janssen and Heijerick 2003).
Therefore, the bulk of this analysis focused on effect estimates derived from the population
growth rate parameter. The number of assays, time intervals, and samples needed for these
7
and other (Chapters 2, 3) experiments could not have been completed using time-intensive
cell counts via light microscopy, and required a more rapid way to quantify D. tertiolecta
abundance. Optical density (OD) measurements previously had been shown to have a high
correlation with cell numbers (Kasai et al. 1993; Pennington and Scott 2001), is far less
labor-intensive, and has better precision than manual cell counting (Geis et al. 2000). Thus,
confirming the strength of the relationship between cell counts and OD (r2 > 0.98; Fig. 1.2),
we used OD techniques modified from Sorokin (1973) to quantify cell densities for
population growth rates.
To address the ecologically relevant condition of altered nutrient regimes and
herbicide co-exposures, a modified version of the standard 96-hr bioassay platform was
assessed over an extended duration, which was required (i) to allow for complete nutritional
deficits to develop when using test inoculations from the same parent culture, and (ii) to
ensure that test rigor was maintained when the methodology was applied to more slowly
growing harmful algal species. Evaluations of complete nutritional defects are more
appropriately examined following more than one week of growth in culture (Nyholm and
Lyngby 1988). Therefore, the assay was amended to assess 10-d or 20-d endpoints, which
has been reported to be an appropriate time frame for examining atrazine impacts on
phytoplankton (Tang et al. 1997). The atrazine concentration was analytically confirmed at
both the beginning and end of tests, and did not decrease over the extended duration (data not
shown).
1.2.ii Culture Conditions
A non-axenic strain of Dunaliella tertiolecta from the University of Texas Culture
Collection (UTEX LB 999; Fig. 1.1) was grown in batch cultures within 250-mL Erlenmeyer
flasks containing 100 mL of salinity 20-modified L1–Si medium (Guillard and Hargraves
1993). Light was measured using a 4π Biospherical QSL 101 quantum lab sensor (BSI, San
Diego, California, USA) and a 16-hr : 8-hr light:dark cycle. Cultures were mixed by gentle
hand agitation once per day and sterile-transferred using aseptic techniques under a laminar
8
flow hood every 7 to 10 days as needed to maintain log phase growth. All media (culture and
test) were prepared by adjusting the salinity of ultrapure Milli-Q water (18 MΩ cm−1 at 25
°C) with Instant Ocean® artificial sea salts (Aquarium Systems, Blacksburg, VA, USA) to
salinity 20. Salinity was measured using a YSI model 3200 conductivity/salinity bridge and
cell (YSI, Yellow Springs, Ohio, USA). The pH was maintained at 8.1 + 0.2 using HCl or
NaOH, and was measured using an Orion Versa Star Pro multi-parameter meter equipped
with an Orion ROSS Sure-Flow pH electrode (Thermo Scientific, Waltham, Massachusetts,
USA). Trace mineral solution was added before autoclaving. Vitamin solution was sterilefiltered (Whatman® Puradisc cellulose acetate syringe filters, nominal pore size 0.2 μm,
Sigma-Aldrich, St. Louis, Missouri, USA), added to the autoclaved media aseptically, and
the final media was vacuum-filtered (Corning® cellulose acetate filters, 0.22-µm nominal
pore size, Sigma-Aldrich) before storage at 4°C for no more than 30 days. All equipment was
initially cleaned by scrubbing in hot, soapy tap water, and then cleaned by acid stripping in
10% HCl (v/v). All equipment was sterilized by autoclaving, and all glassware and nonplastic equipment used in culturing and experiments was rinsed with pesticide-grade acetone,
before use.
1.2.iii Optical Density Versus Direct Cell Counts
Samples (1 mL) were randomly drawn from each parent culture and thoroughly
vortexed before absorbance was read on a Thermo Scientific Spectronic GENESYS 2
spectrophotometer (Thermo Scientific) using a freshly wiped Hellma® quartz cuvette
(Sigma-Aldrich) with a 10-mm path length. All readings were performed in duplicate and
adjusted using a media blank as a measurement baseline. All parent cultures were scanned
individually and the three highest peaks were selected as the most appropriate wavelengths
for analyzing a series of 8-point dilution curves, using progressive dilutions of parent stock
culture in growth media. The curve that produced the ΔOD/ΔPD [change in optical density
over change in population density (Sorokin 1973)] closest to unity (1) was used to determine
the growth rates from each stock culture. Cell counts to determine ΔPD were obtained by
collecting random 1-mL samples from the parent culture. Counts were made in duplicate
9
using a Reichert Bright-Line Phase Contrast Hemocytometer with improved Neubauer
rulings (Hausser Scientific No. 1475, Horsham, Pennsylvania, USA) (Guillard 1973). Two
homogenized 100-µL samples were assessed by counting 18 grids or a minimum of 400 cells
(DeLorenzo and Serrano 2003, Weiner et al. 2004). Cells were preserved in 1% (v/v) acidic
Lugol’s solution (Vollenweider 1974) to preserve cells for ~1 week while preparing dilution
curves immediately before each experiment. The wavelength that produced the highest
correlation between absorbance and cell numbers of the stock culture was used to monitor
population density and calculate growth rates for the experimental duration. All modeled OD
: cell number relationships used had a minimum observed 𝑟 2 value > 0.98 (Fig. 1.2).
1.2.iv Population Growth Rate Measurements
Net cell-specific growth rates (k, day-1) were modeled using OD according to
Sorokin (1973). Samples were collected daily (approximately every 24-hours postinoculation) from hand-agitated test tubes and measured in a clean, wiped quartz cuvette as
described above in duplicate, using a media blank as a baseline measurement. OD was
converted to PD, data were log-transformed (log10 +1) and the linear portion of the growth
curve (slope = 𝑎𝑘 ) was used to calculate growth rate:
𝑘 = 𝑎𝑘 ∗ 3.322
1.2.v Algal Bioassays
Toxicity experiments were based on examining the effect of atrazine, alone and in
varying nutrient regimes, on algal population growth rate. Testing procedures were
derivations on the standard 96-hr static algal toxicity bioassay protocols (ASTM 1996) and
were conducted to assess 96-hr, 10-day, and 20-day endpoints to allow for evaluation of
toxicant responses following nutrient depletion from cellular reserves. In preliminary work,
cultures were most sensitive 24-48 hr after atrazine addition and generally decreased in
sensitivity over time. Thus, multiple time points were required to examine toxicant effects.
Tolerance/ sensitivity to toxicants has been shown to be related to previous exposure history
and genetic variation (deNoyelles et al. 1982; Hamala and Kollig 1985; Behra et al. 1999).
10
Therefore, samples from the same parent culture were used to inoculate test cultures in test
controls and treatments. This approach did not allow for immediate observation of depletion
of cellular nutrient reserves in test cultures since parent cultures were nutrient-replete.
Therefore, the test bioassays were extended past 96 hr to assess any responses to nutrient
depletion x atrazine which occurred after additional time. The 10- and 20-day periods were
selected based on preliminary observations of D. tertiolecta growth curves in replete media.
Factors that can affect toxicity over extended test periods include degradation/detoxification
of the substance, pH changes, adsorption of the toxicant to organic matter, and volatilization.
Atrazine has low potential for volatilization and nominal test concentrations were quantified
by enzyme-linked immunosorbent assays (ELISAs, below) at the end of tests to ensure that
media exposure levels had not changed. The data indicated that atrazine levels did not change
over the extended trial periods by more than 20% of nominal concentrations (See appendix).
Analytical grade standards of atrazine (CASRN 1912-12-9, >98.0% purity) were
obtained from Chemservice (Westchester, PA, USA). Stock solutions were prepared in 100%
pesticide-grade acetone (Fisher Scientific, Fair Lawn, NJ, USA), with testing doses
administered to obtain a final concentration of 0.1% (v/v) acetone in each replicate, except
for a media-only control series which was included for all treatments in all assays. Each
control and treatment was replicated (n = 3). Assays were conducted in 50-mL round-bottom
borosilicate glass tubes with Teflon-lined caps (final total volume, 25 mL). Five-point
concentration curves included concentrations at 12.5, 25, 50, 100 and 200 µg atrazine L-1 as
well as the solvent and media controls. Exposure levels were selected based on range-finding
test estimates that were designed to expose the test organism to concentrations with lowest
and highest potential effects on population growth. When stock culture cells were in logphase growth (assessed by direct cell counts), the test culture tubes plus control cultures were
inoculated under aseptic conditions to achieve the same nominal cell densities of
approximately 2.5 x 104 cells mL-1 in all experimental cultures and controls. Chemical stocks
were diluted in test media to achieve desired final concentrations and were added to fresh,
sterile test culture media aseptically under a laminar flow hood. Time-series testing began
11
when algae were added to test media, and OD was measured immediately following
inoculation to assess initial density (0 hr). During the test, each culture tube was measured
twice at equally spaced daily intervals (0, 24, 48, 72, 96-hr) following inoculation during 96hr bioassays, and again on days 7, 10, 15 and 20 for the 20-day bioassays by inverting the
capped culture tube 2-3 times and aseptically removing 1 mL of randomly sampled test
culture. All culture tubes were hand-agitated by inverting the capped tube 3-4 times per day,
and were randomly repositioned (using computer-generated randomization of culture rack
positioning) post-sampling to avoid persistent effects due to differential light exposures. For
each sample, growth was determined using spectrophotometric determination of optical
density and converted to cell density using equations derived from direct cell count : optical
density relationships established using the parent culture within 1 week prior to the test
initiation, as described above.
The percent inhibition (%I) of algal growth at each atrazine concentration was
calculated by comparing mean growth rates for each treatment to the respective solvent
controls for the treatment:
%I = [1 − (𝑥̅
𝑘
𝑘𝑠𝑜𝑙𝑣𝑒𝑛𝑡 𝑐𝑜𝑛𝑡𝑟𝑜𝑙
)] ∙ 100
where 𝑘 is the growth rate of the pertinent sample, calculated as described in Section 1.2.iv.
Nominal concentrations of atrazine were quantified with enzyme-linked immunosorbent
assays (ELISAs) at the beginning and end of each bioassay with enzyme-labeled
paramagnetic particles of atrazine (method detection limit for atrazine = 0.046 μg/L) (SDI,
Delaware, MD, USA). The ELISA testing quality assurance and controls were as follows:
%CV ≤10%, calibration curve correlation r > .990, reproducibility (spiked recovery of
known material) and accuracy (% recovery of nominal concentrations) [ (Cobserved /
Cexpected)*100 ] between 80-120% (see appendix for data). This method has been shown to
have accuracy comparable to quantification of atrazine concentrations in surface water
samples by gas chromatography (Gascón et al. 1997).
12
Atrazine x nutrient effects were assessed by repeating the above assay procedures in
a simultaneously run three-way bioassay format. Each assay included 63 50-mL flasks in
identical setup, inoculation and chemical preparation approaches as described above, except
that they varied in culture media nutrient additions. To assess nutrient imbalanced effects,
three N:P supply ratios (selected based on deviations from the Redfield Ratio - Redfield
1958) were prepared: 1:1, N-limited (“Low N” = 2.0µM NO3, 2.0µM PO4); 160:1, P-limited
(“Low P” = 32µM NO3, 0.2µM PO4); and 16:1, non-limited (“nutrient-replete = 32µM NO3,
2.0µM PO4) media (Table 1.4). Each low-nutrient treatment series contained corresponding
solvent and media controls. Salinity and pH were checked at test initiation and conclusion to
insure changes were not outside acceptable benchmark limits (salinity 20 ±1, pH 8.1 ±0.2).
1.2.vi Statistical Analysis
All statistical analyses were performed using JMP software (version 12.2.0, SAS
Institute, Cary, NC). Normality was checked with Shapiro-Wilk test, and variance checks
were done using a Levene's test. Herbicide effects on growth were determined using a oneway analysis of variance (ANOVA) for normally distributed data and Kruskal-Wallace
where normality assumptions were violated, and where differences were detected among
treatments (p < 0.05), by using Dunnett’s procedure for multiple comparisons to determine
which specific treatments differed significantly from the controls (Zar 1999). Atrazine x
nutrient comparisons were analyzed by two-way ANOVA using nutrient level and herbicide
concentration as the fixed effects, and growth rate as the dependent variable. The F statistic
and associated P value for the independent variables and the interactions between them are
reported. Point estimates of sublethal toxicity (IC50s) were determined based on growth rates,
using the inhibition concentration linear interpolation (ICp) method described by NorbergKing (1993):
ICp  CJ  [ M 1(1  p /100)  M J ]
wherein;
(CJ  1  CJ )
( MJ  1  MJ )
13
CJ
CJ+1
M1
MJ
MJ+1
p
ICp
= tested concentration whose observed mean response is greater than M1 (1-p/100).
= tested concentration whose observed mean response is less than M1(1-p/100).
= smoothed mean response for the control
= smoothed mean response for concentration J
= smoothed response for concentration J+1
= percent reduction in response relative to the control response.
= estimated concentration at which there is a percent reduction from the smoothed
mean control response. The ICp is reported for the test together with the 95%
confidence interval or standard deviation as calculated by the ICPIN.EXE program
(Norberg-King 1993).
Results were generated using bootstrap methods to derive point estimates and confidence
intervals from no fewer than 80 resamplings (Efron 1982).
1.3 RESULTS
The first set of bioassays was performed to establish the 96-hr response of Dunaliella
tertiolecta to atrazine. Biomass was quantified at test conclusion (96 hr) by manual counting
of cells on a hemocytometer. IC50 estimates ranged from 87.7 µg L-1 to 143.3 µg L-1, (𝑥 =
110.1 µg L-1) and the lowest NOEC was 25 µg L-1 (for 50% of tests conducted) and the
highest was 50 µg L-1 (for 50% of tests) (Table 1.5, Fig. 1.3). After the test-specific effective
concentrations were established, the testing system was evaluated using OD techniques to
monitor biomass over time, and cell density was quantified every 24 hr for comparison.
A second set of bioassays was conducted over a 20-day interval to monitor the
response of D. tertiolecta to atrazine as nutrient limitation effects became more pronounced.
The 96-hr effective concentrations for this series were calculated from extrapolated cell
densities to allow direct comparison with manual cell count point estimates, and the range of
observed 96-h IC50 values was 100.6 to 155.4 µg L-1 (𝑥 = 125.9). The NOEC for all ODderived estimates of biomass was 50 µg atrazine L-1 (Table 1.6). The 95% confidence
intervals for these estimates indicated that the OD-derived point estimates were not
significantly different from those calculated using manual cell counts (Fig. 1.4).
14
The OD estimation technique allowed for monitoring changes in growth rates over
time, and point estimates of toxicity were generated for all measured time points (Table 1.7).
The IC50 estimates based on growth rate ranged from 11.7 µg L-1 (at 48 hr post-inoculation)
to > 200 µg L-1 (𝑥 =127.39 µg L-1). The lowest NOEC was < 12.5 µg L-1 (at 72-hours postinoculation) and the highest was >200 µg L-1 (observed at multiple timepoints). The 96-hr
growth rate-based IC50 estimates ranged from 139.3 to 179.7 µg L-1 (𝑥 =159.2 µg L-1). The
96-hr NOEC derived from growth rates ranged from 25 µg L-1 to 100 µg L-1.
Examination of controls for the 20-day nutrient limitation assay indicated that growth
rates in carrier controls were not significantly different from those in nutrient-replete media
controls, and effects of nutrient limitation were not observed until 7 days post-inoculation
(Fig. 1.5). Nutrient-replete controls continued to increase in density until day 15, and did not
decrease in density at any point in the testing period.
Two additional nutrient-effect bioassays were conducted over the standard 96-hr
testing duration to assess Dunaliella tertiolecta growth in nutrient x herbicide assays (Fig.
1.6). Growth was increasingly stimulated over time at the lowest atrazine concentration in the
nutrient-replete treatment, and at intermediate atrazine concentrations in the low phosphate
treatment, in comparison to controls (Fig. 1.7). Stimulation by atrazine was also observed in
the two highest herbicide concentrations x low nitrogen, but this response did not appear to
increase over time. Inhibition of growth by atrazine could not be quantified in any of the
nutrient-limited treatments using linear interpolation methods, as any observed effects did not
appear to be concentration-dependent and growth rates in low-nutrient x atrazine treatments
were generally similar to or increased over controls rather than inhibited.
For these bioassays, ANOVA models indicated that both nutrient treatment and
herbicide concentration were significant factors influencing growth rates, at both the 96-hr
(Table 1.8) and 20 day endpoints (Table 1.9). The concentration x nutrients interaction term
was also significant in all models; however, it was much less influential at 96-hr and was
15
ranked higher than nutrient treatment variables in the 20-day assay (p values 0.022 and <
0.00001, respectively).
1.4 DISCUSSION
Atrazine toxicity to algal groups represents the largest body of research of a single
organic pollutant on algae (Carder and Hoagland 1998), but in our survey of the ECOTOX
database (U.S. EPA 2017), of the 9,067 records for aquatic toxicity studies with atrazine
(pulled from database by CAS number 1912-24-9),
•
31% (2,812 records) were filed under the species group “Algae, Moss, Fungi” (2,238)
or “Algae, Moss, Fungi; Standard Test Species” (574);
•
19% (523) were labeled as estuarine/marine species (“SW” media type); and
•
only 3% of those (79, or 0.9% of total) were studies with the standard marine test
organism Dunaliella tertiolecta; 14 of these records provide half-maximal effect
concentrations after a minimum of 24 hr of growth; and 3 of the 14 studies which
provided half-maximal concentrations used population growth rate as the measured
endpoint (DeLorenzo and Serrano 2003; DeLorenzo et al. 2004; De Lorenzo and
Serrano 2006).
This relative lack of emphasis on algae may have occurred in part due to a general
assumption in the toxicological literature that there is little concern for off-target impacts
resulting from atrazine exposures (Huber 1993; Giddings 2005; Solomon et al. 2008; Van
Der Kraak et al. 2014). Yet, this assumption is not supported by ecotoxicological
assessments (e.g., Eisler 1989, DeLorenzo 1999, Pennington et al. 2001; DeLorenzo 2001).
The ECOTOX database 96-hr atrazine IC50s based on population changes (estimates of
growth rate or biomass) range from 66.35 to 180 µg/L for Dunaliella tertiolecta (U.S. EPA
2017). For green algae as a group, the effect concentrations ranged from 1.9 µg for Chlorella
vulgaris to 745 mg L-1 for Chlorella saccharophila, exemplifying the range of variability
seen within the same genus (U.S. EPA 2017).
16
The responses of D. tertiolecta to atrazine were quantified using modifications to
standard 96-hr algal bioassays (ASTM 1994) and are reported here using benchmarks which
are useful in inter-laboratory comparisons of sensitivity. The results of the 96-hr atrazine
bioassay tests were within the range of other reported experimental values (above). The 96-hr
ICr50 values were slighter higher than previously reported, but had less scatter (lower SD)
than ICb50s. Nyholm (1985) also noted that there was less scatter among growth-based values
from different laboratories than among estimates derived from biomass or from the area
under growth curves (which typically do not differ significantly from each other [Nyholm
and Källqvist 1989; Nyholm 1990]). Chao and Chen (2001) also found that EC50 estimates
based on growth rates were greater than those based on final yields. Growth rate has been
regarded as a more appropriate response variable for estimating effective concentration
(Nyholm 1990; Stauber 1995), since “using relative biomass as the response variable results
in numerical toxicity estimates and dose-response curve slopes which depend on the test
duration, the absolute magnitude of the maximum specific growth rate µm (and thus particular
system parameters), and which are also influenced by differences in inoculated cell
concentrations and by the occurrence of lag phase” (Nyholm 1990). When algal growth rate
is an exponential rather than a linear (biomass) function, it is less sensitive to these
potentially confounding factors (Nyholm 1985). Growth rate has also been shown to be 10fold more sensitive than ATP determinations after 4-hr exposures of Selenastrum
capricornutum to metals and organic compounds (Hickey et al. 1991).
Several studies have examined the potential for increased sensitivity or induced
tolerance in phytoplankton following short-term (< 5 day) exposures to atrazine. Proposed
mechanisms include mutations in the algal analog of the psbA gene, which mutates the D1
protein binding site and increases or decreases atrazine binding affinity (deNoyelles et al.
1982; Hamala and Kollig 1985; Pennington and Scott 2001). Changes in susceptibility can
also occur, resulting from changes in photoadaptation capacity or reorganization of the
photosynthetic apparatus which is an adaptive acclimation similar to low-light tolerance
(Hatfield et al. 1989; Millie et al. 1992). It has been suggested that the D1 protein may play a
17
role as a “light meter” in natural shade-adaption processes, and that its degradation and resynthesis is a possible signaling mechanism triggering photoadaptation (Koenig 1990). Low
light shunts carbon reserves into the photosynthetic apparatus, and may initiate a change in
the size or in the number of photosynthetic units. When shading (or PSII-inhibition by
herbicide exposure) occurs, the increase in photosynthetic units may, in turn, increase the
availability of D1-binding sights, thus increasing atrazine sensitivity or resulting in acquired
resistance through mutation of the algal psbA gene analog (Cheung et al. 1988).
Slight stimulatory responses (statistically different from controls) at intermediate
concentrations of atrazine were observed in all cultures at some time point during the 96-hr
or 20-day duration. The stimulatory response was most pronounced in low -nutrient
treatments during the first 24-72 hr of testing, wherein growth was elevated in comparison to
growth in nutrient-replete control cultures. In some assays, this stimulatory effect persisted
longer than 96 hr, postponing observations of growth limitation attributable to nutrient
depletion. Stimulation effects following atrazine exposure have been reported elsewhere
(Tang et al. 1997), and may be related to mechanisms for low-light adaptation (Hatfield et al.
1989; Koenig 1990; Gustavson and Wängberg 1995). Alternatively, the increased OD in lowlevel atrazine exposures may have reflected the fact that inhibition by this herbicide primarily
affects cell division rather than photosynthesis at these doses (Chao and Chen 2001). Other
research has shown that algal assemblages from well-shaded, low-light river systems were
less sensitive to atrazine exposures and had “a certain degree of protection against this
herbicide” (Guasch and Sabater 1998).
The presence of atrazine may also mimic nutrient-limited conditions (Weiner et al.
2007). Increased nutrient uptake has been associated with decreased levels of toxicant uptake
(Sanders 1979) and some pesticides have been shown to be more toxic to green algal cultures
under phosphate-limited conditions (Van Donk et al. 1992). Pannard et al. (2009) observed
that cyanobacteria with increased nutrient supply were more tolerant to atrazine, Hamala and
Kollig (1985) also reported that cyanobacteria (blue-green algae, Cyanophyta) dominated
atrazine-contaminated periphyton communities. Cyanobacteria have a competitive advantage
18
in light-limited ecosystems due to their accessory phycobilin pigments (Koenig 1990). They
appear to be relatively insensitive to atrazine (Fairchild et al. 1998). Thus, as other
researchers have suggested, the “increasing prevalence of cyanobacterial blooms may result
from a combination of unbalanced nutrient enrichment and selective pressures from multiple
toxicants” (Pannard et al. 2009). Nutrient enrichment appears to increase the toxicity of
atrazine to other phytoplankton groups (deNoyelles et al. 1982; DeLorenzo et al. 2001).
Nontarget organisms in herbicide exposures, such as algae, are the main primary
producers of many aquatic ecosystems, and potential water quality indicators (Blaise et al.
1997). Increasing loads of nutrients and other chemical contaminants to coastal areas are
altering phytoplankton growth and succession (Cloern 1996). These chemical contaminants,
although abiotic stressors, have been described as functionally similar to competition or
predation in their capacity to structure aquatic communities (Rohr et al. 2006). Nontarget
exposures of many toxicants, and nontarget exposures to co-ocurring contaminants such as
pesticides and nutrients, is a major research need. It is important to examine which factors
might play selective roles and alter community structure if toxicity assessments are to be
informative about how ecosystems exposed to toxicants can be expected to respond. Atrazine
alters phytoplankton assemblage structure (deNoyelles et al. 1982; Hamala and Kollig 1985;
Hamilton et al. 1987) and direct effects of atrazine on algal assemblages have been found at
concentrations as low as 20 µg L-1 (Graymore et al. 2001).
In this work, environmentally relevant concentrations of atrazine were assessed,
together with the two key macronutrients, nitrogen and phosphorus, which limit
phytoplankton growth and are increasingly found in imbalanced levels in aquatic ecosystems
(Glibert et al. 2011, Burkholder and Glibert 2013, and references therein). Many other
common, co-occurring chemical contaminants should be assayed, as well, to increase
confidence in chemical safety. New in silico techniques such as QSAR and chemical readacross offer methodologies to quantitatively group substances based on their mode of action
or functional groups (Hewitt et al. 2010; Patlewicz et al. 2013). These techniques can be
combined with microscale testing procedures (Gaggi et al. 1995; Wells et al. 1997; Wood et
19
al. 2014) which are adapted for use with small test volumes, while still providing the control
performance levels described here (based on the standardized guideline assays). Such
approaches may be useful in beginning to produce studies that screen many more species
within a series.
Photosystem II inhibition by atrazine can act similarly to light limitation and exert
selective control over phytoplankton assemblages, favoring species which have the capacity
to adapt to low light through accessory pigment composition, e.g. cyanobacteria (Koenig
1990), or which have alternative carbon fixation pathways and carbon concentrating
mechanisms (Pannard et al. 2009). The assay platform developed in this research will allow
the testing of atrazine effects on other estuarine microalgae, including ecologically important
mixotrophic harmful algal species (Chapters 2 and 3).
20
1.5 FIGURES AND TABLES
Table 1.1. List of species used for toxicity testing in the ECOTOX database (U.S. EPA 2017) ranked
by the number of different studies on all chemicals.
Species
Norway Rat
Rainbow Trout
Water Flea
Fathead Minnow
House Mouse
Bluegill
Common Carp
Zebra Danio
Domestic Chicken
Japanese Medaka
Goldfish
Green Algae
Rattus norvegicus
Oncorhynchus mykiss
Daphnia magna
Pimephales promelas
Mus musculus
Lepomis macrochirus
Cyprinus carpio
Danio rerio
Gallus domesticus
Oryzias latipes
Carassius auratus
Pseudokirchneriella subcapitata
Count of CASN
31,870
27,932
20,201
16,162
14,113
10,058
7,651
7,215
5,929
4,785
4,230
4,134
Figure 1.1. Light micrograph of Dunaliella tertiolecta, the benign test species used to
design the testing protocol. Scale bar = 50 µm.
21
Table 1.2. List of algal species used for aquatic atrazine toxicity testing in the ECOTOX database
(U.S. EPA 2017) ranked by number of different studies.
Algal Group
# of Studies
Green Algae
997
Algae
719
Diatoms
505
Blue-Green Algae
367
Cryptomonads
29
Haptophytes
23
Flagellate Euglenoids
20
Chrysophytes
15
Yellow Green Algae
14
Green Flagellates
13
Marine Microalgae
13
Red Algae
9
Prasinophyte
8
Microalgae
7
Phytoplankton
7
Golden-brown Algae
5
Filamentous Green Algae 1
Golden Algae
1
Hornwort
1
Species
Algae
Pseudokirchneriella subcapitata
Chlamydomonas reinhardtii
Microcystis aeruginosa
Dunaliella tertiolecta
Oophila sp.
Chlorella vulgaris
Scenedesmus quadricauda
Chlorella pyrenoidosa
Scenedesmus acutus var. acutus
# of Studies
543
265
96
88
79
79
74
48
44
41
22
Table 1.3. Marine algal species recommended for use in guideline testing protocols by indicated
organizations.
Species
Method
Chlorophyta
Dunaliella tertiolecta
ASTM 1990, APHA 1995, U.S. EPA 1974
Haptophyta
Isochrysis galbana
OECD 1998
Heterokontophyta, Bacillariophyceae
Skeletonema costatum
ASTM 1990, APHA 1995, ISO 1995, OECD 1998, U.S. EPA 1985
Thalassiosira pseudonana
ASTM 1990, APHA 1995, OECD 1998, U.S. EPA 1974
Phaeodactylum tricornutum
ISO 1995, OECD 1998
23
Table 1.4. Media recipes used in D. tertiolecta culture and testing.
L1 Macro-nutrient concentrations
Replete:
NaNO3
NaH2PO4
low N:
NaNO3
NaH2PO4
low P:
NaNO3
NaH2PO4
Stock Solution
Volume added to
final medium
Final concentration in
L1 media (µM)
2.7198 g L-1 H2O
0.2400 g L-1 H2O
1 mL
1 mL
32.00
2.00
0.1700 g L-1 H2O
0.2400 g L-1 H2O
1 mL
1 mL
2.00
2.00
2.7198 g L-1 H2O
0.0240 g L-1 H2O
1 mL
1 mL
32.00
0.20
L1 Trace Element Solution
Na2EDTA · 2H2O
FeCl3 · 6H2O
MnCl2 · 4H2O
ZnSO4 · 7H2O
CoCl2 · 6H2O
CuSO4 · 5H2O
Na2MoO4 · 2H2O
H2SeO3
NiSO4 · 6H2O
NaVO4
K2CrO4
Primary stock
solution
Quantity to add to
final medium
----178.10 g L-1 H2O
23.00 g L-1 H2O
11.90 g L-1 H2O
2.5 g L-1 H2O
19.9 g L-1 H2O
1.29 g L-1 H2O
2.63 g L-1 H2O
1.84 g L-1 H2O
1.94 g L-1 H2O
4.36 g
3.15 g
1 mL
1 mL
1 mL
1 mL
1 mL
1 mL
1 mL
1 mL
1 mL
Molar concentration in
final medium (M)
1.17 x 10-5
1.17 x 10-5
9.09 x 10-7
8.00 x 10-8
5.00 x 10-8
1.00 x 10-8
8.22 x 10-8
1.00 x 10-8
1.00 x 10-8
1.00 x 10-8
1.00 x 10-8
L1 Vitamin Solution
thiamine · HCl
(Vitamin B)
biotin (Vitamin H)
cyanocobalamin
(Vitamin B12)
Primary Stock
Solution
Quantity to Add
to Final Medium
Molar Concentration in
Final Medium (M)
---
200 mg
2.96 x 10-7
0.1 g L-1 H2O
10 mL
2.05 x 10-9
1.0 g L-1 H2O
1 mL
3.69 x 10-10
24
Figure 1.2. Correlation statistics between optical density and manual cell counts describing
procedures for selection of most appropriate wavelength for use in monitoring test activity. Where
multiple wavelengths produced similarly strong correlations (as displayed in the series above) the
wavelength with the highest R2 for the dilution curve was selected (in the displayed scenario, the
testing wavelength was 450 nm,)
25
Table 1.5. The 96-hr point estimates (µg atrazine L-1) based on effects on biomass estimated by
manual cell counts. Standard deviations (SDs) of point estimates were calculated by linear
interpolation (Norberg-King 1993).
Test Date
4 Apr
29 May
19 Jun
4 Jul
IC01 (SD)
7.1 (8.2)
1.2 (4.8)
1.3 (4.2)
6.8 (9.4)
IC10 (SD)
29.2 (12.4)
11.7 (5.1)
13.6 (8.5)
30.0 (5.3)
IC25 (SD)
64.8 (15.4)
23.4 (13.9)
60.8 (5.3)
40.9 (4.1)
IC50 (SD)
143.3 (10.7)
114.1 (20.7)
95.3 (10.8)
87.7 (10.0)
NOEC
50
25
50
25
LOEC
100
50
100
50
26
Table 1.6. The 96-hr point estimates (µg atrazine L-1) based on effects on biomass estimated by OD.
Test Date IC01 (SD)
23-July 0.9 (3.9)
27-Aug 1.0 (10.3)
5-Mar 13.1 (5.5)
IC10 (SD)
9.2 (10.3)
IC25 (SD)
43.7 (8.5)
IC50 (SD)
121.8 (27.3)
NOEC
50
LOEC
100
9.9 (18.8)
18.8 (9.6)
70.6 (15.6)
38.6 (16.4)
155.4 (25.6)
100.6 (27.0)
50
50
100
100
27
Figure 1.3. Results of 96 hr atrazine exposure for Dunaliella tertiolecta. Data are given as means + 1
standard error (SE); stars indicate significant differences from control).
28
Figure 1.4. Distribution of biomass derived effect estimates determined by manual cell counts (4 Apr,
29 May, 19 Jun, and 4 Jul) and OD (23 Jul, 27 Aug, and 5-Mar). Data are given as means + 1 SE.
29
Table 1.7. The atrazine effect estimates (µg L-1) over time based on D. tertiolecta growth rates
(divisions day-1) as estimated by OD. Note that BD ≡ beyond detection. Data include means, ranges,
and standard deviations (SDs), and the NOEC and LOEC values.
Test Hour
23 JUL
24
48
72
96
27 AUG
48
72
96
5 MAR
24
48
72
96
168
240
480
IC50 (95% CI)
133.9 (23.1 - 147.8)
137.1 (79.0 - 189.3)
175.5 (BD)
179.7 (BD)
11.7 (8.7 - 31.5)
49.3 (35.5 - 116.4)
139.3 (104.6 - 163.5)
BD
BD
BD
158.5 (96.3 - 186.3)
161.7 (141.9 - 184.1)
BD
BD
SD
24.6
33.1
15.8
13.3
8.7
30.8
17.0
------23.2
12.5
-----
IC10 (95% CI)
8.9 (3.5 - 108.2)
12.2 (3.1 - 103.5)
46.6 (7.9 - 102.4)
36.8 (7.3 - 69.3)
2.3 (1.7 - 3.8)
3.6 (2.6 - 6.5)
5.3 (3.9 - 10.3)
76.87 (BD)
BD
121.1 (BD)
20.0 (11.4 - 61.4)
19.1 (14.9 - 21.4)
24.5 (19.3 - 75.2)
155.9 (BD)
SD
29.1
26.7
24.3
17.9
0.6
1.2
2.9
27.1
--55.2
12.4
1.7
18.9
66.2
NOEC LOEC
100
200
>200
--100
200
100
200
25
50
<12.5 12.5
25
50
>200
-->200
-->200
--50
100
50
100
100
200
>200
---
30
Figure 1.5. Test control performance over time for the 20-day D. tertiolecta atrazine bioassays. Note that replete ≡ nutrient-replete.
31
*
*
*
**
*
*
*
*
*
Figure 1.6. Growth response of Dunaliella tertiolecta to atrazine exposure under different nutrient regimes. Data are given as
means +1 SE; stars indicate treatments that are significantly different from solvent controls).
32
Figure 1.7. Atrazine-induced changes in Dunaliella tertiolecta growth under different nutrient regimes over time. Data are
given as means + 1 SE. Note that R ≡ nutrient-replete.
33
Table 1.8. Influence of treatment variables on D. tertiolecta growth at 96 hr for the three nutrient x
atrazine bioassays. Model performance statistics: 23 July - summary of fit = 0.79, ANOVA F Ratio
7.9027, Prob > F < 0.0001; 27 Aug - summary of fit = 0.81, ANOVA F Ratio 9.1668, Prob > F <
0.0001; 5 March - summary of fit = 0.91, ANOVA F Ratio 22.0286, Prob > F < 0.0001). Trt ≡
treatment, Conc ≡ concentration, and DF ≡ degrees of freedom.
Source
23 July
Trt
Conc
DF
LogWorth
F Ratio
P Value
6
2
8.268
6.793
11.3801
30.9897
0.00000
0.00000
Conc*Trt
27 Aug
Conc
Trt
Conc*Trt
5 Mar
Trt
Conc*Trt
Conc
12
1.651
2.3163
0.02233
6
2
12
9.283
7.073
1.795
17.4558
24.6047
2.4493
0.00000
0.00000
0.01605
6
2
12
19.385
6.550
4.124
6.4097
154.9180
7.6899
0.00000
0.00000
0.00008
34
Table 1.9. Influence of treatment variables on D. tertiolecta growth at day 20. Model performance
statistics: 5 March, day 20 - summary of fit = 0.78, ANOVA F Ratio 7.7827, Prob > F < 0.0001).
Source
DF
LogWorth
F Ratio
P Value
Trt
6
13.239
5.4878
0.00000
Conc*Trt
2
5.626
50.4942
0.00000
12
3.540
6.4865
0.00029
Conc
35
1.6 REFERENCES
Alegria HA, d'Autel JP, Shaw TJ. 2000. Offshore transport of pesticides in the South
Atlantic Bight: preliminary estimate of export budgets. Mar. Pollut Bull 40:1178-1185.
American Public Health Association (APHA). 1995. Standard Methods for the
Examination for Water and Wastewater (19th edition). Byrd Prepess Springfield,
Washington, DC.
American Society for Testing and Materials (ASTM). 1996. Standard Guide for
Conducting Static 96-hour Toxicity Tests with Microalgae. West Conshohocken, PA:
A. Vol. 11.05, 573 -589.
Behra R, Genoni G, Joseph A. 1999. Effect of atrazine on growth, photosynthesis, and
between-strain variability in Scenedesmus subspicatus (Chlorophyceae). Arch. Environ.
Contam. Toxicol. 37:36-41.
Blaise C, Férard, JF, Vasseur, P. 1997. Microplate toxicity tests with microalgae: a review.
In: Microscale Aquatic Toxicology — Advances, Techniques and Practice (Wells PG,
Lee K, Blaise C, eds.). Boca Raton, FL: CRC-Lewis Press, 269-288.
Blanck H, Wallin G, Wangberg SA. 1984. Species-dependent variation in algal sensitivity
to chemical compounds. Ecotoxicol. Environ. Saf. 8:339-351.
Bricker SB, Longstaff B, Dennison W, Jones A, Boicourt K, Wicks C, Woerner J. 2008.
Effects of nutrient enrichment in the nation's estuaries: a decade of change. Harmful
Algae 8:21-32.
Burkholder JM, Glibert PM. 2013. Eutrophication and oligotrophication. In: Encyclopedia
of Biodiversity, 2nd edition, Volume 3 (Levin S, ed.). Waltham, MA: Academic Press,
347-371.
Byrd SM, Burkholder JM, Zimba PV. 2016. Environmental stressors and lipid production
by Dunaliella spp. I. Salinity. J. Exp. Mar. Biol. Ecol. 487:18-32.
Carder JP, Hoagland KD. 1998. Combined effects of alachlor and atrazine on benthic algal
communities in artificial streams. Environ. Toxicol. Chem. 17:1415-1420.
Chalifour A, Juneau P. 2011. Temperature-dependent sensitivity of growth and
photosynthesis of Scenedesmus obliquus, Navicula pelliculosa and two strains of
Microcystis aeruginosa to the herbicide atrazine. Aquat. Toxicol. 103:9-17.
36
Chao M-R, Chen C-Y. 2001. Discrepancies between different response parameters in batch
and continuous algal toxicity tests. J. Hazard. Mater. 82:129-136.
Cheung AY, Bogorad L, Van Montagu M, Schell J. 1988. Relocating a gene for herbicide
tolerance: a chloroplast gene is converted into a nuclear gene. Proc. Nat. Acad. Sci.
(USA) 85:391-395.
Choi CJ, Berges JA, Young EB. 2012. Rapid effects of diverse toxic water pollutants on
chlorophyll a fluorescence: variable responses among freshwater microalgae. Water
Research 46:2615-2626.
Cloern JE. 1996. Phytoplankton bloom dynamics in coastal ecosystems: a review with
some general lessons from sustained investigation of San Francisco Bay, California.
Rev. Geophys. 34:127-168.
Crossett K, Ache B, Pacheco P, Haber K. 2014. National coastal population report,
population trends from 1970 to 2020. National Oceanic and Atmospheric
Administration, Department of Commerce, developed in partnership with the US
Census Bureau Retrieved in pdf from: http:// stateofthecoast noaa gov/ features/
coastal-population-report pdf, last Accessed on 10 January 2017.
DeLorenzo ME. 2009. Utility of Dunaliella in ecotoxicity testing. In: The Alga Dunaliella:
Biodiversity, Physiology, Genomics and Biotechnology Science Publishers (BenAmotz A, Polle J, Suba Rao DV, eds.). Enfield, NH: Science Publishers, 495-512.
DeLorenzo M, Lauth J, Pennington P, Scott G, Ross P. 1999. Atrazine effects on the
microbial food web in tidal creek mesocosms. Aquatic Toxicology 46:241-251.
DeLorenzo ME, Leatherbury M, Weiner JA, Lewitus AJ, Fulton MH. 2004. Physiological
factors contributing to the species-specific sensitivity of four estuarine microalgal
species exposed to the herbicide atrazine. Aquat. Ecosyst. Health Manag. 7:137-146.
DeLorenzo ME, Scott GI, Ross PE. 2001. Toxicity of pesticides to aquatic
microorganisms: a review. Environ. Toxicol. Chem. 20:84-98.
DeLorenzo ME, Serrano L. 2003. Individual and mixture toxicity of three pesticides;
atrazine, chlorpyrifos, and chlorothalonil to the marine phytoplankton species
Dunaliella tertiolecta. J. Environ. Sci. Health, Part B 38:529-538.
DeLorenzo ME, Serrano L. 2006. Mixture toxicity of the antifouling compound irgarol to
the marine phytoplankton species Dunaliella tertiolecta. J. Environ. Sci. Health, Part B
41:1349-1360.
37
deNoyelles F, Kettle WD, Sinn DE. 1982. The responses of plankton communities in
experimental ponds to atrazine, the most heavily used pesticide in the United States.
Ecology:1285-1293.
Efron B. 1982. The jackknife, the bootstrap and other resampling plans. Vol. 38: Society
for Industrial and Applied Mathematics (SIAM).
Eisentraeger A, Dott W, Klein J, Hahn S. 2003. Comparative studies on algal toxicity
testing using fluorometric microplate and Erlenmeyer flask growth-inhibition assays.
Ecotoxicol. Environ. Saf. 54:346-354.
Eisler R. 1989. Atrazine Hazards to Fish, Wildlife, and Invertebrates: A Synoptic Review.
Biological Report 85(1.8), Contaminant Hazard Reviews Report No. 18. Laurel, MD:
U.S. Fish and Wildlife Service.
Eklund BT, Kautsky L. 2003. Review on toxicity testing with marine macroalgae and the
need for method standardization––exemplified with copper and phenol. Mar. Pollut.
Bull. 46:171-181.
Fahl GM, Kreft L, Altenburger R, Faust M, Boedeker W, Grimme LH. 1995. pHdependent sorption, bioconcentration and algal toxicity of sulfonylurea herbicides.
Aquatic Toxicology 31:175-187.
Fairchild JF, Ruessler DS, Carlson AR. 1998. Comparative sensitivity of five species of
macrophytes and six species of algae to atrazine, metribuzin, alachlor, and metolachlor.
Environ. Toxicol. Chem. 17:1830-1834.
Faust M, Altenburger R, Backhaus T, Blanck H, Boedeker W, Gramatica P, Hamer V,
Scholze M, Vighi M, Grimme LH. 2003. Joint algal toxicity of 16 dissimilarly acting
chemicals is predictable by the concept of independent action. Aquatic Toxicology
63:43-63.
Fleeger JW, Carman KR, Nisbet RM. 2003. Indirect effects of contaminants in aquatic
ecosystems. Sci. Total Environ. 317:207-233.
Forney DR, Davis DE. 1981. Effects of low concentrations of herbicides on submersed
aquatic plants. Weed Science 29:677-685.
Fuerst EP, Norman MA. 1991. Interactions of herbicides with photosynthetic electron
transport. Weed Science 39:458-464.
38
Gaggi C, Duccini M, Bacci E, Sbrilli G, Bucci M, Naby A. 1995. Toxicity and hazard
ranking of s-triazine herbicides using microtox® two green algal species and a marine
crustacean. Environ. Toxicol. Chem. 14:1065-1069.
Gascón J, Oubiña A, Barceló D. 1997. Detection of endocrine-disrupting pesticides by
enzyme-linked immunosorbent assay (ELISA): application to atrazine. TrAC (Trends
in Analytical Chemistry) 16:554-562.
Geis SW, Fleming KL, Korthals ET, Searle G, Reynolds L, Karner DA. 2000.
Modifications to the algal growth inhibition test for use as a regulatory assay. Environ.
Toxicol. Chem. 19:36-41.
Giddings JM. 2005. Atrazine in North American surface waters : a probabilistic aquatic
ecological risk assessment. Pensacola, FL: Society of Environmental Toxicology and
Chemistry.
Gilliom RJ, Barbash JE, Crawford CG, Hamilton PA, Martin JD, Nakagaki N, Nowell LH,
Scott JC, Stackelberg PE, Thelin GP. 2006. Pesticides in the Nation's Streams and
Ground Water, 1992-2001. Reston, VA: U.S. Geological Survey.
Glibert PM, Burkholder JM. 2017. Causes of blooms. In: Shumway SE, Burkholder JM,
Morton SL (eds.), Harmful Algal Blooms and Their Management: A Compendium
Desk Reference. Elsevier, New York, New York.
Glibert PM, Fullerton D, Burkholder JM, Cornwell JC, Kana TM. 2011. Ecological
stoichiometry, biogeochemical cycling, invasive species and aquatic food webs: San
Francisco Estuary and comparative systems. Rev. Fish. Sci. 19:358-417.
Glotfelty D, Taylor A, Isensee A, Jersey J, Glenn S. 1984. Atrazine and simazine
movement to Wye River estuary. J. Environ. Qual. 13:115-121.
Graymore M, Stagnitti F, Allinson G. 2001. Impacts of atrazine in aquatic ecosystems.
Environment International 26:483-495.
Guasch H, Muñoz I, Rosés N, Sabater S. 1997. Changes in atrazine toxicity throughout
succession of stream periphyton communities. J. Appl. Phycol. 9:137-146.
Guasch H, Sabater S. 1998. Light history influences the sensitivity to atrazine in periphytic
algae. J. Phycol. 34:233-241.
Guillard R, Hargraves P. 1993. Stichochrysis immobilis is a diatom, not a chrysophyte.
Phycologia 32:234-236.
39
Guillard RR. 1973. Division rates. Handbook of Phycological Methods 1:289-312.
Gustavson K, Wängberg S-Å. 1995. Tolerance induction and succession in microalgae
communities exposed to copper and atrazine. Aquatic Toxicology 32:283-302.
Hamala JA, Kollig HP. 1985. The effects of atrazine on periphyton communities in
controlled laboratory ecosystems. Chemosphere 14:1391-1408.
Hamilton PB, Jackson GS, Kaushik NK, Solomon KR. 1987. The impact of atrazine on
lake periphyton communities, including carbon uptake dynamics using track
autoradiography. Environmental Pollution 46:83-103.
Hatfield PM, Guikema JA, John JBS, Gendel SM. 1989. Characterization of the adaptation
response of Anacystis nidulans to growth in the presence of sublethal doses of
herbicide. Current Microbiology 18:369-374.
Hewitt M, Ellison C, Enoch S, Madden J, Cronin M. 2010. Integrating (Q) SAR models,
expert systems and read-across approaches for the prediction of developmental toxicity.
Reproductive Toxicology 30:147-160.
Hickey C, Blaise C, Costan G. 1991. Microtesting appraisal of ATP and cell recovery
toxicity end points after acute exposure of Selenastrum capricornutum to selected
chemicals. Environ. Toxicol. Water Qual. 6:383-403.
Howarth RW. 2008. Coastal nitrogen pollution: A review of sources and trends globally
and regionally. Harmful Algae 8:14-20.
Huber W. 1993. Ecotoxicological relevance of atrazine in aquatic systems. Environ.
Toxicol. Chem. 12:1865-1881.
Hughes JS, Alexander MM, Balu K. 1988. An evaluation of appropriate expressions of
toxicity in aquatic plant bioassays as demonstrated by the effects of atrazine on algae
and duckweed. In: Aquatic Toxicology and Hazard Assessment, 10th Volume.
Philadelphia, PA: ASTM.
Hylland K, Vethaak AD. 2011. Impact of contaminants on pelagic ecosystems, Chapter 10.
Ecological Impacts of Toxic Chemicals, 212-224.
International Organization for Standardization (ISO). 1995. Water quality––marine algal
growth inhibition test with Skeletonema costatum and Phaeodactylum tricornutum. ISO
International Organization of Standardization (ISO) 10253.
40
Jablonowski ND, Koeppchen S, Hofmann D, Schaeffer A, Burauel P. 2008. Spatial
distribution and characterization of long-term aged 14C-labeled atrazine residues in
soil. J. Agric. Food Chem. 56:9548-9554.
Jablonowski ND, Köppchen S, Hofmann D, Schäffer A, Burauel P. 2009. Persistence of 14
C-labeled atrazine and its residues in a field lysimeter soil after 22 years.
Environmental Pollution 157:2126-2131.
Jablonowski ND, Schäffer A, Burauel P. 2011. Still present after all these years:
persistence plus potential toxicity raise questions about the use of atrazine. Environ.
Sci. Pollut. Res. 18:328-331.
Janssen CR, Heijerick DG. 2003. Algal toxicity tests for environmental risk assessments of
metals. In: Rev. Environ. Contam. Toxicol. 178:23-52.
Kasai F, Takamura N, Hatakeyama S. 1993. Effects of simetryne on growth of various
freshwater algal taxa. Environmental Pollution 79:77-83.
Kemp W, Boynton W, Cunningham J, Stevenson J, Jones T, Means J. 1985. Effects of
atrazine and linuron on photosynthesis and growth of the macrophytes, Potamogeton
perfoliatus L. and Myriophyllum spicatum L. in an estuarine environment. Marine
Environmental Research 16:255-280.
Koenig F. 1990. Shade adaptation in cyanobacteria. Photosynthesis Research 26:29-37.
Lakshminarayana JS, O'Neill HJ, Jonnavithula SD, Léger DA, Milburn PH. 1992. Impact
of atrazine-bearing agricultural tile drainage discharge on planktonic drift of a natural
stream. Environmental Pollution 76:201-210.
Lehotay SJ, Harman-Fetcho JA, McConnell LL. 1998. Agricultural pesticide residues in
oysters and water from two Chesapeake Bay tributaries. Mar. Pollut. Bull. 37:32-44.
Lewis MA. 1990a. Are laboratory-derived toxicity data for freshwater algae worth the
effort? Environ. Toxicol. Chem. 9:1279-1284.
Lewis MA. 1990b. Chronic toxicities of surfactants and detergent builders to algae: a
review and risk assessment. Ecotoxicol. Environ. Saf. 20:123-140.
Lewis MA. 1995. Use of freshwater plants for phytotoxicity testing: a review.
Environmental Pollution 87:319-336.
Lockert C, Hoagland KD, Siegfried BD. 2006. Comparative sensitivity of freshwater algae
to atrazine. Bull. Environ. Contam. Toxicol. 76:73-79.
41
Lytle JS, Lytle TF. 2001. Use of plants for toxicity assessment of estuarine ecosystems.
Environ. Toxicol. Chem. 20:68-83.
McCarthy AM, Bales JD, Cope WG, Shea D. 2007. Modeling pesticide fate in a small tidal
estuary. Ecological Modelling 200:149-159.
McLachlan J. 1960. The culture of Dunaliella tertiolecta Butcher—a euryhaline organism.
Canadian Journal of Microbiology 6:367-379.
Millie DF, Hersh CM, Dionigi CP. 1992. Simazine-induced inhibition in photoacclimated
populations of Anabaena circinalis (Cyanophyta). J. Phycol. 28:19-26.
Mohan B, Hosetti B. 1999. Aquatic plants for toxicity assessment. Environmental Research
81:259-274.
Moreland D, Hill K. 1962. Interference of herbicides with the Hill reaction of isolated
chloroplasts. Weeds 10:229-236.
Norberg-King TJ. 1993. A linear interpolation method for sublethal toxicity: The inhibition
concentration (ICp) approach. NETEC (National Effluent Toxicity Assessment Center)
Technical Report 39:3-93.
Nyholm N. 1985. Response variable in algal growth inhibition tests—biomass or growth
rate? Water Research 19:273-279.
Nyholm N. 1990. Expression of results from growth inhibition toxicity tests with algae.
Arch. Environ. Contam. Toxicol. 19:518-522.
Nyholm N, Källqvist T. 1989. Methods for growth inhibition toxicity tests with freshwater
algae. Environ. Toxicol. Chem. 8:689-703.
Nyholm N, Lyngby JE. 1988. Algal bioassays in eutrophication research—A discussion in
the framework of a mathematical analysis. Water Research 22:1293-1300.
Organization for Economic Cooperation and Development (OECD). 1998. "Alga, Growth
Inhibition Test". OECD Guidelines for Testing of Chemicals.
Pannard A, Le Rouzic B, Binet F. 2009. Response of phytoplankton community to lowdose atrazine exposure combined with phosphorus fluctuations. Arch. Environ.
Contam. Toxicol. 57:50-59.
42
Patlewicz G, Ball N, Booth ED, Hulzebos E, Zvinavashe E, Hennes C. 2013. Use of
category approaches, read-across and (Q) SAR: general considerations. Regulatory
Toxicology and Pharmacology 67:1-12.
Pennington PL, Daugomah JW, Colbert AC, Fulton MH, Key PB, Thompson BC, Strozier
ED, Scott GI. 2001. Analysis of pesticide runoff from mid-Texas estuaries and risk
assessment implications for marine phytoplankton. J. Env. Sci. Health, Part B 36:1-14.
Pennington PL, Scott GI. 2001. Toxicity of atrazine to the estuarine phytoplankter Pavlova
sp. (Prymnesiophyceae): Increased sensitivity after long-term, low-level population
exposure. Environ. Toxicol. Chem. 20:2237-2242.
Pereira WE, Rostad CE. 1990. Occurrence, distributions, and transport of herbicides and
their degradation products in the lower Mississippi River and its tributaries. Environ.
Sci. Technol. 24:1400-1406.
Peterson HG, Boutin C, Martin PA, Freemark KE, Ruecker NJ, Moody MJ. 1994. Aquatic
phytotoxicity of 23 pesticides applied at expected environmental concentrations.
Aquatic Toxicology 28:275-292.
Peterson HG, Healey FP, Wagemann R. 1984. Metal toxicity to algae: a highly pH
dependent phenomenon. Canadian Journal of Fisheries and Aquatic Sciences 41:974979.
Readman J, Albanis T, Barcelo D, Galassi S, Tronczynski J, Gabrielides G. 1993.
Herbicide contamination of Mediterranean estuarine waters: results from a MED POL
pilot survey. Mar. Pollut. Bull. 26:613-619.
Rohr JR, Kerby JL, Sih A. 2006. Community ecology as a framework for predicting
contaminant effects. TREE 21:606-613.
Rohr JR, McCoy KA. 2010. A qualitative meta-analysis reveals consistent effects of
atrazine on freshwater fish and amphibians. Environ. Health Perspect. 118:20-32.
Rojíčková R, Maršálek B. 1999. Selection and sensitivity comparisons of algal species for
toxicity testing. Chemosphere 38:3329-3338.
Ryberg KR, Vecchia AV, Martin JD, Gilliom RJ. 2010. Trends in Pesticide Concentrations
in Urban Streams in the United States, 1992-2008. Reston, VA: U. S. Geological
Survey.
Sanders JG. 1979. Effects of arsenic speciation and phosphate concentration on arsenic
inhibition of Skeletonema costatum (Bacillariophyceae). J. Phycol. 15:424-428.
43
Scott GI, Holland AF, Sandifer PA. 2006. Afterword—managing coastal urbanization and
development in the twenty-first century: The Need for a New Paradigm. In: Changing
Land Use Patterns in the Coastal Zone (Kleppel GS, DeVoe MR, Rawson MV, eds.).
Springer, New York, NY: Springer Science + Business Media, 285-299.
Sjollema SB, Vavourakis CD, Geest H, Vethaak AD, Admiraal W. 2014. Seasonal
variability in irradiance affects herbicide toxicity to the marine flagellate Dunaliella
tertiolecta. Front. Mar. Sci. 1: article 13, 1-5. Doi: 10.3389/fmars.2014.00023.
Solomon KR, Baker DB, Richards RP, Dixon KR, Klaine SJ, La Point TW, Kendall RJ,
Weisskopf CP, Giddings JM, Giesy JP. 1996. Ecological risk assessment of atrazine in
North American surface waters. Environ. Toxicol. Chem. 15:31-76.
Solomon KR, Carr JA, Du Preez LH, Giesy JP, Kendall RJ, Smith EE, Van Der Kraak GJ.
2008. Effects of atrazine on fish, amphibians, and aquatic reptiles: a critical review.
Critical Reviews in Toxicology 38:721-772.
Sorokin C. 1973. Dry weight, packed cell volume and optical density. Handbook of
phycological methods: Culture methods and growth measurements. In: Handbook of
Phycological Methods (Stein JR, ed.). Cambridge, MA: Cambridge University Press,
321-343.
Southwick L, Grigg B, Fouss J, Kornecki T. 2003. Atrazine and metolachlor in surface
runoff under typical rainfall conditions in southern Louisiana. Journal of Agricultural
and Food Chemistry 51:5355-5361.
Starr AV, Bargu S, Maiti K, DeLaune RD. 2016. The Effect of Atrazine on Louisiana Gulf
Coast Estuarine Phytoplankton. Archives of Environmental Contamination and
Toxicology. 1-11.
Stauber J. 1995. Toxicity testing using marine and freshwater unicellular algae.
Australasian J. Ecotoxicol. 1:15-24.
Stephan CE, Mount DI, Hansen DJ, Gentile J, Chapman GA, Brungs WA. 1985.
Guidelines for Deriving Numerical National Water Quality Criteria for the Protection
of Aquatic Organisms and their Uses. Report PB-227049. Duluth, MN: U.S.
Environmental Protection Agency.
Stratton GW, Giles J. 1990. Importance of bioassay volume in toxicity tests using algae and
aquatic invertebrates. Bull. Environ. Contam. Toxicol. 44:420-427.
Swanson SM, Rickard CP, Freemark KE, MacQuarrie P. 1991. Testing for pesticide
toxicity to aquatic plants: recommendations for test species. In: Plants for Toxicity
44
Assessment, Vol. 2 (Gorsuch JW, Lower WR, Wang W, Lewis MA, eds.).
Philadelphia, PA: ASTM, 77-97.
Tang JX, Hoagland KD, Siegfried BD. 1997. Differential toxicity of atrazine to selected
freshwater algae. Bull. Environ. Contam. Toxicol. 59:631-637.
Tang J, Hoagland KD, Siegfried BD. 1998. Uptake and bioconcentration of atrazine by
selected freshwater algae. Environ. Toxicol. Chem. 17:1085-1090.
United States Environmental Protection Agency (U.S. EPA). 2017. ECOTOX User Guide:
ECOTOXicology Knowledgebase System. Version 4.0. Available at:
http:/www.epa.gov/ecotox/ (last accessed February 2017).
United States Environmental Protection Agency (U.S. EPA). 2007. Atrazine: Toxicity and
Exposure Assessment for Children's Health (TEACH) Chemical Summary. Available
at: https://archive.epa.gov/region5/teach/web/pdf/atrazine_summary.pdf (last accessed
March 2017).
United States Environmental Protection Agency (U.S. EPA). 2003. Ambient Aquatic Life
Water Quality Criteria for Atrazine – Revised Draft. Report #EPA-822-R-03-023.
Washington, DC: U.S. EPA. Available at:
https://nepis.epa.gov/Exe/ZyPDF.cgi/100047EJ.PDF?Dockey=100047EJ.PDF (last
accessed 2 February 2017).
United States Environmental Protection Agency (U.S. EPA). (2007b). National Water
Quality Inventory: Report to Congress, 2002 Reporting Cycle. Washington, DC: U.S.
EPA. (EPA 841-R-07-001). Available from:
https://www.epa.gov/sites/production/files/201509/documents/2007_10_15_305b_2002report_report2002305b.pdf (accessed 2 March
2017).
United States Environmental Protection Agency (U.S. EPA). (2002). National Water
Quality Inventory: Report to Congress, 2000 Reporting Cycle. Washington, DC: U.S.
EPA. (EPA 841-R-02-001). Available from:
https://www.epa.gov/sites/production/files/201509/documents/2000_national_water_quality_inventory_report_to_congress.pdf
(accessed 2 March 2017).
United States Environmental Protection Agency (U.S. EPA). 1985. "Toxic Substances
Control Act Test Guidelines; Final Rules," Environmental effects testing guidelines,
Algal acute toxicity test. Federal Register, 40 CFR Part 797.1050, pp. 39321-39331.
Available at: https://www.law.cornell.edu/cfr/text/40/797.1050 (last accessed 2
February 2017)
45
United States Environmental Protection Agency (U.S. EPA). 1971. Algal Assay Procedure
Bottle Test. National Eutrophication Research Program. Corvallis, OR: Pacific
Northwest Environmental Research Laboratory.
Van Der Kraak GJ, Hosmer AJ, Hanson ML, Kloas W, Solomon KR. 2014. Effects of
atrazine in fish, amphibians, and reptiles: an analysis based on quantitative weight of
evidence. Critical Reviews in Toxicology 44:1-66.
Van Donk E, Abdel-Hamid M, Faafeng B, Källqvist T. 1992. Effects of Dursban® 4E and
its carrier on three algal species during exponential and P-limited growth. Aquat.
Toxicol. 23:181-191.
Vollenweider RA (ed.). 1974. A Manual on Methods for Measuring Primary Production in
Aquatic Environments, second edition. Oxford, United Kingdom: Blackwell Scientific.
Walsh GE. 1988. Principles of toxicity testing with marine unicellular algae. Environ.
Toxicol. Chem. 7:979-987.
Walsh GE, Bahner LH, Horning WB. 1980. Toxicity of textile mill effluents to freshwater
and estuarine algae, crustaceans and fishes. Environ. Pollut. Ecol. Biol. 21:169-179.
Walsh GE, Merrill R. 1984. Algal bioassays of industrial and energy process effluents. In:
Algae as Ecological Indicators (Shubert, L.E., ed.). New York, NY: Academic Press,
329-360.
Weiner JA, DeLorenzo ME, Fulton MH. 2004. Relationship between uptake capacity and
differential toxicity of the herbicide atrazine in selected microalgal species. Aquatic
toxicology 68:121-128.
Weiner JA, DeLorenzo ME, Fulton MH. 2007. Atrazine induced species-specific
alterations in the subcellular content of microalgal cells. Pest. Biochem. Physiol. 87:4753.
Wells PG, Lee K, Blaise C. 1997. Microscale testing in aquatic toxicology: advances,
techniques, and practice. Boca Raton, FL: CRC Press.
Wood RJ, Mitrovic SM, Kefford BJ. 2014. Determining the relative sensitivity of benthic
diatoms to atrazine using rapid toxicity testing: A novel method. Sci. Total Environ.
485:421-427.
Zar JH. 1999. Biostatistical Analysis, 4th edition. Upper Saddle River, NJ: Prentice Hall.
46
CHAPTER 2: Persistence Under Sub-Optimal Conditions; Imbalanced Nutrient
Regimes Increase Prymnesium parvum Resilience to Herbicide Exposure
2.1 INTRODUCTION
Within the past few decades there has been a global increase in the frequency,
magnitude and duration of harmful algal blooms (HABs; Smayda 1990, Hallegraeff 1993,
Burkholder 1998, Van Dolah 2000, Glibert et al. 2005, Anderson et al. 2008, Heisler et al.
2008, Lewitus et al. 2012). These blooms are often toxic and threaten aquatic biota, wildlife,
and human health in inland and coastal waters (Burkholder 1998, 2009; Anderson et al.
2008; Lewitus et al. 2012; Brooks et al. 2016). Many if not most harmful algae are
mixotrophs rather than strict phototrophs; that is, they derive nutrition from both
photosynthesis and consumption of dissolved or particulate organic carbon (Burkholder et al.
2008). Thus, their behavior and physiology are markedly different from those of strict
phototrophs. Observed increases in HABs mainly have been related to anthropogenic nutrient
enrichment which can stimulate harmful algae directly or indirectly by stimulating growth of
prey, and also to range expansion due to global climate change and increased dispersal via
global transportation (Heisler et al. 2008, Pitcher 2012). Despite increasing pollutant inputs
to coastal areas from rapid, ongoing watershed development and human population growth
(Scott 2006; U.S. EPA 2016), the impacts of chemical contamination on coastal
estuarine/marine ecosystems are poorly understood, especially the influences of interacting
chemical contaminants on harmful algae (Burkholder and Shumway 2017, and references
therein).
The relatively few available studies about effects of chemical pollutants on estuarine
and marine phytoplankton generally have involved the assessment of direct effects of
individual stressors on one or two model phototrophic species, selected for their high growth
rates and “hardiness” in captivity rather than their ecological importance (Kimball and Levin
1985). In a review of toxicant impacts on marine ecosystem function, only 10% (26 of 264)
of the available studies assessed the impact of herbicides, and less than 40% assessed
phytoplankton responses (Johnston et al. 2015). Yet, species sensitivity comparisons
47
routinely have indicated that estuarine/marine microalgae are among the most sensitive
organisms in response to toxicant exposures, and are especially sensitive to herbicides
(Stauber 1995, Mohan and Hosetti 1999, Geis et al. 2000, Lytle and Lytle 2001). This
sensitivity is not surprising, as algae would be expected to be more susceptible to the main
mode of action of most herbicides (inhibition of photosynthesis) than other marine organisms
(Peterson et al. 1994, Solomon et al. 1996, Giddings 2005). Microalgae are the energetic
base of marine food webs, and therefore largely control the ecosystem carrying capacity and
stability (Lembi and Waaland 1988). Restriction of toxicological assessments to culture
studies of a few “hardy” algae (U.S. EPA 1974, Stauber 1995, ASTM 1996), with little
ecological relevance, has not enabled assessment of the potential for pervasive chemical
contamination to function as an important structuring mechanism in estuarine and marine
communities (Johnston 2016).
Phytoplankton responses to toxicants vary widely by species (Walsh 1972, Blanck et
al. 1984, Rojíčková-Padrtová et al. 1998, Rojíčková and Maršálek 1999, Ma 2005), strain,
and even sub-strain (Kasai et al. 1993, Burkholder and Glibert 2006, and references therein).
No single algal species can be selected as universally protective for all “algae” - a term which
includes widely disparate phyla ecologically and physiologically (Graham et al. 2016).
Current bioassay assessment protocols are designed to test response endpoints of
predominantly phototrophic species under optimal growth conditions, and resulting “safe”
levels likely do not provide an adequate framework for protecting against effects of
pollutants interacting with harmful algal species. “Most sensitive” or “hardiest” species
responses may not be indicative of the “most ecologically important” response. Low toxicity
may not equate with low ecological risk because abundant, ubiquitous chemical contaminants
commonly influence phytoplankton population dynamics through indirect mechanisms
(Burkholder and Shumway 2017, and references therein). Toward the goal of developing
improved management strategies to mitigate impacts from harmful algae, studies should be
focused on species with the highest potential to affect ecosystem structure, and on the
contaminants to which they are commonly exposed. There is an urgent need to evaluate how
48
high tolerance to a chemical contaminant could confer a selective advantage and lead to
increased abundance or even dominance of harmful populations. Changes in algal
assemblages can result in altered ecosystem function (McCormick and Cairns 1994), and
increased abundance of harmful species can result in complete ecological collapse (Sunda et
al. 2006 and references therein). These considerations, considered collectively, indicate that
bioassay data derived from model phototrophic species cannot be used to extrapolate
pollutant effects realistically to mixotrophic HAB species.
Prior to this research, interactions between nutrient levels and herbicide exposure in
affecting estuarine HAB species had not been examined. Since both nutrients and herbicides
commonly co-occur in stormwater runoff (Eisler 1989, Gilliom et al. 2006), a series of
experiments was designed to investigate the effects of herbicide exposure under imbalanced
nutrient regimes on the growth and toxicity of a widespread, common harmful phytoplankter,
the haptophyte Prymnesium parvum Carter 1937. Atrazine (2-chloro-4-ethylamino-6isopropylamino-s-triazine) was the toxicant selected for study because of its specific
mechanism of action as a reversible inhibitor of photosystem II (Solomon et al. 2013), the
amount of comparable information available regarding effects on marine microalgae
(DeLorenzo et al. 2001), and (due to its recognized mobility) its prevalence and persistence
in estuarine ecosystems in the U.S. (Eisler 1989, Giddings 2005, Gilliom et al. 2006,
Jablonowski et al. 2011). This herbicide was also selected in consideration of a recent study
which suggested that atrazine exposure can lead to dominance of P. parvum in phytoplankton
assemblages in situ (Yates and Rogers 2011).
This research was designed to address the following questions: What is the effect of
exposure to atrazine on P. parvum growth? To what extent does atrazine exposure promote
production of hemolytic substances in this HAB species? Is that response affected by the
nutrient or salinity regimes? How do interactions between nutrient limitation, salinity, and
herbicide exposure influence growth rates and hemolytic activity in P. parvum? And, are
these responses consistent for multiple strains (populations or isolates) of P. parvum?
49
2.2 MATERIALS AND METHODS
A series of factorial experiments was designed to expose Prymnesium parvum
(Haptophyta) to a range of co-stressors including salinity, nutrients, and atrazine, using
modifications of standard algal bioassay procedures (Miller and Greene 1978, ASTM 1996)
and examining hemolytic activity using a modified erythrocyte lysis assay (Eschbach et al.
2001). These assays were designed to test the null hypothesis that neither herbicide activity,
nutrient limitation, salinity, strain origin, nor their interactions would have a net effect on
growth rates or relative production of hemolytic compounds in static batch cultures of P.
parvum.
2.2.i Experimental Organism, Strains, and Culture Conditions
2.2.i.a. Toxic Prymnesium parvum
The cosmopolitan (with exception of Antarctica), toxigenic estuarine/marine
phytoplankter, Prymnesium parvum (Fig. 2.1), has an extensive historical record of forming
nearly monospecific, high-biomass, ecosystem-disrupting blooms (Sunda et al. 2006). This
organism forms large blooms and causes major fish kills in brackish waters and aquaculture
facilities in many regions (Reich and Aschner 1947, Shilo and Shilo 1953, Holdway et al.
1978, Kaartvedt et al. 1991, Moestrup 1994, Edvardsen and Paasche 1998). It has also
caused kills of fish and other gill-breathing organisms in high-conductivity inland waters of
the American West and other areas of the U.S. (James and De La Cruz 1989, Sager et al.
2008, Hambright et al. 2010, Southard et al. 2010, VanLandeghem et al. 2015b). This alga
can grow at salinities of 0.8 to100, with optimal growth at salinity 10-20 (Edvardsen and
Imai 2006), but most associated fish kills in U.S. waters have occurred at low to moderate
salinities (< 10 to 20) (Baker et al. 2007, 2009; Sager et al. 2008; Roelke et al. 2011).
Toxicity has been inversely related to salinity (Baker et al. 2007), and has been highest in
eutrophic waters under low-nutrient (nitrogen or phosphorus) stress (Johansson and Granéli
50
1999, Granéli and Johansson 2003b, Baker et al. 2009, Granéli et al. 2012, Roelke et al.
2016).
Growth of P. parvum occurs by photosynthesis, and also by heterotrophy (organic
carbon acquisition) as uptake of dissolved organics from the environment or from prey that it
immobilizes with toxins (Tillmann 1998, Skovgaard and Hansen 2003), or phagotrophy
(consumption of particulate matter) (Nygaard and Tobiesen 1993, Tillmann 1998, MartinCereceda et al. 2003, Skovgaard and Hansen 2003, Tillmann 2003, Graneli 2006, Remmel et
al. 2011, Remmel and Hambright 2012). The toxins can also inhibit or kill competing
phytoplankton or grazer populations, by cell lysis or by suppressing growth, feeding and/or
reproduction (Fistarol et al. 2003, Granéli and Johansson 2003a, Skovgaard and Hansen
2003).
More than 15 toxins or highly bioactive substances have been identified from P.
parvum cells and exudates, including proteolipids (Ulitzur and Shilo 1970), saponins (Yariv
and Hestrin 1961), various fatty acids (Henrikson et al. 2010), and fatty acid amides
[specifically oleamide and linoliamide (Bertin et al. 2012)], although some of this diversity
may reflect differences in the extraction techniques employed (Manning and La Claire 2010).
Some toxic substances from P. parvum extracts have not yet been structurally characterized,
but the impacts to biological systems identified thus far have cytotoxic, hemolytic,
hepatotoxic and/or neurotoxic properties (Shilo 1981, Granéli and Johansson 2003b,
Manning and La Claire 2010). Substances such as hemolysins are ichthyotoxic (Shilo 1967).
Most work to chemically characterize P. parvum toxins has focused on two high-molecularweight, brevetoxin-like, cyclic polyether substances called prymnesin-1 and prymnesin-2
(Ulitzur and Shilo 1970; Shilo 1981; Igarashi et al. 1996, 1999; Manning and La Claire
2010). The available evidence suggests that each toxin has a distinct mode of action, and that
the toxins interact with each other and with pH-dependent co-factors (likely divalent metals
such as Ca+2 or Mg+2) in the aquatic environment which modulate toxicity (Shilo and
Rosenberger 1960, Manning and La Claire 2010, Roelke et al. 2011). Various other
51
environmental factors (e.g., pH, total dissolved solids) influence toxicity as well (Baker et al.
2007, 2009; Valenti et al. 2010; Prosser et al. 2012; VanLandeghem et al. 2015a,b).
2.2.i.b. Strains and Culturing
Strains of Prymnesium parvum were obtained from the University of Texas culture
collection (UTEX LB 2797 and LB 2827, here referred to as the Texas (TX) strain and the
South Carolina (SC) strain, respectively) and from the Center for Marine Science at UNC
Wilmington (laboratory of Dr. C. Tomas), referred to here as the North Carolina (NC) strain.
Background information about collection of these strains is given in Table 2.1.
Non-axenic, unialgal batch cultures of all strains were maintained in 250-mL
Erlenmeyer flasks with 100 mL modified f/2 (-Si) medium (Guillard and Ryther 1962) (see
Table 2.2 for culture and test media recipes). The UTEX strains arrived in salinity 30
Erdschreiber’s Media, and were gradually transferred into f/2 media over two weeks and split
into two subcultures. Each subculture was slowly acclimated over 2-4 weeks to either salinity
20 or 10 for experiments. The NC strain was transferred from the source lab in salinity 4 f/2Si media, and was acclimated into salinity 10 and 20 as described above. All cultures were
maintained at 24°C (ambient air temperature) with light (photosynthetically active radiation)
at 106.31 (±26.16) µmol s-1 m-2 (cool white fluorescent tubes) and a 16:8 hr light:dark cycle.
Light was measured using a Biospherical QSL 101 Quantum lab sensor (BSI, San Diego,
CA, USA). Cultures were sterile-transferred every 7-10 days as needed to maintain log-phase
growth. All media (culture and test) were prepared by adjusting the salinity of ultrapure
Milli-Q water (18 MΩ cm−1 at 25°C) with Instant Ocean® artificial sea salts (Aquarium
Systems, Blacksburg, VA, USA) to the desired level, also adjusting the pH as necessary to
8.1 (± 0.2 unit) using HCl or NaOH. Salinity was measured using a YSI model 3200
conductivity/salinity bridge and cell (YSI, Yellow Springs, OH, USA). The pH was
measured using an Orion Versa Star Pro multi-parameter meter equipped with an Orion
ROSS Sure-Flow pH electrode, Thermo Scientific, Waltham, MA, USA). Trace mineral and
vitamin solutions were sterile-filtered (Whatman® Puradisc cellulose acetate syringe filter,
52
nominal pore size 0.2 μm, Sigma-Aldrich, St. Louis, MO, USA) and added to the autoclaved
media aseptically.
The final medium was vacuum-filtered (Corning® cellulose acetate filter, 0.22 µm
nominal pore size; Sigma-Aldrich) under a laminar flow hood before storage in darkness at
4°C for no more than 30 days. All equipment was initially cleaned by scrubbing in hot, soapy
tap water. All glassware and non-plastic equipment used in culturing and experiments was
rinsed with pesticide-grade acetone before use and all glassware and any non-metallic
equipment used was cleaned by acid stripping in 10% HCl (v/v). All equipment was
sterilized by autoclaving before use. All algal observation and testing methods were initially
developed and optimized using the benign estuarine/marine microalga, Dunaliella tertiolecta,
as described in Chapter 1 of this work.
Stock cultures were maintained at 100 µmol photons m-2 s-1 throughout the
experiments, and were gently swirled once daily. Bioassays (controls and test cultures) were
maintained on a culture rack under a fluorescent light bank (2.5 m long by .75 m deep). All
cultures in the bioassays were gently swirled by hand and randomly repositioned daily on the
culture rack following a computer-generated randomized shelf mapping, to avoid any effects
from possible differential light exposure.
2.2.ii Optical Density Versus Direct Cell Counts
1-mL samples were randomly withdrawn from each parent culture and thoroughly
vortexed before absorbance was read on a Thermo Scientific Spectronic GENESYS 2
spectrophotometer (Thermo Scientific, Waltham, MA, USA), using a freshly wiped Hellma®
quartz cuvette (Sigma-Aldrich, St. Louis, MO, USA) with a 10-mm pathlength. All reads
were performed in duplicate and adjusted using a media blank as a measurement baseline.
All parent cultures were scanned individually, and the three highest peak absorbencies were
selected as the most appropriate wavelengths for analysis of a series of 8-point dilution
curves, using progressive dilutions of parent stock culture in growth media. The curve which
53
produced the ΔOD/ΔPD [change in OD divided by change in population density; Sorokin
(1973)] closest to unity was used to determine the growth rates from each stock culture for
each strain tested. Cell counts to determine ΔPD were obtained by collecting random 1-mL
samples from the parent culture. Cell counts were made in duplicate using a Reichert BrightLine phase contrast hemocytometer with improved Neubauer rulings (Hausser Scientific No.
1475, Horsham, PA, USA) (Guillard 1973). Two homogenized 100-µL samples were
assessed by counting 18 grids or a minimum of 400 cells (DeLorenzo and Serrano 2003;
Weiner et al. 2003). Cell preservation techniques (0.1% Lugol’s solution / 0.05%
gluteraldhyde, v/v) were developed and used to preserve cells for about one week without
appreciable cell distortion or loss while preparing dilution curves immediately before each
experiment. The wavelength which produced the highest correlation between absorbance and
cell number of the stock culture was used to monitor population density and calculate growth
rates for the duration of the experiment. All modeled OD : cell number relationships that
were used had a minimum observed 𝑟 2 value greater than 0.98) (Fig. 2.2).
2.2.iii Population Growth Rate Measurements
Net cell-specific growth rates (day-1) were modeled using OD following Sorokin
(1973). Samples were collected daily from hand-agitated test tubes and measured in duplicate
in a clean, wiped quartz cuvette as described above, using a media blank as a baseline
measurement. The OD was converted to PD, data was log transformed (log10 + 1) and the
linear portion of the growth curve (slope = 𝑎𝑘 ) was used to calculate growth rate k (Sorokin
1973):
𝑘 = 𝑎𝑘 ∗ 3.322
2.2.iv Algal Bioassays
Toxicity experiments were based on examining the effect of atrazine alone and in
varying nutrient regimes on the population growth rate of Prymnesium parvum. Testing
procedures were derivations on the standard 96-hr static algal toxicity bioassay protocols
54
(ASTM 1996), and were conducted to assess both 4- and 10-day endpoints to allow for
evaluation of toxicant responses following nutrient depletion from cellular reserves.
Analytical grade standards of atrazine (CASRN 1912-12-9, >98.0% purity) were obtained
from Chemservice (Westchester, PA, USA). Stock solutions were prepared in 100%
pesticide-grade acetone, with testing doses administered to obtain a final concentration of
0.1% (v/v) acetone in each replicate, except for a medium-only control series (prepared
without acetone additions) which was included in all assays. There were three replicate
culture tubes for each treatment and control series incubated in 50-mL round-bottomed
borosilicate glass tubes with Teflon lined caps. Each tube contained a total final volume of 25
mL. Five-point concentration curves were developed based on geometrically spaced atrazine
concentrations (12.5, 25, 50, 100 and 200 µg/L) and media and solvent controls, with
exposure levels selected based on range-finding test estimates designed to expose the test
organism to concentrations with lowest and highest effects on population growth. When
stock culture concentrations were in log phase growth (as assessed by periodic direct cell
counts) the test culture tubes plus control cultures were inoculated under aseptic conditions to
achieve the same initial cell densities of approximately 5.0 x 104 cells mL-1 in all
experimental cultures and controls. Chemical stocks were diluted to achieve desired final
concentrations and were introduced into fresh, sterile culture media aseptically under a
laminar flow hood. Time-series testing began when algae were added to test media, and OD
was measured immediately following inoculation to assess initial density (0 hr). During the
test, each culture tube is measured twice at equally spaced daily intervals (0, 24, 48, 72, 96-h)
following inoculation during 96-hr bioassays, and again on days 7 and 10 for 10-day
bioassays by inverting the capped culture tube 2-3 times and aseptically removing 1mL of
randomly sampled test culture. All culture tubes were hand agitated by inverting the capped
tube 3-4 times and were randomly repositioned (using computational randomization of
culture rack positioning) post-sampling to avoid persistent effects due to differential light
exposures. For each sample, growth was determined using spectrophotometric determination
of optical density and converted to cell density using equations derived from direct cell count
: optical density relationships established using the parent culture within 1 week prior to the
55
test initiation, as described above. The percent inhibition (%I) of algal growth for each
atrazine concentration was calculated by comparing mean growth rates for each treatment to
the respective solvent controls for the treatment, as follows:
%I = [1 − (𝑥̅
𝑘
𝑘𝑠𝑜𝑙𝑣𝑒𝑛𝑡 𝑐𝑜𝑛𝑡𝑟𝑜𝑙
)] ∙ 100
where 𝑘 is the growth rate of the pertinent sample, calculated as described above. Nominal
concentrations of pesticide were quantified with enzyme-linked immunosorbent assay
(ELISA) at the initiation and conclusion of each bioassay with enzyme-labeled paramagnetic
particles of atrazine (Method Detection Limit for atrazine = 0.046 μg/L) (SDI, Delaware,
MD, USA) (see appendix for raw ELISA data and QA/QC determination). This method has
been shown to have accuracy comparable to GC quantification of atrazine concentrations in
surface water samples (Gascón et al. 1997).
Nutrient x herbicide effects were assessed by repeating the above assay procedures in
a simultaneously run three-way bioassay format. Each assay employed identical set-up,
inoculation and chemical preparation approaches as described above, but varied in culture
media nutrient additions. To assess nutrient effects, three N:P supply ratios (selected based
on deviations from the Redfield Ratio, Redfield 1958) were prepared, 1:1, N-limited (“Low
N” = 2.0µM NO3-, 2.0µM PO4-3), 160:1, P-limited (“Low P” = 32 µM NO3-, 0.2 µM PO4-3)
and 16:1, non-limited (nutrient-replete or “Replete” = 32µM NO3-, 2.0 µM PO4-3) media
(Table 2.2). NO3-N was the only introduced nitrogen source. Each low-nutrient treatment
series contained corresponding solvent and media controls.
2.2.v Hemolytic Activity Assay Protocols
Lack of commercial standards for prymnesin toxins, and the potential involvement of
multiple components in Prymnesium toxicity, prevented assessment of toxicity by direct
chemical analyses. Therefore, toxicological bioassays were conducted using freshly prepared
56
fish erythrocytes to characterize the hemolytic potential of these Prymnesium parvum strains
under various experimental conditions.
2.2.v.a Blood Collection and Erythrocyte Preparation
Adult tilapia (Oreochromis niloticus) were purchased live from a local Asian fish
market and blood was collected from fish anesthetized with 0.02% (w/w) solution of
aminobenzoic acid ethyl ester [TricaineTM (MS-222) (Sigma-Aldrich, St. Louis, MO, USA)]
in water from their respective holding tanks by gill puncture within 3 hr of purchase.
Anaesthetized fish were placed on wet towels and 5 mL of blood were collected using a
Pravaz No. 1 needle and 10-mL syringe prefilled with 5 mL of pH-adjusted (7.2) RPMI 1640
(without phenol red), diluted 10% (v/v) with Milli-Q water to adjust for fish serum
osmolarity, and supplemented with 50 IU sodium heparin salt (Sigma-Aldrich, St. Louis,
MO, USA) as an anticoagulant (Table 2.3). Blood samples were transferred to 15-mL
centrifuge tubes and gently hand-agitated by inverting two or three times. Red bloods cells
were separated from serum by gentle centrifugation (2,100 rpm at 4°C for 5 min) and washed
by replacing supernatant with fresh media until the supernatant appeared clear following
gentle mixing. Whole washed erythrocytes were diluted 1:10 with RPMI 1640 culture media
containing 22.5 IU mL-1 sodium heparin anticoagulant. Erythrocytes (~107 cells mL-1) were
stored at 4°C for no more than 10 days; periodic microscopic and spectrophotometric
examinations showed no noticeable lytic activity over that duration. The erythrocytes were
resuspended daily until use.
2.2.v.b Preparation of Algal Extracts
Following 10-day (240-hr) OD measurements for atrazine x nutrient assays, algal
cells were immediately collected for hemolytic analysis by centrifugation using a CentraCL2 centrifuge (Needham Heights, MA, USA). All hemolytic assay preparation, extraction
and incubation procedures were conducted at 4°C in darkness to prevent photolytic effects on
measured toxin levels, as some toxins of Prymnesium parvum are light-sensitive (P. Moeller,
57
personal communication). An initial volume of 14 mL of remaining test culture was
transferred to a 15-mL centrifuge tube and spun at 3,600 rpm for 15 minutes. After the cell
pellet had accumulated, the supernatant was removed by pipetting and the remaining volume
of test culture was added and pelletized by centrifugation as described until there was
virtually complete separation of test media and biomass. The resultant pellet was then gently
aspirated and rinsed using cold, pH-adjusted, sterile assay buffer (150 mM NaCl, 3.2 mM
KCl, 1.25 mM MgSO4, 3.75 mM CaCl2 and 12.2 mM TRISMA base, pH adjusted to 7.4 with
1N HCl; Table 2.3) and re-consolidated. Next, any supernatant was again carefully removed,
and the pellet was resuspended and rinsed in cold assay buffer for a total of three rinses, to
remove any residual test medium. The cell pellet was again resuspended in 5 mL of assay
buffer, and the tube contents were transferred to a sterile 20-mL glass beaker. An additional
10 mL of assay buffer were added to the centrifuge tube to rinse any remaining cells into the
beaker, and the mixture was sonicated for 30 continuous seconds at amplitude 45 on a 20kHz sonic dismembrator (Fisher Scientific Model 550, Fisher Scientific, Pittsburgh, PA,
USA). The sonicated pellet was used immediately in hemolytic assays.
2.2.v.c Hemolytic Assays
Modified erythrocyte lysis assay (ELA, Eschbach et al. 2001, Schug et al. 2010)
techniques were adapted to work with larger sample volumes, and hemolytic incubations
were conducted in 1-mL microcentrifuge tubes instead of 96-well microtiter plates.
Hemolytic components of algal cells were extracted as described above, and equal volumes
of erythrocytes and algal extracts (300 µL each) were added to microcentrifuge tubes and
incubated for 24 hr at 4°C in darkness. Following the test incubation period, microcentrifuge
tubes were spun at 1,250 rpm for 10 minutes at 4°C in a refrigerated microcentrifuge
(Micromax RF, IEC, Needham Heights, MA, USA). Erythrocytes were also incubated in an
equal volume of assay buffer alone as a negative control. In addition, erythrocytes were
incubated in an equal volume of 20 µg saponin solution mL-1 (Sigma-Aldrich, St. Louis, MO,
USA; a known hemolytic compound prepared from the soap-bark tree, Quillaja saponaria),
which caused 100% lysis of red blood cells and served as a positive control. The entire cell
58
contents were transferred to a quartz cuvette and absorbance was measured at 414 nm. The
highest absorbance peaks in scans of initial preparations of lysed erythrocytes was measured
at that wavelength, in accord with Eschbach et al. 2001). The change in absorption at 414 nm
following exposure to sonicated cells was used to measure hemoglobin released by lysed
erythrocytes as a result of toxic activity of Prymnesium parvum, using the following
normalization equations:
[(𝑐𝑒𝑙𝑙𝑠⁄𝑚𝐿) ∙ (𝑉𝑐𝑒𝑛𝑡𝑟𝑖𝑓𝑢𝑔𝑒𝑑 ) ∙ (𝑉𝑤𝑒𝑙𝑙 )] = 𝑁
𝑠𝑎𝑚𝑝𝑙𝑒 − 𝑥 𝑛𝑒𝑔𝑎𝑡𝑖𝑣𝑒 𝑐𝑜𝑛𝑡𝑟𝑜𝑙
(
) ∙ 100 = % ℎ𝑒𝑚𝑜𝑙𝑦𝑠𝑖𝑠
𝑥𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝑐𝑜𝑛𝑡𝑟𝑜𝑙
(
% ℎ𝑒𝑚𝑜𝑙𝑦𝑠𝑖𝑠
) ∙ 106 = 𝑁𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑒𝑑 𝐿𝑦𝑡𝑖𝑐 𝑉𝑎𝑙𝑢𝑒
𝑁
where (𝑉𝑐𝑒𝑛𝑡𝑟𝑖𝑓𝑢𝑔𝑒𝑑 ) was the total volume of culture centrifuged to form assay pellets, and
(𝑉𝑤𝑒𝑙𝑙 ) (volume added to each well) was always 0.300 mL.
2.2.vi Statistical Analysis
All statistical analyses were performed using JMP software (version 12.2.0, SAS
Institute, Cary, NC, USA). Homogeneity of variance was tested using a Levene test.
Herbicide effects on growth were determined using a one-way analysis of variance
(ANOVA). Where differences were detected among treatments (p <0.05), Dunnett’s
procedure for multiple comparisons was used to determine which specific treatments differed
significantly from the controls (Zar 1999). Atrazine • nutrient • salinity • strain comparisons
were analyzed by multifactor ANOVA procedures using nutrient levels, herbicide exposure,
salinity and strain as the fixed effects, and growth rate or normalized hemolytic activity as
the dependent variable. The F statistic and associated P value for the independent variables
and the interactions between them were also reported. Point estimates of sublethal toxicity
(IC50s) were determined based on growth rates, using the Inhibition Concentration linear
interpolation (ICp) method as described by Norberg-King (1993):
59
ICp  CJ  [ M 1(1  p /100)  M J ]
(CJ  1  CJ )
( MJ  1  MJ )
wherein:
CJ
= tested concentration whose observed mean response is greater than M1 (1-p/100);
CJ+1 = tested concentration whose observed mean response is less than M1(1-p/100);
M1
= smoothed mean response for the control;
MJ
= smoothed mean response for concentration J;
MJ+1 = smoothed response for concentration J+1;
p
= percent reduction in response relative to the control response; and
ICp = estimated concentration at which there is a percent reduction from the smoothed
mean control response. The ICp is reported for the test together with the 95%
confidence interval calculated by the ICPIN.EXE program.
Results were generated using bootstrap methods to derive point estimates and confidence
intervals from no fewer than 80 resamplings (Efron 1982).
2.3 RESULTS
2.3.i Atrazine x Nutrient Bioassays
These assays were designed to test interactions among atrazine exposure, nutrient
limitation, and salinity on growth and hemolytic potential in three geographically distinct
strains of Prymnesium parvum. Atrazine concentrations were quantified at the beginning and
end of assays to ensure that the test compound was not depleted over time (see appendix for
measured values and enzyme-linked immunosorbent assay [ELISA] quality assurance/quality
control data). Generally, effects in nutrient-replete cultures showed increasing tolerance to
atrazine over time for all strains (Table 2.5). Measured atrazine half-maximal inhibition
concentrations (IC50 values) for all strains ranged from 6.7 µg L-1 (at 48-hr for the TX strain
at lower salinity) to 203.5 µg L-1 (for the TX strain at optimal salinity). There were multiple
time points at which no concentration-related effects could be discerned at the tested
concentrations (Table 2.5). The median IC50 for all strains combined was 96.9 µg L-1.
Individually, the SC strain appeared to be the most tolerant to atrazine exposure, and no IC50
could be detected in the optimal salinity (20) treatment for that isolate. The most sensitive
60
strains to atrazine exposure were the TX and NC strains (in lower salinity) (median IC50s
were 74.5 and 73.0 µg L-1, respectively). Considering median IC50 estimates, in all strains the
lower salinity (10) treatment had an overall more sensitive response to atrazine exposure, and
sensitivity decreased over a slower time period, whereas the median response in the highersalinity (20) treatment was tolerance that increased rapidly over a 48-hr period, and was
maintained for the remaining 10-day test duration (Table 2.6).
Bioassays were initially conducted for 96-hr, and preliminary results indicated that
nutrient depletion did not noticeably limit growth during that short duration. In fact, the lownutrient treatment control cultures had some of the highest growth rates measured over the
first four days of testing (Table 2.7). The highest maximum growth rate in all controls was
2.5 divisions day-1, measured 24 hr after test initiation in the salinity 20 N-limited TX culture.
Thus, test durations were extended to 10-day bioassays. Atrazine concentrations did not
appreciably degrade during that time, based on ELISA quantification assays. All subsequent
bioassays were conducted over a 10-day testing interval, with optical density observations
made every 24 hr for the first 96 hr, and thereafter on day 7 and day 10. This duration was
adequate for nutrient effects to be discerned (Fig. 2.3 and 2.4), and also allowed for closer
examination of the potential for induced atrazine tolerance dynamics. For example, the SC
strain maintained similar growth at the highest and lowest atrazine concentrations, but was
slightly stimulated by intermediate atrazine concentrations over time (Fig. 2.5). Moreover, its
growth rates were higher than those of controls in atrazine concentrations of 50 µg/L by day
10 in low salinity, low nutrient (both N and P) treatments (Dunnett’s, p < 0.05) (Figs. 2.5 and
2.6). Nutrient-limited cultures grew slowly but consistently, and atrazine had reduced
inhibitory effect on growth rates for all of the strains in either N- or P- limited treatment (Fig.
2.6 and 2.7). Thus, in some assays the low-nutrient treatments had decreased sensitivity to
the herbicide (relative to sensitivity in replete cultures) and by day 10, the low salinity, low
nutrient treatments in the SC strain had significantly higher growth rates than controls at
intermediate concentrations of herbicide (25-100 µg L-1) (Figs. 2.6 and 2.7).
61
The results of the multifactor ANOVA using either 96-hr or 10-day growth rates as
the response variable and concentration (Atra), nutrient treatment (Nutr), salinity and strain
as the predictor are shown in Table 2.8. The model R2 for the 96-hr and 10-day assay
endpoints was 0.87 and 0.98, respectively. The only insignificant effects in the model were
interaction terms in the 96-hr duration; it should be noted, however, that at 96-hr the
interaction between strain and salinity had the highest LogWorth. The interaction term of the
treatment variables ranked highest at the 10-day endpoint. These results otherwise mirrored
the initial observations; the effect of atrazine concentration was strongest at the 96-hr
endpoint, and nutrient treatment had an increased effect at the 10-day endpoint. An analysis
of variable importance using independent resamples ranked the main effects for the 96-hr test
as follows: Atrazine > Strain ≈ Salinity > Nutrients. The main effects for the 10-day assay
were ranked as: Nutrients >> Atrazine ≈ Strain > Salinity (see appendix for tables of variable
importance).
Because strain had a relatively large interaction effect in both assay durations, the
ANOVA analyses were re-run strain-by-strain as 3-way ANOVAs in an attempt to remove
that effect (Table 2.9). The 96-hr model for the TX strain at high salinity had an R2 of 0.93,
and all other models produced even higher correlations between actual and predicted
variables (R2 > 0.98). The whole-model variable importance analysis for the TX strain at 96
hr was ranked as: Atrazine >> Nutrients >> Salinity. The analysis at 10 days was ranked as:
Nutrients >> Salinity > Atrazine.
The main effects for the NC strain at 96 hr were ranked as: Atrazine >> Salinity >
Nutrients; at 10 days the main effects were ranked as: Nutrients > Atrazine >> Salinity. The
main effects for the SC strain at 96 hr were Salinity > Nutrients >> Atrazine; the main effects
at 10 days were ranked as: Nutrients >> Atrazine ≈ Salinity. The two top ranking main
effects were nutrient treatment effects in all 3-way models, and only for the NC strain at 10
days (atrazine * salinity) and for the SC strain at 96 hr (nutrients * salinity) was an
interaction term ranked higher than any main variable treatment effects.
62
Concentration response curves, showing the effects of nutrient limitation on growth
inhibition by strain and salinity, illustrated the decreased influence of media nutritional
availability over time (Fig. 2.6). Several assays showed either similar levels of growth
inhibition or slightly stimulated responses to atrazine concentrations at 10 days. At that time,
as well, atrazine exposure only had a strong impact on the SC strain at low salinity, and the
NC strain at optimal salinity. Most of the other responses had become muted along the
herbicide concentration gradient at 10 days. Concentration response curves that highlighted
the comparison between atrazine effects at low and optimal salinities are shown in Fig. 2.7.
Many of these relationships showed similar trends, expected for increased sensitivity to
intermediate doses by the NC strain in the optimal-salinity, low-P treatment at 96 hr.
Salinity-based generalizations could not be made across the three P. parvum strains.
2.3.ii Erythrocyte Lysis Assays
Modified erythrocyte lysis assays (ELAs) were conducted to examine the potential
for experimental stressors to modulate toxicity in static P. parvum cultures. The relative
effect of the nutrient treatments on hemolytic activity in control cultures is shown in Fig. 2.8.
Multiple comparison with Dunnett’s procedures indicated significantly increased hemolytic
activity in nutrient-limited treatments for all strains and salinities when compared to
hemolytic activity in nutrient-replete media (p < 0.05) (Fig. 2.8). Results of 3-way ANOVAs
using normalized hemolytic activity as the dependent variable and nutrient treatment,
atrazine concentration and strain origin as the fixed effects are shown in Table 2.10. The
main effects of the ANOVA tables are presented in order of decreasing LogWorth. Threeway ANOVA ranking of variable importance of main effects indicated a rank of: Nutrients >
Strain >> Atrazine (Total Effects = 0.64, 0.50 and 0.07, respectively; Table 2.10). The ELAs
were only conducted on the SC strain at low salinity (due to time constraints) so 4-way
ANOVA procedures were conducted to model the 4-way interaction between nutrient
treatment, atrazine concentration, strain origin and salinity by removing the SC strain from
the analysis (Table 2.11). The 4-way ANOVA incorporating salinity showed that the relative
variable importance was ranked as Nutrients >> Strain > Salinity > Atrazine (Total effects =
63
0.80, 0.36, 0.25, and 0.16, respectively; Table 2.11). Both ELA ANOVA models had high R2
values when comparing observed to predicted values, but the 3-way model without the
salinity term had both a higher R2 (0.96) and a lower AIC value, indicating that it was better
for predicting potential hemolytic activity. The strains were also analyzed separately to try to
examine the interaction of Herbicide x Nutrients x Salinity while avoiding any confounding
influences of strain variability on the model (Table 2.12, also ordered in decreasing
LogWorth). Atrazine exposure had a stronger influence on predicted hemolytic activity than
nutrient limitation for the SC strain (salinity was not included as a model variable). In
contrast, nutrient limitation was the strongest predictor of hemolytic activity for both the TX
and NC strains. Atrazine exposure had a stronger effect than salinity for the TX strain, but for
the NC strain, hemolytic activity was largely a function of nutrient availability, salinity level,
and their interaction with atrazine exposures. The model for the TX strain also indicated that,
for that strain, herbicide exposure was a stronger predictor of hemolytic activity than salinity.
The relative production of hemolytic substances (normalized to extracted cell density)
by the three strains at different salinities in a range of atrazine concentrations are shown in
Fig. 2.9. The nutrient-limited treatments were all significantly more toxic than the nutrientreplete treatments, and this effect generally increased with increasing atrazine concentration.
With exception of the TX strain at optimal salinity, the strains all were more hemolytic in
low-N treatments. The NC and SC strains exhibited a general trend toward increasing
toxicity with increasing atrazine exposure in low- nutrient treatments. The SC strain in the
low-salinity, low-N, high-atrazine treatment was the most hemolytic culture tested.
There was a log-normal relationship between herbicide exposure and hemolytic
activity (Figs. 2.10 [salinity 20] and 2.11 [salinity 10]). While the majority of hemolytic
responses were more strongly predicted by nutrient limitation, all three strains showed a
relationship between herbicide exposure and hemolytic activity. In nutrient-replete
treatments, cultures had a generalized response of decreasing hemolytic activity with
increasing atrazine concentration, and this relationship was moderately significant for the TX
and SC strains at low salinity (r2 = 0.69 and 0.65, respectively), and for the SC strain at high
64
salinity as well (r2 = 0.87). Yet, in all low-nutrient treatments the SC strain showed
increasing hemolytic activity with increasing atrazine concentration, and increasing
hemolytic activity with increasing atrazine also was exhibited by the NC and TX strains in
the low-N treatments (salinity 20 and 10, respectively). There were no significant
relationships between hemolytic activity and atrazine concentration for the NC strain at low
salinity (data not shown).
The relationship between hemolytic activity and growth rates (in divisions day-1) was
also explored [Figs. 2.12 (salinity 10) and 2.13 (salinity 20)], and was inconstant among
different strains and treatments. R2 values >0.65 were observed at salinity 10 for the TX
strain in nutrient-replete and low-N treatments. In nutrient-replete treatments, growth rates
were positively related to toxicity, but in low-N treatments the highest hemolytic activity
corresponded to the lowest growth rates. The NC strain in low-N media and the SC strain in
nutrient-replete media responded similarly. Growth rates were less well associated with
hemolytic activity in optimal-salinity cultures, and only the NC strain under nutrient-replete
conditions showed a significant relationship between growth rate and hemolytic activity at
salinity 20 (R2 = 0.77).
2.4 DISCUSSION
This study examined interactions between co-occurring growth stressors in a stresstolerant HAB species, and it showed the following: (1) Prymnesium parvum growth
responses to atrazine exposure were influenced by salinity, nutrient status and strain, (2)
sensitivity to herbicide exposure decreased over short time frames (less than 10 days) and (3)
the relative production of hemolytic compounds could be enhanced in low nutrient systems
by atrazine exposures. Competitive ability previously has been shown to be influenced by
toxicants, and contaminant-induced impacts on R* values under resource limitation can shift
a superior competitor for the limiting resource to an inferior one (Tilman 1982, Landis and
Ho 1995, Interlandi 2002). Here, interactions were explored which might shift this
established dynamic in the opposite direction, wherein an inferior competitor could become a
65
superior one in a toxicant-contaminated, nutrient-limited system. The slowly growing P.
parvum can become dominant at sub-optimal growth conditions, and forms nearly
monospecific blooms during the winter season in low-salinity rivers and reservoirs in Texas,
with devastating impacts to fisheries and recreational economies (Southard et al. 2010,
VanLandeghem et al. 2013). In September 2009, P. parvum was the causative agent of a
~48-km- (30-mile)-long bloom in an Appalachian stream in West Virginia, USA, involving
sudden death of thousands of fish and salamanders (Renner 2009). The most diverse
population of highly endangered freshwater mussels in the Monongahela River basin was
also decimated (Renner 2009). The diverse toxins of P. parvum have been tentatively
implicated (by geographic proximity) in the death of birds and other animals (MoustakaGouni et al. 2004), but are not thought to be toxic to humans (Manning and La Claire 2010).
The ecological persistence and potential for harm is understudied in this group of
compounds; indeed, some (many?) toxins from P. parvum have not yet been structurally
characterized (Roelke et al. 2016). The environmental factors influencing P. parvum bloom
initiation, persistence and toxicity in the field are complex, and likely involve a dynamic
interplay among multiple factors including salinity (Roelke et al. 2011, Israël et al. 2014),
hydrology (Roelke et al. 2010), nutrient status (Errera et al. 2008, Patiño et al. 2014), pH,
water hardness, and total dissolved solids (VanLandeghem et al. 2012, VanLandeghem et al.
2015b; also see Brooks et al. 2011, Granéli et al. 2012, Roelke et al. 2016, and references
therein). Confounding the ability to model and predict the distribution of P. parvum impacts
at local scales is the fact that many of these responses appear to vary from basin to basin
(Israël et al. 2014, VanLandeghem et al. 2015a).
The results presented here show that growth and toxicity are highly strain-dependent.
The growth response to optimal and lowered salinity in the short term was strongly
influenced by herbicide exposure and strain, and by the interactions between these factors
and salinity (p < 0.05) (Fig. 2.7, Table 2.8). These effects were overcome, however, by the
influence of nutrient limitation. By the end of the 10-day testing period, all strains of P.
parvum appeared to exhibit some ‘stress-stimulation’ effect, whereby populations in low-
66
nutrient treatments had increased growth relative to nutrient-rich controls when exposed to
intermediate levels of atrazine. Both the NC and SC strains were most stimulated in lowsalinity media, whereas the Texas strain increased growth at salinity 20 (its optimal salinity
for growth) in low nutrient x atrazine treatments. The Texas strain was also more resilient to
atrazine stress in salinity 20 media under nutrient limitation, but it was the only strain tested
that became more sensitive to growth-inhibiting effects of atrazine over time in nutrientreplete media at salinity 20.
These observations make sense if considered as a potential function of increased
predatory activity by P. parvum on bacteria. Bacterial consumption was not tracked, but the
studies were performed on unialgal batch cultures containing bacteria. The potential for
mixotrophic activity cannot be excluded since P. parvum is known to consume bacteria (e.g.,
see Martin-Cereceda et al. 2003; Carvalho and Granéli 2010). If bacterivory in Prymnesium
parvum is increased by reduced nutrient availability and light exposure as has been
previously suggested (Carvalho and Granéli 2010, Lundgren et al. 2016), that might suggest
a mechanism by which ecologically stressed, toxin-producing strains could have an
advantage over strains that produce little or no toxins and do not engage in bacterivory
(Glibert et al. 2009, Brutemark and Granéli 2011). The TX strain in the optimal-salinity,
nutrient-replete treatment may have taken more time to become physiologically stressed from
atrazine exposure, without a concomitant increase in bacterivory – which would be supported
by the observation that the TX strain in salinity 20 nutrient-replete culture had negligible
production of hemolytic substances (Figs. 2.8 and 2.10). Moreover, the strain and treatments
which exhibited, overall, the highest tolerance to atrazine had the highest production of
hemolytic substances (SC strain in low-nutrient treatments). Only the SC strain showed a
strong correlation between atrazine exposure and hemolytic activity, but the direction of this
relationship was reversed between replete and low nutrient treatments. The low-salinity SC
strain also had the steepest dose-response curve, which can indicate strong affinity of atrazine
for the mode-of-action receptor – of interest considering the large hemolytic response that
was observed in that culture. A few cultures showed a strong relationship between growth
67
rate and toxicity, but the majority did not. Moreover, there was a positive linear correlation
between growth rate and toxicity for the three strains (NC – salinity 20; TX, SC – salinity 10)
in nutrient-replete treatments and a negative linear correlation for the TX and NC strains in
low-nitrogen treatments (salinity 10).
While this work supports the reports that few generalizations can be made across
phytoplankton strains (Burkholder and Glibert 2006, and references therein), it is worthwhile
to examine the differences shown by these three P. parvum strains in response to
environmental stressors. The relative ranking of hemolytic activity in nutrient-replete, lowsalinity media was TX ≈ NC < SC, whereas in nutrient-replete, optimal-salinity media, NC >
TX. In low-N, low-salinity cultures, hemolytic activity was ordered TX < NC < SC, whereas
in low-N, optimal-salinity cultures, TX ≈ NC. In low-P, optimal-salinity cultures, the TX
strain showed the highest hemolytic activity, but in low-salinity, low-P conditions, the NC
and SC strains were more toxic than the TX strain. These data indicate that hemolytic
responses are influenced by a complex interaction of environmental factors. Efforts to
reliably predict the toxic potential of P. parvum blooms may involve development of an
intricate “mandala”-like model of niche factors (e.g., see Glibert 2016).
Mixotrophy is an under-appreciated mechanism in marine phytoplankton ecological
models (Tittel et al. 2003, Burkholder et al. 2008, Moore 2013, Caron 2016). Prymnesium
parvum has been referred to as a “predatory phytoflagellate” (Caron 2016), a name which is
appropriately descriptive depending on resource quality in the habitat. The prymnesin toxins
are high molecular weight compounds rich in carbon. Production of such energetically
expensive compounds would be disadvantageous unless there was an equitable return in
benefits – likely in the form of enhanced access to resources. Molecular analyses of P.
parvum have revealed that a gene which encodes an ABC-type phosphate transport system is
one of the most highly expressed genes in the P. parvum transcriptome (La Claire II 2006),
and is preferentially upregulated under P limitation (Beszteri et al. 2012). These putative
phosphate transporters include enzymes which either hydrolyze extracellular organically
bound phosphate or relocate intracellular phosphate, and P. parvum is capable of utilizing a
68
variety of organic and inorganic P sources in dissolved or particulate forms (Manning and La
Claire 2010). It also has cell-surface L-amino oxidases which oxidize amino acids and
primary amines, producing ammonium compounds which P. parvum can then consume
(Palenik and Morel 1991). And, P. parvum is capable of assimilating dissolved organic
nitrogen from sewage as well (Lindehoff et al. 2009). Considering its extracellular toxins
(Blossom et al. 2014) and its assimilation of bacterial carbon stores (Lundgren et al. 2016),
these features could indicate a mechanism by which P. parvum is uniquely equipped to
succeed in low-nutrient environments. Toxin production facilitates prey capture by P.
parvum, and toxicity increases when nutrients or light become limited (Dafni et al. 1972;
Uronen et al. 2005; Carvalho and Granéli 2010, Lundgren et al. 2016; see also Graneli et al.
2012, and references within). As atrazine is a reversible inhibitor of photosystem II electron
transport proteins, it also reasonable that phagotrophy would be enhanced under high levels
of herbicide exposure. This research has shown that atrazine sensitivity in P. parvum is
influenced by nutrient availability and salinity, but the magnitude of the effect is straindependent, which is indicative of a genetically-dictated response variability.
These data also indicate a potential for induced tolerance of P. parvum to atrazine
over time. The IC50 estimates of growth inhibition were consistently lower at the beginning
of the tests and increased by the end of the tests. A possible mechanism to explain these
observations is that the populations gradually increased production of hemolytic compounds
and increased reliance on phagotrophic assimilation of carbon. It is in line with
considerations of dynamic energy budget theory to suggest that carbon storage products (long
chain cyclic polyethers, e.g., prymnesins) would only be excreted if reliance on
photosynthetic carbon assimilation was reduced. In contrast, if nutrients and light are not
growth-limiting factors, these products should be increasingly retained, and reliance on
autotrophic carbon storage should again be favored (Kooijman 1993, Nisbet et al. 2012).
Here, P. parvum increased production of toxic substances under environmental stress, as has
been reported by other researchers (Nygaard and Tobiesen 1993, Barreiro et al. 2005, Uronen
et al. 2005, Errera et al. 2008). These results also support the findings of Yates and Rogers
69
(2011), who showed that P. parvum was less susceptible to atrazine impacts than cooccurring phytoplankton, and that P. parvum is becoming dominant in atrazine-contaminated
waters of Texas.
This microalga is regarded as among the deadliest harmful algae to fish and shellfish
(Moestrup 1994). While other studies have examined the linkages between eutrophication
and the “global HAB epidemic” (Smayda 1990, Burkholder 2001, Glibert and Burkholder
2006, Anderson et al. 2008), less work has focused on the dynamics of highly competitive
microbial communities resulting from the selective macro- or micro-nutrient depletion that is
caused by rapidly growing blooms – which, themselves, are a natural consequence of nutrient
over-enrichment (Heisler et al. 2008, Glibert et al. 2011, Burkholder and Glibert 2013).
Other researchers have described P. parvum as “physiologically pre-adapted to invade Plimited brackish waters” (Baker et al. 2009) due to its broad salinity tolerance and its high
competitive ability under P limitation. The latter characteristic results from high expression
of putative phosphate transporter genes in P. parvum, and its increased consumption of
relatively P-rich bacteria in P-limited waters (Carvalho and Granéli 2010, Lundgren et al.
2016; note that bacteria can sequester high levels of P relative to background concentrations
in P-limited systems – Bratbak and Thingstad 1985, Andersen et al. 1986). Thus, increased
expansion of P. parvum blooms in eutrophic freshwaters is expected, especially waters
increasingly threatened by salinization (Williams 2001, Kaushal et al. 2005) and coastal
organic nitrogen pollution leading to N:P stoichiometric imbalance (Glibert et al. 2006,
Burkholder et al. 2007, Bricker et al. 2008, Burkholder and Glibert 2013, Glibert et al.
2014). The findings from this research suggest that increasing chemical contamination of
inland surface waters is helping to promote ecosystem-disruptive, strongly mixotrophic algal
blooms.
70
2.5 FIGURES AND TABLES
haptonema
Figure 2.1. Light micrograph Prymnesium parvum (Texas strain), taken using an Olympus AX70
research microscope with 60x water-immersion lens. Scale bar = 10 µm. Arrow indicates the
characteristic haptonema, structurally intact following preservation techniques.
71
Table 2.1. Collection locations in the U.S. for the tested Prymnesium culture isolates.
Collection ID
Collection Site
Collection
Date
Date
Acquired
P. parvum (TX)
UTEX LB
2797
Texas Colorado River at
US Hwy 183
12/01
04/30/2009
P. parvum (SC)
UTEX LB
2827
Oyster Rake Pond near
Charleston, South Carolina
05/01
06/18/2009
P. parvum (NC)
UNC-W
Elizabeth City, NC
05/02
01/16/2013
Species
72
Table 2.2. Media recipes used in Prymnesium testing and culture.
f/2 Macronutrient concentrations
Stock solution
Volume added to final
medium
Final concentration
in f/2 media (µM)
NaNO3
NaH2PO4
2.72 g L-1 H2O
0.24 g L-1 H2O
1 mL
1 mL
32.00
2.00
Low-N
NaNO3
NaH2PO4
0.17 g L-1 H2O
0.24 g L-1 H2O
1 mL
1 mL
2.00
2.00
Low-P
NaNO3
NaH2PO4
2.72 g L-1 H2O
0.02 g L-1 H2O
1 mL
1 mL
32.00
0.20
Nutrient-Replete
f/2 Trace Element Solution
Primary stock solution
Na2EDTA · 2H2O
FeCl3 · 6H2O
ZnSO4 · 7H2O
CoCl2 · 6H2O
CuSO4 · 5H2O
Na2MoO4 · 2H2O
MnCl2 · 4H2O
----22.00 g L-1 H2O
10.00 g L-1 H2O
9.80 g L-1 H2O
6.30 g L-1 H2O
180.00 g L-1 H2O
Quantity added to final
medium
4.36 g
3.15 g
1 mL
1 mL
1 mL
1 mL
1 mL
Molar concentration in
final medium (M)
1.17 X 10-5
1.17 X 10-5
7.65 x 10-8
4.20 x 10-8
3.93 x 10-8
2.60 x 10-8
9.10 x 10-7
Quantity added to
final medium
200 mg
1 mL
1 mL
Molar concentration in
final medium (M)
2.96 x 10-7
2.05 X 10-9
3.69 x 10-10
f/2 Vitamin Solution
thiamine · HCL (Vitamin B)
biotin (Vitamin H)
cyanocobalamin (Vitamin B12)
Primary stock
solution
--0.10 g L-1 H2O
1.00 g L-1 H2O
Table 2.3. Erythrocyte lysis assay buffer preparation. *
NaCl
KCl
MgSO4
CaCl2
TRISMA base
Mass (g L-1) Concentration (mM)
8.7700
150
0.2386
3.2
0.3081
1.25
0.4162
3.75
1.4779
12.2
*pH adjusted to 7.4 at 4°C with 1N HCl
73
Table 2.4. Blood storage medium preparation. *
RPMI-1640**
NaHCO3 L-1
Sodium Heparin
Mass (g L-1)
10.40 g L-1
2.00 g L-1
0.05 mg mL-1
*Diluted 10% (v/v) with Milli-Q
** RPMI-1640 *with L-glutamine *without phenol red *without NaHCO3
74
Figure 2.2. Example of the relationship between optical density (OD) and Prymnesium parvum
population density (PD) Similar relationships were determined for all strains prior to use of
absorbance measurement methods in toxicity endpoint estimates.
75
Table 2.5. Changes in atrazine effects in nutrient-replete media over time (µg/L) and maximum growth
rate kmax (units, divisions day-1) for all strains tested. BD = beyond detection, ND = not determined.
Strain
(salinity)
TX (20)
Test date
30-Jul-12
TX (20)
28-Jan-13
TX (10)
20-Aug-12
TX (10)
4-Mar-13
NC (20)
3-May-13
NC (10)
25-Mar-13
SC (20)
7-Jun-13
SC (10)
21-Jun-13
Hour
24
72
96
24
48
72
96
168
240
24
48
72
96
48
72
96
168
240
24
72
96
168
240
24
48
72
96
168
240
72
96
168
240
72
96
168
240
IC50 (SD)
96.9 (30.6)
151.2 (4.4)
142.2 (4.7)
>200
169.2 (BD)
131.2 (20.1)
119.0 (23.9)
>200
>200
19.5 (4.6)
6.7 (0.2)
9.3 (0.3)
80.2 (6.3)
>200
78.2 (3.3)
92.3 (3.3)
177.9 (5.9)
189.6 (BD)
20.7 (2.7)
70.6 (8.7)
118.3 (8.0)
104.4 (10.2)
155.4 (5.4)
ND
43.4 (18.3)
88.3 (BD)
73.0 (8.7)
>200
>200
>200
>200
>200
>200
67.3 (22.3)
88.2 (9.9)
148.0 (4.1)
161.0 (2.78)
IC10 (SD)
20.5 (9.8)
40.3 (4.4)
39.9 (5.5)
23.4 (13.0)
41.9 (17.8)
24.4 (6.8)
23.0 (14.5)
50.1 (21.8)
92.3 (29.6)
10.4 (4.7)
1.3 (0.1)
1.9 (0.1)
4.9 (1.0)
24.2 (7.2)
17.4 (4.7)
14.9 (2.9)
38.7 (19.6)
63.8 (14.2)
14.1 (0.4)
25.6 (13.1)
44.8 (13.6)
28.3 (14.5)
54.2 (14.2)
ND
9.0 (6.5)
18.7 (8.2)
19.2 (7.4)
58.9 (13.5)
120.2(11.7)
123.3 (5.2)
99.7 (8.6)
25.3 (8.8)
42.0 (8.9)
15.1 (3.3)
22.6 (22.6)
69.0 (2.9)
81.3 (8.3)
kmax
(controls)
0.6
0.8
0.8
1.4
0.7
0.9
0.8
0.7
0.6
0.5
0.7
0.8
0.6
1.3
0.9
0.9
0.6
0.5
0.5
0.5
0.6
0.6
0.6
2.4
1.4
1.0
0.9
0.7
0.5
1.2
1.1
0.7
0.5
0.6
0.6
0.5
0.4
76
Table 2.6. Median half-maximal activity estimates (units, µg L-1) over time for all strains in nutrientreplete media.
Salinity
10
20
Strain
NC
SC
TX
NC
SC
TX
IC50 =
73.0
118.1
74.5
100.8
BD
138.9
IC10 =
53.7
44.9
12.9
31.1
69.4
36.0
Salinity Hour
10 24
48
72
96
168
240
20 24
48
72
96
168
IC50 =
19.5
25.0
72.7
84.2
162.9
175.3
58.8
169.2
131.2
119.0
153.9
77
Table 2.7. Heat map of kmax values for the growth response of P. parvum (three strains – TX, NC, and
SC) in different treatments over time (in hours) (Nutr = Nutrient treatment).
Strain Salinity Test Date
TX
20
30-Jul-12
28-Jan-13
10
20-Aug-12
4-Mar-13
NC
SC
20
3-May-13
10
25-Mar-13
20
7-Jun-13
10
21-Jun-13
Nutr
R
Low N
Low P
R
Low N
Low P
R
Low N
Low P
R
Low N
Low P
R
Low N
Low P
R
Low N
Low P
R
Low N
Low P
R
Low N
Low P
24
0.6
1.0
0.8
1.4
2.5
2.4
0.5
0.7
0.8
48
0.7
1.5
1.5
0.7
0.9
0.9
1.3
1.3
1.3
0.5
0.9
1.4
2.4 1.4
1.1 1.1
2.1 1.3
1.1
72
0.8
0.8
0.7
0.9
1.0
1.1
0.8
0.8
0.8
0.9
0.7
1.0
0.5
0.8
0.8
1.0
0.9
1.2
1.2
0.6
0.8
0.6
0.4
0.7
96
0.8
0.6
0.6
0.8
0.7
0.8
0.6
0.6
0.6
0.9
0.5
0.8
0.6
0.7
0.8
0.9
0.7
1.0
1.1
0.5
0.8
0.6
0.4
0.5
168 240
0.7
0.4
0.5
0.6
0.3
0.4
0.6
0.2
0.4
0.6
0.5
0.5
0.7
0.4
0.6
0.7
0.3
0.5
0.5
0.2
0.3
0.5
0.1
0.2
0.6
0.3
0.4
0.5
0.3
0.4
0.5
0.2
0.3
0.4
0.1
0.2
78
Figure 2.3. Cell densities of control cultures for the Texas strain of P. parvum at salinity 20. Growth rates in low-nutrient versus
nutrient-replete treatments were comparable until after 96 hr.
79
*
*
*
*
*
*
*
Figure 2.4. Growth rates of the NC strain of P. parvum at different salinities in all treatments at
day 10. Growth in nutrient-replete treatments was more sensitive to atrazine effects than growth in
either nutrient-limited treatment.
80
Figure 2.5. Atrazine inhibition curves in salinity 10 for the South Carolina strain of P. parvum over
time. At timepoints > 168 (day 7) growth was slightly stimulated at intermediate atrazine
concentrations (25-50 µg/L) in comparison to growth in controls with no atrazine, whereas %
inhibition at the highest atrazine level (200 µg/L) did not vary significantly over time. Data are
given as means + 1 SE.
81
Figure 2.6. Concentration inhibition curves for all atrazine assays over time. Colors represent
different nutrient regime treatments. Data are given as means + 1 SE.
82
Figure 2.7. Concentration inhibition curves for all atrazine assays over time point. Colors represent different
levels of salinity. Data are given as means + 1 SE.
83
Table 2.8. Results of 4-way ANOVA testing growth rates of P. parvum (all three strains) at 96-hr or
10-day endpoints as dependent variables. Fixed effects are atrazine concentration (Atra), nutrient
regime (Nutr), geographic origin (strain) and salinity. DF ≡ degrees of freedom. Significance is
indicated by P values < 0.0088 to < 0.001 (bold), or < 0.0126 to < 0.0292 (bold/italics).
96-Hr Growth Effects
Parameter
Day-10 Growth Effects
DF F Ratio P Value Parameter
2
257.22 < 0.0001 Nutr
DF F Ratio
Strain*Salinity
Atra
5
92.58
2
Strain
2
176.97 < 0.0001 Salinity
1
Nutr*Salinity
2
45.49
< 0.0001
Atra
5
1047.05 0.0088
108.36 < 0.0001
Nutr*Strain*Salinity
4
23.94
< 0.0001
Strain*Salinity
2
209.49
10 37.02
4 33.76
< 0.0001
< 0.0001
< 0.0001
Strain
Nutr
2
41.86
< 0.0001
Atra*Nutr
Salinity
1
62.93
< 0.0001
Nutr*Strain
2
P Value
1486.45 0.0009
618.06 < 0.0001
Atra*Nutr
10
7.75
Atra*Nutr*Salinity
10
4.04
Nutr*Strain
4
4.12
< 0.0001 Atra*Strain*Salinity
10 15.21
< 0.0001 Atra*Nutr*Strain*Salinity 20 4.12
0.0029
Nutr*Strain*Salinity
4 9.33
Atra*Strain
10
2.48
0.0072
Atra*Nutr*Strain
Atra*Nutr*Strain*Salinity 20
Atra*Strain*Salinity
10
1.89
0.0126
Atra*Salinity
1.07
0.3847
Atra*Nutr*Salinity
Atra*Salinity
5
1.01
0.4108
Atra*Nutr*Strain
20
0.87
0.6291
< 0.0001
< 0.0001
< 0.0001
< 0.0001
20 3.30
5 5.58
< 0.0001
0.0292
Nutr*Salinity
10 3.14
2 4.84
Atra*Strain
10 2.06
< 0.0001
< 0.0001
< 0.0001
84
Table 2.9. Results of a 3-way ANOVAs for each of the three P. parvum strains (TX, NC, SC) testing
growth rates at 96-hr or 10-day endpoints as dependent variables. Fixed effects are atrazine
concentration (Atra), nutrient regime (Nutr), and salinity. Significance is indicated by P values < 0.0067
to < 0.0001 (bold), or < 0.0117 (bold/italics).
TX
End Parameter
NC
SC
DF
F Ratio
P Value
F Ratio P Value
F Ratio P Value
96
Atra
5
72.04
< 0.0001
64.34
< 0.0001
94.55
hr
Nutr
2
74.59
< 0.0001
35.91
< 0.0001
74.67
< 0.0001
255.19
< 0.0001
1
31.91
Atra*Nutr
10
6.06
< 0.0001
7.53
< 0.0001
< 0.0001
2258.92 < 0.0001
11.48
< 0.0001
Atra*Salinity
5
3.52
0.0067
0.76
0.5837
5.70
0.0002
Nutr*Salinity
2
34.97
< 0.0001
23.17
< 0.0001
377.98
< 0.0001
Atra*Nutr*Salinity
10
0.99
0.4580
7.59
< 0.0001
8.22
< 0.0001
Atra
5
38.87
< 0.0001
25.04
< 0.0001
64.72
< 0.0001
< 0.0001
243.10
< 0.0001
Salinity
10
< 0.0001
days Nutr
2
835.93
< 0.0001
670.13
Salinity
1
9.86
0.0025
630.87
< 0.0001
Atra*Nutr
10
1121.16 < 0.0001
14.43
< 0.0001
5.29
< 0.0001
36.55
< 0.0001
Atra*Salinity
5
3.19
0.0117
18.50
< 0.0001
10.41
< 0.0001
Nutr*Salinity
2
0.74
0.4811
0.97
0.3832
35.08
< 0.0001
Atra*Nutr*Salinity
10
1.18
0.3217
3.82
0.0004
7.84
< 0.0001
85
Figure 2.8. Differences in relative production of hemolytic compounds by P. parvum strains (TX,
NC) in control treatments at both salinity levels (10, 20). Data are given as means + 1 SE.
86
Table 2.10. Results from a 3-way ANOVA testing normalized hemolytic activity as the dependent
variable and atrazine concentration (Atra), nutrient regime (Nutr), and geographic origin (Strain) as
the fixed effects. The table is ordered in decreasing LogWorth. All strains were tested at salinity 10.
Significance is indicated by P values < 0.0001 (bold), or 0.0256 to 0.0281 (bold/italics).
Parameter
LogWorth DF
Nutr
59.393
2
Strain
53.275
2
Nutr*Strain
33.644
4
Atra*Nutr
11.425 10
Atra*Nutr*Strain
6.454 20
Atra*Strain
1.592 10
Atra
1.551
5
F Ratio
625.6194
469.569
94.5214
10.4307
4.3046
2.1609
2.6211
P Value
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
0.0256
0.0281
Whole Model
R2 0.962779
AICc 1050.116
Analysis of variable importance for 3-way ANOVA, random independent resampling
Column
Nutr
Strain
Atra
Main Effect
Main Effect
SE
Total Effect
Total Effect
SE
0.458
0.345
0.026
0.009
0.009
0.004
0.641
0.503
0.074
0.01
0.009
0.004
87
Table 2.11. Results from a 4-way ANOVA testing hemolytic effects at all salinities (NC and TX
strains only) against atrazine concentration (Atra), nutrient regime (Nutr), geographic origin (Strain),
and salinity. Significance is indicated by P values < 0.0001 to 0.0007 (bold), or 0.0435 (bold/italics).
Hemolytic activity in the South Carolina strain was only tested at salinity 10, so that strain was
excluded from this analysis.
Parameter
LogWorth DF
F Ratio
Nutr
69.258
2 587.5598
Strain
32.29
1 246.6901
Nutr*Salinity*Strain
29.161
2 110.9587
Salinity*Strain
16.669
1 93.8258
Nutr*Strain
14.53
2 42.5887
Atra*Nutr*Salinity*Strain
8.815 10
7.4987
Atra*Nutr*Salinity
8.766 10
7.4604
Atra*Nutr
7.718 10
6.6541
Atra
6.768
5
9.041
Nutr*Salinity
6.496
2
16.626
Atra*Strain
3.734
5
5.2518
Atra*Nutr*Strain
3.334 10
3.4321
Atra*Salinity
3.149
5
4.5419
Salinity
1.362
1
4.1501
Atra*Salinity*Strain
0.453
5
1.1195
P Value
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
0.0002
0.0005
0.0007
0.0435
0.3527
Whole Model
R2 0.938837
AICc 1355.593
Analysis of variable importance for hemolytic activity - 4-way ANOVA, random independent
resampling
Column
Main Effect
Nutr
Strain
Salinity
Atra
0.548
0.112
0.036
0.047
Main Effect
SE
0.008
0.009
0.006
0.006
Total Effect Total Effect
SE
0.8
0.01
0.358
0.007
0.245
0.006
0.158
0.005
88
Table 2.12. Results from a 3-way ANOVA testing normalized hemolytic activity in individual
strains as the dependent variable and atrazine concentration (Atra), nutrient regime (Nutr), and
salinity as the fixed effects. Significance is indicated by P values < 0.0007 (bold), or 0.0315
(bold/italics).
Source
TX Nutr
Nutr*Salinity
Atra
Salinity
Atra*Nutr*Salinity
Atra*Salinity
Atra*Nutr
NC Nutr
Salinity
Atra*Nutr*Salinity
Nutr*Salinity
Atra*Nutr
Atra
Atra*Salinity
SC Atra
Atra*Nutr
Nutr
LogWorth
47.484
38.617
17.481
15.146
12.606
9.82
6.603
32.497
7.695
6.287
4.536
4.512
1.502
0.231
3.142
11.588
30.512
DF
2
2
5
1
10
5
10
2
1
10
2
10
5
5
5
10
2
F Ratio
714.4232
389.5878
35.722
106.8492
13.9804
15.937
6.7638
251.7324
39.8104
6.4484
12.117
4.7702
2.6139
0.7519
5.5128
21.5613
874.0726
P Values
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
0.0315
0.5874
0.0007
<.0001
<.0001
89
Nutrient-Replete IC50 = 189.6
Low N IC50 = BD
Nutrient-Replete IC50 = BD
Nutrient-Replete IC50 = BD
Low N IC50 = BD
Nutrient-Replete IC50 = BD
Replete IC50 = 155.8
Low N IC50 = 165.7
Replete IC50 = 176.3
Replete IC50 = 129.5
Low N IC50 = BD
Replete IC50 = BD
SC (10 psu)
Replete IC50 = 161.3
Low N IC50 = BD
Replete IC50 = BD
Figure 2.9. Relative production of hemolytic activity plotted against atrazine concentration for all
tested strains, showing line-of-fit for non-linear relationships and corresponding R2, and confidence
intervals as colored zones corresponding to different nutrient treatments (half maximal growth effect
concentrations for atrazine are provided as insets with each assay display). Data are given as means
+ 1 SE; BD ≡ below detection.
90
Figure 2.10. Log-normal relationships between hemolytic activity and atrazine concentration, in
salinity 20 f/2 media: Panels (a, b) are from in nutrient-replete media; panels (c, d) are from lownitrogen media; and panels (e-f) are from low-phosphorus media. Panels (a, c, e) are for the TX
strain, and panels (b, d, f) are for the NC strain.
91
Figure 2.11. Log-normal relationships between hemolytic activity and atrazine concentration, in
salinity 10 f/2 media. Panels (a-b) are from nutrient-replete media; panels (c-d) are from lownitrogen media; and panels (e-f) are from low-phosphorus media. Panels (a, c, e) are for the TX
strain, and panels (b, d, f) are for the SC strain.
92
Figure 2.12. Linear relationships between growth rates (divisions day-1) and hemolytic activity in P. parvum strains tested in salinity 10 f/2 media.
Panels (a-c) were in nutrient-replete media; panels (d-f) were in low-nitrogen media, and panels (g-i) were in low-phosphorus media. Panels (a, d and
g) are for TX strain; panels (d, e, h) are for the NC strain; and panels (c, f, i) are for the SC strain.
93
b
a
r 2 = 0.77
r 2 = 0.55
c
d
r 2 = 0.00
r 2 = 0.54
e
r 2 = 0.03
f
r 2 = 0.46
Figure 2.13. Linear relationships between hemolytic activity and growth rates (divisions day-1),
at salinity 20 f/2 media. Panels (a, b) are from nutrient-replete media; panels (c, d) are from
low-nitrogen media; and panels (e, f) are from low-phosphorus media. Panels (a, c, e) are for
the TX strain, and panels (b, d, f) are for the NC strain.
94
2.6 REFERENCES
Andersen O, Goldman JC, Caron DA, Dermett, MR. 1986. Nutrient cycling in a
microflagellate food chain III. Phosphorus dynamics. Mar. Ecol. Prog. Ser. 31:47-55.
Anderson DM, Burkholder JM, Cochlan WP, Glibert PM, Gobler CJ, Heil CA, Kudela RM,
Parsons ML, Rensel JJ, Townsend DW. 2008. Harmful algal blooms and
eutrophication: examining linkages from selected coastal regions of the United States.
Harmful Algae 8:39-53.
ASTM. 1996. Standard Guide for Conducting Static 96-h Toxicity Tests with Microalgae.
West Conshohocken, PA: American Society for Testing and Materials.
Baker JW, Grover JP, Brooks BW, Ureña‐Boeck F, Roelke DL, Errera R, Kiesling RL. 2007.
Growth and toxicity of Prymnesium parvum (Haptophyta) as a function of salinity,
light, and temperature. J. Phycol. 43:219-227.
Baker JW, Grover JP, Ramachandrannair R, Black C, Valenti Jr TW, Brooks BW, Roelke
DL. 2009. Growth at the edge of the niche: an experimental study of the harmful alga
Prymnesium parvum. Limnol. Oceanogr. 54:1679.
Barreiro A, Guisande C, Maneiro I, Lien TP, Legrand C, Tamminen T, Lehtinen S, Uronen
P, Granéli E. 2005. Relative importance of the different negative effects of the toxic
haptophyte Prymnesium parvum on Rhodomonas salina and Brachionus plicatilis.
Aquat. Microb. Ecol. 38:259-267.
Bertin MJ, Zimba PV, Beauchesne KR, Huncik KM, Moeller PD. 2012. The contribution of
fatty acid amides to Prymnesium parvum Carter toxicity. Harmful Algae 20:117-125.
Beszteri S, Yang I, Jaeckisch N, Tillmann U, Frickenhaus S, Glöckner G, Cembella A, John
U. 2012. Transcriptomic response of the toxic prymnesiophyte Prymnesium parvum
(N. Carter) to phosphorus and nitrogen starvation. Harmful Algae 18:1-15.
Blanck H, Wallin G, Wangberg SA. 1984. Species-dependent variation in algal sensitivity to
chemical compounds. Ecotoxicol Environ Saf 8:339-351.
Blossom HE, Andersen NG, Rasmussen SA, Hansen PJ. 2014. Stability of the intra-and
extracellular toxins of Prymnesium parvum using a microalgal bioassay. Harmful
Algae 32:11-21.
Bratbak G, Thingstad T. 1985. Phytoplankton-bacteria interactions: an apparent paradox?
Analysis of a model system with both competition and commensalism. Mar. Ecol.
Prog. Ser. 25:23-30.
95
Bricker SB, Longstaff B, Dennison W, Jones A, Boicourt K, Wicks C, Woerner J. 2008.
Effects of nutrient enrichment in the nation's estuaries: a decade of change. Harmful
Algae 8:21-32.
Brooks BW, James SV, Valenti Jr TW, Urena‐Boeck F, Serrano C, Berninger JP, Schwierzke
L, Mydlarz LD, Grover JP, Roelke DL. 2010. Comparative Toxicity of Prymnesium
parvum in Inland Waters. JAWRA 46:45-62.
Brooks BW, Lazorchak JM, Howard MD, Johnson MVV, Morton SL, Perkins DA, Reavie
ED, Scott GI, Smith SA, Steevens JA. 2016. Are harmful algal blooms becoming the
greatest inland water quality threat to public health and aquatic ecosystems? Environ.
Toxicol. Chem. 35:6-13.
Brutemark A, Granéli E. 2011. Role of mixotrophy and light for growth and survival of the
toxic haptophyte Prymnesium parvum. Harmful Algae 10:388-394.
Burkholder J. 1998. Implications of harmful marine microalgae and heterotrophic
dinoflagellates in management of sustainable marine fisheries. Ecol. Appl. 8: S37S62.
Burkholder JM. 2001. Beyond algal blooms, oxygen deficits and fish kills: Chronic, longterm impacts of nutrient pollution on aquatic ecosystems. In: Waters in Peril
(Bendell-Young L, Gallaugher P, eds.). Dordrecht, The Netherlands: Kluwer
Academic Publishers, 103-125.
Burkholder JM. 2009. Harmful algal blooms, pp. 264-285. In: Encyclopedia of Inland
Waters, Volume 1 (Likens GE, ed.). Oxford, United Kingdom: Elsevier, 264-285.
Burkholder J, Glibert P. 2006. Intraspecific variability: an important consideration in forming
generalizations about toxigenic algal species. Afr. J. Mar. Sci. 28:177-180.
Burkholder J, Glibert PM. 2013. Eutrophication and oligotrophication, pp. 347-371. In:
Encyclopedia of Biodiversity, 2nd edition, Volume 3 (Levin S, ed.). by Levin S (ed.).
Waltham, MA: Academic Press, 649-670.
Burkholder JM, Glibert PM, Skelton HM. 2008. Mixotrophy, a major mode of nutrition for
harmful algal species in eutrophic waters. Harmful Algae 8:77-93.
Burkholder J, Libra B, Weyer P, Heathcote S, Kolpin D, Thorne PS, Wichman M. 2007.
Impacts of waste from concentrated animal feeding operations on water quality.
Environ. Health Perspect. 115:308-312.
96
Burkholder J, Shumway SE. 2017. Food web and ecosystem impacts of harmful algae. In:
Harmful Algal Blooms and Their Management: A Compendium Desk Reference
(Shumway SE, Burkholder JM, Morton SL, eds.). New York, NY: Elsevier.
Caron DA. 2016. Mixotrophy stirs up our understanding of marine food webs. Proc. Nat.
Acad. Sci. (USA) 113:2806-2808.
Carvalho WF, Granéli E. 2010. Contribution of phagotrophy versus autotrophy to
Prymnesium parvum growth under nitrogen and phosphorus sufficiency and
deficiency. Harmful Algae 9:105-115.
Clements WH, Rohr JR. 2009. Community responses to contaminants: using basic ecological
principles to predict ecotoxicological effects. Environ. Toxicol. Chem. 28:1789-1800.
Dafni Z, Ulitzur S, Shilo M. 1972. Influence of light and phosphate on toxin production and
growth of Prymnesium parvum. Microbiology 70:199-207.
DeLorenzo ME, Scott GI, Ross PE. 2001. Toxicity of pesticides to aquatic microorganisms: a
review. Environ. Toxicol. Chem. 20:84-98.
DeLorenzo ME, Serrano L. 2003. Individual and mixture toxicity of three pesticides;
atrazine, chlorpyrifos, and chlorothalonil to the marine phytoplankton species
Dunaliella tertiolecta. Journal of Environmental Science and Health, Part B 38:529538.
Edvardsen B, Imai I. 2006. The ecology of harmful flagellates within Prymnesiophyceae and
Raphidophyceae. In: Ecology of Harmful Algae (Granéli, E., Turner, J., eds.). New
York, NY: Springer, 67-79.
Edvardsen B, Paasche E. 1998. Bloom dynamics and physiology of Prymnesium and
Chrysochromulina. NATO ASI Series G Ecol. Sci. 41:193-208.
Efron B. 1982. The jackknife, the bootstrap and other resampling plans. Vol. 38: Society for
Industrial and Applied Mathematics (SIAM).
Eisler R. 1989. Atrazine Hazards to Fish, Wildlife, and Invertebrates: A Synoptic Review.
Biological Report 85(1.8), Contaminant Hazard Reviews Report No. 18. Laurel, MD:
U.S. Fish and Wildlife Service.
Errera RM, Roelke DL, Kiesling RL, Brooks BW, Grover JP, Schwierzke L, Ureña-Boeck F,
Baker JW, Pinckney JL. 2008. Effect of imbalanced nutrients and immigration on
Prymnesium parvum community dominance and toxicity: results from in-lake
microcosm experiments. Aquat. Microb. Ecol. 52:33-44.
97
Eschbach E, Scharsack JP, John U, Medlin LK. 2001. Improved erythrocyte lysis assay in
microtitre plates for sensitive detection and efficient measurement of haemolytic
compounds from ichthyotoxic algae. J. Appl. Toxicol. 21:513-519.
Fistarol GO, Legrand C, Granéli E. 2003. Allelopathic effect of Prymnesium parvum on a
natural plankton community. Mar. Ecol. Prog. Ser. 255:115-125.
Gascón J, Oubiña A, Barceló D. 1997. Detection of endocrine-disrupting pesticides by
enzyme-linked immunosorbent assay (ELISA): application to atrazine. Trends Anal.
Chem. 16:554-562.
Geis SW, Fleming KL, Korthals ET, Searle G, Reynolds L, Karner DA. 2000. Modifications
to the algal growth inhibition test for use as a regulatory assay. Environ. Toxicol.
Chem. 19:36-41.
Giddings JM. 2005. Atrazine in North American Surface Waters: A Probabilistic Aquatic
Ecological Risk Assessment. Pensacola, FL: Society of Environmental Toxicology
and Chemistry.
Gilliom RJ, Barbash JE, Crawford CG, Hamilton PA, Martin JD, Nakagaki N, Nowell LH,
Scott JC, Stackelberg PE, Thelin GP. 2006. Pesticides in the nation's streams and
ground water, 1992-2001. Reston, VA: U.S. Geological Survey.
Glibert PM. 2016. Margalef revisited: A new phytoplankton mandala incorporating twelve
dimensions, including nutritional physiology. Harmful Algae 55:25-30.
Glibert PM, Anderson DM, Gentien P, Graneli E, Sellner KG. 2005. The global, complex
phenomena of harmful algal blooms. Oceanography 18:136-147
Glibert P, Burkholder J. 2006. The complex relationships between increases in fertilization of
the earth, coastal eutrophication and proliferation of harmful algal blooms. In:
Ecology of Harmful Algae (Granéli, E., Turner, J., eds.). New York, NY: Springer,
341-354.
Glibert PM, Burkholder JM, Kana TM, Alexander J, Skelton H, Shilling C. 2009. Grazing by
Karenia brevis on Synechococcus enhances its growth rate and may help to sustain
blooms. Aquat. Microb. Ecol. 55:17-30.
Glibert PM, Fullerton D, Burkholder JM, Cornwell JC, Kana TM. 2011. Ecological
stoichiometry, biogeochemical cycling, invasive species and aquatic food webs: San
Francisco Estuary and comparative systems. Rev. Fish. Sci. 19: 358-417.
98
Glibert PM, Harrison J, Heil C, Seitzinger S. 2006. Escalating worldwide use of urea–a
global change contributing to coastal eutrophication. Biogeochemistry 77:441-463.
Glibert PM, Maranger R, Sobota DJ, Bouwman L. 2014. The Haber Bosch–harmful algal
bloom (HB–HAB) link. Environ. Res. Lett. 9:105001.
Graneli E. 2006. Kill your enemies and eat them with the help of your toxins: an algal
strategy. Afr. J. Mar. Sci. 28:331-336.
Granéli E, Edvardsen B, Roelke DL, Hagström JA. 2012. The ecophysiology and bloom
dynamics of Prymnesium spp. Harmful Algae 14:260-270.
Granéli E, Johansson N. 2003a. Effects of the toxic haptophyte Prymnesium parvum on the
survival and feeding of a ciliate: the influence of different nutrient conditions. Mar.
Ecol. Prog. Ser. 254: 49-56.
Granéli E, Johansson N. 2003b. Increase in the production of allelopathic substances by
Prymnesium parvum cells grown under N-or P-deficient conditions. Harmful Algae
2:135-145.
Granéli E, Salomon PS. 2010. Factors Influencing Allelopathy and Toxicity in Prymnesium
parvum. JAWRA 46:108-120
Guillard RR. 1973. Division rates. In: Handbook of Phycological Methods, Vol. 1. New
York, NY: Cambridge University Press, 289-312.
Guillard RR, Ryther JH. 1962. Studies of marine planktonic diatoms: I. Cyclotella Nana
Hustedt, and Detonula confervacea (Cleve.) Gran. Can. J. Microbiol. 8:229-239.
Hallegraeff GM. 1993. A review of harmful algal blooms and their apparent global increase.
Phycologia 32:79-99.
Hambright KD, Zamor RM, Easton JD, Glenn KL, Remmel EJ, Easton AC. 2010. Temporal
and spatial variability of an invasive toxigenic protist in a North American subtropical
reservoir. Harmful Algae 9:568-577.
Heisler J, Glibert PM, Burkholder JM, Anderson DM, Cochlan W, Dennison WC, Dortch Q,
Gobler CJ, Heil CA, Humphries E. 2008. Eutrophication and harmful algal blooms: a
scientific consensus. Harmful Algae 8:3-13.
Henrikson JC, Gharfeh MS, Easton AC, Easton JD, Glenn KL, Shadfan M, Mooberry SL,
Hambright KD, Cichewicz RH. 2010. Reassessing the ichthyotoxin profile of
cultured Prymnesium parvum (golden algae) and comparing it to samples collected
99
from recent freshwater bloom and fish kill events in North America. Toxicon
55:1396-1404.
Holdway P, Watson R, MOSS B. 1978. Aspects of the ecology of Prymnesium parvum
(Haptophyta) and water chemistry in the Norfolk Broads, England. Freshwater Biol.
8:295-311.
Igarashi T, Satake M, Yasumoto T. 1996. Prymnesin-2: a potent ichthyotoxic and hemolytic
glycoside isolated from the red tide alga Prymnesium parvum. Journal of the
American Chemical Society 118:479-480.
Igarashi T, Satake M, Yasumoto T. 1999. Structures and partial stereochemical assignments
for prymnesin-1 and prymnesin-2: potent hemolytic and ichthyotoxic glycosides
isolated from the red tide alga Prymnesium parvum. Journal of the American
Chemical Society 121:8499-8511.
Interlandi SJ. 2002. Nutrient–toxicant interactions in natural and constructed phytoplankton
communities: results of experiments in semi-continuous and batch culture. Aquat.
Toxicol. 61:35-51.
Israël NM, VanLandeghem MM, Denny S, Ingle J, Patino R. 2014. Golden alga presence and
abundance are inversely related to salinity in a high-salinity river ecosystem, Pecos
River, USA. Harmful Algae 39:81-91.
Jablonowski ND, Schäffer A, Burauel P. 2011. Still present after all these years: persistence
plus potential toxicity raise questions about the use of atrazine. Environ. Sci. Pollut.
R. 18:328-331.
James T, De La Cruz A. 1989. Prymnesium Parvum Carter (Chrysophyceae) as a suspect of
mass mortalities of fish and shellfish communities in western Texas. Tex. J. Sci
41:429-430.
Johansson N, Graneli E. 1999. Influence of different nutrient conditions on cell density,
chemical composition and toxicity of Prymnesium parvum (Haptophyta) in semicontinuous cultures. J. Exp. Mar. Biol. Ecol. 239:243-258.
Johnston EL, Mayer‐Pinto M, Crowe TP. 2015. REVIEW: Chemical contaminant effects on
marine ecosystem functioning. J. Appl. Ecol. 52:140-149.
Kaartvedt S, Johnsen T, Aksnes D, Lie U, Svendsen H. 1991. Occurrence of the toxic
phytoflagellate Prymnesium parvum and associated fish mortality in a Norwegian
fjord system. Can. J. Fish. Aquat. Sci. 48:2316-2323.
100
Kapustka L. 2008. Limitations of the current practices used to perform ecological risk
assessment. Integr. Environ. Assess. Manag. 4:290-298.
Kasai F, Takamura N, Hatakeyama S. 1993. Effects of simetryne on growth of various
freshwater algal taxa. Environmental Pollution 79:77-83.
Kaushal SS, Groffman PM, Likens GE, Belt KT, Stack WP, Kelly VR, Band LE, Fisher GT.
2005. Increased salinization of fresh water in the northeastern United States. Proc.
Nat. Acad. Sci. (USA)102:13517-13520.
Kimball KD, Levin SA. 1985. Limitations of laboratory bioassays: the need for ecosystemlevel testing. BioScience 35:165-171.
Kooijman SALM. 1993. Dynamic Energy Budgets in Biological Systems. New York, NY:
Cambridge University Press.
La Claire II JW. 2006. Analysis of expressed sequence tags from the harmful alga,
Prymnesium parvum (Prymnesiophyceae, Haptophyta). Marine Biotechnology 8:534546.
Landis WG, Ho YM. 1995. Introduction to Environmental Toxicology. Boca Raton, FL:
Lewis Publisher, CRC Press, Inc.
Lembi CA, Waaland JR. 1988. Algae and Human Affairs. New York: Cambridge University
Press.
Lewitus AJ, Horner RA, Caron DA, Garcia-Mendoza E, Hickey BM, Hunter M, Huppert
DD, Kudela RM, Langlois GW, Largier JL. 2012. Harmful algal blooms along the
North American west coast region: History, trends, causes, and impacts. Harmful
Algae 19:133-159.
Lindehoff E, Granéli E, Granéli W. 2009. Effect of tertiary sewage effluent additions on
Prymnesium parvum cell toxicity and stable isotope ratios. Harmful Algae 8:247-253.
Lundgren VM, Glibert PM, Granéli E, Vidyarathna NK, Fiori E, Ou L, Flynn KJ, Mitra A,
Stoecker DK, Hansen PJ. 2016. Metabolic and physiological changes in Prymnesium
parvum when grown under, and grazing on prey of, variable nitrogen: phosphorus
stoichiometry. Harmful Algae 55:1-12.
Lutz‐Carrillo DJ, Southard GM, Fries LT. 2010. Global genetic relationships among isolates
of golden alga (Prymnesium parvum) 1. JAWRA 46:24-32.
101
Lytle JS, Lytle TF. 2001. Use of plants for toxicity assessment of estuarine ecosystems.
Environ. Toxicol. Chem. 20:68-83.
Ma J. 2005. Differential sensitivity of three cyanobacterial and five green algal species to
organotins and pyrethroids pesticides. Sci. Total Environ. 341:109-117.
Manning SR, La Claire JW. 2010. Prymnesins: Toxic metabolites of the golden alga,
Prymnesium parvum Carter (Haptophyta). Marine Drugs 8:678-704.
Martin-Cereceda M., Novarino G, Young JR. 2003. Grazing by Prymnesium parvum on
small planktonic diatoms. Aquat. Microb. Ecol. 33: 191-199.
McCormick PV, Cairns Jr J. 1994. Algae as indicators of environmental change. J. Appl.
Phycol. 6:509-526.
Miller WE, Greene JC. 1978. The Selenastrum capricornutum Printz algal assay bottle test:
Experimental design, application, and data interpretation protocol: Environmental
Protection Agency, Office of Research and Development, Corvallis Environmental
Research Laboratory.
Moestrup Ø. 1994. Economic aspects: ‘blooms’, nuisance species and toxins. In: Haptophyte
Algae. Systematics Association Special Volume No. 51: Clarendon Press.
Mohan B, Hosetti B. 1999. Aquatic plants for toxicity assessment. Environmental Research
81:259-274.
Moore LR. 2013. More mixotrophy in the marine microbial mix. Proc. Nat. Acad. Sci.
(USA) 110:8323-8324.
Moustaka-Gouni M, Cook CM, Gkelis S, Michaloudi E, Pantlelidakis K, Pyrovetsi M,
Lanaras T. 2004. The coincidence of a Prymnesium parvum bloom and the mass kill
of birds and fish in Lake Koronia. Harmful Algae News:1-3.
Nisbet RM, Jusup M, Klanjscek T, Pecquerie L. 2012. Integrating dynamic energy budget
(DEB) theory with traditional bioenergetic models. J. Exper. Biol. 215:892-902.
Norberg-King TJ. 1993. A linear interpolation method for sublethal toxicity: The inhibition
concentration (ICp) approach. National Effluent Toxicity Assessment Center
Technical Report 39:3-93.
Nygaard K, Tobiesen A. 1993. Bacterivory in algae: a survival strategy during nutrient
limitation. Limnol. Oceanogr. 38:273-279.
102
Palenik B, Morel FM. 1991. Amine oxidases of marine phytoplankton. Appl. Environ.
Microbiol. 57:2440-2443.
Patiño R, Dawson D, VanLandeghem MM. 2014. Retrospective analysis of associations
between water quality and toxic blooms of golden alga (Prymnesium parvum) in
Texas reservoirs: Implications for understanding dispersal mechanisms and impacts
of climate change. Harmful Algae 33:1-11.
Peterson HG, Boutin C, Martin PA, Freemark KE, Ruecker NJ, Moody MJ. 1994. Aquatic
phytotoxicity of 23 pesticides applied at expected environmental concentrations.
Aquat. Toxicol. 28:275-292.
Pitcher GC. 2012. Harmful algae—the requirement for species-specific information. Harmful
Algae 14:1-4.
Prosser KN, Valenti TW, Hayden NJ, Neisch MT, Hewitt NC, Umphres GD, Gable GM,
Grover JP, Roelke DL, Brooks BW. 2012. Low pH preempts bloom development of a
toxic haptophyte. Harmful Algae 20:156-164.
Redfield AC. 1958. The biological control of chemical factors in the environment. American
Scientist 46:230A-221.
Reich K, Aschner M. 1947. Mass development and control of the phytoflagellate
Prymnesium parvum in fish ponds in Palestine. Palestine J. Bot. 4:14-23.
Remmel EJ, Hambright KD. 2012. Toxin‐assisted micropredation: experimental evidence
shows that contact micropredation rather than exotoxicity is the role of Prymnesium
toxins. Ecology Letters 15:126-132.
Remmel EJ, Kohmescher N, Larson JH, Hambright K. 2011. An experimental analysis of
harmful algae-zooplankton interactions and the ultimate defense. Limnol. Oceanogr.
56:461-470.
Renner R. 2009. Salt-loving algae wipe out fish in Appalachian stream. Environ. Sci.
Technol. 43:9046-9047.
Roelke DL, Barkoh A, Brooks BW, Grover JP, Hambright KD, LaClaire II JW, Moeller PD,
Patino R. 2016. A chronicle of a killer alga in the west: ecology, assessment, and
management of Prymnesium parvum blooms. Hydrobiologia 764:29-50.
Roelke DL, Gable GM, Valenti TW, Grover JP, Brooks BW, Pinckney JL. 2010. Hydraulic
flushing as a Prymnesium parvum bloom-terminating mechanism in a subtropical
lake. Harmful Algae 9:323-332.
103
Roelke DL, Grover JP, Brooks BW, Glass J, Buzan D, Southard GM, Fries L, Gable GM,
Schwierzke-Wade L, Byrd M. 2011. A decade of fish-killing Prymnesium parvum
blooms in Texas: roles of inflow and salinity. J. Plankton Res. 33:243-253.
Rojíčková-Padrtová R, Maršálek B, Holoubek I. 1998. Evaluation of alternative and standard
toxicity assays for screening of environmental samples: selection of an optimal test
battery. Chemosphere 37:495-507.
Rojíčková R, Maršálek B. 1999. Selection and sensitivity comparisons of algal species for
toxicity testing. Chemosphere 38:3329-3338.
Sager D, Barkoh A, Buzan D, Fries L, Glass J, Kurten G, Ralph J, Singhurst E, Southard G,
Swanson E. 2008. Toxic Prymnesium parvum: a potential threat to US reservoirs. In:
Proceedings of Balancing Fisheries Management and Water Uses for Impounded
River Systems. American Fisheries Society, Symposium, Vol. 62, 261-278.
Schug KA, Skingel TR, Spencer SE, Serrano CA, Le CQ, Schug CA, Valenti Jr TW, Brooks
BW, Mydlarz LD, Grover JP. 2010. Hemolysis, Fish Mortality, and LC‐ESI‐MS of
Cultured Crude and Fractionated Golden Alga (Prymnesium parvum) 1. JAWRA
46:33-44.
Shilo M. 1967. Formation and mode of action of algal toxins. Bacteriological Reviews
31:180.
Shilo M. 1981. The toxic principles of Prymnesium parvum. In: The Water Environment:
Algal Toxins and Health (Carmichael WW, ed.). New York, NY: Plenum Press, 3747.
Shilo M, Rosenberger RF. 1960. Studies on the toxic principles formed by the chrysomonad
Prymnesium Parvum Carter. Ann. NY Acad. Sci. 90:866-876.
Shilo M, Shilo M. 1953. Conditions which determine the efficiency of ammonium sulphate
in the control of Prymnesium parvum in fish breeding ponds. Applied Microbiology
1:330.
Skovgaard A, Hansen PJ. 2003. Food uptake in the harmful alga Prymnesium parvum
mediated by excreted toxins. Limnol. Oceanogr. 48:1161-1166.
Smayda T. 1990. Novel and nuisance phytoplankton blooms in the sea: evidence for a global
epidemic. In: Toxic Marine Phytoplankton (Granéli E, Sundstrøm B, Edler L,
Anderson DM, eds.). New York, NY: Elsevier, 29-40.
104
Solomon KR, Baker DB, Richards RP, Dixon KR, Klaine SJ, La Point TW, Kendall RJ,
Weisskopf CP, Giddings JM, Giesy JP. 1996. Ecological risk assessment of atrazine
in North American surface waters. Environ. Toxicol. Chem. 15:31-76.
Solomon KR, Giesy JP, LaPoint TW, Giddings JM, Richards RP. 2013. Ecological risk
assessment of atrazine in North American surface waters. Environ. Toxicol. Chem.
32:10-11.
Sorokin C. 1973. Dry weight, packed cell volume and optical density. Handbook of
Phycological Methods, Vol. 1 (Stein JR, ed.). 321-343. New York, NY: Cambridge
University Press, 289-312.
Southard GM, Fries LT, Barkoh A. 2010. Prymnesium parvum: The Texas Experience.
JAWRA 46:14-23.
Stauber J. 1995. Toxicity testing using marine and freshwater unicellular algae. Australasian
J. Ecotoxicol. 1:15-24.
Sunda WG, Graneli E, Gobler CJ. 2006. Positive feedback and the development and
persistence of ecosystem disruptive algal blooms1. J. Phycol. 42:963-974.
Tillmann U. 1998. Phagotrophy by a plastidic haptophyte, Prymnesium patelliferum. Aquat.
Microb. Ecol. 14:155-160.
Tillmann U. 2003. Kill and eat your predator: a winning strategy of the planktonic flagellate
Prymnesium parvum. Aquat. Microb. Ecol. 32:73-84.
Tilman D. 1982. Resource Competition and Community Structure. Princeton, NH: Princeton
University Press.
Tittel J, Bissinger V, Zippel B, Gaedke U, Bell E, Lorke A, Kamjunke N. 2003. Mixotrophs
combine resource use to outcompete specialists: implications for aquatic food webs.
Proc. Nat. Acad. Sci. (USA) 100:12776-12781.
Ulitzur S, Shilo M. 1970. Procedure for purification and separation of Prymnesium parvum
toxins. Biochim. Biophys. Acta 201:350-363.
United States Environmental Protection Agency (U.S. EPA). 1974. Marine Algal Assay
Procedure: Bottle Test. Corvallis, OR: U.S. EPA National Environmental Research
Center.
105
Uronen P, Lehtinen S, Legrand C, Kuuppo P, Tamminen T. 2005. Haemolytic activity and
allelopathy of the haptophyte Prymnesium parvum in nutrient-limited and balanced
growth conditions. Mar. Ecol. Prog. Ser. 299:137-148.
Van Dolah FM. 2000. Marine algal toxins: origins, health effects, and their increased
occurrence. Environ. Health Perspect. 108:133.
VanLandeghem MM, Denny S, Patiño R. 2015a. Predicting the risk of toxic blooms of
golden alga from cell abundance and environmental covariates. Limnol. Oceanogr:
Methods 13:568-586.
VanLandeghem MM, Farooqi M, Farquhar B, Patiño R. 2013. Impacts of Golden Alga
Prymnesium parvum on fish populations in reservoirs of the upper Colorado River
and Brazos River Basins, Texas. Trans. Am. Fish. Soc. 142:581-595.
VanLandeghem MM, Farooqi M, Southard GM, Patiño R. 2015b. Associations between
water physicochemistry and Prymnesium parvum presence, abundance, and toxicity
in west Texas reservoirs. JAWRA 51:471-486.
VanLandeghem MM, Meyer MD, Cox SB, Sharma B, Patiño R. 2012. Spatial and temporal
patterns of surface water quality and ichthyotoxicity in urban and rural river basins in
Texas. Water Research 46:6638-6651.
Walsh GE. 1972. Effects of herbicides on photosynthesis and growth of marine unicellular
algae. Hyacinth Control J. 10:45-48.
Weiner JA, DeLorenzo ME, Fulton MH. 2004. Relationship between uptake capacity and
differential toxicity of the herbicide atrazine in selected microalgal species. Aquatic
Toxicology 68:121-128
Williams W. 2001. Anthropogenic salinisation of inland waters. Hydrobiologia 466:329-337.
Yariv J, Hestrin S. 1961. Toxicity of the extracellular phase of Prymnesium parvum cultures.
Microbiology 24:165-175.
Yates BS, Rogers WJ. 2011. Atrazine selects for ichthyotoxic Prymnesium parvum, a
possible explanation for golden algae blooms in lakes of Texas, USA. Ecotoxicology
20:2003-2010.
Zar JH. 1999. Biostatistical Analysis, 4th edition. Upper Saddle River, NJ: Prentice Hall.
106
Chapter 3: Examination of Chattonella subsalsa (Raphidophyceae) Growth and
Hemolytic Activity Responses to Co-Occurring Environmental Stressors
3.1 INTRODUCTION
Coastal zone estuaries and wetland habitats are now recognized as being among the
most highly stressed natural systems due to rapid urbanization and development of
coastlands, and concomitant pollution from land-based runoff (Scott et al. 2006, U.S. EPA
2016a). About two-thirds of the nation's coastal areas and more than one-third of the nation's
estuaries showed impairment from nutrient (nitrogen, N, and phosphorus, P) pollution (U.S.
EPA 2016b). As anthropogenic nutrient pollution increases, many estuarine and marine
coastal waters are sustaining harmful algal blooms (HABs) which are increasing in duration
and frequency (Hallegraeff 1993, Glibert 2005, Anderson et al. 2008, Heisler et al. 2008).
Eutrophication is now recognized as one of the most important factors contributing to the
global expansion of HAB species (Burkholder 1998, Glibert 2005; Heisler et al. 2008). High
levels of phosphate loading relative to dissolved inorganic nitrogen have been related to the
increased occurrence of harmful algal species such as toxigenic raphidophytes
(Heterokontophyta, Raphidophyceae; Lewitus et al. 2003 a,b).
The majority of chemical contaminants reaching estuaries are from agriculture, and
many of them are herbicides (Gilliom et al. 2006), which commonly co-occur in runoff with
nutrients from fertilizers and animal wastes (Zimba 2000; Hapeman et al. 2002; Peters et al.
2005; Burkholder et al. 2007). Atrazine (1-chloro-3-ethylamino-5-isopropylamino-2,4,6triazine) is the second most commonly used herbicide in the U.S., and the most common
surface water contaminant (Gilliom et al. 2006, Ryberg et al. 2010). This herbicide is a
photosynthetic inhibitor which works by blocking electron transport in photosystem II
(DeLorenzo et al. 2001, Vencill 2002), and can elicit effects similar to nutrient dificiency in
exposed algal populations (Weiner et al. 2007). It has been found at environmental
concentrations as high as 1,000 µg L-1 in surface waters adjacent to treated fields (deNoyelles
et al. 1982, Eisler 1989, Penninigton et al. 2001).
107
Recently, blooms of the toxigenic raphidophyte Chattonella subsalsa B. Biecheler
were linked to fish kills in eutrophic waters of coastal South Carolina in the southeastern
U.S. (Lewitus et al. 2003b, 2008). This and other species of Chattonella are widely
distributed in temperate and subtropical/tropical estuarine and marine waters, and frequently
have been linked to fish kills (Imai et al. 1998, Cortés-Altamirano et al. 2006, Zhang et al.
2006, Hallegraeff et al. 1998, Lewitus et al. 2012). The species C. subsalsa appears to be
especially well-adapted to thrive in shallow, eutrophic habitats (Zhang et al. 2006, BandSchmidt et al. 2012). Several mechanisms have been suggested for Chattonella spp.
ichthyotoxicity, including production of neurotoxic brevetoxins or brevetoxin-like substances
(Ahmad 1995; Khan et al. 1995, 1996); a synergistic interaction between reactive oxygen
species (ROSs) and free fatty acids (FFAs) (Marshall et al. 2003, 2005); and gill damage by
unidentified hemolytic substances (Shimada et al. 1983) or polyunsaturated fatty acids
(Marshall et al. 2002) which can have hemolytic effects (Fu et al. 2004); and physical
clogging of gills (Matsusato and Kobayashi 1974). Hemolytic and hemagglutinating effects
are common (Shimada et al. 1983, Ahmed et al. 1995, Kuroda et al. 2005, Pistocchi et al.
2012). The first signs of physiological disturbance in exposed fish include a decrease in the
oxygen partial pressure of arterial blood (Ishimatsu et al. 1996b), and many affected fish
have shown production of mucus-like substances and gill injury (Hiroishi et al. 2005, Shen et
al. 2011). Suffocation appears to be the ultimate cause of death (Kim et al. 2001, Tiffany et
al. 2001, Imai and Yamaguchi 2012). Bioactive substances released by Chattonella spp. have
also adversely affected diatoms (Matsuyama et al. 2000) and dinoflagellates (FernándezHerrera et al. 2016), allowing these relatively large and slowly growing flagellates to become
dominant under conducive conditions.
This study examined the interactive effects of nutrients and atrazine exposure on the
toxicity of Chattonella subsalsa. Little is known about herbicidal effects on harmful algal
species, which can be difficult to culture and are rarely used in toxicity testing (Thursby et al.
1993; Lytle and Lytle 2001; and see Chapter 1 and references therein). Autecological and
ecotoxicological studies of harmful algal species are needed to gain insights about the
108
mechanisms which lead to toxic HABs and which regulate toxicity in estuarine waters that
are increasingly contaminated by chemical substances from land-based runoff (Cloern 1996;
Pelley 1998; Zimba 2000; Hapeman et al. 2002; Bricker et al. 2008).
3.2 MATERIALS AND METHODS
3.2.i. Experimental Organism
The strain of Chattonella subsalsa used in this study (CCMP2191) was obtained in
unialgal, non-axenic culture from the National Center for Marine Algae and Microbiota on
11 February 2009. It was originally isolated from on 13 August 2001 from the Indian River
Bay, Delaware, USA.
This raphidophyte species is a relatively large flagellate (length and width up to 50
µm and 25 µm, respectively; Hallegraeff and Hara 2003), with highly variable shape due to
the lack of a rigid cell wall (Band-Schmidt et al. 2012, Graham et al. 2016). The fragile cells
are morphologically plastic and the cell shape is frequently lost with fixation, making
identification difficult if based solely on morphological characteristics of preserved samples
(Band-Schmidt et al. 2012). This species has broad environmental tolerances; it grows well
at temperatures ranging from 10o to 30oC, and salinities of 5-30, with optima (although
strain-dependent) reported at 20-30oC and salinities of 15-25 (Zhang et al. 2006). Like many
other microalgae under environmentally unfavorable conditions, C. subsalsa forms cysts
(benthic stage), which can remain dormant for months to years. This ability to remain a
member of the “hidden flora” (Smayda 2002) means that previous blooms can leave “seed”
populations as cysts for years until conditions become favorable for germination. Thus, C.
subsalsa may remain undetected and then suddenly become dominant in the plankton and
potentially lethal to aquatic life (Imai et al. 1991; Imai et al. 1998; Imai and Yamaguchi
2012).
Other characteristics of C. subsalsa (based, however, on assessment of few strains)
that were important considerations for this work are its slow growth, its preference of
109
ammonium over nitrate as a nitrogen (N) source, its mixotrophic capabilities, and its toxin
production. The maximum nutrient-saturated growth rate of this species is ~0.2 to 1.26
divisions day-1 (e.g., Zhang et al. 2006, Band-Schmidt et al. 2012, Imai and Yamaguchi
2012), and it is generally considered to be a large, slowly growing flagellate. Half-saturation
constants (Ks) for C. subsalsa have been reported at 0.84 µM for PO43, 8.98 µM for NO3- and
1.46 µM for NH4+, indicating that this raphidophyte can use low concentrations of NH4+
more efficiently than NO3- (Zhang et al. 2006). Moreover, it appears to attain higher biomass
(~double) when NH4+ is the major inorganic N source (Zhang et al. 2006). Organic forms of
N (e.g., glutamic acid) and phosphorus (P, e.g. ATP/ADP, adenosine triphosphate and
adenosine diphosphate, respectively) can be used by C. subsalsa, as well as inorganic N and
P forms (Zhang et al. 2006, Yamaguchi et al. 2008). In addition to photosynthetic carbon
assimilation, C. subsalsa can consume coccoid unicellular cyanobacteria, and possibly other
small (< 2 µm) organisms, but phagotrophy apparently is limited to the size of the mucocyst
openings on the cell surface which are believed to be the sites of ingestion (Jeong et al. 2010,
Jeong 2011). Regarding toxicity, Chattonella spp. can produce ROSs (e.g., hydrogen
peroxide, superoxide, and hydroxyl radicals) at levels about 100-fold higher than other
microalgal species (Marshall 2005a), and ROSs can greatly enhance the toxicity of hemolytic
free fatty acids such as eicosapentaenoic acid, which occurs at high levels in C. subsalsa and
other raphidophytes (Marshall et al. 2002).
3.2.ii Culture Conditions
Batch cultures of Chattonella subsalsa strain CCMP2191 were maintained in 500-mL
Erlenmeyer flasks containing 250 mL of salinity 20-modified L1–Si medium (Guillard and
Hargraves 1993) at 24°C (ambient air temperature) and 106.31 (±26.16) µmol photons of
photosynthetically active radiation (PAR) m-2 s-1 (cool white fluorescent tubes) under a 16:8
hr light:dark cycle. Cultures were sterile-transferred under a laminar flow hood every 10-15
days as needed to maintain log growth phase. All media (culture and test) were prepared by
adjusting the salinity of ultrapure Milli-Q water (18 MΩ cm−1 at 25 °C) with Instant Ocean®
artificial sea salts (Aquarium Systems, Blacksburg, VA, USA) to the desired level, adjusting
110
the pH as necessary to 8.1 (± 0.2) using HCl or NaOH, and adding trace mineral and vitamin
solutions following autoclaving. Light was measured using a Biospherical QSL 101 quantum
lab sensor (BSI, San Diego, CA, USA); salinity was measured using a YSI model 3200
conductivity/salinity bridge and cell (YSI [Yellow Springs Instruments], Yellow Springs,
OH, USA); and pH was measured using an Orion Versa Star Pro multi-parameter meter
equipped with an Orion ROSS Sure-Flow pH electrode (Thermo Scientific, Waltham, MA,
USA). Trace mineral and vitamin solutions were sterile filtered (Whatman® Puradisc
cellulose acetate syringe filter, nominal pore size 0.2 μm, Sigma-Aldrich, St. Louis, MO,
USA), added to the autoclaved media aseptically, and the final media was vacuum filtered
(Corning® cellulose acetate filter, 0.22 µm nominal pore size, Sigma-Aldrich) before storage
at 4°C for no more than 30 days. All equipment was initially cleaned by scrubbing in hot,
soapy tap water. All glassware and non-plastic equipment used in culturing and experiments
was rinsed with pesticide-grade acetone before use, and all glassware and any non-metallic
equipment used was cleaned by acid stripping in 10% HCl (v/v). All equipment was
sterilized by autoclaving before use. Testing and culture transfers and media prep were
conducted using aseptic techniques, and algal observation and testing methods were initially
developed and optimized using Dunaliella tertiolecta as described in Chapter 1 of this work.
Stock cultures were maintained at 106.31 (±26.16) µmol photons m-2 s-1 and a 16:8-hr
light:dark period throughout the experiments, and were gently swirled once daily. Bioassays
(controls and test cultures) were maintained on a culture rack under a light bank (2.5 m long
by .75 m deep). All cultures in the bioassays were gently swirled by hand and randomly
repositioned daily on the culture rack following a computer-generated randomized shelf
mapping, to avoid any effects from possible differential light exposure.
3.2.iii Optical Density Versus Direct Cell Counts
Samples (1 mL) were randomly withdrawn from each parent culture and thoroughly
vortexed before absorbance was read on a Thermo Scientific Spectronic GENESYS 2
spectrophotometer (Thermo Scientific) using a freshly wiped Hellma® quartz cuvette
111
(Sigma-Aldrich) with a 10-mm pathlength. All readings were performed in duplicate and
adjusted using a media blank as a measurement baseline. Preceding all testing procedures,
the absorbance of the parent culture of C. subsalsa was scanned, and the three highest peaks
were selected as the most appropriate wavelengths for analysis of a series of 8-point dilution
curves, using progressive dilutions of stock culture in growth media. The curve that produced
the ΔOD/ΔPD [the change in optical density divided by the change in population density;
Sorokin 1973)] closest to unity (1) was used to determine the growth rates from each stock
culture for each strain tested. Cell counts to determine ΔPD were obtained by collecting
random 1-mL samples from the parent culture. Cells were quantified from duplicate
preserved samples using a Reichert bright-line phase contrast hemocytometer with improved
Neubauer rulings (Hausser Scientific No. 1475, Horsham, PA, USA) (Guillard 1973). Two
well mixed 100-µL samples were assessed by counting 18 grids or a minimum of 400 cells.
Because C. subsalsa cells lyse quickly upon exposure to standard preservatives, a
preservative which minimized lysis was used (1% HEPES-buffered paraformaldehyde with
1% glutaraldehyde; Katano et al. 2009) (Fig. 3.1). Preservation was adequate to preserve
cells for ~1 week without appreciable cell distortion or loss while preparing dilution curves
immediately before each experiment. The wavelength that produced the highest correlation
between absorbance and cell numbers of the stock culture was used to monitor population
density and calculate growth rates for the duration of the experiment. All modeled OD:cell
number relationships used had a minimum observed 𝑟 2 value >0.95 (e.g., Fig. 3.2).
3.2.iv Population Growth Rate
Net cell-specific growth rates (day-1) were modeled using OD following Sorokin
(1973). Samples were collected in duplicate daily from gently hand-agitated test tubes and
measured in a clean, wiped quartz cuvette as described above. A media blank served as a
baseline measurement. The OD was converted to population density (PD), data were logtransformed (log10 + 1), and the linear portion of the growth curve (slope = 𝑎𝑘 ) was used to
calculate growth rate k (Sorokin 1973):
𝑘 = 𝑎𝑘 ∗ 3.322
112
3.2.v Bioassays
Concentration-response bioassays were conducted to describe the effects of atrazine
exposures and imbalanced versus balanced nutrient levels on the population growth rate of
the Chattonella subsalsa strain, using the Redfield ratio (Redfield 1934, 1956; Harris 1986)
as balanced. Three nutrient regimes were used including low N, low P, and N- plus P-replete
media (below).
Analytical-grade atrazine standards (CASRN 1912-12-9, >98% purity) were obtained
from Chemservice (Westchester, PA, USA), and stock solutions were prepared in 100%
pesticide-grade acetone (Fisher Scientific, Fair Lawn, NJ, USA). Chemical stocks were
diluted in pesticide-grade acetone to desired final concentrations and were introduced into
fresh, sterile culture media aseptically under a laminar flow hood. Testing doses were
administered to obtain a final concentration of 0.1% (v/v) acetone in each replicate, with
exception of a media-only control series which was included for all treatments (n = 3) in all
assays. Controls and treatments were incubated in 50-mL Erlenmeyer flasks, filled to 25 mL
and capped with foil. Five concentrations of atrazine were used to develop response curves
(12.5, 25, 50, 100 and 200 µg atrazine L-1, in addition to the media and solvent controls).
Exposure levels were selected based on range-finding test estimates in preliminary work, and
were designed to expose C. subsalsa to concentrations with 0% and 100% growth inhibition
effects.
The three nutrient regimes were as follows: To assess low-nutrient effects, three
inorganic N : inorganic P supply ratios (molar basis; selected based on deviations from the
Redfield Ratio as explained above) were prepared, including 1:1, N-limited (“Low N” = 2.0
µM NO3, 2.0 µM PO4); 160:1, P-limited (“Low P” = 32 µM NO3, 0.2 µM PO4); and 16:1,
nutrient-replete (“Replete” = 32 µM NO3, 2.0 µM PO4) media (Table 3.1). Each low-nutrient
treatment series contained corresponding solvent and media controls.
113
Nutrient x herbicide effects were assessed by repeating the above assay procedures in
a simultaneously run, three-way bioassay format. Each assay included 63 50-mL flasks in
identical set-up, inoculation and chemical preparation approaches as described above, but
varied in culture media nutrient additions.
Testing procedures were derivations on the standard 96-h static algal toxicity
bioassay protocols (ASTM 1996), and were extended to assess 10- and 28-day endpoints.
The 10-day endpoint was used to evaluate growth responses of C. subsalsa to treatments
during early exponential growth phase. In a second series of assays, for hemolytic activity
(below), a 28-day endpoint was used to enable assessment of treatment effects during late
exponential growth under nutrient depletion, which can take 14 days or longer to observe in
Chattonella spp. (Kim et al. 2004; Yamaguchi et al. 2008). Growth of C. subsalsa was slow
in comparison to bioassays conducted with the small green flagellate, Dunaliella tertiolecta
(Chlorophyta) (replete media Kmax = 0.95 div day-1 versus 2.04, respectively), in work to
develop the general test platform for these assays (see Chapter1). In assays with C. subsalsa,
the lag phase continued for 48 to 72 hr post-inoculation, and exponential growth did not
begin to occur until about 96-hr post-inoculation. Therefore, 96-hr bioassay protocols
commonly used in standard toxicity testing were not considered appropriate to assess effects
of atrazine x nutrient stress, and were modified to accommodate the slow growth of this test
species. The 96-hr lowest-observed (LOEC) and half-maximal effective concentrations (IC50)
were calculated in this work to facilitate comparisons of atrazine tolerance and sensitivity
reported in the literature for other organisms, but these values were presented with the caveat
that the 96-hr endpoint may not effectively capture the response of C. subsalsa to the
treatments. These analyses instead emphasized responses to treatments at endpoints that were
10 days and 28 days post-exposure.
For all experimental assays, when stock culture of Chattonella subsalsa was in
exponential growth phase (assessed by periodic direct cell counts), test and control cultures
were inoculated into foil-capped 50 mL Erlenmeyer flasks (with each flask containing a total
final volume of 25 mL) under aseptic conditions to attain initial cell densities of ~1.5 x 103
114
cells mL-1 (10-day assays) or 1.0 x 103 cells mL-1 (28-day assays) in all experimental cultures
and controls. Initially in experiments, the cultures were not nutrient-limited (as they had been
taken from nutrient-replete stock culture), the cells were in exponential growth phase, and all
cells had the same nutritional history. The inoculum was less than 5% of the total test culture
volume (less than 1.25 mL) to minimize the influence from the high nutrient levels in the
stock culture on controls and treatments. Time-series testing began when C. subsalsa was
added to testing media, and samples were collected immediately following inoculation (0 hr)
to assess the initial OD. Samples (n = 2) for OD measurements were taken during the assays
at equally spaced daily intervals for the first 4 days following inoculation (0, 24, 48, 72, and
96 hr), and then every third day for the remainder of the test duration. Immediately before
each sample was taken, tubes were gently swirled by hand for 30-60 s, and then 1 mL was
gently, randomly siphoned using a sterile transfer pipette and aseptically removed for OD
measurement.
For each sample taken from control and test replicates, growth of C. subsalsa was
determined by spectrophotometric determination of OD. The OD was then converted to cell
density using equations derived from direct cell count : optical density relationships that had
been established using the parent culture within one week prior to the test initiation, as
described above. The percent inhibition (%I) of algal growth at each atrazine concentration
was calculated by comparing mean growth rates for each treatment to the respective solvent
controls for the treatment, as follows:
%I = [1 − (𝑥̅
𝑘
𝑘𝑠𝑜𝑙𝑣𝑒𝑛𝑡 𝑐𝑜𝑛𝑡𝑟𝑜𝑙
)] ∙ 100
where 𝑘 is the growth rate of the pertinent sample, calculated as described previously.
Nominal concentrations of pesticide were quantified with enzyme-linked immunosorbent
assay (ELISA) at the initiation and conclusion of each bioassay with enzyme-labeled
paramagnetic particles of atrazine (method detection limit for atrazine = 0.046 μg/L; SDI,
Delaware, MD, USA; see appendix for raw ELISA data and QA/QC determination). The
115
accuracy of this method has been shown to be comparable to gas chromatographic
quantification of atrazine concentrations in surface water samples (Gascón et al. 1997).
Salinity and pH were checked at test initiation and conclusion to ensure that changes
were within acceptable limits (salinity 20 ±1, pH 8.1 ±0.2) based on the test platform
developed in Chapter 1. At assay initiation, all cultures were inoculated from the same the
same parent culture. Solvent controls showed similar growth as the media-only controls for
all tests (Fig. 3.3), so the solvent controls were used as the baseline activity for analytical
comparisons.
3.2.vi Protocols for Assays of Hemolytic Activity
The lack of chemical structure analysis for toxins from C. subsalsa prevented
analysis of toxicity by toxin quantification. As a proxy for the toxic potential of the control
and test cultures, erythrocyte lysis assays (ELAs, developed for 96-well microtiter plates with
well volume 120 μl; Eschbach et al. 2001; Kuroda et al. 2005) were modified for use with a
larger sample volume (1 mL), and then ELAs were performed using freshly prepared fish
erythrocytes to characterize the hemolytic activity (HA) of C. subsalsa.
3.2.vi.a Blood Collection and Erythrocyte Preparation
Adult tilapia (Oreochromis niloticus) were purchased live from a local Asian fish
market and transported to CAAE in ~20 l of water from their holding tanks. Fish were
anesthetized with a pH-adjusted 0.02% (w/v) solution of aminobenzoic acid ethyl ester
[TricaineTM (MS-222) (Sigma-Aldrich, St. Louis, MO, USA)] in transport water. Blood was
collected by gill puncture within 3 hr of purchase. To collect blood samples, anaesthetized
fish were placed on wet towels and 5 mL of blood were collected using a Pravaz No. 1
needle and a 10-mL syringe prefilled with 5 mL of pH-adjusted (7.2) RPMI 1640 media
(without phenol red), diluted to 10% (v/v) with Milli-Q water (to adjust for fish serum
osmolarity), and supplemented with 50 IU sodium heparin salt (Sigma-Aldrich) added as an
anticoagulant. Blood samples were transferred to 15-mL centrifuge tubes and gently hand-
116
agitated by inverting two or three times. Red bloods cells were then separated from serum by
gentle centrifugation (2,100 rpm at 4°C for 5 min) and washed by replacing supernatant with
fresh media until the supernatant appeared clear after gentle mixing. Whole, washed
erythrocytes were diluted 1:10 with RPMI 1640 culture medium containing 22.5 IU sodium
heparin anticoagulant mL-1. Erythrocytes (~107 cells mL-1) were stored at 4°C for no more
than 10 days, based on preliminary work wherein microscopic observations and
spectrophotometric measurements (λ = 415) indicated no noticeable lytic activity over that
duration. The erythrocytes were resuspended daily until use.
3.2.vi.b Preparation of Algal Extracts
Because the hemolytic component of Chattonella is light-dependent (Kuroda et al.
2005), all hemolytic (ELA) preparation and procedures were conducted at 4oC (Eschbach et
al. 2001) under continuous fluorescent lighting (30 µmol photons PAR m-2 s-1). Following
10-day or 28-day assay endpoint OD measurements, Chattonella subsalsa cells were
immediately centrifuged (IEC Centra-CL2, Needham Heights, MA, USA) for hemolytic
analysis. Because of the low cell densities in high herbicide treatments, treatment replicates
were pooled prior to extraction procedures as described below. Before replicates were
pooled, they were checked to ensure that the variation in cell density between individual
replicates was less than the variation between treatments (F-test p <0.05). Thus, all remaining
culture from two of the three replicates was transferred to a 50-mL centrifuge tube (which
could fit the volume of only two replicates at a time) and spun at 2,000 rpm for 10 minutes,
during which time a cell pellet accumulated. The supernatant was carefully removed by
pipetting and the volume of third test replicate was added and centrifuged, thus combining all
replicates of a control or a treatment into one pellet (n = 21 different pellets for assessment of
treatment effects, including 5 concentrations and solvent and media controls for 3 different
nutrient regimes).
Once the three replicates had been combined into a single pellet, the pellet was gently
aspirated and rinsed using cold, pH-adjusted, sterile ELA buffer (150 mM NaCl, 3.2 mM
117
KCl, 1.25 mM MgSO4, 3.75 mM CaCl2, and 12.2 mM Trisma base [TRIS], pH-adjusted to
7.4 with 1N HCl), and re-consolidated by centrifugation as above. The supernatant was again
carefully removed, and the pellet was resuspended and rinsed in cold assay buffer (for a total
of three rinses) to remove any residual test medium. The cell pellet was then resuspended in
5 mL of assay buffer, and the tube contents were transferred to a sterile, acid-stripped, 20-mL
glass beaker. An additional 10 mL of assay buffer was added to the centrifuge tube to rinse
any remaining cells into the beaker. The beaker contents were sonicated for 30 continuous
seconds at amplitude 45 on a 20kHz Sonic Dismembrator (Fisher Scientific Model 550,
Fisher Scientific, Pittsburgh, PA, USA). The sonicated pellet was used immediately in ELAs.
3.2.vi.c Hemolytic Assays
The modified ELAs used for analysis of hemolytic activity by Chattonella subsalsa
were conducted in 1-mL conical microcentrifuge tubes. Algal cellular contents were
extracted as described above, and equivolume amounts of erythrocytes and algal sample (300
µL) were added to a microcentrifuge tube (to enhance pellet formation; n = 3) and incubated
for 24 hr at 4°C and 30 µmol photons PAR m-2 s-1. Microcentrifuge tubes were then spun at
1250 rpm for 10 minutes in a refrigerated microcentrifuge (4oC; Micromax RF, IEC,
Needham Heights, MA, USA). Erythrocytes were also incubated in an equal volume of assay
buffer alone as a negative control, and in an equal volume of 20 µg mL-1 saponin solution
(Sigma-Aldrich preparation from the soap-bark tree Quillaja saponaria, a known hemolytic
compound) which was microscopically verified to cause 100% lysis of red blood cells and
served as a positive control. The microcentrifuge tube supernatant was transferred to a quartz
cuvette for measurement of sample absorbance (wavelength 414 nm, based on preliminary
scans of initial preparations of lysed erythrocytes which showed the highest hemoglobin
absorbance peaks, following Eschbach et al. 2001). The change in absorption at 414 nm
following exposure to sonicated cells was used to measure hemoglobin released by lysed
erythrocytes as a result of toxic activity of C. subsalsa, using the following normalization
equations:
[(𝑐𝑒𝑙𝑙𝑠⁄𝑚𝐿) ∙ (𝑉𝑐𝑒𝑛𝑡𝑟𝑖𝑓𝑢𝑔𝑒𝑑 ) ∙ (𝑉𝑤𝑒𝑙𝑙 )] = 𝑁
118
𝑠𝑎𝑚𝑝𝑙𝑒 − 𝑥 𝑛𝑒𝑔𝑎𝑡𝑖𝑣𝑒 𝑐𝑜𝑛𝑡𝑟𝑜𝑙
(
) ∙ 100 = % ℎ𝑒𝑚𝑜𝑙𝑦𝑠𝑖𝑠
𝑥𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝑐𝑜𝑛𝑡𝑟𝑜𝑙
(
% ℎ𝑒𝑚𝑜𝑙𝑦𝑠𝑖𝑠
) ∙ 106 = 𝑁𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑒𝑑 𝐿𝑦𝑡𝑖𝑐 𝑉𝑎𝑙𝑢𝑒
𝑁
where (𝑉𝑐𝑒𝑛𝑡𝑟𝑖𝑓𝑢𝑔𝑒𝑑 ) was the total volume of culture centrifuged to form assay pellets, and
(𝑉𝑤𝑒𝑙𝑙 ) (volume added to well) was 0.300 mL (300 µL).
3.2.vii Statistical Analysis
All statistical analyses were performed using JMP software (version 12.2.0, SAS
Institute, Cary, NC). Normality of the data was checked with the Shapiro-Wilke test, and
Bartlett's test for unequal variance was used to check variation. The only sample which failed
normality checks was the low P series on day 4 of the 10-day assays. Analysis of variance
(ANOVA) comparing nutrient and herbicide effects on growth considered the day 10 data
from that assay series, and days 10 and 28 from the 28-d assay series. The 10-day and 28-day
bioassays were independently analyzed.
Herbicide effects on growth were determined using a one-way analysis of variance
(ANOVA). No-observed effect concentrations (NOECs) and lowest-observed effect
concentrations (LOECs) were determined (p < 0.05) using Dunnett’s procedure for multiple
comparisons to determine which specific treatments differed significantly from controls (Zar
1999). Atrazine x nutrient comparisons were analyzed by two-way ANOVAs using nutrient
levels and herbicide concentration as the fixed effects, and growth rate or normalized
hemolytic activity as the dependent variable. The F statistic and associated P value for the
independent variables and the interactions between them were also reported. Point estimates
of sublethal toxicity (IC50s) were based on growth rates, and were calculated using the
Inhibition Concentration linear interpolation (ICp) method of Norberg-King (1993):
119
ICp  CJ  [ M 1(1  p /100)  M J ]
(CJ  1  CJ )
( MJ  1  MJ )
where:
CJ
≡ tested concentration whose observed mean response is greater than M1 (1-p/100);
CJ+1
≡ tested concentration whose observed mean response is less than M1(1-p/100);
M1
≡ smoothed mean response for the control;
MJ
≡ smoothed mean response for concentration J;
MJ+1
≡ smoothed response for concentration J +1;
P
≡ percent reduction in response relative to the control response; and
ICp
≡ estimated concentration at which there is a percent reduction from the smoothed
mean control response. The ICp is reported for the test together with the 95%
confidence interval calculated by the ICPIN.EXE program.
Results were generated using bootstrap methods to derive point estimates and confidence
intervals from no fewer than 80 resamplings (Efron 1982).
3.3 RESULTS
3.3.i Growth
Media controls in short-term (10-d) assay had cell density had more than doubled by
96 hours (1341 ±23 cells mL-1 versus 2794 ±261 cells mL-1) and had an almost 200%
increase in density by day 10 (4005 ±299 cells mL-1) (Fig. 3.3). Controls in the longerduration bioassays (28 days) began to decrease in cell density after day 23. Controls in
nutrient-replete (Redfield ratio) media remained in lag phase for the first 48 to 72 hr, and
then increased in cell density until day 23. Exponential growth occurred approximately
during days 4 through 20 and stationary growth began thereafter. Growth in nutrient-replete
cultures was similar to that in “nutrient-limited” cultures for the first seven days, reflecting
the influence of the nutrient supplies added with the initial parent stock solution of
Chattonella subsalsa. After that time, effects of nutrient depletion were detected.
120
3.3.ii Atrazine Effects
Atrazine concentration significantly affected the growth of Chattonella subsalsa in
nutrient-replete media at all time points after 96 hr, as shown by LOECs and IC50s (Tables
3.2 and 3.3). In the 28-day assays, overall sensitivity to atrazine decreased until day 15,
although there was no herbicide degradation based on ELISA checks of test samples on day
28. Effective dose estimates for C. subsalsa in nutrient-replete media leveled off, and
responses of C. subsalsa to atrazine concentrations were similar on days 15-28 (mean IC50 =
122.6 µg l-1). Atrazine effects on cell growth in nutrient-replete media followed typical doseresponse curves after 7-10 days. In the 10-day assays, nutrient-replete C. subsalsa cultures
were most resistant to atrazine at 96 hr (IC50 = 186.5 µg L-1), and sensitivity to atrazine
increased continuously until day 10 (IC50 = 46.9 µg L-1; Table 3.2, Fig. 3.4). In the 28-day
assays with nutrient-replete media, higher atrazine doses (100 to 200 µg L-1) had stronger
growth-limiting effects on C. subsalsa over time, whereas responses in mid-range atrazine
doses (25-50 µg L-1) were similar to responses of control cultures without atrazine by day 28
and the lowest dose (12.5 µg L-1) had slightly denser cell numbers than controls by day 28
(Fig. 3.4).
Effective concentration estimations based on growth rate inhibition were most precise
from days 4 through 10 in the 28-day assay (Fig. 3.5) and post-day 4 in the 10-day assay
(data not shown). Dose-response curves showed the increasing potency of atrazine from days
4 to 10 in the 10-day assay (Fig. 3.6), and between days 2-4 in the 28 day assay. In the
ANOVA model, atrazine concentration had the strongest effect on cell growth at day-10
endpoints in both assays, but all model variables were significant at that endpoint, including
the nutrient x atrazine interaction term (which did not outrank either treatment effect). At the
28-day endpoint there also were significant effects from all tested variables, but the effect of
nutrient limitation and the interaction term had much stronger influence on modeled growth
rates than the effect of herbicide concentration (two-way ANOVA, p ≤ 0.05, Table 3.4),
resulting in greatly diminished influence of herbicide inhibition on cell growth relative to
nutrient depletion by day 28.
121
Nitrogen and phosphorus limitation had approximately equal effects on growth rates
until day 10 in the assays, when nutrient limitation began to have observable effects on cell
growth (the actual point of initial growth depression effects was between days 7 and 10, but
observation was restricted to the measured endpoints). In the 10-day assay series, nutrientlimited cultures of C. subsalsa showed atrazine sensitivity in both nutrient limited treatments
was highest at the 24 hour mark (43.2µg L-1 in low N and 11.5 µg L-1 in Low P), but again,
following a period of wide ranging point estimates, effect estimates became more consistent
and precise by day 10 (Table 3.2). There were similar growth inhibition relationships by
atrazine on day 10 in the 10-day assays and 28-day assays (Figs. 3.7 and 3.8a). In the 28-day
assay series, however, there was significant growth stimulation by atrazine in both the N- and
P-limited cultures (Fig. 3.8b) relative to the low-nutrient controls. Thus, IC50 values could
not be determined for C. subsalsa in the low-nutrient treatments during late exponential
growth.
3.3.iii Toxicity as Hemolytic Activity
Considering nutrient regimes without atrazine, nutrient depletion, especially of P,
increased hemolytic activity in the Chattonella subsalsa no-atrazine controls at the 28-day
timepoint (Fig. 3.9), in agreement with previous reports of increased toxicity under nutrient
depletion in other harmful algal species (Johansson and Graneli 1999; Granéli and Johansson
2003b; Granéli et al. 2012). The 10-day series, by contrast, measured hemolytic activity
(HA) was significantly higher in nutrient-replete cultures, and HA was reduced in the low-N
and low-P treatments (Fig. 3.9). These findings are in contrast to reports that ROS production
(and, by proxy, toxicity) is not related to nutrient availability in Chattonella (Liu et al. 2007)
and potentially offers some context for reports that hemolytic activity is not increased by
nutrient limitation in other harmful raphidophytes (de Boer et al. 2004). As nutrient depletion
progressed over time, P-limited C. subsalsa exhibited significantly higher hemolytic activity
than the other treatments, and C. subsalsa hemolytic activity in both N- and P-limited
cultures was significantly higher than in nutrient-replete cultures in the late-exponential
growth phase. In the atrazine + nutrient-replete treatments, the toxicity response of C.
122
subsalsa differed in the 10-day versus 28-day assay series as well. In the 10-day assays,
highest hemolytic activity occurred in nutrient replete cultures at an atrazine concentration of
25 µg L-1, and toxicity decreased at higher atrazine concentrations. Hemolytic activity was
comparable to that of controls at intermediate atrazine concentrations (12.5 µg L-1 and 50 µg
L-1), and hemolytic activity in controls was significantly higher than in the highest atrazine
concentrations (100 µg L-1 and 200 µg L-1) (Fig. 10a). In the atrazine x low-nutrient
treatments in the 10-day assays, C. subsalsa became increasingly toxic with increasing
atrazine concentration except for in the 200 µg L-1 treatment (Fig. 3.10a); at that
concentration, cell density and hemolytic activity were not significantly different from zero
in any treatment. The low-N and low-P cultures showed similar responses to atrazine in
terms of hemolytic activity (Fig. 3.10a), but were much lower in hemolytic activity than the
nutrient-replete controls without atrazine. An exception occurred at 100 µg atrazine L-1; at
that herbicide concentration, the low-nutrient cultures exhibited significantly higher
hemolytic activity than nutrient-replete cultures.
In the 28-day assay series, there was no apparent effect of atrazine concentration on
hemolytic activity in nutrient-replete cultures, which had low toxicity in comparison to
nutrient-limited cultures at all atrazine concentrations. At the 28-d endpoint, the end-of-test
cell density in replete culture was negligible at 200µg L-1, as noted above, however all low
nutrient • atrazine treatments had higher cell density by day 28 than 200 µg atrazine L-1
replete cultures (see appendix Table A3 for 28-d nutrient limited culture cell density). With
the exception of the 200 µg L-1 treatment, cultures in the low-P treatment were comparable in
hemolytic activity across the atrazine concentration gradient, whereas in the low-N treatment,
the hemolytic activity of C. subsalsa was significantly higher at 100 µg atrazine L-1 than at
other herbicide levels (Fig. 3.10b). Both N- and P-limited C. subsalsa in late exponential
growth had significantly higher hemolytic activity at atrazine concentrations above 25 µg L-1
(again, with exception of the 200 µg L-1 level). Overall, hemolytic activity in C. subsalsa was
more influenced by atrazine concentration in the 10-day assay series, whereas the nutrient
regime was the best predictor of hemolytic activity in the 28-day series (Table 5). Log-
123
normal relationships between hemolytic activity (normalized to cell density and sample
volume, and omitting the 200 µg atrazine L-1 treatment in low nutrient regimes) and atrazine
concentration had high r2 values (0.82 to 0.92) across nutrient treatments in 10-day assay
(Fig. 3.11) but the strength of these relationships declined over time (see appendix Fig. A4
for modeled relationships at 28-d endpoint). Hemolytic activity was inversely related to logatrazine concentration in nutrient-replete cultures, but positively related to the log-atrazine
concentration in low nutrient treatments. The r2 values were somewhat lower for hemolytic
activity versus growth rate in the low-nutrient cultures (0.64 to 0.77) than in the nutrientreplete regime (0.85), but the relationships still were clear: In nutrient-replete cultures,
hemolytic activity decreased with increasing atrazine concentration, but under nutrientlimited conditions hemolytic activity increased with increasing exposure to herbicide.
Hemolytic activity also increased up to intermediate growth rates and then declined
(parabolic response curve) in nutrient replete media, whereas in nutrient-limited cultures,
hemolytic activity was inversely related to growth rate (Fig. 3.11).
3.5 DISCUSSION
This study was the first to test atrazine sensitivity in a toxigenic raphidophyte species,
wherein the toxicity (as hemolytic activity) of Chattonella subsalsa was exposed to a range
of photosystem II-inhibiting atrazine concentrations under different nutrient regimes.
Inhibition by atrazine was the dominant predictor of growth rates for this strain in early
exponential growth phase (10-day assay series), whereas the nutrient regime was a more
important predictor of C. subsalsa population growth (as cell density) during late exponential
growth (28-day assay series). Hemolytic activity in the absence of atrazine was highest in
low-P cultures among all controls and treatments. Hemolytic activity in the presence of > 25
µg atrazine L-1 was highest in low-N cultures, and increased with increasing atrazine
concentration.
Research with other harmful algae, such as the haptophyte Prymnesium parvum and
the dinoflagellate Karenia brevis, have shown increasing toxin production under nutrient
124
depletion or nutrient limitation (Johansson and Graneli 1999a,b; Granéli and Johansson 2003;
Hardison et al. 2013). These findings are consistent with the “chemical ecology hypothesis,”
which predicts enhanced toxicity of nutrient-limited HAB species (Ianora et al. 2011). That
hypothesis was derived, in turn, from a “carbon:nutrient balance hypothesis” (Bryant et al.
1983; Lambers et al. 2008), which predicted increased production of toxins and other
defenses in terrestrial plants under low resource conditions wherein grazing is reduced. The
latter hypothesis may not hold up well in the terrestrial systems for which it was designed
(Hamilton et al. 2001), but it has been used successfully to explain increased toxicity with
nutrient limitation in unicellular algae (e.g., Ianora et al. 2006; Ianora et al. 2011; Hardison et
al. 2012). Production of toxic compounds by phytoplankton has frequently been associated
with allelopathic interactions (Legrand 2001; Fistarol et al. 2005; Uronen et al. 2005; Granéli
et al. 2008; Driscoll et al. 2013; Fernández-Herrera et al. 2016), which may help reduce
grazing and competition for scarce resources and promote survival in competitive (nutrientimbalanced) systems.
Here, increasing toxicity as hemolytic activity was also observed at low to moderate
growth rates in nutrient-replete C. subsalsa cultures, and at low growth rates in nutrientlimited cultures. Others have reported that the closely related species, Chattonella marina, is
most toxic during late exponential growth phase and toxicity decreases over stationary phase
(measured as putative brevetoxins affecting fish; Khan et al. 1995; Marshall et al. 2003). The
toxicity of C. marina has been highly correlated with growth phase, but not with cell density
(Shen et al. 2010); or, lower cell densities of C. marina (below 104 cells mL-1) produced
more superoxide on a per-cell basis than denser cultures (Marshall et al. 2005b). These
differences may relate to the tested strain or other factors. Similarly, other studies on
individual phytoplankton species responses to chronic atrazine exposures have yielded mixed
results. Some have reported that algal populations with a history of exposure became more
sensitive to atrazine upon repeated exposure (Pennington and Scott 2001), whereas others
have reported that phytoplankton became more tolerant over time (Baxter et al. 2013). These
discrepancies may be attributed to the wide range of phylogenic diversity which is often
125
mistakenly “lumped” under the umbrella of “phytoplankton responses.” The ecological
diversity of the huge, extremely diverse organisms referred to as algae cannot be considered
as one functional group (although they frequently are so considered), or even as one trophic
level (e.g., phototrophs versus predatory mixotrophs).
Phytoplankton assemblages have been investigated for response to repeated atrazine
exposure – also with conflicting results, including increased sensitivity (Hamala and Kollig
1985; Guasch et al. 2007), no effect (Nyström et al. 2000; Pinckney et al. 2002), and
increased tolerance (deNoyelles et al. 1982; Hamilton et al. 1987). These findings are not
surprising, as the ability of a phytoplankton assemblage to recover from herbicide exposure
following photosystem II inhibition depends on the species composition of the original
assemblage, their physiological state, the herbicide concentration (Gustavson and Wängberg
1995; Kasai 1999), the season (Guasch et al. 1997; Lorente et al. 2015), and other
environmental factors. Several consistent findings, however, are that (i) phytoplankton
species and strains vary greatly in response to atrazine exposure (Blanck et al. 1984; Choi et
al. 2012); (ii) intraspecific variation in responses can be very high (Kasai et al. 1993); and
(iii) atrazine exposure can alter phytoplankton assemblage composition as sensitive species
are eliminated and tolerant species gain a selective advantage (Hamala and Kollig 1985;
Hamilton et al. 1987; Graymore et al. 2001). Thus, it is important to strengthen
understanding about how harmful, potentially ecosystem-disruptive phytoplankton species
(Sunda et al. 2006) respond to common environmental contaminants, despite the challenges
involved in attempting to apply standard ecotoxicological methods to organisms which may
be more challenging to culture than standard model organisms. Autecological and ecotoxicological studies of potentially toxic species are relevant for understanding the
mechanisms that lead to the formation of harmful algal blooms and the influence of common
contaminants such as atrazine on toxin production. Regarding Chattonella subsalsa, this
research offers insights about the complexities of nutrient interactions with atrazine in
influencing growth and toxicity, yet – as only one strain was examined – it is only a first step
toward understanding C. subsalsa response.
126
In the past few decades there has been a dramatic increase in the frequency,
magnitude, and duration of HABs, with devastating impacts on global fisheries stocks,
coastal economies and freshwater and marine ecosystems (Hallegraeff 1993, Burkholder
1998, Van Dolah 2000, Glibert et al. 2005, Heisler et al. 2008, Lewitus et al. 2012). It has
also been more than two decades since Cloern (1996) discussed the potential for toxic
contaminants and nutrient pollution to alter natural cycles of estuarine phytoplankton blooms,
yet there still are very few studies about how these combined stressors affect HAB species.
This research indicates that imbalanced nutrient regimes can act in conjunction with
herbicide exposures to promote hemolytic activity in C. subsalsa, a toxigenic algal species
that is expected to continue to thrive in increasingly urbanized coastal zones.
127
3.6 FIGURES AND TABLES
Table 3.1. Culture media recipes used for culturing and assays of Chattonella subsalsa.
L1 Macro-nutrient concentrations
Replete:
NaNO3
NaH2PO4
low N:
NaNO3
NaH2PO4
low P:
NaNO3
NaH2PO4
Stock Solution
Volume added to
final medium
Final concentration in
L1 media (µM)
2.7198 g l-1 H2O
0.2400 g l-1 H2O
1 mL
1 mL
32.00
2.00
0.1700 g l-1 H2O
0.2400 g l-1 H2O
1 mL
1 mL
2.00
2.00
2.7198 g l-1 H2O
0.0240 g l-1 H2O
1 mL
1 mL
32.00
0.20
L1 Trace Element Solution
Primary stock
solution
Na2EDTA ·
2H2O
FeCl3 · 6H2O
MnCl2 · 4H2O
ZnSO4 · 7H2O
CoCl2 · 6H2O
CuSO4 · 5H2O
Na2MoO4 · 2H2O
H2SeO3
NiSO4 · 6H2O
NaVO4
K2CrO4
Quantity to add to
final medium
Molar
concentration in
final medium (M)
---
4.36 g
1.17 x 10-5
--178.10 g L-1 H2O
23.00 g L-1 H2O
11.90 g L-1 H2O
2.5 g L-1 H2O
19.9 g L-1 H2O
1.29 g L-1 H2O
2.63 g L-1 H2O
1.84 g L-1 H2O
1.94 g L-1 H2O
3.15 g
1 mL
1 mL
1 mL
1 mL
1 mL
1 mL
1 mL
1 mL
1 mL
1.17 x 10-5
9.09 x 10-7
8.00 x 10-8
5.00 x 10-8
1.00 x 10-8
8.22 x 10-8
1.00 x 10-8
1.00 x 10-8
1.00 x 10-8
1.00 x 10-8
L1 Vitamin Solution
Primary Stock
Solution
Quantity to Add
to Final Medium
Molar Concentration in
Final Medium (M)
---
200 mg
2.96 x 10-7
biotin (Vitamin H)
0.1 g L-1 H2O
10 mL
2.05 x 10-9
cyanocobalamin (Vitamin B12)
1.0 g L-1 H2O
1 mL
3.69 x 10-10
thiamine · HCl (Vitamin B)
128
Table 3.2. Growth inhibition and lowest-observed effect concentrations (µg atrazine L-1) over time in
10-day bioassays with Chattonella subsalsa. BD ≡ beyond detection.
Replete
Low N
Low P
1
3
4
7
10
1
3
4
7
10
1
3
4
7
10
IC50:
18.8
88.6
186.5
74.4
46.9
43.2
147.8
99.5
107.7
119.7
11.5
102.5
88.2
50.0
64.7
CI(lower)
166.7
44.0
40.4
6.3
84.5
105.7
7.8
48.2
39.3
58.5
CI(upper)
BD
BD
208.6
114.2
56.9
104.5
BD
BD
151.7
140.2
82.1
BD
126.4
57.1
68.0
IC10: CI(lower) CI(upper)
13.8
BD
2.9
57.7
65.5
5.6
65.8
62.8
9.9
35.2
29.5
7.5
21.3
12.9
1.3
55.0
28.4
4.1
72.5
40.2
3.6
54.4
8.3
4.5
26.7
7.1
8.4
25.6
9.9
1.5
30.5
2.3
17.8
62.4
21.8
16.5
20.9
18.0
5.2
12.6
7.3
6.0
25.7
8.5
LOEC:
200
BD
BD
50
25
BD
BD
200
100
100
BD
BD
200
25
50
129
Table 3.3. Growth inhibition and lowest-observed effect concentrations (µg atrazine L-1) over time in
28-day replete media bioassays with Chattonella subsalsa. BD ≡ beyond detection; ND ≡ not
determined.
Day
1
2
3
4
7
10
15
20
23
28
IC50 (CI)
146.2 (BD)
ND
42.3 (32.8 - 8.7)
36.1 (27.7- 43.2)
46.5 (36.9 - 55.3)
69.1 (56.5 - 81.0)
105.3 (89.0 - 128.2)
130.8 (100.2 - 147.4)
124.3 (97.8 - 142.3)
129.7 (96.4 - 154.7)
IC10 (CI)
61.5 (5.1 – 122.9)
ND
10.6 (3.8 – 10.4)
7.5 (3.5 - 28.3)
6.3 (3.9 - 25.3)
8.8 (5.4 - 18.5)
11.69 (8.9 - 21.2)
36.7 (12.4 - 58.2)
55.6 (10.1 - 62.3)
61.9 (51.2 - 64.6)
LOEC
BD
ND
100
50
50
25
50
100
100
100
130
Table 3.4. Results of two-way ANOVAs examining atrazine concentration (Conc) and nutrient
regime (Trt) effects on growth rates of Chattonella subsalsa in bioassays at different endpoints.
Source
Day 10; 10-day assay Conc
Trt
Conc*Trt
Day 10; 28-day assay Conc
Trt
Conc*Trt
Day 28; 28-day assay Conc
Trt
Conc*Trt
DF
F Ratio
Prob > F
5
2
10
5
2
10
5
2
10
150.80
15.64
7.19
84.76
41.25
3.03
4.44
25.59
25.33
<.0001
<.0001
<.0001
<.0001
<.0001
0.0070
0.0030
<.0001
<.0001
131
Table 3.5. Results of two-way ANOVAs examining atrazine concentration (Conc) and nutrient
regime (Trt) effects on hemolytic activity of Chattonella subsalsa in bioassays at the 10-day and 28day endpoints.
10-day assay
28-day assay
Source
DF
F Ratio
Prob > F
Conc
Trt
Conc*Trt
Conc
Trt
Conc*Trt
5
2
10
5
2
10
118.90
191.05
55.79
70.95
297.94
13.69
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
132
a
b
Figure 3.1. a) Chattonella subsalsa cells fixed with 1% PFA + 1%glut (as described in
procedures). b) Chattonella subsalsa cells fixed with acidic Lugol’s solution. (Scale bar = 50
µm).
133
Figure 3.2. Example of the modeled relationships between optical density (OD) and C. subsalsa cell
density that was used to monitor population growth for atrazine inhibition assays.
134
Figure 3.3. Growth of C. subsalsa in control cultures during 10-day (a) and 28-day (b) bioassays.
Data are given as means + 1 SE.
135
a
b
Figure 3.4. Growth curves for Chattonella subsalsa in (a) 10-day and (b) 28-day bioassays with
atrazine under nutrient-replete conditions. Data are given as means + 1 SE.
136
Figure 3.5. Variation in atrazine half-maximal effective concentrations for Chattonella subsalsa in
28-day nutrient-replete bioassays over time. Confidence intervals (bars) around the means are shown.
137
Figure 3.6. Dose-response relationships showing increasing potency of atrazine in 10day, nutrient-replete growth bioassays (+ sign, dashed line ≡ day 4; x sign, solid line ≡
day 7; * sign, solid line ≡ day 10).
138
Figure 3.7. Inhibition concentration curves for different nutrient regimes tested in 10-day atrazine
exposure assay. Data are given as means + 1 SE.
139
10 days
28 days
Figure 3.8. Inhibition concentration curves for different nutrient regimes tested in 28-day
atrazine exposure assays, and measured at 10 and 28 days (upper and lower panel, respectively).
Error bars are 1 SEM.
140
Figure 3.9. Normalized hemolytic activity in control cultures, measured at end of 10- and 28-day
assays. Data are given as means + 1 SE.
141
a
b
Figure 3.10. Growth rates (divisions day-1) and normalized hemolytic activity in all tested atrazine
concentrations, nutrient treatments and controls. Data are given as means + 1 SE.
142
Nutrient-replete
Low N
Low P
Figure 3.11. Log-normal relationships between hemolytic activity (normalized to cell density and
sample volume) and atrazine concentration (a, c, e) or growth rates (b, d, f) at 10-day endpoint. Panels
a and b ≡ nutrient-replete media; and panels c and d ≡ low N; panels e and f ≡ low P. Note that panels
c-f (low-nutrient treatments) exclude 200 µg atrazine L-1 cultures due to near-complete cell loss (see
text).
143
Figure 3.12. Hemolytic activity at different test endpoints, normalized to cell density and sample volume, in various
nutrient regimes under a range of atrazine concentrations. Data are given as means + 1 SE
144
3.7 REFERENCES
Ahmad S. 1995. Antioxidant mechanisms of enzymes and proteins. In: Oxidative
Stress and Antioxidant Defenses in Biology (Ahmad, S., ed.). New York, NY:
Springer, 238-272.
Ahmed MS, Khan S, Arakawa O, Onoue Y. 1995. Properties of hemagglutinins newly
separated from toxic phytoplankton. Biochim. Biophys. Acta 1243: 509-512.
American Society for Testing and Materials (ASTM). 1996. Standard Guide for
Conducting Static 96-Hour Toxicity Tests with Microalgae. West
Conshohocken, PA: ASTM.
Band-Schmidt CJ, Martinez-Lopez A, Bustillos-Guzmán JJ, Carréon-Palau L,
Morquecho L, Olguín-Monroy N, Zenteno-Savín T, Mendoza-Flores A,
González-Acosta B, Hernández-Sandoval F. 2012. Morphology, biochemistry,
and growth of raphidophyte strains from the Gulf of California. Hydrobiologia
693:81-97.
Baxter L, Brain R, Prosser R, Solomon K, Hanson M. 2013. Sensitivity of a green alga
to atrazine is not enhanced by previous acute exposure. Environ. Pollut.
181:325-328.
Bérard A, Benninghoff C. 2001. Pollution-induced community tolerance (PICT) and
seasonal variations in the sensitivity of phytoplankton to atrazine in nanocosms.
Chemosphere 45:427-437.
Blanck H. 2002. A critical review of procedures and approaches used for assessing
pollution-induced community tolerance (PICT) in biotic communities. HERA
(Human and Ecological Risk Assessment) 8:1003-1034.
Blanck H, Wallin G, Wangberg SA. 1984. Species-dependent variation in algal
sensitivity to chemical compounds. Ecotoxicol. Environ. Saf. 8:339-351.
Blanck H, Wängberg S-Å, Molander S. 1988. Pollution-induced community
tolerance—a new ecotoxicological tool. In: Functional Testing of Aquatic Biota
for Estimating Hazards of Chemicals (Cairns J Jr, Pratt JR, eds.). STP 988.
Philadelphia, PA: ASTM, 219-230.
145
Bricker SB, Longstaff B, Dennison W, Jones A, Boicourt K, Wicks C, Woerner J.
2008. Effects of nutrient enrichment in the nation's estuaries: a decade of
change. Harmful Algae 8:21-32.
Bryant JP, Chapin III FS, Klein DR. 1983. Carbon/nutrient balance of boreal plants in
relation to vertebrate herbivory. Oikos:357-368.
Burkholder JM. 1998. Implications of harmful marine microalgae and heterotrophic
dinoflagellates in management of sustainable marine fisheries. Ecol. Appl. 8:
S37-S62.
Burkholder JM, Glibert PM, Skelton HM. 2008. Mixotrophy, a major mode of nutrition
for harmful algal species in eutrophic waters. Harmful Algae 8:77-93.
Burkholder J, Libra B, Weyer P, Heathcote S, Kolpin D, Thome PS, Wichman M.
2007. Impacts of waste from concentrated animal feeding operations on water
quality. Environmental Health Perspectives:308-312.
Caron DA. 2016. Mixotrophy stirs up our understanding of marine food webs. Proc.
Nat. Acad. Sci. (USA) 113:2806-2808.
Choi CJ, Berges JA, Young EB. 2012. Rapid effects of diverse toxic water pollutants
on chlorophyll a fluorescence: variable responses among freshwater
microalgae. Water Research 46:2615-2626.
Cloern JE. 1996. Phytoplankton bloom dynamics in coastal ecosystems: a review with
some general lessons from sustained investigation of San Francisco Bay,
California. Rev. Geophys. 34:127-168.
Cortés-Altamirano R, Alonso-Rodríguez R, Sierra Beltrán A. 2006. Fish mortality
associated with Chattonella marina and C. cf. ovata (Raphidophyceae) blooms
in Sinaloa (Mexico). Harmful Algae News:7-8.
de Boer MK, Tyl MR, Vrieling EG, van Rijssel M. 2004. Effects of salinity and
nutrient conditions on growth and haemolytic activity of Fibrocapsa japonica
(Raphidophyceae). Aquatic Microbial Ecology 37:171-181.
DeLorenzo ME, Scott GI, Ross PE. 2001. Toxicity of pesticides to aquatic
microorganisms: a review. Environ. Toxicol. Chem. 20:84-98.
146
deNoyelles F, Kettle WD, Sinn DE. 1982. The responses of plankton communities in
experimental ponds to atrazine, the most heavily used pesticide in the United
States. Ecology:1285-1293.
Driscoll WW, Espinosa NJ, Eldakar OT, Hackett JD. 2013. Allelopathy as an
emergent, exploitable public good in the bloom-forming microalga Prymnesium
parvum. Evolution 67:1582-1590.
Efron B. 1982. The jackknife, the bootstrap and other resampling plans. Vol. 38:
Society for Industrial and Applied Mathematics (SIAM).
Eisler R. 1989. Atrazine hazards to fish, wildlife, and invertebrates: a synoptic review.
Biological Report 85(1.8), Contaminant Hazard Reviews Report No. 18.
Laurel, MD: United States Fish and Wildlife Service.
Eschbach E, Scharsack JP, John U, Medlin LK. 2001. Improved erythrocyte lysis assay
in microtitre plates for sensitive detection and efficient measurement of
haemolytic compounds from ichthyotoxic algae. J. Appl. Toxicol. 21:513-519.
Fernández-Herrera LJ, Band-Schmidt CJ, López-Cortés DJ, Hernández-Guerrero CJ,
Bustillos-Guzmán JJ, Núñez-Vázquez E. 2016. Allelopathic effect of
Chattonella marina var. marina (Raphidophyceae) on Gymnodinium catenatum
(Dinophyceae). Harmful Algae 51:1-9.
Fistarol GO, Legrand C, Graneli E. 2005. Allelopathic effect on a nutrient-limited
phytoplankton species. Aquat. Microb. Ecol. 41:153-161.
Fu M, Koulman A, van Rijssel M, Lützen A, de Boer MK, Tyl MR, Liebezeit G. 2004.
Chemical characterisation of three haemolytic compounds from the microalgal
species Fibrocapsa japonica (Raphidophyceae). Toxicon 43:355-363.
Gascón J, Oubiña A, Barceló D. 1997. Detection of endocrine-disrupting pesticides by
enzyme-linked immunosorbent assay (ELISA): application to atrazine. Trends
Anal. Chem. 16:554-562.
Gilliom RJ, Barbash JE, Crawford CG, Hamilton PA, Martin JD, Nakagaki N, Nowell
LH, Scott JC, Stackelberg PE, Thelin GP. 2006. Pesticides in the nation's
streams and ground water, 1992-2001. Reston, Virginia: United States
Geological Survey.
147
Glibert PM, S. Seitzinger, C.A. Heil, J.M. Burkholder, M.W. Parrow, L.A. Codispoti,
and V. Kelly. 2005. The role of eutrophication in the global proliferation of
harmful algal blooms. Oceanography 18:198-209.
Graham LE, J.M. Graham JM, Wilcox LW, Cook ME. 2015. Algae, 3rd edition.
Madison, WI: LJLM Press, LLC.
Granéli E, Edvardsen B, Roelke DL, Hagström JA. 2012. The ecophysiology and
bloom dynamics of Prymnesium spp. Harmful Algae 14:260-270.
Granéli E, Johansson N. 2003. Increase in the production of allelopathic substances by
Prymnesium parvum cells grown under N-or P-deficient conditions. Harmful
Algae 2:135-145.
Graymore M, Stagnitti F, Allinson G. 2001. Impacts of atrazine in aquatic ecosystems.
Environment International 26:483-495.
Guasch H, Lehmann V, Van Beusekom B, Sabater S, Admiraal W. 2007. Influence of
phosphate on the response of periphyton to atrazine exposure. Arch. Environ.
Contam. Toxicol. 52:32-37.
Guasch H, Muñoz I, Rosés N, Sabater S. 1997. Changes in atrazine toxicity throughout
succession of stream periphyton communities. J. Appl. Phycol. 9:137-146.
Guillard R, Hargraves P. 1993. Stichochrysis immobilis is a diatom, not a chrysophyte.
Phycologia 32:234-236.
Guillard RR. 1973. Division rates. In: Handbook of Phycological Methods (Stein JR,
ed). Cambridge, United Kingdom: Cambridge University Press, 290-311.
Gustavson K, Wängberg S-Å. 1995. Tolerance induction and succession in microalgae
communities exposed to copper and atrazine. Aquat. Toxicol. 32:283-302.
Hallegraeff G, Hara Y. 2003. Taxonomy of harmful marine raphidophytes. In:
Hallegraeff GM, Anderson DM, Cembella AD (eds) Manual on Harmful
Marine Microalgae, Vol 11, 2nd edn. IOC-UNESCO, Paris, pp 511–522
Hallegraeff G, Munday B, Baden D, Whitney P. Chattonella marina raphidophyte
bloom associated with mortality of cultured bluefin tuna (Thunnus maccoyii) in
South Australia. In: Harmful Algae, VIII Conference on Harmful Algae, Vigo,
148
Spain, June 25-29, 1997 (Reguera B, Blanco J, Fernández M, Wyatt T, eds.).
Paris, France: International Oceanographic Commission of UNESCO, 93-96.
Hallegraeff GM. 1993. A review of harmful algal blooms and their apparent global
increase. Phycologia 32:79-99.
Hamala JA, Kollig HP. 1985. The effects of atrazine on periphyton communities in
controlled laboratory ecosystems. Chemosphere 14:1391-1408.
Hamilton J, Zangerl A, DeLucia E, Berenbaum M. 2001. The carbon–nutrient balance
hypothesis: its rise and fall. Ecology Letters 4:86-95.
Hamilton PB, Jackson GS, Kaushik NK, Solomon KR. 1987. The impact of atrazine on
lake periphyton communities, including carbon uptake dynamics using track
autoradiography. Environ. Pollut. 46:83-103.
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. Aquat. Microb.
Ecol. 40:121-132.
Hapeman CJ, Dionigi CP, Zimba PV, McConnell LL. 2002. Agrochemical and nutrient
impacts on estuaries and other aquatic systems. Journal of Agricultural and
Food Chemistry 50:4382-4384.
Hardison DR, Sunda WG, Shea D, Litaker RW. 2013. Increased toxicity of Karenia
brevis during phosphate limited growth: ecological and evolutionary
implications. PLoS ONE 8(3): e58545. doi:10.1371/journal.pone.0058545.
Hardison DR, Sunda WG, Litaker RW, Shea D, Tester PA. 2012. Nitrogen limitation
increases brevetoxins in Karenia brevis (Dinophyceae): implications for bloom
toxicity. J. Phycol. 48:844-858.
Harris G P. 1986. Phytoplankton Ecology: Structure, Function and Fluctuation.
London, United Kingdom: Chapman and Hall.
Heisler J, Glibert PM, Burkholder JM, Anderson DM, Cochlan W, Dennison WC,
Dortch Q, Gobler CJ, Heil CA, Humphries E. 2008. Eutrophication and harmful
algal blooms: a scientific consensus. Harmful Algae 8:3-13.
149
Hiroishi S, Okada H, Imai I, Yoshida T. 2005. High toxicity of the novel bloomforming species Chattonella ovata (Raphidophyceae) to cultured fish. Harmful
Algae 4:783-787.
Ianora A, Bentley MG, Caldwell GS, Casotti R, Cembella AD, Engström-Öst J,
Halsband C, Sonnenschein E, Legrand C, Llewellyn CA. 2011. The relevance
of marine chemical ecology to plankton and ecosystem function: An emerging
field. Marine Drugs 9:1625-1648.
Ianora A, Boersma M, Casotti R, Fontana A, Harder J, Hoffmann F, Pavia H, Potin P,
Poulet S, Toth G. 2006. New trends in marine chemical ecology. Estuar. Coasts
29:531-551.
Imai I, Itakura S, Itoh K. 1991. Life cycle strategies of the red tide causing flagellates
Chattonella (Raphidophyceae) in the Seto Inland Sea. Marine Pollution
Bulletin 23:165-170.
Imai I, Yamaguchi M. 2012. Life cycle, physiology, ecology and red tide occurrences
of the fish-killing raphidophyte Chattonella. Harmful Algae 14:46-70.
Imai I, Yamaguchi M, Watanabe M. 1998. Ecophysiology, life cycle, and bloom
dynamics of Chattonella in the Seto Inland Sea, Japan. In: Physiological
Ecology of Harmful Algal Blooms, (Anderson DM, Cembella AD, Hallegraeff
GM, eds.). Berlin, Germany: Springer-Verlag, 95-112.
Ishimatsu A, Oda T, Yoshida M, Ozaki M. 1996a. Oxygen radicals are probably
involved in the mortality of yellowtail by Chattonella marina. Fish. Sci.
62:836-837.
Ishimatsu A, Sameshima M, Tamura A, Oda T. 1996b. Histological Analysis of the
Mechanisms of Chattonella-Induced Hypoxemia in Yellowtail. Fish. Sci.
62:50-58.
Jeong HJ. 2011. Mixotrophy in red tide algae raphidophytes. J. Eukaryot. Microbiol.
58: 215-222.
Jeong HJ., Seong KA, Kang NS, Yoo YD, Nam SW, Park JY, Shin W, Glibert PM,
John D. 2010. Feeding by raphidophytes on the cyanobacterium Synechococcus
sp. Aquat. Microb. Ecol. 58:181-195.
150
Johansson N, Graneli E. 1999a. Cell density, chemical composition and toxicity of
Chrysochromulina polylepis (Haptophyta) in relation to different N: P supply
ratios. Mar. Biol. 135:209-217.
Johansson N, Graneli E. 1999b. Influence of different nutrient conditions on cell
density, chemical composition and toxicity of Prymnesium parvum
(Haptophyta) in semi-continuous cultures. J. Exp. Mar. Biol. Ecol. 239:243258.
Kasai F. 1999. Shifts in herbicide tolerance in paddy field periphyton following
herbicide application. Chemosphere 38:919-931.
Kasai F, Takamura N, Hatakeyama S. 1993. Effects of simetryne on growth of various
freshwater algal taxa. Environ. Pollut. 79:77-83.
Katano T, Yoshida M, Lee J, Han M-S, Hayami Y. 2009. Fixation of Chattonella
antiqua and C. marina (Raphidophyceae) using Hepes-buffered
paraformaldehyde and glutaraldehyde for flow cytometry and light microscopy.
Phycologia 48:473-479.
Khan S, Ahmed M, Arakawa O, Onoue Y. 1995. Properties of neurotoxins separated
from a harmful red tide organism Chattonella marina. Isr. J. Aquacult.-Bamid.
47:137-141.
Khan S, Arakawa O, Onoue Y. 1996. A toxicological study of the marine
phytoflagellate, Chattonella antiqua (Raphidophyceae). Phycologia 35:239244.
Kim D, Okamoto T, Oda T, Tachibana K, Lee KS, Ishimatsu A, Matsuyama Y, Honjo
T, Muramatsu T. 2001. Possible involvement of the glycocalyx in the
ichthyotoxicity of Chattonella marina (Raphidophyceae): Immunological
approach using antiserum against cell surface structures of the flagellate. Mar.
Biol. 139: 625–632.
Kim D, Watanabe M, Nakayasu Y, Kohata K. 2004. Production of superoxide anion
and hydrogen peroxide associated with cell growth of Chattonella antiqua.
Aquat. Microb. Ecol. 35:57-64.
Kim D, Watanabe M, Nakayasu Y, Kohata K. 2005. Changes in O2–and H2O2
production by Chattonella antiqua during diel vertical migration under nutrient
stratification. Aquat. Microb. Ecol. 39:183-191.
151
Kim D, Yamaguchi K, Oda T. 2006. Nitric oxide synthase-like enzyme mediated nitric
oxide generation by harmful red tide phytoplankton, Chattonella marina. J.
Plankton Res. 28:613-620.
Kuroda A, Nakashima T, Yamaguchi K, Oda T. 2005. Isolation and characterization of
light-dependent hemolytic cytotoxin from harmful red tide phytoplankton
Chattonella marina. Comp. Biochem. Physiol. C, Toxicol. Pharmacol. 141:297305.
Lambers H, Chapin III FS, Pons TL. 2008. Ecological biochemistry: allelopathy and
defense against herbivores. In: Plant Physiological Ecology: Springer, 445-477.
Legrand C. 2001. Phagotrophy and toxicity variation in the mixotrophic Prymnesium
patelliferum (Haptophyceae). Limnol. Oceanogr. 46:1208-1214.
Lewitus AJ, Brock LM, Burke MK, DeMattio KA, Wilde SB. 2008. Lagoonal
stormwater detention ponds as promoters of harmful algal blooms and
eutrophication along the South Carolina coast. Harmful Algae 8:60-65.
Lewitus AJ, Horner RA, Caron DA, Garcia-Mendoza E, Hickey BM, Hunter M,
Huppert DD, Kudela RM, Langlois GW, Largier JL. 2012. Harmful algal
blooms along the North American west coast region: History, trends, causes,
and impacts. Harmful Algae 19:133-159.
Lewitus AJ, Holland AF. 2003a. Initial results from a multi-institutional collaboration
to monitor harmful algal blooms in South Carolina. Environmental Monitoring
and Assessment 81:361-371.
Lewitus AJ, Schmidt LB, Mason LJ, Kempton JW, Wilde SB, Wolny JL, Williams BJ,
Hayes KC, Hymel SN, Keppler CJ. 2003b. Harmful algal blooms in South
Carolina residential and golf course ponds. Popul. Environ. 24:387-413.
Liu W, Au DW, Anderson DM, Lam PK, Wu RS. 2007. Effects of nutrients, salinity,
pH and light: dark cycle on the production of reactive oxygen species in the
alga Chattonella marina. Journal of Experimental Marine Biology and Ecology
346:76-86.
Lorente C, Causape J, Glud RN, Hancke K, Merchan D, Muniz S, Val J, Navarro E.
2015. Impacts of agricultural irrigation on nearby freshwater ecosystems: The
seasonal influence of triazine herbicides in benthic algal communities. Sci.
Total Environ. 503:151-158.
152
Lytle JS, Lytle TF. 2001. Use of plants for toxicity assessment of estuarine ecosystems.
Environmental Toxicology and Chemistry 20:68-83.
Marshall J-A, Nichols PD, Hallegraeff GM. 2002. Chemotaxonomic survey of sterols
and fatty acids in six marine raphidophyte algae. J. Appl. Phycol. 14:255-265.
Marshall J-A, Nichols PD, Hamilton B, Lewis RJ, Hallegraeff GM. 2003.
Ichthyotoxicity of Chattonella marina (Raphidophyceae) to damselfish
(Acanthochromis polycanthus): the synergistic role of reactive oxygen species
and free fatty acids. Harmful Algae 2:273-281.
Marshall J-A, Salas M, Oda T, Hallegraeff G. 2005a. Superoxide production by marine
microalgae I. Mar. Biol. 147:541-549.
Marshall J-A, Ross T, Pyecroft S, Hallegraeff G. 2005b. Superoxide production by
marine microalgae II. Mar. Biol. 147:541-549.
Matsusato T, Kobayashi H. 1974. Studies on death of fish caused by red tide. Bull.
Nansei Reg. Fish. Res. Lab. 7: 43-67.
Matsuyama Y, Uchida T, Kotani Y. 2000. Effect of culture filtrate of raphidophytes
Heterosigma akashiwo and Chattonella antiqua on the growth of diatom
Skeletonema costatum. Bull. Environ. Inland Sea 2: 57-66.
Moore LR. 2013. More mixotrophy in the marine microbial mix. Proc. Nat. Acad. Sci.
(USA) 110:8323-8324.
Norberg-King TJ. 1993. A linear interpolation method for sublethal toxicity: The
inhibition concentration (ICp) approach. National Effluent Toxicity Assessment
Center Technical Report 39:3-93.
Nygaard K, Tobiesen A. 1993. Bacterivory in algae: a survival strategy during nutrient
limitation. Limnol. Oceanogr. 38:273-279.
Nyström B, Paulsson M, Almgren K, Blank H. 2000. Evaluation of the capacity for
development of atrazine tolerance in periphyton from a Swedish freshwater site
as determined by inhibition of photosynthesis and sulfolipid synthesis. Environ.
Toxicol. Chem. 19:1324-1331.
153
Oda T, Ishimatsu A, Shimada M, Takeshita S, Muramatsu T. 1992. Oxygen-radicalmediated toxic effects of the red tide flagellate Chattonella marina on Vibrio
alginolyticus. Mar. Biol. 112:505-509.
Oda T, Nakamura A, Shikayama M, Kawano I, Ishimatsu A, Muramatsu T. 1997.
Generation of reactive oxygen species by raphidophycean phytoplankton.
Biosci. Biotech. Biochem. 61:1658-1662.
Pelley J. 1998. What is causing toxic algal blooms? Environmental Science &
Technology 32:26A-30A.
Pennington PL, Daugomah JW, Colbert AC, Fulton MH, Key PB, Thompson BC,
Strozier ED, Scott GI. 2001. Analysis of pesticide runoff from mid-Texas
estuaries and risk assessment implications for marine phytoplankton. J.
Environ. Sci. Heal. B 36:1-14.
Pennington PL, Scott GI. 2001. Toxicity of atrazine to the estuarine phytoplankter
Pavlova sp. (Prymnesiophyceae): Increased sensitivity after long-term, lowlevel population exposure. Environ. Toxicol. Chem. 20:2237-2242.
Peters NE, Meybeck M, Chapman DV. 2005. Effects of human activities on water
quality. Encyclopedia of Hydrological Sciences.
Pinckney JL, Örnólfsdóttir EB, Lumsden SE. 2002. Estuarine phytoplankton groupspecific responses to sublethal concentrations of the agricultural herbicide,
atrazine. Mar. Pollut. Bull. 44:1109-1116.
Pistocchi R, Guerrini F, Pezzolesi L, Riccardi M, Vanucci S, Ciminiello P,
Dell’Aversano C, Forino M, Fattorusso E, Tartaglione L, Milandri A, Pompei
M, Cangini M, Pigozzi S, Riccardi E. 2012. Toxin levels and profiles in
microalgae from the north-western Adriatic Sea – 15 years of studies on
cultured species. Mar. Drugs 10:140-162.
Redfield AC. 1934. On the proportions of organic derivatives in sea water and their relation
to the composition of plankton. In: James Johnstone Memorial Volume. Liverpool:
University of Liverpool Press, p.176-192.
Redfield AC. 1958. The biological control of chemical factors in the environment. Am. Sci.
46:205-221.
154
Ryberg KR, Vecchia AV, Martin JD, Gilliom RJ. 2010. Trends in Pesticide
Concentrations in Urban Streams in the United States, 1992-2008. Reston, VA:
United States Geological Survey.
Scott GI, Holland AF, Sandifer PA. 2006. Afterword—Managing Coastal Urbanization
and Development in the Twenty-First Century: The Need for a New Paradigm.
In: Changing Land Use Patterns in the Coastal Zone: Managing Environmental
Quality in Rapidly Developing Regions (Kleppel GS, ed.). New York, NY:
Springer, 285-299.
Seguin F, Le Bihan F, Leboulanger C, Bérard A. 2002. A risk assessment of pollution:
induction of atrazine tolerance in phytoplankton communities in freshwater
outdoor mesocosms, using chlorophyll fluorescence as an endpoint. Water
Research 36:3227-3236.
Shen M, Xu J, Chiang MW, Au DW. 2011. Unravelling the pathway of respiratory
toxicity in goldlined seabream (Rhabdosargus sarba) induced by the harmful
alga Chattonella marina. Aquat. Toxicol. 104:185-191.
Shen M, Xu J, Tsang TY, Au DW. 2010. Toxicity comparison between Chattonella
marina and Karenia brevis using marine medaka (Oryzias melastigma):
evidence against the suspected ichthyotoxins of Chattonella marina.
Chemosphere 80:585-591.
Shimada M, Murakami TH, Imahayashi T, Ozaki HS, Toyoshima T, Okaichi T. 1983.
Effects of sea bloom, Chattonella antiqua, on gill primary lamellae of the
young yellowtail, Seriola quinqueradiata. Acta Histochem. Cytochem. 16:232244.
Smayda TJ. 2002. Adaptive ecology, growth strategies and the global bloom expansion
of dinoflagellates. J. Oceanogr. 58:281-294.
Sorokin C. 1973. Dry weight, packed cell volume and optical density. Handbook of
Phycological Methods, Culture Methods and Growth Measurement (Stein JR,
ed.). Cambridge, United Kingdom: Cambridge University Press, 321-343.
Thursby GB, Anderson BS, Walsh GE, Steele RL. 1993. A review of the current status
of marine algal toxicity testing in the United States. In: Environmental
Toxicology and Risk Assessment: ASTM International.
155
Tittel J, Bissinger V, Zippel B, Gaedke U, Bell E, Lorke A, Kamjunke N. 2003.
Mixotrophs combine resource use to outcompete specialists: implications for
aquatic food webs. Proc. Nat. Acad. Sci. (USA) 100:12776-12781.
Uronen P, Lehtinen S, Legrand C, Kuuppo P, Tamminen T. 2005. Haemolytic activity
and allelopathy of the haptophyte Prymnesium parvum in nutrient-limited and
balanced growth conditions. Mar. Ecol. Prog. Ser. 299:137-148.
United States Environmental Protection Agency (U.S. EPA). 2016a. National Coastal
Condition Report IV. Report EPA-842-R-10-003. Washington, DC: U.S. EPA.
United States Environmental Protection Agency (U.S. EPA). 2016b. National Lakes
Assessment 2012: A Collaborative Survey of Lakes in the United States. Report
EPA 841-R-16-113. Washington, DC: U.S. EPA.
Vencill WK. 2002. Herbicide Handbook, 10th edition. Lawrence, KS: Weed Science
Society of America.
Watanabe M, Kohata K, Kimura T. 1991. Diel vertical migration and nocturnal uptake
of nutrients by Chattonella antiqua under stable stratification. Limnol.
Oceanogr. 36:593-602.
Weiner JA, DeLorenzo ME, Fulton MH. 2007. Atrazine induced species-specific
alterations in the subcellular content of microalgal cells. Pest. Biochem.
Physiol. 87:47-53.
Yamaguchi H, Sakamoto S, Yamaguchi M. 2008. Nutrition and growth kinetics in
nitrogen-and phosphorus-limited cultures of the novel red tide flagellate
Chattonella ovata (Raphidophyceae). Harmful Algae 7:26-32.
Zar JH. 1999. Biostatistical Analysis, 4th edition. Upper Saddle River, NJ: Prentice
Hall, 389-394.
Zhang Y, Fu F-X, Whereat E, Coyne KJ, Hutchins DA. 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 5:310-320.
Zimba PV. 2000. Role of nutrient runoff in altering downstream ecosystems: Florida
Bay, FL, USA. In: Proceedings of the 220th National Meeting of the American
Chemical Society; American Chemical Society. Washington DC. p. 159
156
APPENDIX
157
Table A1. Effect concentrations for all tested strains of Prymnesium parvum, over time. (BD
=Beyond Detection)
Strain Salinity
TX
10
Nutrient
R
Low N
Low P
20
R
Low N
Low P
NC
10
R
Low N
Low P
NC
20
R
Point Estimate
IC01
0.5 (0.3 - 0.7)
IC10
4.9 (3.3 - 7.1)
IC25
24.1 (9.4 - 55.9)
IC50
80.2 (67.2 - 91.2)
IC01
0.8 (0.5 - 1.5)
IC10
6.9 (4.9 - 13.7)
IC25
54.5 (12.4 - 73.1)
IC50
158.9 (148.3 - 168.6)
IC01
0.7 (0.5 - 0.8)
IC10
6.5 (5.3 - 8.2)
IC25
37.1 (28.8 - 49.8)
IC50
162.5 (140.1 - 175.3)
IC01
7.0 (2.4 - 26.2)
IC10
40.2 (31.5 - 52.1)
IC25
75.8 (66.5 - 84.6)
IC50
142.2 (132.5 - 149.0)
IC01
32.7 (15.7 - 52.8)
IC10
72.8 (55.0 - 101.5)
IC25
113.3 (82.6 - 133.6)
IC50
174.3 (156.5 - 192.2)
IC01
21.2 (14.3 - 30.1)
IC10
55.1 (46.8 - 60.5)
IC25
87.1 (74.6 - 102.2)
IC50
150.5 (132.1 - 166.6)
IC01
3.9 (0.7 - 14.8)
IC10
19.2 (7.3 - 29.7)
IC25
38.8 (31.4 - 47.5)
IC50
73.0 (50.6 - 83.8)
IC01
29.8 (7.3 - 54.9)
IC10
87.8 (49.5 - 106.6)
IC25
152.0 (132.9 - 170.4)
IC50
BD
IC01
5.1 (2.2 - 14.6)
IC10
36.6 (20.1 - 54.0)
IC25
79.6 (72.3 - 86.5)
IC50
162.2 (154.8 - 170.6)
IC01
IC10
IC25
14.7 (1.8 - 1.8)
44.8 (22.0 - 60.1)
71.8 (60.7 - 79.5)
158
Table A1 (continued)
Low N
Low P
SC
10
R
Low N
Low P
SC
20
R
Low N
Low P
IC50
IC01
IC10
IC25
IC50
IC01
IC10
IC25
IC50
IC01
IC10
IC25
IC50
IC01
IC10
IC25
IC50
IC01
IC10
IC25
IC50
IC01
IC10
IC25
IC50
IC01
IC10
IC25
IC50
IC01
IC10
IC25
IC50
118.3 (103.4 - 133.9)
4.9 (1.1 - 31.9)
36.4 (10.7 - 59.0)
64.2 (47.6 - 75.3)
108.7 (94.6 - 125.3)
10.0 (0.6 - 13.2)
14.3 (4.7 - 20.2)
21.3 (11.8 - 32.7)
66.0 (24.3 - 141.0)
13.8 (13.0 - 16.7)
22.6 (18.2 - 31.9)
41.9 (34.4 - 53.7)
88.2 (69.8 - 106.0)
54.1 (52.4 - 61.3)
86.8 (73.3 - 118.3)
137.5 (111.5 - 159.0)
BD
24.2 (13.9 - 51.6)
61.4 (53.1 - 66.9)
88.2 (82.6 - 91.6)
161.7 (151.0 - 175.7)
10.8 (4.1 - 57.2)
101.6 (86.1 - 115.7)
151.4 (143.1 - 159.6)
BD
3.8 (1.0 - 17.7)
26.8 (10.1 - 58.0)
115.7 (100.5 - 146.6)
BD
6.2 (0.9 - 27.6)
57.0 (10.4 - 122.5)
BD
BD
159
Figure A1. Distribution of P. parvum effective concentrations
Table A2. 4-way ANOVA modeling the effects of herbicide concentration, nutrient regime,
salinity and strain on P. parvum growth:
96-hour variable importance; independent resamples
Column
Conc
Strain
Salinity
Nutr
Main Effect
0.402
0.066
0.088
0.081
Main Effect
Std Err
0.009
0.008
0.009
0.008
Total Effect
0.533
0.381
0.372
0.265
Total Effect
Std Err
0.01
0.009
0.009
0.008
day 10 variable importance:
Column
Nutr
Conc
Strain
Salinity
Main Effect
0.509
0.101
0.106
0.078
Main Effect
Std Err
0.008
0.007
0.007
0.006
Total Effect
0.622
0.205
0.199
0.151
Total Effect
Std Err
0.01
0.006
0.005
0.005
160
Table A3. 3-way ANOVA modeling the effects of herbicide concentration, nutrient regime,
salinity on P. parvum growth, strains are analyzed separately:
TX variable importance: independent resamples
Column;
Overall
Nutr
Conc
Salinity
Column:
96 -hour
Conc
Nutr
Salinity
Column:
day-10
Nutr
Salinity
Conc
Main Effect
0.372
0.304
0.118
Main Effect
0.557
0.116
0.052
Main Effect
0.628
0.183
0.052
Main Effect
Std Err
0.004
0.004
0.003
Main Effect
Std Err
0.007
0.007
0.004
Main Effect
Std Err
0.005
0.005
0.004
Total Effect
0.54
0.451
0.188
Total Effect
0.781
0.362
0.146
Total Effect
0.718
0.23
0.12
Total Effect
Std Err
0.005
0.005
0.003
Total Effect
Std Err
0.01
0.007
0.004
Total Effect
Std Err
0.007
0.004
0.003
NC variable importance: independent resamples
Column;
overall
Conc
Nutr
Salinity
Column;
96-hour
Conc
Salinity
Nutr
Column;
day 10
Nutr
Conc
Salinity
Main Effect
0.42
0.262
0.128
Main Effect
0.542
0.225
0.054
Main Effect
0.469
0.298
0.031
Main Effect
Std Err
0.005
0.004
0.004
Main Effect
Std Err
0.006
0.007
0.004
Main Effect
Std Err
0.008
0.009
0.003
Total Effect
0.581
0.399
0.226
Total Effect
0.684
0.365
0.187
Total Effect
0.61
0.478
0.087
Total Effect
Std Err
0.006
0.005
0.003
Total Effect
Std Err
0.008
0.006
0.004
Total Effect
Std Err
0.01
0.009
0.003
161
Table A3. (continued)
SC variable importance: independent resamples
Column;
overall
Nutr
Salinity
Conc
Column;
96-hour
Salinity
Nutr
Conc
Column;
day 10
Nutr
Conc
Salinity
Main Effect
0.441
0.259
0.111
Main Effect
0.422
0.226
0.128
Main Effect
0.657
0.094
0.096
Main Effect
Std Err
0.005
0.005
0.005
Main Effect
Std Err
0.008
0.009
0.007
Main Effect
Std Err
0.007
0.007
0.006
Total Effect
0.59
0.37
0.195
Total Effect
0.586
0.409
0.193
Total Effect
0.77
0.196
0.154
Total Effect
Std Err
0.006
0.005
0.004
Total Effect
Std Err
0.009
0.008
0.005
Total Effect
Std Err
0.01
0.006
0.005
151
Table A4. ELISA sample QA/QC checks to quantify measurement of nominal doses.
Samples:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Sample ID
Test ID
PpNC20 d10
PpNC20 d10
PpNC20 d10
PpNC20 d10
PpNC20 d10
PpNC20 d10
PpNC20 d10
PpNC20 d10
PpNC20 d10
PpNC20 d10
Csub d28
Csub d28
Csub d28
Csub d28
Csub d28
Csub d28
Csub d28
Csub d28
Csub d28
Csub d28
PpNC20 d10
PpNC20 d10
PpNC20 d10
PpNC20 d10
PpNC20 d10
Csub d28
Csub d28
Csub d28
Csub d28
Csub d28
PpNC20 d10
PpNC20 d10
Csub d28
Csub d28
Treatment
R
R
R
R
R
-P
-N
-P
-N
-P
R
R
R
R
R
-N
-P
-N
-P
-N
R
R
R
R
R
R
R
R
R
R
-N
-P
-N
-P
Actual conc
12.5
25
50
100
200
12.5
25
50
100
200
12.5
25
50
100
200
12.5
25
50
100
200
12.5
25
50
100
200
12.5
25
50
100
200
25
100
100
100
Accuracy:
Measured conc
(% recovery)
10.45
84
24.99
100
35.31
71
95.51
96
149.02
75
11.88
95
20.20
81
42.84
86
111.70
112
183.80
92
13.31
106
26.53
106
44.48
89
89.96
90
197.88
99
13.16
105
27.96
112
55.85
112
98.04
98
166.98
83
13.56
109
20.20
81
53.01
106
102.53
103
219.22
110
19.23
154
18.82
75
72.84
146
119.46
119
209.05
105
24.99
100
80.97
81
96.58
97
84.71
85
152
Table A4. (continued)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
1
2
3
4
5
6
7
8
9
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
STANDARD
STANDARD
STANDARD
Csub6hr
Csub6hr
Csub6hr
Csub6hr
Csub6hr
Csub6hr
Csub6hr
Csub6hr
Csub6hr
Csub6hr
Csub6hr
Csub6hr
Csub6hr
Csub6hr
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
PpNC10 day 10
R
R
-N
-N
-N
-N
-P
-P
-P
-P
-P
R
R
R
R
R
-N
-N
-N
-N
-N
-P
-P
-P
-P
R
R
-N
-N
-N
-N
-P
-P
-P
100
100
12.5
25
100
200
12.5
25
50
100
200
0
1
3
12.5
25
50
100
200
12.5
25
50
100
200
12.5
50
100
200
100
100
12.5
25
100
200
12.5
25
50
99.76
103.34
10.11
20.26
110.65
163.79
14.68
30.91
56.56
120.61
200.89
0.24
1.04
2.62
3.28
2.99
2.21
2.36
3.58
3.46
2.81
2.52
2.59
3.91
3.96
2.49
2.43
4.20
99.56
103.36
10.01
20.35
110.48
161.16
14.40
30.56
56.31
100
103
81
81
111
82
117
124
113
121
100
104
87
105
120
111
118
90
111
112
126
130
98
127
125
122
105
100
103
80
81
110
81
115
122
113
153
Table A4. (continued)
10
11
12
13
14
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
PpNC10 day 10
PpNC10 day 10
STANDARD
STANDARD
STANDARD
PpTX10 d10
PpTX10 d10
PpTX10 d10
PpTX10 d10
PpTX10 d10
PpTX10 d10
PpTX10 d10
PpTX10 d10
PpTX10 d10
PpTX10 d10
Dt d10
Dt d10
Dt d10
Dt d10
Dt d10
-P
-P
R
R
R
R
R
-N
-P
-N
-P
-P
R
-N
-N
R
-P
100
200
0
1
3
12.5
25
50
100
200
12.5
25
50
100
200
12.5
25
50
100
200
120.09
196.81
0.23
1.05
2.61
8.73
22.64
58.26
123.72
102.62
8.23
21.95
65.49
111.50
100.79
8.44
22.57
62.92
124.22
110.28
120
98
105
87
94
107
125
107
86
110
116
93
139
90
102
108
103
106
72
154
Figure A2. Growth curves for D. tertiolecta over test duration of 20-d atrazine exposure.
155
Figure A3. Growth curves for C. subsalsa over test duration of 28-d atrazine exposure.
156
Figure A4. Relationship between hemolytic activity and atrazine or growth in C. subsalsa at 28-days post inoculation