Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Application of evolutionary theory and methods to aquatic ecotoxicology
(USC Thesis Other)
Application of evolutionary theory and methods to aquatic ecotoxicology
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Application of evolutionary theory and methods to aquatic ecotoxicology
by
Alice L. Coleman
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOLOGY (MARINE BIOLOGY AND
BIOLOGICAL OCEANOGRAPHY))
December 2024
ii
Acknowledgements
Firstly, I would like to thank my advisor, Suzanne Edmands for responding to my very
first email to her in August 2018 and to every silly little email of mine since. Her patience,
understanding and encouragement of my independent research interests has helped me grow
tremendously as a scientist. Thanks are also due to the members of my committee Carly Kenkel,
Melissa Guzman and Ian Ehrenreich for their insightful and constructive feedback on my work.
Thank you to all of the past and present members of the Edmands Lab with whom I
overlapped, including Eric Watson, Ben Flanagan, Murad Jah, Scott Applebaum, and particularly
my partners-in-crime Jake Denova, Kimberly Schoenberger and Eliza Kirsch. The fun and jovial
community we cultivated in our lab has meant the world to me, and I am grateful to each of them
for their friendship and support during the exceptionally challenging final stretch of my work. I
would also like to thank the dedicated team of USC undergraduate students that assisted me in
the wet lab, without whom my Tigriopus project would never have gotten off the ground. I hope
that I was the mentor to them that they deserved.
I am also indebted to the USC Sea Grant Program, the Southern California Academy of
Sciences and the Southern California chapter of the Society of Environmental Toxicology and
Chemistry for providing me with funding support throughout the various stages of my research.
Lastly, I would like to thank my family and friends for their unwavering support of me over the
years, as listening to the complaints of a cranky graduate student is not easy.
iii
Table of Contents
Acknowledgements .....................................................................................................................................ii
List of Tables...............................................................................................................................................v
List of Figures.............................................................................................................................................vi
Abstract.....................................................................................................................................................viii
Chapter 1: Introduction .............................................................................................................................1
References................................................................................................................................................9
Chapter 2: Data and diversity in the development of acute water quality criteria in the United
States..........................................................................................................................................................13
Abstract..................................................................................................................................................13
Introduction...........................................................................................................................................14
Methods..................................................................................................................................................17
Results....................................................................................................................................................22
Discussion ..............................................................................................................................................25
Conclusions............................................................................................................................................35
References..............................................................................................................................................35
Figures and Tables................................................................................................................................39
Supplementary Materials.....................................................................................................................45
Chapter 3: Phylogeny predicts sensitivity in aquatic animals for only a minority of chemicals.......58
Abstract..................................................................................................................................................58
Introduction...........................................................................................................................................59
Methods..................................................................................................................................................62
Results....................................................................................................................................................66
Discussion ..............................................................................................................................................69
Conclusions............................................................................................................................................73
References..............................................................................................................................................73
Figures and Tables................................................................................................................................80
Supplementary Materials.....................................................................................................................95
Chapter 4: Heat exposure improves copper tolerance in the intertidal copepod Tigriopus
californicus...............................................................................................................................................132
Abstract................................................................................................................................................132
Introduction.........................................................................................................................................133
Methods................................................................................................................................................137
Results..................................................................................................................................................143
Discussion ............................................................................................................................................146
iv
References............................................................................................................................................153
Figures and Tables..............................................................................................................................160
Supplementary Materials...................................................................................................................171
Chapter 5: Conclusions ..........................................................................................................................179
References............................................................................................................................................181
Comprehensive References....................................................................................................................183
v
List of Tables
Table 2-1: Properties of analyzed chemicals..............................................................................................41
Table 2-2: Comparison of mean criteria from datasets that were assembled using the minimum data
requirements to the mean criteria from randomly assembled datasets........................................................42
Table 2-3: Linear model summary for the Shannon diversity index of a quality-controlled toxicity
dataset as a predictor of criterion value ......................................................................................................43
Table 2-4: Kruskal-Wallis comparison of phyla ........................................................................................44
Table 3-1: Properties of chemicals evaluated in analyses..........................................................................90
Table 3-2: Phylogenetic signal results for all dataset variations................................................................91
Table 3-3: Results of PGLS regressions for temperature and pH with toxicity data .................................93
Table S3-1: Toxicity dataset sample sizes before and after filtering for availability of experimental
life stage, temperature, pH and hardness information.................................................................................95
Table S3-2: Summary of NOEC effects information in chronic toxicity datasets.....................................96
Table 4-1: The number of differentially expressed antioxidant, heat shock protein, stress-induced,
chitin and cuticle genes in each treatment. ...............................................................................................170
Table S4-1: Approximate coordinates for each of the sampled Tigriopus californicus populations.......173
Table S4-2: Number of raw reads and percentage of reads mapped to the San Diego reference
genome in each sample .............................................................................................................................174
Table S4-3: Copepod mortalities after a thermal stress test at 35°C and after 24 hours in 60 mg/L Cu..174
Table S4-4: Differentially expressed genes shared by all four treatment groups.....................................175
vi
List of Figures
Figure 1-1: Dose-response curve. ................................................................................................................2
Figure 1-2: Conceptual diagram of cross-protection responses.. .................................................................8
Figure 2-1: Taxonomic composition of the trimmed acute toxicity datasets from marine and
freshwater species for twelve chemicals that were available in 1985 and 2021. ........................................39
Figure 2-2: Distribution of acute toxicity data for endosulfan from saltwater and freshwater species......40
Figure S2-1: Taxonomic composition of acute toxicity data before application of quality control
parameters from saltwater and freshwater species for twelve chemicals that were available in 1985
and 202........................................................................................................................................................45
Figure S2-2: Histograms of Δ+ values from 1000 random subsamples of saltwater and freshwater
species data. ................................................................................................................................................46
Figure S2-3: Distribution of acute toxicity data for ammonia from saltwater and freshwater species. .....47
Figure S2-4: Distribution of acute toxicity data for atrazine from saltwater and freshwater species ........48
Figure S2-5: Distribution of acute toxicity data for cadmium from saltwater and freshwater species......49
Figure S2-6: Distribution of acute toxicity data for copper from saltwater and freshwater species..........50
Figure S2-7: Distribution of acute toxicity data for saltwater from saltwater and freshwater species.......51
Figure S2-8: Distribution of acute toxicity data for nickel from saltwater and freshwater species. ..........52
Figure S2-9: Distribution of acute toxicity data for 4-nitrophenol from saltwater and freshwater
species.........................................................................................................................................................53
Figure S2-10: Distribution of acute toxicity data for pentachlorophenol from saltwater and freshwater
species.........................................................................................................................................................54
Figure S2-11: Distribution of acute toxicity data for phenol from saltwater and freshwater species........55
Figure S2-12: Distribution of acute toxicity data for tributyltin oxide from saltwater and freshwater
species.........................................................................................................................................................56
Figure S2-13: Distribution of acute toxicity data for toluene from saltwater and freshwater species.......57
Figure 3-1: Phylogenetic tree and toxicity data heatmap for the acute chlorine dataset............................80
Figure 3-2: Phylogenetic tree and toxicity data heatmap for the acute guthion dataset.............................81
Figure 3-3: Phylogenetic tree and toxicity data heatmap for the acute lindane dataset .............................83
Figure 3-4: Phylogenetic tree and toxicity data heatmap for the acute pentachlorophenol dataset ...........85
Figure 3-5: Combined phylogenetic tree and toxicity data heatmap for the chronic cadmium dataset .....85
Figure 3-6: Combined phylogenetic tree and toxicity data heatmap for the chronic copper dataset .........86
Figure 3-7: Combined phylogenetic tree and toxicity data heatmap for the chronic diazinon dataset ......87
Figure 3-8: Combined phylogenetic tree and toxicity data heatmap for the chronic lindane dataset ........88
Figure 3-9: Combined phylogenetic tree and toxicity data heatmap for the chronic phenol dataset .........89
Figure S3-1: Phylogenetic tree and toxicity data heatmap for the complete acute ammonia dataset ........97
Figure S3-2: Phylogenetic tree and toxicity data heatmap for the complete acute atrazine dataset ..........98
Figure S3-3: Phylogenetic tree and toxicity data heatmap for the complete chronic atrazine dataset.....100
Figure S3-4: Phylogenetic tree and toxicity data heatmap for the complete acute cadmium dataset ......102
Figure S3-5: Phylogenetic tree and toxicity data heatmap for the complete acute chlorpyrifos dataset..104
Figure S3-6: Phylogenetic tree and toxicity data heatmap for the complete chronic chlorpyrifos
dataset .......................................................................................................................................................105
Figure S3-7: Phylogenetic tree and toxicity data heatmap for the complete acute copper dataset ..........107
Figure S3-8: Phylogenetic tree and toxicity data heatmap for the complete acute DDT dataset ...........109
Figure S3-9: Phylogenetic tree and toxicity data heatmap for the complete acute diazinon dataset .......111
Figure S3-10: Phylogenetic tree and toxicity data heatmap for the complete acute dieldrin dataset.......113
Figure S3-11: Phylogenetic tree and toxicity data heatmap for the complete acute endosulfan dataset..115
vii
Figure S3-12: Phylogenetic tree and toxicity data heatmap for the complete chronic endosulfan dataset
..................................................................................................................................................................115
Figure S3-13: Phylogenetic tree and toxicity data heatmap for the complete acute endrin dataset.........117
Figure S3-14: Phylogenetic tree and toxicity data heatmap for the complete acute glyphosate dataset..118
Figure S3-15: Phylogenetic tree and toxicity data heatmap for the complete chronic glyphosate
dataset .......................................................................................................................................................118
Figure S3-16: Phylogenetic tree and toxicity data heatmap for the complete acute malathion dataset ...120
Figure S3-17: Phylogenetic tree and toxicity data heatmap for the complete chronic malathion dataset 120
Figure S3-18: Phylogenetic tree and toxicity data heatmap for the complete acute mercury dataset......121
Figure S3-19: Phylogenetic tree and toxicity data heatmap for the complete acute nickel dataset .........122
Figure S3-20: Phylogenetic tree and toxicity data heatmap for the complete acute 4-nitrophenol
dataset .......................................................................................................................................................123
Figure S3-21: Phylogenetic tree and toxicity data heatmap for the complete acute parathion dataset ....124
Figure S3-22: Phylogenetic tree and toxicity data heatmap for the complete chronic
pentachlorophenol dataset.........................................................................................................................125
Figure S3-23: Phylogenetic tree and toxicity data heatmap for the complete acute phenol dataset. .......127
Figure S3-24: Phylogenetic tree and toxicity data heatmap for the complete acute TBTO dataset.........128
Figure S3-25: Phylogenetic tree and toxicity data heatmap for the complete acute toluene dataset. ......129
Figure S3-26: Phylogenetic tree and toxicity data heatmap for the complete acute zinc dataset ............131
Figure S3-27: Phylogenetic tree and toxicity data heatmap for the complete chronic zinc dataset.........131
Figure 4-1: Map of the seven Tigriopus californicus populations collected from the California
coastline ....................................................................................................................................................160
Figure 4-2: Linear regressions between median lethal temperatures and median lethal concentrations
with latitude and as well as with each other..............................................................................................161
Figure 4-3: Median lethal temperature of seven T. californicus populations with and without a brief
exposure to copper prior to thermal testing. .............................................................................................162
Figure 4-4: Median lethal concentration of seven T. californicus populations with and without a brief
exposure to heat stress prior to toxicity testing.........................................................................................163
Figure 4-5: Survival probability of five populations over a 17 day exposure in the five chronic stress
exposure experiments, grouped by treatment. ..........................................................................................164
Figure 4-6: Linear regressions of acute and chronic tolerance for heat and copper.................................165
Figure 4-7: Principal components analysis (PCA) of the top 500 most variable genes from all twelve
samples, and for the samples from each population considered separately. .............................................166
Figure 4-8: Venn diagram of the number of differentially expressed genes shared between all
population-treatment combinations...........................................................................................................167
Figure 4-9: Hierarchical clustering of biological process (BP) gene ontology (GO) categories in
Bodega Bay copepods exposed to heat and copper. .................................................................................168
Figure 4-10: Hierarchical clustering of biological process (BP) gene ontology (GO) categories in
La Jolla copepods exposed to heat and copper. ......................................................................................169
Figure S4-1: Distribution of the number of offspring at first count from five populations in the five
chronic stress exposure experiments.........................................................................................................171
Figure S4-2: Survival probability of five populations over a 17 day exposure in the five chronic
stress exposure experiments, grouped by population................................................................................172
Figure S4-3: Heatmap of the rlog transformed expression values of the top 250 most variable genes
across all groups........................................................................................................................................173
viii
Abstract
Chemical pollution represents one of the most potent anthropogenic forces of change in
the aquatic environment. Precise understanding of the toxic effects of chemicals on aquatic
organisms is therefore crucial to pollution mitigation efforts, however, the role of evolutionary
processes in shaping chemical sensitivity has been understudied. This dissertation seeks to
address this gap in knowledge by integrating theory and methods from evolutionary biology into
two distinct areas of research in aquatic ecotoxicology: the development of pollution regulations
and the study of multiple stressors. Firstly, I performed two broad meta-analyses to evaluate the
taxonomic composition of the toxicity data used to derive water quality criteria and to quantify
the extent to which phylogeny determines chemical sensitivity. I then utilized in vivo stress
experiments and transcriptomic sequencing to evaluate the interpopulation variation in copper
tolerance in the intertidal copepod Tigriopus californicus and to investigate whether the coppertolerant phenotype is an exaptation derived from the heat stress response mechanism. Toxicity
data were found to be lacking in biological diversity, although the taxonomic composition of
datasets did not quantitatively affect water quality criteria in most cases. Broad taxonomic
patterns of sensitivity were also evident in these data, although phylogeny was a significant
predictor of sensitivity for only a minority of chemicals. Additionally, I found that copper
tolerance in T. californicus populations was positively correlated with heat tolerance, and that
there was overlap between the gene expression responses for both stressors that could confer
fitness benefits during multiple stressor exposure. Overall, this work advances the integration of
evolutionary biology and ecotoxicology and highlights the importance of evolutionary processes
in both a regulatory and ecological context.
1
Chapter 1: Introduction
Aquatic ecotoxicology
The aquatic environment is the final destination for nearly all anthropogenic
contaminants, with more than 100,000 substances introduced into the environment in recent
history (Amiard-Triquet 2015). These contaminants pose a serious threat to both aquatic life and
the valuable services derived from these ecosystems, creating an urgent need to develop robust
management strategies and strict legislation that can mitigate the deleterious effects of pollution.
In order to do so, comprehensive knowledge of the effects of toxic chemicals on aquatic
biological systems is required, which constitutes the field of aquatic ecotoxicology.
Pollutants are typically transported into aquatic ecosystems by one of three main routes:
direct dumping, runoff and atmospheric deposition (Beiras 2018). The sources of pollutants can
be further classified as point (single discrete origin) or nonpoint (multiple possible origins)
depending on the degree to which the chemicals are dispersed in the environment. Once present,
the residence times of chemical pollutants vary, with some readily degraded by chemical,
biological or photolytic processes while others remain in the environment for years. These
persistent pollutants can undergo long-range transport and accumulate in areas distant from
where they were used, and are capable of affecting entire food webs through the processes of
biomagnification and bioaccumulation. Exposure in aquatic organisms occurs by either the
ingestion of a contaminated food source or the direct absorption of waterborne chemicals via a
non-dietary route such as respiration (Barron 2002), and the consequences of such exposure can
vary. Organism mortality is often the most easily identifiable symptom of pollutant exposure, but
other major effects include inhibition of reproduction, immune system suppression, disruption of
the endocrine system and cellular/molecular damage, among others (Shahidul Islam and Tanaka
2
2004). One of the main objectives of ecotoxicology is to identify which concentrations of
hazardous chemicals trigger hazardous effects in different species using toxicity tests.
The most common test in ecotoxicology is the single species acute toxicity test.
Performed in a controlled laboratory setting, these tests involve exposing a population of test
organisms to a dilution series of a suspected toxicant for a short period on a single occasion. The
measured endpoint in these tests can be a lethal or sublethal response, although in practice
mortality tends to be the most common outcome. When testing is complete, chemical
concentrations and the corresponding biological responses are plotted together and fit with a
distribution to create a dose-response curve (Figure 1-1). Various parameters of toxicity can be
calculated from the curve such as the threshold or median effective concentrations, which
respectively describe the minimum
concentration of a chemical that causes
harmful effects and the concentration
that causes a 50% reduction in the
biological response variable. These
toxicity data are then used to develop
environmental standards which are
intended to limit the exposure of organisms to harmful concentrations of pollutants. However,
the acute toxicity database generated through these means suffers from at least two major flaws:
a lack of taxonomic breadth and an inability to describe the effects of multiple stressors. This
dissertation seeks to address these gaps in knowledge by applying theory and methods from
evolutionary biology.
Figure 1-1: Dose-response curve. From the Toxicology
Education Foundation.
3
Evolutionary processes in ecotoxicology
Chemical pollution is a potent force of ecological and evolutionary change. Exposure to
pollutants can reduce population size and decrease genetic diversity, rendering that population
more vulnerable to the effects of other environmental changes and at greater risk of extinction.
Chemical effects can also be transmitted across generations through either epigenetic changes
(Vandegehuchte and Janssen 2014) or by the induction of heritable mutations in germ cells
(Bickham et al. 2000), which may have profound evolutionary consequences. One of the most
prominent examples of chemical pollution initiating evolutionary change is the repeated
observation of enhanced tolerance in populations with histories of pollution exposure in a wide
range of species (Brown 1978; Klerks and Levinton 1989; Barata et al. 2002; Ownby et al. 2002;
Nacci et al. 2010). The development of a tolerant phenotype depends both on the characteristics
of the exposed population (generation time, size, level of genetic variation) and the chemical
agent (mode of action, intensity), and additionally may impose a cost to fitness that could
constrain adaptation (Whitehead et al. 2017). Notably, some species have evolved similar
tolerance for chemicals with the same mode of action (Whitehead et al. 2017), which hints at the
possibility of overlap between stress response mechanisms that could improve fitness in multiple
stressor scenarios. In Chapter 4 of this dissertation, I evaluate the variation in copper (Cu)
tolerance between allopatric populations of the intertidal copepod Tigriopus californicus and
whether the response mechanism for copper may be shared by another stressor as part of a larger
multiple stressor experiment.
The relationship between chemical pollution, sensitivity and evolutionary processes has
influence beyond the population level. Sensitivity to a chemical is a complex phenotype
determined by a network of interacting morphological, physiological and behavioral traits that
4
regulate the uptake, metabolism and excretion of pollutants by an organism. Dependence on a
high number of subordinate traits can generate phylogenetic autocorrelation (Guénard et al.
2011), which is especially likely for chemical sensitivity, given that many of its determining
factors such as body size and metabolic rate exhibit strong phylogenetic dependence (Kamilar
and Cooper 2013; Hylton et al. 2018). The phylogenetic component of chemical sensitivity may
result in broad taxonomic patterns of sensitivity, where entire clades or taxonomic groups are
more sensitive to a given chemical than others. For example, freshwater mussels (family
Unionidae) are known to be extremely sensitive to ammonia, while fish are considered
moderately sensitive and crustaceans more tolerant than both groups (Arthur et al. 1987;
Augspurger et al. 2003). If such patterns are significant and common in toxicity data and not
sufficiently accounted for, they could potentially have major downstream impacts on the
effectiveness of pollution regulations. Chapters 2 and 3 of this dissertation explore this idea in
the context of water quality criteria development and toxicity testing. In Chapter 2, I perform a
broad meta-analysis of toxicity data at the phylum level to test for significant differences in
chemical sensitivity, while Chapter 3 contains a second, larger meta-analysis of toxicity data that
quantifies the phylogenetic contribution to both acute and chronic sensitivity for a wide array of
chemicals.
Water quality criteria
One of the key applications of aquatic toxicity data is in the development of water quality
criteria (WQC). These criteria are numeric values that represent the maximum concentrations of
chemicals in surface waters that should not be exceeded to avoid harming the majority of aquatic
life. Because WQC are considered scientific recommendations rather than enforceable
regulations, most environmental authorities will take an additional step to develop environmental
5
quality standards (EQS) on the basis of these criteria that are backed by laws and enforced as
compulsory maximum concentrations.
WQC are calculated using the toxicity data that are available for a given chemical. There
are several published criteria methodologies that vary slightly, but most follow the same general
procedure in which toxicity data from multiple species for a particular chemical are ranked by
descending value and are fit with a cumulative distribution. The desired criterion is then
calculated using the left-tail of the fitted distribution (Stephan et al. 1985; Posthuma et al. 2001).
There is no universally agreed upon minimum sample size or ideal taxonomic composition of the
datasets used to derive WQC, so the various regulatory agencies that develop criteria, such as the
U.S. Environmental Protection Agency (EPA), set their own minimum data requirement
(MDRs). For example, the EPA’s MDRs for the derivation of an acute criterion for saltwater
taxa call for data from the following groups:
a) Two families from the phylum Chordata
b) One family in a phylum other than Arthropoda or Chordata
c) One species from either the Mysidae or Penaeidae family
d) Three species from families not in the phylum Chordata (may include either Mysidae or
Penaeidae if not used previously)
e) One species from any other family
Crucially, the sample size and specific taxa stated in the EPA’s MDRs were selected on
the basis that most chemicals did not have more than eight acceptable toxicity values (Stephan
1984; Forbes and Calow 2002) rather than statistical or ecological reasoning. Most toxicity
testing has traditionally been conducted for only a small number of model species (Seegert et al.
1985; Anderson and Phillips 2016; Buchwalter et al. 2017), and as a result it is common practice
6
in ecotoxicology to extrapolate the chemical effects data from this limited group of species to
entire ecosystems (Schäfer et al. 2023). Recent advocates of modernizing WQC procedures have
called for increasing biological diversity during criteria development and by extension, in
toxicity testing (Buchwalter et al. 2017). However, multiple challenges arise when considering
how to implement these changes.
Firstly, the present makeup of the pool of toxicity data that can be used in WQC
development is uncertain. While toxicity testing has continued since the publication of the EPA’s
MDRs and criteria methodology, there is no indication as to whether the biological diversity of
these data has increased over time. Similarly, it is unclear whether the amount of biological
diversity or even if the specific taxonomic groups represented in a toxicity dataset significantly
affect criteria values. Together, these uncertainties make the benefits and feasibility of altering
MDRs unclear. In Chapter 2, I address these questions by conducting a meta-analysis of aquatic
toxicity data and evaluate the impacts of data availability, biological diversity and dataset
assembly rules on the development of water quality criteria.
Efforts to increase biological diversity in WQC development are also constrained by the
ongoing push in the field of ecotoxicology to move away from using traditional laboratory
testing (Sun et al. 2012; LaLone et al. 2021). While data obtained from in vivo testing is
considered the gold standard in toxicology (Merlot 2010; Sun et al. 2012), there are significant
costs associated with generating these data in terms of time, resources and animal usage. The
low-throughput of such methods additionally means that only a limited number of compounds
have been tested, creating a clear need for alternative methods of chemical hazard evaluation.
Phylogenetic methods of cross-species extrapolation are one such possible alternative, but the
feasibility of using such tools on a large scale is unclear. Chapter 3 of this dissertation estimates
7
the frequency and magnitude of phylogenetic signal in aquatic toxicity data with a second metaanalysis as a preliminary step towards the implementation of phylogenetic models for toxicity
data extrapolation.
Multiple stressors
Multiple stressors are the norm in aquatic ecosystems, and chemical pollution is just one
among many anthropogenic stressors that threaten aquatic biodiversity, joining others such as
habitat degradation, overexploitation and climate change. Stressors frequently co-occur spatially
and temporally, and much of the global ocean has been found to be impacted by multiple
stressors (Halpern et al. 2015). As such, studying and predicting the responses of biological
systems to multiple stressors is a critical area of research. In aquatic ecotoxicology, this work
typically focuses on combinations of chemicals and environmental variables with the aim of
predicting how species will tolerate chemical exposure in future climate scenarios.
Stress occurs when the deviation of one or more environmental factors from an
organism’s specific optimum creates a physiological disturbance and decrease in fitness
(Sokolova 2013). Given that different stressors have different physiological and molecular
mechanisms that can interact in complex ways, studying the effects of multiple stressors is a
major challenge. Classic multiple stressor theory predicts that organisms exposed to multiple
stressors will display either an additive, antagonistic or synergistic response (Folt et al. 1999;
Todgham and Stillman 2013; Gunderson et al. 2016). Stressors are considered additive when the
effect of a group of stressors equals the sum of the effects of each individual component.
Antagonistic effects occur when the impact of multiple stressors is less than the predicted
additive effect of the group, while synergistic stressors are predicted to exceed the additive
effect. Because synergistic stressors present the greatest threat to life, many studies have focused
8
on identifying potential synergies in nature (Breitburg and Riedel 2005; Crain et al. 2008), but
understanding of when all three interaction types occur is valuable when designing and
implementing management plans. Knowledge of antagonistic stressor combinations is
particularly important in this context as management plans might be ineffective or even
detrimental if a set of stressors assumed to be synergistic are actually antagonistic (Brown et al.
2013).
Antagonisms between stressors can arise through multiple means. One such possibility is
that the first stressor is so dominant that it removes the most sensitive individuals from a
population, leaving the more tolerant individuals behind that are less responsive to the secondary
stressor. Alternatively, exposure to one stressor could mitigate the effects of others, or multiple
stressors could share protective mechanisms or signaling pathways (Breitburg and Riedel 2005;
Rodgers and Gomez Isaza 2021), leading to an antagonistic outcome. This third scenario refers
to the phenomenon known as cross-protection, which occurs when exposure to an initial stressor
temporarily confers increased resistance to a second stressor (Figure 1-2; Rodgers and Gomez
Isaza 2023). For example, Todgham et al. (2005) found that the survival of tidepool sculpins in
severe osmotic and hypoxic conditions increased when the fish underwent a preliminary heat
shock. This phenomenon
is not limited to a select
group of taxa or certain
stressor combinations
either, as cross-protection
has been observed in over
50 animal species and for
Figure 1-2: Conceptual diagram of cross-protection responses. From
Rodgers and Gomez Isaza (2021).
9
more than 16 heterologous stressors (Rodgers and Gomez Isaza 2023). Mechanistically, this
phenomenon arises when stressors share either protective mechanisms (referred to as ‘crosstolerance’) or cellular signaling pathways that activate independent protective mechanisms
(referred to as ‘cross-talk’) (Rodgers and Gomez Isaza 2023). Because individuals experiencing
cross-protection may have fitness advantages during extreme climactic events or exposure to
novel anthropogenic stressors, cross-protection has been hypothesized to serve as a possible
“pre-adaptation” that could shield individuals from emerging stressors, including chemical
pollution (Ramegowda et al. 2020; Rodgers and Gomez Isaza 2023). In Chapter 4, I evaluate the
potential for cross-protection to occur between copper and heat stresses in T. californicus with a
series of acute and chronic toxicity tests as well as gene expression profiling.
References
Amiard-Triquet C. 2015. Chapter 1 - Introduction. In: Amiard-Triquet C, Amiard J-C,
Mouneyrac CBT-AE, editors. Academic Press. p. 1–23.
https://www.sciencedirect.com/science/article/pii/B9780128009499000012.
Anderson B, Phillips B. 2016. Chapter 6 - Saltwater Toxicity Tests. In: Blasco J, Chapman PM,
Campana O, Hampel M, editors. Marine Ecotoxicology. Academic Press. p. 167–197.
https://www.sciencedirect.com/science/article/pii/B9780128033715000060.
Arthur JW, West CW, Allen KN, Hedtke SF. 1987. Seasonal Toxicity of Ammonia to Five Fish
and Nine Invertebrate Species. Bull Environ Contam. 38:324–331.
Augspurger T, Keller A, Black M, Cope G, Dwyer J. 2003. Water Quality Guidance for
Protection of Freshwater Mussels (Unionidae) from Ammonia Exposure. Environ Sci Technol.
22(11):2569–2575.
Barata C, Baird DJ, Mitchell SE, Soares AMVM. 2002. Among- and within-population
variability in tolerance to cadmium stress in natural populations of Daphnia magna: Implications
for ecological risk assessment. Environ Toxicol Chem. 21(5):1058–1064.
doi:10.1002/etc.5620210523. https://doi.org/10.1002/etc.5620210523.
Barron MG. 2002. Bioaccumulation and Bioconcentration in Aquatic Organisms. In: Hoffman
DJ, Rattner BA, Burton Jr, G, Cairns Jr. J, editors. Handbook of Ecotoxicology. 2nd ed. Boca
Raton: CRC Press. p. 877–892.
Beiras R. 2018. Marine Pollution: Sources, Fate and Effects of Pollutants in Coastal Ecosystems.
Elsevier Science. https://books.google.com/books?id=sOtlDwAAQBAJ.
10
Bickham JW, Sandhu S, Hebert PDN, Chikhi L, Athwal R. 2000. Effects of chemical
contaminants on genetic diversity in natural populations: implications for biomonitoring and
ecotoxicology. Mutat Res Mutat Res. 463(1):33–51. doi:10.1016/S1383-5742(00)00004-1.
https://www.sciencedirect.com/science/article/pii/S1383574200000041.
Breitburg DL, Riedel GF. 2005. Multiple Stressors in Marine Systems. In: Marine Conservation
Biology: The Science of Maintaining the Sea’s Biodiversity. p. 167–182.
Brown BE. 1978. Lead detoxification by a copper-tolerant isopod. Nature. 276(5686):388–390.
doi:10.1038/276388a0. https://doi.org/10.1038/276388a0.
Brown CJ, Saunders MI, Possingham HP, Richardson AJ. 2013. Managing for Interactions
between Local and Global Stressors of Ecosystems. PLoS One. 8(6):e65765.
https://doi.org/10.1371/journal.pone.0065765.
Buchwalter DB, Clements WH, Luoma SN. 2017. Modernizing Water Quality Criteria in the
United States: A Need to Expand the Definition of Acceptable Data. Environ Toxicol Chem.
36(2):285–291. doi:10.1002/etc.3654.
Crain CM, Kroeker K, Halpern BS. 2008. Interactive and cumulative effects of multiple human
stressors in marine systems. Ecol Lett. 11:1304–1315. doi:10.1111/j.1461-0248.2008.01253.x.
Folt CL, Chen CY, Moore M V, Burnaford J. 1999. Synergism and antagonism among multiple
stressors. Limnol Oceanogr. 44(3part2):864–877. doi:10.4319/lo.1999.44.3_part_2.0864.
https://doi.org/10.4319/lo.1999.44.3_part_2.0864.
Forbes VE, Calow P. 2002. Species Sensitivity Distributions Revisited: A Critical Appraisal.
Hum Ecol Risk Assess An Int J. 8(3):473–492. doi:10.1080/10807030290879781.
https://doi.org/10.1080/10807030290879781.
Guénard G, von der Ohe PC, de Zwart D, Legendre P, Lek S. 2011. Using phylogenetic
information to predict species tolerances to toxic chemicals. Ecol Appl. 21(8):3178–3190.
doi:10.1890/10-2242.1. https://doi.org/10.1890/10-2242.1.
Gunderson AR, Armstrong EJ, Stillman JH. 2016. Multiple Stressors in a Changing World: The
Need for an Improved Perspective on Physiological Responses to the Dynamic Marine
Environment. Ann Rev Mar Sci. 8(1):357–378. doi:10.1146/annurev-marine-122414-033953.
https://doi.org/10.1146/annurev-marine-122414-033953.
Halpern BS, Frazier M, Potapenko J, Casey KS, Koenig K, Longo C, Lowndes JS, Rockwood
RC, Selig ER, Selkoe KA, et al. 2015. Spatial and temporal changes in cumulative human
impacts on the world’s ocean. Nat Commun. 6(1):7615. doi:10.1038/ncomms8615.
https://doi.org/10.1038/ncomms8615.
Klerks PL, Levinton JS. 1989. Rapid Evolution of Metal Resistance in a Benthic Oligochaete
Inhabiting a Metal-polluted Site. Biol Bull. 176(2):135–141. doi:10.2307/1541580.
https://doi.org/10.2307/1541580.
LaLone CA, Basu N, Browne P, Edwards SW, Embry M, Sewell F, Hodges G. 2021.
International Consortium to Advance Cross-Species Extrapolation of the Effects of Chemicals in
Regulatory Toxicology. Environ Toxicol Chem. 40(12):3226–3233. doi:10.1002/etc.5214.
https://doi.org/10.1002/etc.5214.
11
Merlot C. 2010. Computational toxicology—a tool for early safety evaluation. Drug Discov
Today. 15(1):16–22. doi:https://doi.org/10.1016/j.drudis.2009.09.010.
https://www.sciencedirect.com/science/article/pii/S1359644609003353.
Nacci DE, Champlin D, Jayaraman S. 2010. Adaptation of the Estuarine Fish Fundulus
heteroclitus (Atlantic Killifish) to Polychlorinated Biphenyls (PCBs). Estuaries and Coasts.
33(4):853–864. doi:10.1007/s12237-009-9257-6. https://doi.org/10.1007/s12237-009-9257-6.
Ownby DR, Newman MC, Mulvey M, Vogelbein WK, Unger MA, Arzayus LF. 2002. Fish
(Fundulus heteroclitus) populations with different exposure histories differ in tolerance of
creosote-contaminated sediments. Environ Toxicol Chem. 21(9):1897–1902.
doi:https://doi.org/10.1002/etc.5620210917. 10.1002/etc.5620210917.
Posthuma L, Suter GW, Traas TP. 2001. Species Sensitivity Distributions in Ecotoxicology.
Lewis Publishers.
Ramegowda V, Da Costa MVJ, Harihar S, Karaba NN, Sreeman SM. 2020. Chapter 17 - Abiotic
and biotic stress interactions in plants: A cross-tolerance perspective. In: Hossain MA, Liu F,
Burritt DJ, Fujita M, Huang BBT-P-MS and C-ST in CP, editors. Priming-mediated stress and
cross-tolerance in crop plants. Academic Press. p. 267–302.
https://www.sciencedirect.com/science/article/pii/B9780128178928000179.
Rodgers EM, Gomez Isaza DF. 2021. Harnessing the potential of cross-protection stressor
interactions for conservation: a review. Conserv Physiol. 9(1). doi:10.1093/conphys/coab037.
https://doi.org/10.1093/conphys/coab037.
Rodgers EM, Gomez Isaza DF. 2023. The mechanistic basis and adaptive significance of crosstolerance: a ‘pre-adaptation’ to a changing world? J Exp Biol. 226(11):jeb245644.
doi:10.1242/jeb.245644. https://doi.org/10.1242/jeb.245644.
Schäfer RB, Jackson M, Juvigny-Khenafou N, Osakpolor SE, Posthuma L, Schneeweiss A,
Spaak J, Vinebrooke R. 2023. Chemical Mixtures and Multiple Stressors: Same but Different?
Environ Toxicol Chem. 42(9):1915–1936. doi:10.1002/etc.5629.
https://doi.org/10.1002/etc.5629.
Seegert G, Fava J, Cumbia P. 1985. How Representative Are the Data Sets Used to Derive
National Water Quality Criteria ? Aquat Toxicol Hazard Assess Seventh Symp. 854:527–537.
doi:10.1520/STP36288S.
Shahidul Islam M, Tanaka M. 2004. Impacts of pollution on coastal and marine ecosystems
including coastal and marine fisheries and approach for management: a review and synthesis.
Mar Pollut Bull. 48(7):624–649. doi:10.1016/j.marpolbul.2003.12.004.
https://www.sciencedirect.com/science/article/pii/S0025326X03005459.
Sokolova IM. 2013. Energy-Limited Tolerance to Stress as a Conceptual Framework to Integrate
the Effects of Multiple Stressors. Integr Comp Biol. 53(4):597–608. doi:10.1093/icb/ict028.
https://doi.org/10.1093/icb/ict028.
Stephan C, Mount D, Hansen D, Gentile J, Chapman G, Brungs W. 1985. Guidelines for
Deriving Numerical National Water Quality Criteria for the Protection of Aquatic Organisms and
Their Uses. US Environmental Protection Agency.
12
Stephan CE. 1984. Are the “National Guidelines” Based on Sound Judgements? Aquat Toxicol
Hazard Assess Seventh Symp.
Sun H, Xia M, Austin CP, Huang R. 2012. Paradigm Shift in Toxicity Testing and Modeling.
AAPS J. 14(3):473–480. doi:10.1208/s12248-012-9358-1. https://doi.org/10.1208/s12248-012-
9358-1.
Todgham AE, Schulte PM, Iwama GK. 2005. Cross‐Tolerance in the Tidepool Sculpin: The
Role of Heat Shock Proteins. Physiol Biochem Zool. 78(2):133–144. doi:10.1086/425205.
https://doi.org/10.1086/425205.
Todgham AE, Stillman JH. 2013. Physiological Responses to Shifts in Multiple Environmental
Stressors: Relevance in a Changing World. Integr Comp Biol. 53(4):539–544.
doi:10.1093/icb/ict086. https://doi.org/10.1093/icb/ict086.
Vandegehuchte MB, Janssen CR. 2014. Epigenetics in an ecotoxicological context. Mutat Res
Toxicol Environ Mutagen. 764–765:36–45. doi:10.1016/j.mrgentox.2013.08.008.
https://www.sciencedirect.com/science/article/pii/S1383571813002325.
Whitehead A, Clark BW, Reid NM, Hahn ME, Nacci D. 2017. When evolution is the solution to
pollution: Key principles, and lessons from rapid repeated adaptation of killifish (Fundulus
heteroclitus) populations. Evol Appl. 10(8):762–783. doi:10.1111/eva.12470.
https://doi.org/10.1111/eva.12470.
13
Chapter 2: Data and diversity in the development of acute water quality criteria in the
United States
Alice L. Coleman and Suzanne Edmands
This chapter appears as published in Environmental Toxicology and Chemistry.
Abstract
The United States Environmental Protection Agency (EPA) is responsible for the
development of water quality criteria, regulatory standards that protect aquatic organisms from
harmful chemical exposure. Although these criteria are intended to be broadly protective of
aquatic life, the data used to derive criteria do not necessarily reflect the actual diversity of
natural communities nor are they available for most chemicals. Additionally, while the EPA’s
current procedures emphasize using toxicity data with a certain minimum amount of biological
diversity, the quantitative impact of such diversity on criteria is unclear. In the present study we
assessed the changes to acute toxicity data over time, determined the prevalence of significant
taxonomic differences in sensitivity and investigated the effect of biological diversity on criteria.
We found major gaps in existing toxicity data that we hypothesize have contributed to the
absence of acute criteria for the majority of chemical pollutants. Taxonomic patterns of
sensitivity in these data are abundant, although the resolution of the patterns is relatively poor.
Additionally, we found that the amount of biological diversity in a toxicity dataset and dataset’s
taxonomic composition does not quantitatively affect criteria in most cases. Given that the EPA
has published acute criteria for fewer than 20% of priority pollutants and the persistence of major
gaps in toxicity data over the last thirty-six years, we recommend the EPA consider revisions to
their water quality criteria guidelines that will expedite the criteria development process and
advance the responsible management of pollutants in the aquatic environment.
14
Introduction
The Clean Water Act requires the United States Environmental Protection Agency (EPA)
to derive and publish numeric water quality criteria (WQC) that are intended to protect aquatic
organisms and their uses from the adverse effects of chemical pollutants. These criteria are
recommended national standards that describe the maximum concentration of a pollutant in
surface water expected not to harm the majority (95%) of aquatic life, but do not serve as legally
binding requirements. The EPA’s water quality criteria methodology (the “National Guidelines”)
directs the development of these values by defining the “materials of concern” for criteria,
issuing data collection and quality requirements, and establishing calculation procedures for
acute and chronic criteria. The first document of its kind, the National Guidelines have been
repeatedly adopted by groups outside the EPA for water quality criteria development (Wu et al.
2015; David B. Buchwalter et al. 2017; Liu et al. 2019). However, several alternate criteria
methodologies have since been established by other regulatory agencies that incorporate the
advances in aquatic toxicology and additional toxicity data that emerged after the publication of
the National Guidelines in 1985 (Canadian Council of Ministers of the Environment 2007;
Tenbrook et al. 2010; Warne et al. 2015). Given the advances in the field, rise of divergent
methodologies and perpetual need for effective criteria, it is necessary to consider whether the
National Guidelines warrant revisions.
Review of the National Guidelines by the EPA began as early as 1990 and remains an
active priority for the agency (Wilcut et al. 2015; Ankley et al. 2017). While the aim of the
modernization efforts is to produce a comprehensive update of the National Guidelines, most
proposed revisions have thus far maintained the original minimum data requirements (MDRs) for
water quality criteria set by Stephan et al. (1985). These requirements describe the minimum set
of taxa that must have available toxicity data in order to derive a criterion. For example, the
15
minimum data requirements for acute saltwater criteria call for acceptable acute tests with at
least one species from at least eight different families, specifically including:
f) Two families from the phylum Chordata
g) One family in a phylum other than Arthropoda or Chordata
h) One species from either the Mysidae or Penaeidae family
i) Three species from families not in the phylum Chordata (may include either Mysidae or
Penaeidae if not used previously)
j) One species from any other family
Similarly, the minimum data requirements for acute freshwater criteria call for acceptable tests
from at least one species in at least eight different families, specifically including:
a) One species from the family Salmonidae
b) One additional family from Osteichthyes
c) A third family from Chordata (may be in Osteichthyes)
d) One planktonic crustacean
e) One benthic crustacean
f) One insect
g) One family in a phylum other than Arthropoda or Chordata
h) One family in any order of insect or phylum not already represented
While the purpose of water quality criteria is to protect the majority of aquatic life, the
MDRs do not impose specific diversity requirements aside from ensuring the representation of at
least eight families and three phyla in a dataset. Given that the sample size for these minimum
data requirements was based on the fact that most chemicals did not have more than eight
16
acceptable toxicity values (Stephan 1984), it is reasonable to question whether data from just
eight families can confer the desired level of protection. Additionally, by 1985 it had been
established that the majority of available toxicity data at the time were obtained from a limited
group of species and did not adequately sample from the full range of sensitivities expected in a
natural species assemblage (Seegert et al. 1985). The pool of toxicity data that can be used to
develop criteria has grown over the past thirty-six years, so it is important to assess how the
taxonomic composition of these data has changed over time and to determine how many
chemicals are now able to fulfill either set of minimum data requirements.
Aside from these concerns over the impact of data availability on criteria, it is also
unclear whether the amount of biological diversity or specific taxonomic groups represented in a
toxicity dataset significantly affect criterion value. Previous studies in aquatic toxicology have
identified generalized taxonomic patterns of sensitivity for some pollutants. For example,
phytoplankton are considered to be the most sensitive group to the herbicide atrazine, followed
by benthic invertebrates, planktonic invertebrates and then fish (Solomon et al. 1996). Other
studies report a different pattern for ammonia, wherein freshwater mussels (family Unionidae)
are extremely sensitive to ammonia, fish are moderately sensitive and crustaceans are generally
more tolerant than both groups (Arthur et al. 1987; Augspurger et al. 2003). These taxonomic
sensitivity patterns could affect criteria development if biological diversity is limited during
toxicity testing, resulting in a multimodal dataset. For example, a toxicity dataset for ammonia
composed primarily of values from freshwater molluscs and crustaceans would likely have two
significantly different modes. The value of a criterion from such multimodal datasets would be
influenced by the number of species from each mode, and thus may not reflect the sensitivity of a
natural aquatic community (Giddings et al. 2019).
17
In the present study we assessed the impact of data availability, biological diversity and
the minimum data requirements on the development of water quality criteria with the aim of
informing the EPA’s revisions of the National Guidelines. Using data from twelve aquatic
pollutants, we analyzed the shift in the abundance and taxonomic composition of acute toxicity
data since 1985, quantified the relationship between the amount of biological diversity in a
toxicity dataset and criterion value and tested for taxonomic differences in acute tolerance. We
additionally estimated the proportion of chemicals with sufficient data to satisfy the EPA’s
minimum data requirements for both saltwater and freshwater acute criteria and provide a brief
overview of relevant potential changes to the National Guidelines.
Methods
Data collection
We collected and analyzed acute toxicity test results from saltwater and freshwater
species for twelve aquatic pollutants that span a range of chemical classes and toxic modes of
action (Table 2-1). Candidate chemicals were drawn from literature and the Clean Water Act’s
Priority Pollutant List (PPL), the set of chemicals for which the EPA is required to develop water
quality criteria. We conducted a preliminary survey of the EPA’s ECOTOXicology (ECOTOX)
knowledgebase to identify which chemicals potentially had enough data to fulfill the saltwater
and freshwater minimum data requirements and used the results to narrow down our list to
twelve. The primary source of data for the present study, ECOTOX is an online public database
comprised of more than 1,000,000 toxicity test results that cover ~12,300 chemicals and ~13,600
species.
We downloaded acute toxicity data for each chemical from ECOTOX that were obtained
using the following search parameters:
18
1. CAS number
2. Kingdom: Animals
3. Endpoint: Median Lethal Concentration (LC50)
4. Test Location: Lab
5. Exposure Media: Freshwater or Saltwater
When possible, the CAS number used for a chemical was matched to its CAS number given
in the Aquatic Life Criteria Table, an online resource maintained by the EPA that contains the
current set of water quality criteria (US EPA 2021). Freshwater and saltwater datasets were
downloaded separately in May 2021. Additional acute toxicity datapoints were collected with
searches in Google Scholar and Web of Science using the chemical name and keywords “acute
toxicity” and either “marine” or “freshwater” and then appended to the ECOTOX datasets. The
combined datasets were then trimmed according to the following quality control parameters to
emulate the data collection rules listed in the National Guidelines:
1. Test organism was a resident North American species
2. Test organism was not a brine shrimp
3. Test species was not a single-celled organism
4. Tests performed with cladocerans and midges were 48 hours in length
5. Tests performed with all other freshwater and saltwater species were 96 hours in length
When a species had more than one LC50 value available for a chemical, the data were
condensed into a Species Mean Acute Value (SMAV) by calculating the geometric mean of all
datapoints. For each genus with more than one species available, the Genus Mean Acute Value
(GMAV) was calculated as the geometric mean of all datapoints.
19
In addition, we used the EPA’s Aquatic Life Criteria Table to identify which of the
twelve chemicals had official acute criteria and to examine available criteria documents. We also
performed a second survey of ECOTOX in July 2021 to estimate the proportion of chemicals
from the Priority Pollutant List that could satisfy the saltwater and freshwater minimum data
requirements using the same quality control parameters listed above.
The trimmed toxicity datasets were used to determine which chemicals had sufficient
data to satisfy the minimum data requirements for acute saltwater and freshwater criteria. We
further sampled from the trimmed datasets to create the following dataset subtypes for each
chemical:
1. Eight randomly selected genera that satisfy the saltwater MDRs (TS8)
2. Eight randomly selected genera that satisfy the freshwater MDRs (TF8)
3. Eight randomly selected saltwater genera (RS8)
4. Eight randomly selected freshwater genera (RF8)
5. Twenty randomly selected saltwater genera (RS20)
6. Twenty randomly selected freshwater genera (RF20)
A sample size of eight genera was chosen for dataset types one through four to be consistent
with the minimum data requirements. Twenty genera were used for the RF20 and RS20 datasets
to approach the mean sample size (𝑥 ≈ 29) of the datasets used by the EPA to develop criteria
for the chemicals in our set while accounting for the sample size limitations of the datasets we
assembled (Table S2-1).
Criteria derivation
20
Criteria were derived using the method defined by Stephan et al. (1985) in the National
Guidelines. Briefly, GMAVs were ordered by decreasing value, assigned a rank and calculated a
cumulative probability. The four GMAVs with cumulative probabilities closest to 0.05 were
utilized in a series of four equations stated in the National Guidelines to calculate the Final Acute
Value (FAV). The FAV was then divided by two to reach the Criterion Maximum Concentration
(CMC), the EPA’s acute criterion.
We also calculated criteria using species sensitivity distributions (SSDs), a relatively new
derivation technique employed by other criteria methodologies that is under consideration by the
EPA for inclusion in the National Guidelines (Tenbrook et al. 2010; Warne et al. 2015; Wilcut et
al. 2015). SSDs consist of a plot of toxicity data from multiple species that are ranked and
assigned a percentile. A cumulative distribution is then fitted to the data and used to calculate the
hazardous concentration for 5% of taxa (HC5) which is considered equivalent to a Final Acute
Value derived by the EPA method (Wilcut et al. 2015). To approximate a Criterion Maximum
Concentration from an SSD, we calculated SSDs using GMAVs and divided all HC5s by two.
All SSDs were generated in R with the package ssdtools (Thorley and Schwarz 2018; R Core
Team 2019). This package allows the user to fit multiple distributions to each dataset, so we used
the Akaike Information Criterion corrected for sample size (AICC) to select the best-fit
distributions (Burnham and Anderson 2002).
Taxonomic analyses
We assessed the change in our trimmed datasets over time by quantifying the taxonomic
composition of the data that were available in 1985 and comparing it to the makeup of the
complete datasets compiled in May 2021. We further compared the datasets by calculating the
average taxonomic distinctness index (Δ+) of the 1985 and 2021 datasets for both saltwater and
21
freshwater data with the R package vegan (Clarke and Warwick 1998; Oksanen et al. 2007).
Change in Δ+ between 1985 and 2021 was tested with a randomization test for which we
generated 1000 random subsets of the species data with the same sample size as the 1985
datasets (saltwater n = 52; freshwater n = 83) and calculated the corresponding Δ+ values. We
then defined the intervals that contained 95% of the simulated Δ+ values and compared the
intervals to the actual Δ+ values from the 1985 and 2021 datasets, with the latter treated as the
“true” Δ+ for the data. For a broad-scale assessment of diversity, we also used Fisher’s exact
tests with Monte Carlo simulations (B = 1,000,000) to determine whether the taxonomic
composition of the data available in 1985 and those data that were published in 1986 or later (up
to May 2021) were different at the level of phylum. In addition, we used two-proportion z-tests
to assess whether the proportion of the data that met our quality control parameters changed
between 1985 and 2021.
To evaluate the impact of the minimum data requirements on criterion value, we
compared the criteria calculated from datasets that were assembled randomly (RS8/RF8) to those
assembled according to the minimum data requirements (TS8/TF8). Fifty versions of each
dataset type were generated for each chemical, except for those that could not satisfy either of
the MDRs, and then calculated the CMC for each individual dataset. We then compiled the
individual criteria for each group, calculated the mean CMCs and used t-tests to determine
whether the means of the TS8/RS8 and TF81/RF8 datasets were significantly different. MannWhitney U tests were used in place of t-tests when the datasets were not normally distributed as
indicated by a Shapiro-Wilk test. We conducted this analysis twice for each chemical, comparing
criteria calculated by either the EPA or the SSD method.
22
In addition, we performed a linear regression analysis to model the relationship between
the amount of biological diversity in a dataset and criterion value. Like in the criteria
comparison, we generated fifty versions of the RS20 and RF20 datasets for the chemicals with
data from at least twenty genera (Table S2-1) and then calculated the CMC for each dataset using
both the EPA and SSD methods. The diversity of each dataset was measured as the Shannon
diversity index and calculated using the package vegan (Shannon 1948). We then fit a linear
model to the data frame containing the criteria and diversity indices, treating diversity index as
the independent variable and CMC as the dependent variable. This analysis was performed twice
for each chemical, using criteria derived with either the EPA or SSD method to compare the
effect of diversity on the different techniques.
Finally, Kruskal-Wallis tests, which are the non-parametric analog to the one-way
analysis of variance (ANOVA), were used to determine whether the mean sensitivities of any
phyla in a toxicity dataset significantly differed from each other. Pairwise comparisons of phyla
were conducted with post-hoc Dunn’s tests to then identify those significantly different phyla,
the results of which were then used to outline taxonomic sensitivity patterns. These tests were
performed twice per chemical, using each saltwater and freshwater dataset separately. In this
analysis, the phyla with only one datapoint available were removed from a chemical’s dataset
because the mean is an uninformative statistic for a group with a sample size of one.
Results
Dataset composition
In May 2021 we found a total of 1,662 saltwater and 5,359 freshwater acute toxicity
values (LC50s) for our set of twelve pollutants. Of these totals, 558 of the saltwater and 1,662 of
the freshwater datapoints were published in 1985 or earlier. After the application of quality
control parameters, the saltwater and freshwater datasets shrank to 535 and 1,626 values
23
respectively, and most chemicals saw a reduction in the number of species and genera
represented in their individual datasets (Table S2-1). Of the trimmed dataset totals, 224 saltwater
and 668 freshwater values were published in 1985 or earlier. The trimmed saltwater dataset
consisted of 104 species (84 genera) in eight phyla, while the trimmed freshwater dataset
contained data from 172 species (115 genera) in eight phyla. Species from Chordata and
Arthropoda contributed the most data to both the saltwater and freshwater datasets, followed
distantly by species from the phyla Mollusca and Rotifera (Fig. 2-1). The number of phyla and
species represented in each dataset increased slightly over time.
The Δ+ calculations indicated that the average taxonomic distinctness of the saltwater data
significantly decreased over time (1985 Δ+ = 87.23; 2021 Δ+ = 85.097; Fig. S2), while the
average taxonomic distinctness of the freshwater data remained relatively unchanged (1985 Δ+ =
84.34; 2021 Δ+ = 85.28; Fig. S2-2). Fisher’s exact tests indicated that the taxonomic
compositions of the data that were available in 1985 compared to the data that were published
after 1985 (up to June 2021) were significantly different (p < 5E-4) at the phylum level for both
the saltwater and freshwater datasets. In addition, the two-proportion z-tests indicated that the
proportions of both the saltwater (from 40% in 1985 to 32% in 2021; p = 0.00059) and
freshwater datasets (from 40% in 1985 to 31% in 2021; p = 2.061E-13) that met our quality
parameters decreased between 1985 and 2021.
In addition, based solely on the trimmed datasets we assembled, two chemicals (4-
nitrophenol and toluene) could not satisfy the saltwater minimum data requirements and three
chemicals (nickel, 4-nitrophenol and toluene) could not satisfy the freshwater minimum data
requirements. We also estimated that of the 126 chemicals on the Priority Pollutant List, 21 have
24
enough data in ECOTOX to satisfy the saltwater MDRs and 37 have sufficient data to meet the
freshwater MDRs.
Criteria search
To date, there are 29 acute saltwater criteria and 34 acute freshwater criteria available in
the EPA’s Aquatic Life Criteria Table. These official values cover eight of the twelve chemicals
in our set, excluding atrazine, 4-nitrophenol, phenol and toluene which have no criteria listed
(Table 2-1). We treated the EPA’s criteria for tributyltin (TBT) and α/β-endosulfan as the criteria
for tributyltin oxide (TBTO) and endosulfan respectively. Importantly, the criteria for the
endosulfan isomers were developed using the EPA’s precursor document to the National
Guidelines from 1980 which utilized different minimum data requirements than those discussed
here (45 Fed. Reg. 79347 1980).
Additionally, despite their absence from the Aquatic Life Criteria Table, we were able to
retrieve archived EPA criteria documents for atrazine, 4-nitrophenol, phenol and toluene. We
found a complete set of draft criteria for atrazine that appear to have never been published in a
finalized form (Office of Water 2003), and found criteria for 4-nitrophenol, phenol and toluene
that were developed using the precursor to the National Guidelines, similar to the criteria for the
endosulfan isomers. It is unclear why the criteria for these three chemicals are not included in the
Aquatic Life Criteria Table with the endosulfan criteria.
Taxonomic analyses
We found nominal evidence of a quantitative effect of the minimum data requirements on
criterion value. The comparison of criteria showed that using different methods to assemble
toxicity datasets resulted in significantly different values in only five of eighteen tests (Table 2-
2). There appeared to be no effect of the method of dataset assembly on criterion value in these
25
significant cases, as the larger criterion was from an MDR (TS8/TF8) dataset in three cases and
from a random (RS8/RF8) dataset in two cases.
Similarly, we found minimal evidence of a quantitative effect of the diversity in a toxicity
dataset on criterion value. Of the twenty-four models we tested, just two (both
pentachlorophenol) described a significant relationship between a dataset’s Shannon diversity
index and criterion (Table 2-3). The regression coefficient for diversity index in the significant
models was negative in both cases, but the low R2 values for these models (0.17 and 0.10) and
overall low number of significant models in the set we tested suggests that diversity index is a
poor predictor of criterion values. There also appeared to be no effect of condition
(saltwater/freshwater) or calculation method (EPA/SSD) on the diversity-criterion relationship,
as there was one significant model in each category.
The Kruskal-Wallis tests indicated that 12 of the 24 datasets we assembled contained
phyla with significantly different mean sensitivities, covering eight of the twelve chemicals
(Table 2-4). Eight of the datasets with significant differences were freshwater and four were
saltwater. Post-hoc Dunn’s tests identified which specific phyla’s means were significantly
different from each other (Fig. 2-2, S2-2:S2-12). We also found that the phylum that was the
most sensitive to a chemical varied among all twelve chemicals, although we were unable to
determine whether this variation was caused by true differences in the sensitivity of phyla to
chemicals or by differences in dataset composition.
Discussion
In the United States, the development of acute water quality criteria relies on toxicity data
that meet the specific requirements laid out in the EPA’s National Guidelines. Here, we
assembled and analyzed acute toxicity datasets for twelve chemicals and found them to lack data
26
from saltwater and “non-standard” laboratory test species, although the amount of biological
diversity and taxonomic composition of a dataset did not appear to affect criterion value.
The low availability of toxicity data from saltwater species relative to freshwater is
particularly striking. While the disparity between freshwater and saltwater datasets is evident
from the raw counts of the cumulative trimmed datasets we assembled (freshwater n = 1636,
saltwater n = 535), comparison of the single chemical datasets indicate that marine datasets are
generally smaller and contain fewer species and genera than their freshwater counterparts (Table
S2-1). Our observation of fewer saltwater values from fewer taxa is consistent with the findings
of other studies in aquatic toxicology, a trend that has been attributed to the challenges of
calculating the chemical speciation of toxicants in seawater and the historical focus of risk
assessments on freshwater systems (Leung et al. 2001; Pavlaki et al. 2016). As a result of these
differences in sample size, freshwater toxicity datasets are generally considered to be more
representative of the true amount of biological diversity expected in a natural system than
saltwater datasets (Leung et al. 2001). These disparities in dataset sample size and
representativeness pose a major challenge to the development of acute water quality criteria for
saltwater organisms.
The unbalanced taxonomic composition of our toxicity datasets hints at another
significant barrier to the development of acute water quality criteria. The overwhelming majority
of the data we collected are derived from Chordata and Arthropoda (90% of trimmed saltwater
dataset, 94% of trimmed freshwater dataset) with the rest of the datasets made up of small
contributions from eight other phyla (Fig. 2-1). The dominance of Chordata and Arthropoda
persisted during the period between 1985 and 2021 even with the statistically significant changes
in the taxonomic composition of both the saltwater and freshwater datasets during that time. In
27
addition, the average taxonomic distinctness (Δ+) of the saltwater data decreased over that
period, which we speculate was caused by the addition of chordate and arthropod species to the
dataset that were closely related to some of the species that had data available in 1985. Thus, we
hypothesize that the minimal amount of toxicity data available from other faunal groups is the
most important factor preventing the majority of chemicals from meeting one or both sets of the
minimum data requirements for acute criteria.
Importantly, the dominance of chordate and arthropod taxa in toxicity datasets does not
reflect the actual abundance of these species in the world’s aquatic environments. According to
the World Register of Marine Species, there are 57,368 valid marine arthropod species and
24,100 valid marine chordate species. Together, the species from these phyla make up just 40%
of the total number of valid marine species in the Kingdom Animalia (WoRMS Editorial Board
2021). This persistent discrepancy between the diversity found in toxicity datasets and the actual
diversity of aquatic life should be a driving force to expand the coverage of aquatic taxa included
in the criteria development process.
There are multiple factors behind the reliance on chordate and arthropod species in
toxicity testing. Many fish and crustacean species have historically been used in toxicity testing
because they are widely available to researchers and easily maintained in the laboratory, leading
to the development of standard toxicity test procedures for certain species. As a result, common
laboratory species like Cyprinodon variegatus and Palaemonetes pugio are present in the
datasets of multiple chemicals. The language of the National Guidelines may have also
influenced the usage of chordate and arthropod species during testing as they explicitly
emphasize protecting “commercially, recreationally, and socially important species” (Stephan
1984), which in practice tends to translate to fishes and crustaceans.
28
While the continued usage of chordates and arthropods in laboratory testing is a major
determinant of the taxonomic composition of toxicity data, the makeup of the specific datasets
that we assembled could also be linked to our choice of chemicals. Eight of the chemicals we
analyzed have official EPA criteria and two (endrin and endosulfan) are banned or are being
phased out in the United States, meaning that considerable toxicity testing has already been
conducted for these chemicals. With these regulatory measures in place, there may be little
incentive to continue testing a chemical, leading to minimal change in the composition of
toxicity data for that chemical over time. However, when a significant issue with its criterion
emerges, such as in the case of ammonia, further toxicity testing and the expansion of toxicity
datasets is required.
In 2013 the EPA published an updated set of freshwater criteria for ammonia that
represented a major decrease from the previous values set in 1999 (for example, the CMC
decreased from 24 mg/L to 17 mg/L). This change was driven by the publication of studies in
2003 which found that the existing criteria did not protect the highly sensitive Unionidae family
of freshwater mussels from the harmful effects of ammonia exposure (Augspurger et al. 2003;
Office of Water 2013). While the exclusion of any major taxonomic group by a criterion is
concerning, this conclusion was particularly compelling because more than half of the nearly 300
Unionidae species in North America are listed as threatened or endangered (Augspurger et al.
2003). In response, the EPA conducted additional toxicity testing to validate these findings and
ultimately drafted a new set of ammonia criteria in 2009 for bodies of water with and without
mussels present (Office of Water 2013). Following an external peer review process, the EPA
then released a finalized set of updated criteria for ammonia in 2013 that were derived from a
new toxicity dataset that included the previously absent unionid mussels and gill-breathing
29
snails. This update of the ammonia criteria is significant because it demonstrated the EPA’s
ability to integrate taxonomic sensitivity information specific to a single chemical into the
criteria development process, an ability which may be required in the future if other conflicts
between criteria protectiveness and taxonomic sensitivity patterns are exposed.
Here, we investigated the taxonomic patterns of sensitivity for twelve chemicals by
identifying and comparing the marine and freshwater phyla with significantly different mean
acute sensitivities (Fig. 2-2; S2-3:S2-13). Using these results, we can draw tentative conclusions
about the relative sensitivities of different taxa to the twelve chemicals. For example, it is evident
in the endosulfan data that saltwater annelids are more tolerant than both saltwater arthropods
and chordates and that freshwater arthropods and molluscs are more tolerant of endosulfan than
freshwater chordates (Fig. 2-2). Unfortunately, however, our ability to resolve taxonomic
sensitivity patterns to a high degree using the results of the comparisons between phyla was
restricted in several instances by the low statistical power of the Dunn’s test and the low amount
of data available from certain phyla. For example, no conclusions can be made about the relative
sensitivities of marine or freshwater arthropods and chordates to atrazine (Fig. S2-4). Similarly,
for ammonia, no distinctions can be made between the saltwater phyla although in the freshwater
data arthropods seem to be more tolerant of ammonia than chordates (Fig. S2-3) As a result of
this poor degree of resolution, it is difficult to determine whether our findings are consistent with
the trends in sensitivity to atrazine and ammonia documented in the literature. Despite the
reoccurrence of this issue in the toxicity datasets we assembled, the frequency of significant
differences between phyla in sensitivity as indicated by the results of the Kruskal-Wallis tests
suggests that taxonomic relationships in sensitivity to toxicants are a common phenomenon
(Table 2-4; Fig. 2-1, S2-3:S2-13). We anticipate that criteria developed using both the EPA
30
method and species sensitivity distributions could be impacted by taxonomic patterns of
sensitivity. The EPA method primarily uses the four values in a dataset with cumulative
probabilities closest to 0.05 to calculate criteria without imposing requirements on the taxonomic
makeup of those four datapoints. If a certain group of organisms is particularly sensitive to the
target chemical, it is possible that all of the four datapoints used to calculate the criterion could
come from a set of closely related species. Meanwhile, the species sensitivity distribution
approach incorporates all of the values in a toxicity dataset, but if a taxonomic sensitivity pattern
causes a dataset to be multimodal then the cumulative distribution that is fitted to the data may
be a poor fit for the left-tail of the SSD. The HC5 is derived from the left-tail of an SSD, so
multimodality may result in greater uncertainty in the HC5. In light of these potential effects,
further exploration of taxonomic trends in sensitivity within phyla is required.
Ultimately, thirty-six years after the publication of the National Guidelines the vast
majority of aquatic pollutants still do not have acute criteria. While the overall availability of
toxicity data has improved since 1985 which in theory should benefit criteria development, there
are still fewer than 60 official EPA water quality criteria that cover less than 30% of the
chemicals on the Priority Pollutant List. Although the scarcity of criteria can be explained by
multiple scientific and bureaucratic factors, we hypothesize that the cumulative lack of toxicity
data from marine and non-standard test species is the most important constraint on the
development of water quality criteria in the United States. Because criteria provide a crucial
measure of protection for aquatic organisms, the EPA should consider revisions to the National
Guidelines that will facilitate their development. There are several possible changes that could
help achieve this goal, although none of the options presented here can be considered a universal
solution.
31
The simplest means of accelerating criteria development is to increase the amount of
available toxicity data. While the single-species toxicity test is the preferred method of data
collection in aquatic toxicology, the financial and logistical constraints associated with
laboratory testing make it an impractical solution for filling the large data gaps observed here. At
present, data extrapolation techniques such as quantitative structure activity relationship (QSAR)
models and interspecies correlation estimation (ICE) models are the best alternatives to
traditional toxicity testing. QSAR models use molecular descriptors and chemical properties to
predict the toxicity of chemicals to organisms, while ICE models estimate the acute toxicity of a
chemical to a data-deficient species by performing a least-squares regression with data from a
related surrogate species. Species sensitivity distributions can be augmented with ICEextrapolated toxicity values, increasing dataset sample size without significantly affecting HC5
uncertainty (Awkerman et al. 2014). The EPA developed software that can estimate toxicity
values using ICE models and includes an SSD module (Raimondo, Vivian, et al. 2010), but has
not embraced ICE-derived datapoints in official development of official water quality criteria
development. Thus, in light of their statistical support, we recommend the EPA formally
integrate ICE models into the National Guidelines as an acceptable alternate method of toxicity
data production. Alternatively, if the extrapolation of chemical sensitivity data is undesirable or
infeasible, the EPA could explore revising the data quality control parameters in the National
Guidelines to take advantage of a greater portion of the existing pool of toxicity data.
One of the major challenges in water quality criteria development is the conflict between
data quality and quantity. Quality control parameters implemented during data collection ensure
the quality and consistency in the toxicity data used to derive criteria, but may actually hinder the
criteria development process by eliminating many potentially usable toxicity datapoints. Here,
32
the imposition of relatively simple quality parameters similar to those in the National Guidelines
forced a ~70% drop in the size of both our freshwater and saltwater datasets, and we observed a
decrease in the proportion of the data that met those quality parameters over time. The majority
of datapoints were censored because they were either recorded from a toxicity test with an
inappropriate duration or were derived from a test organism that could not be classified as a
North American resident species. As a result of the removal of these data, the biological diversity
in the datasets for most chemicals was severely reduced and likely made the datasets less
representative of the aquatic communities that water quality criteria are intended to protect. Data
losses of similar or greater magnitude are probable regardless of chemical choice, so it may be
necessary to reconsider some of the specific quality requirements in the National Guidelines. For
example, it may be beneficial to allow for the use of data from species that do not have
reproducing wild populations in North America during criteria derivation because the
sensitivities of temperate and tropical species are relatively similar for many chemicals (Wang et
al. 2014). Although the use of region-specific data is certainly preferable in theory when
developing criteria, in practice the limitations in data availability means that a localized approach
may not be feasible for most chemicals. Alternatively, the EPA could pivot to allow for the use
of field or mesocosm-based data during criteria development. Stephan et al. (1985) did not
include these data types in the National Guidelines because of concerns over data complexity and
the feasibility of field testing, however, weight-of-evidence approaches to using field data in
water quality criteria development are now available (Cormier et al. 2008; Warne et al. 2015).
These approaches have allowed newer criteria methodologies to take advantage of field and
mesocosm data, meaning that their integration into the National Guidelines should be possible.
33
Another means of expediting criteria development would adjust how data are grouped
prior to the calculation of criteria. One such option is to combine data from freshwater and
saltwater species during criteria derivation, as Leung et al. (2001) and Wheeler et al. (2002)
demonstrated that freshwater datasets for ammonia and several metals could provide adequate
protection for saltwater taxa. However, these same studies also indicated that freshwater data
would not be protective of saltwater species for pesticides such as chlordane, endosulfan and
chlorpyrifos largely because of differences in dataset sample sizes and taxonomic composition.
Thus, the integration of freshwater data with saltwater may not be appropriate for all chemicals.
The EPA could also consider setting criteria for groups of chemicals with the same adverse
outcome pathway instead of creating criteria for single chemicals (Elias et al. 2019; Giddings et
al. 2019). Group criteria, or normalized hazardous concentrations (HC5n), are derived from
species sensitivity distributions populated with the toxicity data from multiple chemicals
(Giddings et al. 2019). This approach does not require the individual chemicals datasets to satisfy
the minimum data requirements, thereby reducing the need for additional toxicity testing or data
extrapolation during criteria development. Additionally, the higher sample size of a combined
dataset can lead to greater biological diversity in a species sensitivity distribution and increase
the statistical precision of the HC5n estimate (Carr and Belanger 2018). Giddings et al. (2019)
utilized this approach to derive a criterion for a set of nine pyrethroids, a class of synthetic
pesticides, and calculated an HC5n for the entire group that was more statistically robust than the
individual criteria would have been. While these results are promising, the group criteria
approach is limited to those chemicals that share a toxic mode of action. The modes of action
have not been defined for many chemicals, making it difficult to determine when grouping
34
chemicals for criteria derivation is appropriate. As such, revising the National Guidelines to
allow for group criteria is not yet viable.
A final option to consider is a modification of the minimum data requirements in the
National Guidelines to include aquatic plant and algal species. Although aquatic plants and algae
contribute considerable ecological and economic value, they have historically been excluded
from water quality criteria development in the United States because of uncertainty over how to
include them in the process (Lewis and Thursby 2018). There are now numerous studies
available that use plant data in species sensitivity distributions (Song et al. 2015; Ding et al.
2016; Lewis and Thursby 2018) suggesting that this uncertainty should now be less of a factor in
criteria development. In the context of our study, the inclusion of non-animal data would
increase the biological diversity of toxicity datasets and confer greater protection to primary
producers from the effects of chemical contaminants. However, it is important to note that like
many animal phyla, plants and algae also suffer from significant data shortages in aquatic
toxicology. For example, an ECOTOX search for plant and algal data for the twelve chemicals
assessed in the present study returned a total of just 60 saltwater and four freshwater datapoints
that met our quality parameters. While this low amount of data could be caused by our choice of
chemicals, most of which are not typically used as herbicides, the overall low amount of
sensitivity data from aquatic plants and algae is a recognized trend in the literature (Lewis and
Thursby 2018). The lack of qualified data for aquatic plants and algae is striking given their
critical importance to aquatic food webs and ecosystem, and needs to be addressed in water
quality criteria development.
35
Conclusions
Thirty-six years after the publication of the EPA’s National Guidelines, it is critical to
reassess the factors that influence the development of water quality criteria. Here, we identified
the persistence of major gaps in toxicity data since 1985 that are acting collectively to limit the
EPA’s development of acute water quality criteria. While there was minimal evidence to suggest
that the amount of biological diversity of a toxicity dataset can influence the value of a criterion,
we did identify significant taxonomic differences in sensitivity in the toxicity data from a diverse
group of aquatic pollutants which suggests that it may be necessary to consider taxonomic
patterns of sensitivity during future criteria development. This is particularly evident when it
comes to the need to generate and incorporate toxicity data for primary producers. Finally,
because the current rate of criteria development dramatically lags behind the identification of
harmful pollutants in aquatic environments, we recommend the EPA formally incorporate
interspecies correlation estimation models into the National Guidelines and reconsider their
definition of acceptable toxicity data when calculating water quality criteria. Changes such as
these are necessary if the EPA is to meet the goals of the Clean Water Act and keep pace with
the growing number of chemical pollutants that threaten the aquatic environment.
References
Ankley G, Barron M, Bergeron C, Burgess R, Cormier S, Etterson M, Hoff D, Lalone C, Mount
D, Suter G, et al. 2017. Update on US EPA’s Revision to the 1985 Guidelines for Deriving
Aquatic Life Criteria. :1–3.
Arthur JW, West CW, Allen KN, Hedtke SF. 1987. Seasonal Toxicity of Ammonia to Five Fish
and Nine Invertebrate Species. Bull Environ Contam. 38:324–331.
Augspurger T, Keller A, Black M, Cope G, Dwyer J. 2003. Water Quality Guidance for
Protection of Freshwater Mussels (Unionidae) from Ammonia Exposure. Environ Sci Technol.
22(11):2569–2575.
Awkerman J, Raimondo S, Jackson C, Barron M. 2014. Augmenting Aquatic Species Sensitivity
Distributions with Interspecies Toxicity Estimation Models. Environ Toxicol Chem. 33(3):688–
695. doi:10.1002/etc.2456.
36
Board WE. 2021. World Register of Marine Species. doi:10.14284/170.
Buchwalter DB, Clements WH, Luoma SN. 2017. Modernizing Water Quality Criteria in the
United States: A Need to Expand the Definition of Acceptable Data. Environ Toxicol Chem.
36(2):285–291. doi:10.1002/etc.3654.
Burnham K, Anderson D. 2002. Model Selection and Multimodel Inference.
Canadian Council of Ministers of the Environment. 2007. A Protocol for the Derivation of Water
Quality Guidelines for the Protection of Aquatic Life 2007.
Carr GJ, Belanger SE. 2018. SSDs Revisited: Part I—A Framework for Sample Size Guidance
on Species Sensitivity Distribution Analysis. Environ Toxicol Chem. 38(7):1514–1525.
doi:10.1002/etc.4445. http://dx.doi.org/10.1002/etc.4445.
Clarke KR, Warwick RM. 1998. A taxonomic distinctness index and its statistical properties. J
Appl Ecol. 35(4):523–531. doi:https://doi.org/10.1046/j.1365-2664.1998.3540523.x.
https://doi.org/10.1046/j.1365-2664.1998.3540523.x.
Cormier SM, Paul JF, Spehar RL, Shaw-Allen P, Berry WJ, Suter GW. 2008. Using field data
and weight of evidence to develop water quality criteria. Integr Environ Assess Manag.
4(4):490–504. doi:https://doi.org/10.1897/IEAM_2008-018.1.
https://doi.org/10.1897/IEAM_2008-018.1.
Ding C, Ma Y, Li X, Zhang T, Wang X. 2016. Derivation of soil thresholds for lead applying
species sensitivity distribution: A case study for root vegetables. J Hazard Mater. 303:21–27.
doi:https://doi.org/10.1016/j.jhazmat.2015.10.027.
https://www.sciencedirect.com/science/article/pii/S0304389415301552.
Elias M, Ankley G, Barron M, Eignor D, Erickson R, Hoff D, Jarvis L, Justice JR, Lalone C,
Mount D, et al. 2019. Developing Protective Aquatic Life Values for Data-Limited PFAS
Chemicals Criteria Development and Minimum Data Requirements ( MDRs ). :1–18.
Giddings JM, Wirtz J, Campana D, Dobbs M. 2019. Derivation of combined species sensitivity
distributions for acute toxicity of pyrethroids to aquatic animals. Ecotoxicology. 28:242–250.
doi:10.1007/s10646-019-02018-0. http://dx.doi.org/10.1007/s10646-019-02018-0.
Guidelines and Methodology Used in the Preparation of Health Effect Assessment Chapters of
the Consent Decree Water Criteria Documents. 1980. Fed Regist. 45(231).
Leung KMY, Morritt D, Wheeler JR, Whitehouse P, Sorokin N, Toy R, Holt M, Crane M. 2001.
Can saltwater toxicity be predicted from freshwater data? Mar Pollut Bull. 42(11):1007–1013.
doi:10.1016/S0025-326X(01)00135-7.
Lewis M, Thursby G. 2018. Aquatic plants: Test species sensitivity and minimum data
requirement evaluations for chemical risk assessments and aquatic life criteria development for
the USA. Environ Pollut. 238:270–280. doi:10.1016/j.envpol.2018.03.003.
https://doi.org/10.1016/j.envpol.2018.03.003.
37
Liu Q, Xu Xiaoqun, Zeng J, Huang W, Xu Xudan, Shou L, Chen Q. 2019. Development of
marine water quality criteria for inorganic mercury in China based on the retrievable toxicity
data and a comparison with relevant criteria or guidelines. Ecotoxicology. 28(4):412–421.
doi:10.1007/s10646-019-02032-2. http://dx.doi.org/10.1007/s10646-019-02032-2.
Oksanen AJ, Kindt R, Legendre P, Hara BO, Simpson GL, Stevens MHH, Wagner H. 2007. The
vegan Package. Community Ecol Packag.(May 2014):631–637.
Pavlaki MD, Araújo MJ, Cardoso DN, Silva ARR, Cruz A, Mendo S, Soares AMVM, Calado R,
Loureiro S. 2016. Ecotoxicity and genotoxicity of cadmium in different marine trophic levels.
Environ Pollut. 215:203–212. doi:10.1016/j.envpol.2016.05.010.
R Core Team. 2019. R: A language and environment for statistical computing. https://www.rproject.org/.
Raimondo S, Vivian DN, Barron MG. 2010. Web-based Interspecies Correlation Estimation
(Web-ICE) for Acute Toxicity : User Manual. (January).
Seegert G, Fava J, Cumbia P. 1985. How Representative Are the Data Sets Used to Derive
National Water Quality Criteria ? Aquat Toxicol Hazard Assess Seventh Symp. 854:527–537.
doi:10.1520/STP36288S.
Shannon C. 1948. A Mathematical Theory of Communication. Bell Syst Tech J. 27:379–423.
Solomon KR, Baker DB, Richards RP, Dixon KR, Klaine SJ, La Point TW, Kendall RJ,
Weisskopf CP, Giddings JM, Giesy JP, et al. 1996. Ecological risk assessment of atrazine in
North American surface waters. Environ Toxicol Chem. 15(1):31–76.
doi:https://doi.org/10.1002/etc.5620150105. https://doi.org/10.1002/etc.5620150105.
Song W, Chen S, Liu J, Chen L, Song N, Li N, Liu B. 2015. Variation of Cd concentration in
various rice cultivars and derivation of cadmium toxicity thresholds for paddy soil by speciessensitivity distribution. J Integr Agric. 14(9):1845–1854. doi:https://doi.org/10.1016/S2095-
3119(14)60926-6. https://www.sciencedirect.com/science/article/pii/S2095311914609266.
Stephan CE. 1984. Are the “National Guidelines” Based on Sound Judgements? Aquat Toxicol
Hazard Assess Seventh Symp.
Tenbrook PL, Palumbo AJ, Fojut TL, Hann P, Karkoski J, Tjeerdema RS. 2010. The University
of California-Davis methodology for deriving aquatic life pesticide water quality criteria. Rev
Environ Contam Toxicol. 209(August):1–172. doi:10.1007/978-1-4419-6883-8_1.
Thorley J, Schwarz C. 2018. ssdtools : An R package to fit Species Sensitivity Distributions. J
Open Source Softw. 3(31):2–3. doi:10.1007/b97636.
U.S. Environmental Protection Agency. 2013. Aquatic Life Ambient Water Quality Criteria for
Ammonia - Freshwater (2013) Fact Sheet.
U.S. Environmental Protection Agency. 2003. Ambient Aquatic Life Water Quality Criteria for
Atrazine - Revised Draft.
38
U.S. Environmental Protection Agency. 2021. National Recommended Water Quality Criteria -
Aquatic Life Criteria Table. https://www.epa.gov/wqc/national-recommended-water-qualitycriteria-aquatic-life-criteria-table.
Wang Z, Kwok KWH, Lui GCS, Zhou G-J, Lee J-S, Lam MHW, Leung KMY. 2014. The
difference between temperate and tropical saltwater species’ acute sensitivity to chemicals is
relatively small. Chemosphere. 105:31–43.
doi:https://doi.org/10.1016/j.chemosphere.2013.10.066.
https://www.sciencedirect.com/science/article/pii/S0045653513015002.
Warne Ms, Batley G, van Dam R, Chapman J, Fox D, Hickey C, Stauber J. 2015. Revised
Method for Deriving Australian and New Zealand Water Quality Guideline Values for
Toxicants. (August).
Wheeler JR, Leung KMY, Morritt D, Sorokin N, Rogers H, Toy R, Holt M, Whitehouse P,
Crane M. 2002. Freshwater to saltwater toxicity extrapolation using species sensitivity
distributions. Environ Toxicol Chem. 21(11):2459–2467.
doi:https://doi.org/10.1002/etc.5620211127. https://doi.org/10.1002/etc.5620211127.
Wilcut L, Lehmann W, Elias M. 2015. Updating EPA’s Guidelines for Deriving National
Recommended Water Quality Criteria. Invit Expert Meet Revising US EPA’s Guidel Deriving
Aquat Life Criteria.
Wu J, Yan Z, Liu Z, Liu J, Liang F, Wang X, Wang W. 2015. Development of water quality
criteria for phenanthrene and comparison of the sensitivity between native and non-native
species. Environ Pollut. 196:141–146. doi:10.1016/j.envpol.2014.09.024.
http://dx.doi.org/10.1016/j.envpol.2014.09.024.
39
Figures and Tables
Figure 2-1: Taxonomic composition of the trimmed acute toxicity datasets from marine (A/B)
and freshwater species (C, D) for twelve chemicals that were available in 1985 (A, C) and 2021
(B, D). Phylum codes: An = Annelida, Ar = Arthropoda, B = Bryozoa, Ch = Chordata, Cn =
Cnidaria, E = Echinodermata, M = Mollusca, N = Nematoda, P = Platyhelminthes, R = Rotifera.
40
Figure 2-2: Distribution of acute toxicity data for endosulfan from saltwater (A) and freshwater
(B) species. The x-axes of each plot have been log-transformed. The black squares in each box
represent the mean acute sensitivity of a phylum, the means of phyla with a common letter are
not significantly different according to Dunn’s test. Phylum codes: An = Annelida, Ar =
Arthropoda, Ch = Chordata, M = Mollusca.
41
Table 2-1: Properties of analyzed chemicals
Chemical CAS
Number Origin Class Mode of Action
Ammonia† 7664417 Natural Inorganic Osmoregulatory
impairment
Cadmium*† 7440439 Natural Metal
Metallic
iono/osmoregulatory
impairment
Copper*† 7440508 Natural Metal
Metallic
iono/osmoregulatory
impairment
Nickel*† 7440020 Natural Metal
Metallic
iono/osmoregulatory
impairment
Phenol* 108952 Natural Phenol Polar narcosis
Toluene* 108883 Natural Aromatic
Hydrocarbon Nonpolar narcosis
4-Nitrophenol* 100027 Synthetic Nitrophenol Polar narcosis
Atrazine 1912249 Synthetic Triazine Narcosis
Endosulfan*† 115297 Synthetic Organochlorine Neurotoxicity
Endrin*† 72208 Synthetic Organochlorine Neurotoxicity
Pentachlorophenol*† 87865 Synthetic Organochlorine Electron transport
inhibition
Tributyltin Oxide† 56359 Synthetic Organotin Electron transport
inhibition
* Chemical is a member of the Priority Pollutant List
† Chemical has a criterion in the Aquatic Life Criteria Table
42
Table 2-2: Comparison of mean criteria from datasets that were assembled using the minimum
data requirements (TS8/TF8) to the mean criteria from randomly assembled datasets (RS8/RF8)
Chemical
Saltwater Freshwater
Mean TS8
Criterion
(ug/L)
Mean RS8
Criterion
(ug/L)
Mean TF8
Criterion
(ug/L)
Mean RF8
Criterion
(ug/L)
Ammonia 19 19 9.7 11
Atrazine 13† 12† 21 20
Cadmium 4.3 4.3 2.9* 4.2*
Copper 3.9* 3.5* 2.5† 3.2†
Endosulfan 0.22 0.27 0.81 0.83
Endrin 0.053 0.066 0.16 0.17
Nickel - - - -
4-Nitrophenol - - - -
Pentachlorophenol 4.1 4.3 4.1 4,1
Phenol 25 26 31 32
Tributyltin Oxide 0.98† 0.92† 0.74 0.79
Toluene - - - -
* T-test p < 0.05
† Mann-Whitney U-test p < 0.05
43
Table 2-3: Linear model summary for the Shannon diversity index of a quality-controlled
toxicity dataset as a predictor of criterion value
Chemical Conditions
EPA Criteria SSD Criteria
RSquared
Regression
Coefficient p-value RSquared
Regression
Coefficient p-value
Ammonia Saltwater - - - - - -
Freshwater 0.10 31 0.014 -0.020 -12 0.83
Atrazine Saltwater - - - - - -
Freshwater 0.019 -39 0.17 -7.5E-05 -413 0.32
Cadmium Saltwater 0.015 3.7 0.19 -0.0061 16 0.41
Freshwater 0.017 -1.4 0.18 -0.020 -0.46 0.90
Copper Saltwater - - - - - -
Freshwater 0.044 -1.6 0.078 0.010 -2.0 0.22
Endosulfan Saltwater 0.028 0.028 0.13 0.031 0.011 0.11
Freshwater 0.052 0.33 0.087 0.012 0.050 0.21
Endrin Saltwater -0.016 0.0068 0.63 0.0066 0.017 0.26
Freshwater -0.021 0.0073 0.98 -0.020 0.012 0.89
Nickel Saltwater - - - - - -
Freshwater - - - - - -
4-Nitrophenol Saltwater - - - - - -
Freshwater - - - - - -
Pentachlorophenol Saltwater 0.17 -4.5 0.0016* 0.10 -15 0.014*
Freshwater -0.0017 1.0 0.34 0.0029 5.7 0.29
Phenol Saltwater - - - - - -
Freshwater -0.016 53 0.64 -0.015 1124 0.60
Tributyltin Oxide Saltwater - - - - - -
Freshwater - - - - - -
Toluene Saltwater - - - - - -
Freshwater - - - - - -
* p < 0.05
44
Table 2-4: Kruskal-Wallis comparison of phyla
Chemical Conditions
Kruskal-Wallis Test
χ
2
DF p-value
Ammonia Saltwater 1.1 2 0.58
Freshwater 13 3 0.0041*
Atrazine Saltwater 1.6 1 0.20
Freshwater 0.57 1 0.45
Cadmium Saltwater 25 2 3.8E-06*
Freshwater 18 4 0.0012*
Copper Saltwater 4.6 1 0.032*
Freshwater 79 4 2.4E-16*
Endosulfan Saltwater 23 3 3.8E-05*
Freshwater 59 2 1.8E-13*
Endrin Saltwater 3.1 1 0.080
Freshwater 9.1 1 0.0025*
Nickel Saltwater 5.2 2 0.076
Freshwater 6.8 2 0.054
4-Nitrophenol Saltwater 2.3 1 0.13
Freshwater 2.2 1 0.14
Pentachlorophenol Saltwater 4 3 0.26
Freshwater 122 4 <2.2E-16*
Phenol Saltwater 5.6 2 0.062
Freshwater 17 4 0.0024*
Tributyltin Oxide Saltwater 16 3 0.0013*
Freshwater 23 4 0.00016*
Toluene Saltwater 2.4 1 0.12
Freshwater 3.7 3 0.16
* p < 0.05
DF = Degrees of Freedom
45
Supplementary Materials
Figure S2-1: Taxonomic composition of acute toxicity data before application of quality control
parameters from saltwater (A, B) and freshwater species (C, D) for twelve chemicals that were
available in 1985 (A, C) and 2021 (B, D). Phylum codes: An = Annelida, Ar = Arthropoda, B =
Bryozoa, Cg = Chaetognatha, Ch = Chordata, Cn = Cnidaria, E = Echinodermata, M = Mollusca,
N = Nematoda, P = Platyhelminthes, R = Rotifera
46
Figure S2-2: Histograms of Δ+ values from 1000 random subsamples of saltwater (A; n = 52)
and freshwater (B; n = 83) species data. The dashed blue lines represent the interval over which
95% of the simulated Δ+ values are found. The solid red lines represent the Δ+ from the 2021
datasets and the dotted red lines represent the Δ+ values from the 1985 datasets.
47
Figure S2-3: Distribution of acute toxicity data for ammonia from saltwater (A) and freshwater
(B) species. The x-axes of each plot have been log-transformed. The black squares in each box
represent the mean acute sensitivity of a phylum, phyla with a common letter are not
significantly different according to Dunn’s test. Phylum codes: An = Annelida, Ar = Arthropoda,
Ch = Chordata, M = Mollusca.
48
Figure S2-4: Distribution of acute toxicity data for atrazine from saltwater (A) and freshwater
(B) species. The x-axes of each plot have been log-transformed. The black squares in each box
represent the mean acute sensitivity of a phylum, phyla with a common letter are not
significantly different according to Dunn’s test. The NS (not significant) indicates that the
Kruskal-Wallis test detected no difference between groups. Phylum codes: Ar = Arthropoda, Ch
= Chordata.
49
Figure S2-5: Distribution of acute toxicity data for cadmium from saltwater (A) and freshwater
(B) species. The x-axes of each plot have been log-transformed. The black squares in each box
represent the mean acute sensitivity of a phylum, phyla with a common letter are not
significantly different according to Dunn’s test. Phylum codes: An = Annelida, Ar = Arthropoda,
B = Bryozoa, Ch = Chordata, M = Mollusca.
50
Figure S2-6: Distribution of acute toxicity data for copper from saltwater (A) and freshwater (B)
species. The x-axes of each plot have been log-transformed. The black squares in each box
represent the mean acute sensitivity of a phylum, phyla with a common letter are not
significantly different according to Dunn’s test. The NS indicates that the Kruskal-Wallis test
detected no difference between groups. Phylum codes: Ar = Arthropoda, B = Bryozoa, Ch =
Chordata, Cn = Cnidaria, M = Mollusca.
51
Figure S2-7: Distribution of acute toxicity data for saltwater from saltwater (A) and freshwater
(B) species. The x-axes have been log-transformed. The black squares in each box represent the
mean acute sensitivity of a phylum, phyla with a common letter are not significantly different
according to Dunn’s test. The NS (not significant) indicates that the Kruskal-Wallis test detected
no difference between groups. Phylum codes: Ar = Arthropoda, Ch = Chordata.
52
Figure S2-8: Distribution of acute toxicity data for nickel from saltwater (A) and freshwater (B)
species. The x-axes of each plot have been log-transformed. The black squares in each box
represent the mean acute sensitivity of a phylum, phyla with a common letter are not
significantly different according to Dunn’s test. The NS (not significant) indicates that the
Kruskal-Wallis test detected no difference between groups. Phylum codes: Ar = Arthropoda, Ch
= Chordata, M = Mollusca.
53
Figure S2-9: Distribution of acute toxicity data for 4-nitrophenol from saltwater (A) and
freshwater (B) species. The x-axes of each plot have been log-transformed. The black squares in
each box represent the mean acute sensitivity of a phylum, phyla with a common letter are not
significantly different according to Dunn’s test. The NS (not significant) indicates that the
Kruskal-Wallis test detected no difference between groups. Phylum codes: Ar = Arthropoda, Ch
= Chordata.
54
Figure S2-10: Distribution of acute toxicity data for pentachlorophenol from saltwater (A) and
freshwater (B) species. The x-axes of each plot have been log-transformed. The black squares in
each box represent the mean acute sensitivity of a phylum, phyla with a common letter are not
significantly different according to Dunn’s test The NS (not significant) indicates that the
Kruskal-Wallis test detected no difference between groups. The x-axes have been logtransformed. Phylum codes: An = Annelida, Ar = Arthropoda, Ch = Chordata, M = Mollusca, N
= Nematoda, P = Platyhelminthes.
55
Figure S2-11: Distribution of acute toxicity data for phenol from saltwater (A) and freshwater
(B) species. The x-axes of each plot have been log-transformed. The black squares in each box
represent the mean acute sensitivity of a phylum, phyla with a common letter are not
significantly different according to Dunn’s test. The x-axes have been log-transformed. Phylum
codes: An = Annelida, Ar = Arthropoda, Ch = Chordata, M = Mollusca, P = Platyhelminthes.
56
Figure S2-12: Distribution of acute toxicity data for tributyltin oxide from saltwater (A) and
freshwater (B) species. The x-axes of each plot have been log-transformed. The black squares in
each box represent the mean acute sensitivity of a phylum, phyla with a common letter are not
significantly different according to Dunn’s test. The NS (not significant) indicates that the
Kruskal-Wallis test detected no difference between groups. The x-axes have been logtransformed. Phylum codes: An = Annelida, Ar = Arthropoda, Ch = Chordata, Cn = Cnidaria, M
= Mollusca.
57
Figure S2-13: Distribution of acute toxicity data for toluene from saltwater (A) and freshwater
(B) species. The x-axes of each plot have been log-transformed. The black squares in each box
represent the mean acute sensitivity of a phylum, phyla with a common letter are not
significantly different according to Dunn’s test. The NS (not significant) indicates that the
Kruskal-Wallis test detected no difference between groups. The x-axes have been logtransformed. Phylum codes: Ar = Arthropoda, Ch = Chordata, M = Mollusca.
58
Chapter 3: Phylogeny predicts sensitivity in aquatic animals for only a minority of
chemicals
Alice L. Coleman and Suzanne Edmands
This chapter appears as published in Ecotoxicology.
Abstract
There are substantial gaps in our empirical knowledge of the effects of chemical exposure
on aquatic life that are unlikely to be filled by traditional laboratory toxicity testing alone. One
possible alternative of generating new toxicity data is cross-species extrapolation (CSE), a
statistical approach in which existing data are used to predict the effect of a chemical on untested
species. Some CSE models use relatedness as a predictor of chemical sensitivity, but relatively
little is known about how strongly shared evolutionary history influences sensitivity across all
chemicals. To address this question, we conducted a survey of phylogenetic signal in the toxicity
data from aquatic animal species for a large set of chemicals using a phylogeny inferred from
taxonomy. Strong phylogenetic signal was present in just nine of thirty-six toxicity datasets, and
there were no clear shared properties among those datasets with strong signal. Strong signal was
rare even among chemicals specifically developed to target insects, meaning that these chemicals
may be equally lethal to non-target taxa, including chordates. When signal was strong, distinct
patterns of sensitivity were evident in the data, which may be informative when assembling
toxicity datasets for regulatory use. Although strong signal does not appear to manifest in aquatic
toxicity data for most chemicals, we encourage additional phylogenetic evaluations of toxicity
data in order to guide the selection of CSE tools and as a means to explore the patterns of
chemical sensitivity across the broad diversity of life.
59
Introduction
Pollution represents one of the greatest threats to global biodiversity (Novacek and
Cleland 2001). The biological consequences of pollution exposure are numerous, including
reduced fecundity, compromised immune systems, developmental abnormalities and outright
mortality (McKinlay et al. 2008), all of which threaten the health and functioning of organisms,
populations and ecosystems at large. In recognition of this danger, regulatory bodies like the
United States Environmental Protection Agency (EPA) manage pollution in part by setting
numeric criteria that are intended to protect life by limiting the accumulation of hazardous
concentrations of chemicals in the environment. The development of such criteria depends on
empirical data that describe the concentrations of chemicals that inflict adverse effects on
organisms, however, most chemicals do not have robust toxicity datasets (J R Wheeler et al.
2002; Dowse et al. 2013; Coleman and Edmands 2022).
Most toxicity data are derived from single-species toxicity tests that measure the dose of
a substance that induces adverse effects in a test population over a certain period. These tests are
classified as either acute (short-term) or chronic (long-term), with chronic tests typically
performed less frequently because of the high costs associated with long-term testing. In theory,
the species evaluated in such tests should be representative of the biological community a
criterion is intended to protect, but in practice, most testing is performed using a select group of
model species (Seegert et al. 1985; Anderson and Phillips 2016; David B Buchwalter et al.
2017). As a result, our existing toxicity database does not necessarily reflect the broad diversity
of life nor the entire spectrum of sensitivity to any given chemical. The logical means of
addressing these data gaps would be to simply increase testing, but the sheer number of possible
species-chemical combinations means that expanding laboratory efforts is not a viable solution.
As a result, there is a demand for other tools that can reliably estimate how sensitive a species is
60
to any given chemical. At present, computational methods that extrapolate existing toxicity data
to untested species (i.e cross-species extrapolation) represent the most promising alternative to
traditional laboratory testing (van den Berg et al. 2021; LaLone et al. 2021).
There are four main types of cross-species extrapolation (CSE) models: interspecies
correlation, trait, genomic and relatedness-based (van den Berg et al. 2021). Here, we are most
interested in relatedness-based models, which operate under the assumption that evolutionary
relationships can explain the variation in chemical sensitivity across taxa. These models generate
new toxicity data for species by using metrics of relatedness as predictors of sensitivity. One
such metric used in relatedness-based CSE is phylogenetic signal, which is a measure of the
statistical dependence among species trait values that arises from their phylogenetic relationships
(Revell et al. 2008; van den Berg et al. 2021). When phylogenetic signal is high, close relatives
on a phylogeny will tend to exhibit very similar trait values while distantly related species will
not. When signal is low, trait values will tend to be randomly distributed across a phylogeny and
distantly related taxa may resemble each other more than close relatives (Kamilar and Muldoon
2010; Kamilar and Cooper 2013). Chemical sensitivity is regarded as a possible candidate to
exhibit strong phylogenetic signal (Hylton et al. 2018).
Sensitivity to a chemical is a complex phenotype determined by a network of interacting
morphological, physiological and behavioral traits that regulate the uptake, metabolism and
excretion of pollutants by an organism. Many traits linked to sensitivity such as body size and
metabolic rate are known to exhibit phylogenetic signal, suggesting that we are likely to find
some amount of signal in chemical sensitivity data (Kamilar and Cooper 2013; Hylton et al.
2018). To date, only a small number of studies have evaluated toxicity data for phylogenetic
signal, some of which identified meaningful associations between shared evolutionary history
61
and species sensitivity (Buchwalter et al. 2008; Guénard et al. 2011; Hammond et al. 2012;
Chiari et al. 2015; Hylton et al. 2018; Moore et al. 2020; Duque et al. 2023). The scope of this
work has been relatively narrow to date, with emphasis placed on data from either a single
chemical or taxonomic group at a time. These limitations mean that we do not have a full
understanding of the abundance of strong phylogenetic signal in toxicity data and so the
practicality of using relatedness-based CSE to fill data gaps on a large scale remains in question.
Improved knowledge of phylogenetic signal in toxicity data may have other uses aside
from cross-species extrapolation. One option would be to test hypotheses that explain why
phylogenetic signal manifests more strongly in some datasets than others. To this end, we
propose three hypotheses based on qualitative properties of chemicals and toxicity data. First, we
predict that strong phylogenetic signal will be more common among synthetic chemicals than
naturally occurring chemicals. Synthetic chemicals are relative newcomers in the environment,
so potentially only a subset of lineages will have had enough exposure for selection to favor
increased tolerance. Second, we predict that strong phylogenetic signal will be more common in
the data from chemicals with specific toxic modes of action (MOAs) than those that are
nonspecific. This is because specific MOAs interfere with biological processes by precisely
binding to a particular site or molecule, so an adaptive phenotype could in theory arise after only
a small number of molecular changes (Whitehead et al. 2017; Gupta 2018). Chemicals with
nonspecific MOAs tend to induce generic stress responses, so the corresponding adaptive
phenotypes may be complex and thus take much longer to develop (Whitehead et al. 2017).
Lastly, we predict that strong phylogenetic signal will be more common in chronic toxicity
datasets than acute because rapid responses to acute stress are often generalized and
evolutionarily conserved (Kültz 2020). In contrast, responses to chronic stress have been found
62
to be more chemical- and lineage-specific (Kovalchuk et al. 2007; McRae et al. 2022). Testing
hypotheses such as these may provide insight into whether qualitative properties are viable
predictors of strong phylogenetic signal, which would be useful for identifying the datasets to
which relatedness-based CSE models can be applied.
In this study we expand upon previous phylogenetic analyses in ecotoxicology by
considering toxicity data from both a large number of chemical pollutants and a biologically
diverse set of species. Using published laboratory toxicity data and a phylogeny inferred from
taxonomy, we quantified the phylogenetic signal present in these data and investigated whether
various chemical properties and experimental conditions such as temperature and pH affected
phylogenetic signal. When strong signal was identified, we additionally used the phylogenetic
analyses to identify the sensitive and resistant clades within a dataset. The application of CSE
methods to environmental risk assessment and regulation has been a recent focus in aquatic
toxicology (Raimondo, Jackson, et al. 2010; Schlekat et al. 2010; Lewis and Thursby 2018;
Coleman and Edmands 2022), so here our analysis specifically dealt with data similar to those
used in water quality criteria development.
Methods
Data Collection
Toxicity data from aquatic species were collected for twenty-four chemicals (Table 3-1)
that span a range of classes and modes of action (MOA). The MOA of each chemical was
obtained and labeled as either generic or precise using the MOAtox database from Barron et al.
(2015). Acute toxicity data were then obtained from the EPA’s ECOTOXicology
Knowledgebase (ECOTOX; Olker et al. 2022) using these search parameters:
1. Chemical Abstracts Service (CAS) Registry Number
63
2. Endpoint: Median Lethal Concentration (LC50)
3. Kingdom: Animals
4. Test Location: Lab
5. Exposure Media: Freshwater or Saltwater
6. Duration: 48 or 96 hours
An LC50 represents the concentration of a toxic substance that kills half of the test
organisms during the testing period. LC50s from 96-hour tests were used for all animals except
cladocerans and midges, for which we collected 48-hour data. These parameters were
specifically designed to approximate the data collection rules stated in the EPA’s methodology
for water quality criteria development (Stephan et al. 1985).
Chronic toxicity data were similarly mined from ECOTOX for twelve chemicals (Table
3-2) using the following parameters:
1. CAS Registry Number
2. Endpoint: No Observed Effect Concentration (NOEC)
3. Kingdom: Animals
4. Test Location: Lab
5. Exposure Media: Freshwater or Saltwater
6. Duration: 7 – 60 days
The NOEC represents the highest concentration of a chemical that does not induce a
response that differs significantly from the control treatment in a toxicity test. The remaining 12
chemicals from the initial set of 24 were excluded because they did not have NOEC data from at
least ten unique species available in ECOTOX. As there is no universally agreed upon period
64
that constitutes a chronic toxicity test, we allowed for a broad range of durations to maximize the
amount of usable data.
Test organism life stage, experimental temperature, pH and hardness (for metals only) for
each toxicity value were also obtained from ECOTOX when possible because these variables are
known to modify the toxicity of many chemicals to aquatic species (Cairns et al. 1975; Yim et al.
2006; Pinheiro et al. 2021; Kazmi et al. 2022). Organism life stage was standardized to either
“subadult”, “adult” or “NR” (not reported) based on the information provided by ECOTOX. We
then filtered the data based on the availability of each experimental variable to create the
following five variants of each dataset:
1. Complete: all toxicity values obtained from ECOTOX
2. Subadult: only toxicity values derived from subadult test organisms
3. Temperature: only toxicity values with experimental temperature information
4. pH: only toxicity values with experimental pH information
5. Hardness: only toxicity values with experimental hardness information
If more than one toxicity value was available for a given chemical-species pair, the
values were summarized as their geometric mean. Similarly, if the toxicity values for a chemicalspecies pair were derived from tests performed at different temperatures, pH or hardnesses,
information from each variable was summarized as a geometric mean.
Phylogenetic Analyses
We used the National Center for Biotechnology Information’s (NCBI) Taxonomy
database via the phyloT generator (Letunica 2022) to produce a single comprehensive
phylogenetic tree for all of the species that appeared in the toxicity datasets we assembled.
65
Species without a matching entry in the NCBI database were excluded from the tree. We
imported the tree generated by phyloT into R using the package ape (Paradis and Schliep 2019),
where we then computed its branch lengths using Grafen’s transformation (ρ = 1; Grafen and
Hamilton 1989) and annotated its tips with the toxicity data.
Phylogenetic signal was estimated as Pagel’s λ (Pagel 1998). The values of λ range
between zero and one, where a value close to zero indicates that the trait is phylogenetically
independent and a λ near one indicates that the distribution of the trait is consistent with the
Brownian Motion model, wherein closely related lineages are more similar to each other than
those that are distantly related (Münkemüller et al. 2012). Following Hylton et al. (2018), we
considered λ values of 0.5 or greater to be evidence of strong phylogenetic signal. A maximum
of five calculations of λ were performed for each toxicity dataset and its corresponding variants.
The first two were conducted using the function phylosig from the R package phytools (Revell
2012) for the complete dataset while the second was performed for the subset of toxicity data
that were derived from subadult test organisms. When strong signal was identified in a complete
dataset, we plotted the toxicity data alongside the phylogeny (Fig. 1-9) and used it to identify
taxonomic patterns of sensitivity. We used the second calculation of λ to determine whether the
frequency of strong phylogenetic signal in the toxicity data differed in organisms tested at early
(subadult) versus various (subadult, adult and NR) life stages.
Calculations of λ for the temperature, pH and hardness dataset variants were evaluated
using linear phylogenetic generalized least squares (PGLS; Grafen and Hamilton 1989) models.
A form of regression analysis, PGLS models are typically used to test the association between
variables while statistically controlling for the effects of phylogenetic signal in the data
(Symonds and Blomberg 2014). These models also include a measurement of λ in a dataset
66
which is calculated on the model’s residuals using maximum likelihood estimation (Revell 2010;
Symonds and Blomberg 2014; Pagel 1999). Following Chiari et al. (2015) and Hylton et al.
(2018), we used PGLS to calculate phylogenetic signal while statistically controlling for
variation in experimental conditions. PGLS models were implemented using the package caper
(Orme et al. 2018; Grafen and Hamilton 1989), in which toxicity (LC50 or NOEC) was set as the
dependent variable and the environmental information (temperature, pH or hardness) as the
predictor variable. Temperature, pH and hardness were modeled separately because most toxicity
values did not have information for all three variables available.
Results
Data composition
We collected a cumulative total of 9,138 datapoints to create 36 toxicity datasets (24
acute, 12 chronic) for 24 chemicals (9 natural, 15 synthetic; 12 generic MOAs, 12 precise
MOAs). Together, these datasets included values from 595 unique species from ten phyla
(Annelida, Arthropoda, Bryozoa, Chordata, Cnidaria, Echinodermata, Mollusca, Nematoda,
Platyhelminthes and Rotifera). The sizes of the toxicity datasets varied considerably, with the
largest (acute chlorpyrifos) composed of 993 toxicity values from 112 species while the smallest
(chronic lindane) was made up of just 33 toxicity values from 13 species. In general, acute
toxicity datasets were larger than chronic and synthetic chemical datasets were larger those of the
naturally-occurring chemicals (Table S3-1). Dataset sample sizes decreased substantially when
filtered by availability of life stage, temperature, pH or hardness information (Table S3-1).
Phylogenetic signal
Strong phylogenetic signal (λ ≥ 0.5) was identified in nine of the 36 complete toxicity
datasets (acute chlorine, acute guthion, acute lindane, acute pentachlorophenol, chronic
cadmium, chronic copper, chronic diazinon, chronic lindane and chronic phenol) when we did
67
not consider any of life stage, experimental temperature, pH or hardness (Table 3-2).
Convergence on the lower bound of λ was common, as 18 datasets returned a signal value at or
near 7.3E-05. Fisher’s exact tests indicated that there was no difference in frequencies of strong
phylogenetic signal between complete toxicity datasets from chemicals of different origins
(natural vs. synthetic; p = 0.68), classes (see “Class” column in Table 3-1; p = 0.86) or modes of
action (generic vs. precise; p = 1). Similarly, the frequencies of strong phylogenetic signal in
acute and chronic toxicity data were the same (p = 0.085).
Strong phylogenetic signal was also identified in eleven of 36 subadult, five of 36
temperature, eight of 36 pH and five of eight hardness datasets However, the frequencies of
strong signal in each of these dataset variants were not significantly different from the frequency
of strong signal in the complete datasets according to comparisons performed with Fisher’s exact
tests (complete vs. subadult p = 0.79; complete vs. temperature p = 0.37, complete vs. pH p = 1,
complete vs. hardness p = 1).
Patterns of sensitivity
The patterns of relative sensitivity in the complete acute toxicity datasets with strong
phylogenetic signal varied across chemicals. For example, the data for chlorine (λ = 0.9; Fig. 3-
1) suggests that insects are highly sensitive while fishes, crustaceans and molluscs are more
tolerant. Within the crustaceans, the tolerance of decapods appears to exceed that of the
amphipods while in the molluscs, bivalves are more resistant than gastropods. The guthion
dataset (λ = 0.56; Fig. 3-2) indicates that amphibians and Cypriniform fishes are resistant to
guthion exposure, while the crustaceans and fishes from the order Perciformes and family
Salmonidae are highly sensitive. Among the taxa with acute lindane data (λ = 1; Fig. 3-3), the
decapods and amphipods appear to be very sensitive while the molluscs, annelids and frog
68
species are relatively more resistant, while the acute pentachlorophenol dataset (λ = 1; Fig. 3-4)
suggests that most species are highly vulnerable during short-term exposure to the
organochlorine pesticide.
The patterns of relative sensitivity were similarly variable in the chronic toxicity datasets.
The cadmium data (λ = 1; Fig. 3-5) indicates that crustaceans (Malacostraca and Branchiopoda)
are highly sensitive to long-term exposure to the metal while molluscs and fishes are more
tolerant. The majority of species featured in the chronic copper dataset (λ = 1; Fig. 3-6) seem to
experience toxic effects when exposed to low levels of copper, with high relative tolerance
evident in just one species of frog (Pelophylax ridibundus) and a few cases of moderate tolerance
scattered across different clades. Chordates largely appear to be tolerant of chronic diazinon
exposure (λ = 1; Fig. 3-7), while crustaceans are more sensitive. Crustaceans also appear highly
vulnerable to chronic lindane exposure (λ = 1; Fig. 3-8), while bivalves exhibit relatively
moderate tolerance and low sensitivity only evident in one frog species (Rana temporaria).
Similarly, a frog species (Xenopus laevis) was the only taxon to exhibit relatively high resistance
to phenol while the crustaceans seem to be more sensitive (λ = 1; Fig. 3-9).
The phylogenetic trees for datasets without strong phylogenetic signal (λ < 0.5) are
provided in the Supplementary Information (Figs. S3-1:S3-27).
PGLS
Experimental temperature, pH and hardness were significant predictors of toxicity in
PGLS models for six, two and two datasets each respectively (Table 3-3). The adjusted R2
values
of the significant models ranged between 0.04 and 0.47 for temperature, between 0.26 and 0.79
for pH and between 0.94 and 1 for hardness (Table 3-3). Experimental temperature was
positively associated with tolerance (i.e. LC50/NOEC increases when temperature increases) in
69
two datasets and negatively associated with tolerance (i.e. LC50/NOEC decreases when
temperature increases) in the other four with significant models. Experimental pH was positively
associated with tolerance in the acute chlorine data and negatively associated with tolerance in
the chronic zinc data. Hardness was positively associated with tolerance of acute chlorine
exposure, and negatively associated with acute mercury tolerance.
Discussion
Our results indicate that strong phylogenetic signal is rare in toxicity data from aquatic
animals and that its frequency does not appear to be biased by organism life stage, exposure type,
chemical origin, class or mode of action. We also found evidence of significant effects of
experimental temperature, pH and hardness on chemical tolerance in a few datasets, although
there were no consistent generalizable trends in how any of these variables affected toxicity
across all chemicals considered. High variation in phylogenetic signal magnitude across different
chemicals is consistent with the results of the similarly-sized analysis by Hylton et al. (2018),
which found strong signal in just 10 of the 42 toxicity datasets they examined. In general, most
of the previous studies (Buchwalter et al. 2008; Guénard et al. 2011; Hammond et al. 2012;
Chiari et al. 2015; Hylton et al. 2018) that identified strong phylogenetic signal appear to have
less taxonomic breadth in their data than what we used in our work, suggesting that phylogenetic
signal might manifest more strongly at lower taxonomic levels (Carew et al. 2011). The absence
of strong phylogenetic signal among the synthetic chemicals is particularly notable given that
most are selective pesticides that were specifically developed to target insects. Instead, nontarget organisms, including chordates, appear to be equally as sensitive as insects to many of
these chemicals (see Fig. 3-3 for an example) which points to how pesticides can uniformly
threaten all members of aquatic communities.
70
There are several factors that might explain the low frequency of strong phylogenetic
signal in our results. First, it is possible that the sample sizes of some datasets, like chronic
phenol (n = 10 species) and chronic zinc (n = 13 species), were simply too small to provide
adequate statistical power to estimate λ. Sample sizes might increase if data quality collection
parameters were less restrictive, however, this solution is not ideal because of the underlying
variability in our toxicity datasets. This variation arises because an organism’s sensitivity to a
toxicant may be affected by any number of environmental (temperature, pH, salinity, etc.),
biological (size, sex, life stage, etc.) or chemical (mode of action, structure, solubility, etc.)
factors. We aimed to reduce such variation and its effects by setting precise requirements for
certain variables (CAS number, exposure media, test duration) during data collection and
controlling others (life stage, temperature, pH and hardness) in the analyses. However, because
of the sporadic availability of information on toxicity testing conditions in ECOTOX, we left
variables such as chemical purity, dissolved oxygen content and test organism sex unrestricted to
ensure there were enough data to perform the planned analyses. As such, we expect that some
underlying variation remains in our data and may have obscured the phylogenetic signal in some
instances.
Additionally, there is likely substantial statistical noise in the chronic toxicity data that
does not affect acute data because of inconsistencies in the NOEC data we collected. Generated
from post hoc tests after performing an ANOVA, the NOEC represents the highest concentration
of a chemical that does not induce a response that differs significantly from the control. The
biological effect used to determine a NOEC can vary widely between toxicity tests, and such
experimental variation was present in the chronic toxicity datasets we collected (Table S3-2). For
example, the chronic lindane dataset (n = 33 toxicity values) contained toxicity data that were
71
measured using twelve different endpoints, which included but are not limited to mortality,
behavior, population growth, reproduction, enzyme activity and hormone levels. Across all of the
chronic datasets, each of the various effects had a very small average sample size of species (𝑥̅<
4), the smallest unit utilized in our phylogenetic analysis, meaning that utilizing only NOECs
derived from the same biological effect was not a viable approach in our study. Additionally, the
NOEC is generally considered to be a relatively poor indicator of safe chemical concentrations
(Crane and Newman 2000), and as a result there has been a strong push in ecotoxicology to
utilize alternative measures of chronic toxicity (Warne and van Dam 2008; Laskowski 1995;
Kooijman 1996). Given this criticism and substantial issue of experimental variation, NOEC data
does not appear suitable for use in relatedness-based cross-species extrapolation. Opportunities
to work with alternatives, however, are minimal given that chronic data are severely limited or
even nonexistent for many chemicals (de Zwart 2001). The paradox created by this shortage and
the requirement of existing data for typical CSE approaches means we are unlikely to be able to
comprehensively fill gaps in the chronic toxicity database using these methods. Instead,
statistical approaches that extrapolate chronic values using acute data (Duboudin et al. 2004;
Hiki and Iwasaki 2020) appear better suited to this challenge. Given that the endpoints and test
protocols used to measure acute toxicity are more uniform, it is more likely that phylogenetic
methods can be used to address the gaps in the acute database.
The value of phylogenetic approaches to CSE arises from the concept that species data
cannot be considered independent observations (Felsenstein 1985). All species are related within
a hierarchical phylogenetic tree, so similar phenotype values among species could be a product
of limited divergence from a shared common ancestor or convergent evolution (Stone et al.
2011). Most standard statistical tests assume datapoints to be independent, meaning that using
72
such methods to analyze a dataset heavily influenced by evolutionary history could lead to
spurious results. Phylogenetically-informed methods avoid this issue by explicitly accounting for
possible phylogenetic structure in a dataset. In this context, the best-studied method of
phylogenetic CSE is the phylogenetic eigenvector map (PEM; Guénard, Legendre, and PeresNeto 2013; Guénard et al. 2014). A PEM is a set of eigenfunctions derived from a phylogeny
that describe the magnitude of various possible phylogenetic patterns (i.e. phylogenetic signal
values) in a dataset. A subset of these eigenfunctions are then utilized as the independent
variables in a regression model that generates estimates of trait values for species on the
phylogeny that lack data. PEMs have also been combined with descriptors of chemical properties
in a bilinear modelling approach that can predict the toxicity of multiple chemicals to many
species (Guénard et al. 2014), which represents a substantial expansion of conventional
modelling efforts.
Approaching toxicity data with an evolutionary perspective may benefit pollution
management efforts beyond cross-species extrapolation. Here, we identified patterns of
sensitivity in the datasets that exhibited strong phylogenetic signal that correspond with
descriptions from the literature of how these chemicals affect different taxa. For example,
neurotoxins like guthion, an organophosphate insecticide, are typically considered more toxic to
invertebrates than vertebrates (Legradi et al. 2018), which is reflected in our results (Fig. 3-2).
Similarly, freshwater mussels have been found to be more tolerant of chlorine than other aquatic
species (Valenti et al. 2006) which matches with our evaluation of the molluscs in our acute
toxicity dataset for chlorine (Fig. 3-1). These similarities suggest that phylogenetic assessments
of toxicity data can provide reliable insights into patterns of sensitivity which, in a regulatory
context, could aid in water quality criteria development. For instance, by referring to our guthion
73
plot (Fig. 3-2), it becomes clear that amphipods are some of the most acutely sensitive taxa to
guthion. A toxicity dataset for guthion water quality criteria could then be specifically assembled
to include amphipods rather than relying on the EPA’s taxonomic requirements (Stephan et al.
1985) to ensure their representation. While setting unique taxonomic requirements for every
chemical is unrealistic, knowledge of patterns of sensitivity could be used to complement other
efforts to modernize water quality criteria.
Conclusions
Given that understanding of the concentrations of chemicals that adversely affect species
is a vital component of the risk assessment process, it is critical that we address the major gaps in
our toxicity database. Phylogenetic methods of cross-species extrapolation represent a possible
alternative to laboratory testing, provided that chemical sensitivity is strongly influenced by
evolutionary relationships. We found that strong phylogenetic signal is apparent at high
taxonomic levels for only a small subset of chemicals in both acute and chronic data, which was
surprising given the specific targets of many pesticides included in this study. For those acute
toxicity datasets that do feature strong signal, phylogenetic tools provide a framework with
which we can reliably assess patterns in chemical sensitivity and a means of avoiding the
statistical pitfalls associated with phylogenetically structured data. Moving forwards, we
recommend that future efforts in this field evaluate the phylogenetic signal in their data as a
preliminary analytical step to ensure the selection of the appropriate method of cross-species
extrapolation.
References
Anderson B, Phillips B. 2016. Chapter 6 - Saltwater Toxicity Tests. In: Blasco J, Chapman PM,
Campana O, Hampel M, editors. Marine Ecotoxicology. Academic Press. p. 167–197.
https://www.sciencedirect.com/science/article/pii/B9780128033715000060.
74
Barron MG, Lilavois CR, Martin TM. 2015. MOAtox: A comprehensive mode of action and
acute aquatic toxicity database for predictive model development. Aquat Toxicol. 161:102–107.
doi:10.1016/j.aquatox.2015.02.001.
https://www.sciencedirect.com/science/article/pii/S0166445X15000521.
van den Berg SJP, Maltby L, Sinclair T, Liang R, van den Brink PJ. 2021. Cross-species
extrapolation of chemical sensitivity. Sci Total Environ. 753:141800.
doi:10.1016/j.scitotenv.2020.141800.
https://www.sciencedirect.com/science/article/pii/S0048969720353298.
Buchwalter DB, Cain DJ, Martin CA, Xie L, Luoma SN, Garland T. 2008. Aquatic insect
ecophysiological traits reveal phylogenetically based differences in dissolved cadmium
susceptibility. Proc Natl Acad Sci. 105(24):8321–8326. doi:10.1073/pnas.0801686105.
https://doi.org/10.1073/pnas.0801686105.
Buchwalter DB, Clements WH, Luoma SN. 2017. Modernizing Water Quality Criteria in the
United States: A Need to Expand the Definition of Acceptable Data. Environ Toxicol Chem.
36(2):285–291. doi:10.1002/etc.3654. https://doi.org/10.1002/etc.3654.
Cairns JJ, Heath AG, Parker BC. 1975. Temperature influence on chemical toxicity to aquatic
organisms. J Water Pollut Control Fed. 47(2):267–280.
Carew ME, Miller AD, Hoffmann AA. 2011. Phylogenetic signals and ecotoxicological
responses: potential implications for aquatic biomonitoring. Ecotoxicology. 20(3):595–606.
doi:10.1007/s10646-011-0615-3. https://doi.org/10.1007/s10646-011-0615-3.
Chiari Y, Glaberman S, Serén N, Carretero MA, Capellini I. 2015. Phylogenetic signal in
amphibian sensitivity to copper sulfate relative to experimental temperature. Ecol Appl.
25(3):596–602. doi:10.1890/14-0439.1. https://doi.org/10.1890/14-0439.1.
Coleman AL, Edmands S. 2022. Data and Diversity in the Development of Acute Water Quality
Criteria in the United States. Environ Toxicol Chem. 41(5):1333–1343. doi:10.1002/etc.5302.
https://doi.org/10.1002/etc.5302.
Crane M, Newman MC. 2000. What level of effect is a no observed effect? Environ Toxicol
Chem. 19(2):516–519. doi:10.1002/etc.5620190234. https://doi.org/10.1002/etc.5620190234.
Dowse R, Tang D, Palmer CG, Kefford BJ. 2013. Risk assessment using the species sensitivity
distribution method: Data quality versus data quantity. Environ Toxicol Chem. 32(6):1360–1369.
doi:10.1002/etc.2190. https://doi.org/10.1002/etc.2190.
Duboudin C, Ciffroy P, Magaud H. 2004. Acute-to-chronic species sensitivity distribution
extrapolation. Environ Toxicol Chem. 23(7):1774–1785. doi:10.1002/etc.5620230723.
https://doi.org/10.1002/etc.5620230723.
Duque T, Chowdhury S, Isaia M, Pekár S, Riess K, Scherf G, Schäfer RB, Entling MH. 2023.
Sensitivity of spiders from different ecosystems to lambda-cyhalothrin: effects of phylogeny and
climate. Pest Manag Sci. 80(2):857–865. doi:https://doi.org/10.1002/ps.7818.
https://doi.org/10.1002/ps.7818.
75
Felsenstein J. 1985. Phylogenies and the comparative method. Am Nat. 125(1):1–15.
Grafen A, Hamilton WD. 1989. The phylogenetic regression. Philos Trans R Soc London B, Biol
Sci. 326(1233):119–157. doi:10.1098/rstb.1989.0106. https://doi.org/10.1098/rstb.1989.0106.
Guénard G, Carsten von der Ohe P, Carlisle Walker S, Lek S, Legendre P. 2014. Using
phylogenetic information and chemical properties to predict species tolerances to pesticides. Proc
R Soc B Biol Sci. 281(1789):20133239. doi:10.1098/rspb.2013.3239.
https://doi.org/10.1098/rspb.2013.3239.
Guénard G, Legendre P, Peres-Neto P. 2013. Phylogenetic eigenvector maps: a framework to
model and predict species traits. Methods Ecol Evol. 4(12):1120–1131. doi:10.1111/2041-
210X.12111. https://doi.org/10.1111/2041-210X.12111.
Guénard G, von der Ohe PC, de Zwart D, Legendre P, Lek S. 2011. Using phylogenetic
information to predict species tolerances to toxic chemicals. Ecol Appl. 21(8):3178–3190.
doi:10.1890/10-2242.1. https://doi.org/10.1890/10-2242.1.
Gupta PK. 2018. Chapter 3 - Mechanism of toxicity. In: Gupta PKBT-IT, editor. Illustrated
Toxicology. Academic Press. p. 107–129.
https://www.sciencedirect.com/science/article/pii/B9780128132135000031.
Hammond JI, Jones DK, Stephens PR, Relyea RA. 2012. Phylogeny meets ecotoxicology:
evolutionary patterns of sensitivity to a common insecticide. Evol Appl. 5(6):593–606.
doi:10.1111/j.1752-4571.2011.00237.x. https://doi.org/10.1111/j.1752-4571.2011.00237.x.
Hiki K, Iwasaki Y. 2020. Can We Reasonably Predict Chronic Species Sensitivity Distributions
from Acute Species Sensitivity Distributions? Environ Sci Technol. 54(20):13131–13136.
doi:10.1021/acs.est.0c03108. https://doi.org/10.1021/acs.est.0c03108.
Hylton A, Chiari Y, Capellini I, Barron MG, Glaberman S. 2018. Mixed phylogenetic signal in
fish toxicity data across chemical classes. Ecol Appl. 28(3):605–611. doi:10.1002/eap.1698.
https://doi.org/10.1002/eap.1698.
Kamilar JM, Cooper N. 2013. Phylogenetic signal in primate behaviour, ecology and life history.
Philos Trans R Soc B Biol Sci. 368(1618):20120341. doi:10.1098/rstb.2012.0341.
https://doi.org/10.1098/rstb.2012.0341.
Kamilar JM, Muldoon KM. 2010. The Climatic Niche Diversity of Malagasy Primates: A
Phylogenetic Perspective. PLoS One. 5(6):e11073.
https://doi.org/10.1371/journal.pone.0011073.
Kazmi SSUH, Wang YYL, Cai Y-E, Wang Z. 2022. Temperature effects in single or combined
with chemicals to the aquatic organisms: An overview of thermo-chemical stress. Ecol Indic.
143:109354. doi:https://doi.org/10.1016/j.ecolind.2022.109354.
https://www.sciencedirect.com/science/article/pii/S1470160X22008275.
Kooijman SALM. 1996. An Alternative for NOEC Exists, but the Standard Model Has to Be
Abandoned First. Oikos. 75(2):310–316. doi:10.2307/3546255.
http://www.jstor.org/stable/3546255.
76
Kovalchuk I, Molinier J, Yao Y, Arkhipov A, Kovalchuk O. 2007. Transcriptome analysis
reveals fundamental differences in plant response to acute and chronic exposure to ionizing
radiation. Mutat Res Mol Mech Mutagen. 624(1):101–113. doi:10.1016/j.mrfmmm.2007.04.009.
https://www.sciencedirect.com/science/article/pii/S0027510707002011.
Kültz D. 2020. Evolution of cellular stress response mechanisms. J Exp Zool Part A Ecol Integr
Physiol. 333(6):359–378. doi:10.1002/jez.2347. https://doi.org/10.1002/jez.2347.
LaLone CA, Basu N, Browne P, Edwards SW, Embry M, Sewell F, Hodges G. 2021.
International Consortium to Advance Cross-Species Extrapolation of the Effects of Chemicals in
Regulatory Toxicology. Environ Toxicol Chem. 40(12):3226–3233. doi:10.1002/etc.5214.
https://doi.org/10.1002/etc.5214.
Laskowski R. 1995. Some Good Reasons to Ban the Use of NOEC, LOEC and Related Concepts
in Ecotoxicology. Oikos. 73(1):140–144. doi:10.2307/3545738.
http://www.jstor.org/stable/3545738.
Legradi JB, Di Paolo C, Kraak MHS, van der Geest HG, Schymanski EL, Williams AJ,
Dingemans MML, Massei R, Brack W, Cousin X, et al. 2018. An ecotoxicological view on
neurotoxicity assessment. Environ Sci Eur. 30(1):46. doi:10.1186/s12302-018-0173-x.
https://doi.org/10.1186/s12302-018-0173-x.
Letunica I. 2022. phyloT: A phylogenetic tree generator, based on NCBI or GTD taxonomy.
https://phylot.biobyte.de/index.cgi.
Lewis M, Thursby G. 2018. Aquatic plants: Test species sensitivity and minimum data
requirement evaluations for chemical risk assessments and aquatic life criteria development for
the USA. Environ Pollut. 238:270–280. doi:10.1016/j.envpol.2018.03.003.
https://doi.org/10.1016/j.envpol.2018.03.003.
McKinlay R, Plant JA, Bell JNB, Voulvoulis N. 2008. Endocrine disrupting pesticides:
Implications for risk assessment. Environ Int. 34(2):168–183. doi:10.1016/j.envint.2007.07.013.
https://www.sciencedirect.com/science/article/pii/S0160412007001444.
McRae CJ, Keshavmurthy S, Meng P-J, Rosset SL, Huang W-B, Chen CA, Fan T-Y, Côté IM.
2022. Variable responses to chronic and acute elevated temperature of three coral species from
reefs with distinct thermal regimes. Mar Biol. 169(7):97. doi:10.1007/s00227-022-04071-6.
https://doi.org/10.1007/s00227-022-04071-6.
Moore DRJ, Priest CD, Galic N, Brain RA, Rodney SI. 2020. Correcting for Phylogenetic
Autocorrelation in Species Sensitivity Distributions. Integr Environ Assess Manag. 16(1):53–65.
doi:10.1002/ieam.4207. https://doi.org/10.1002/ieam.4207.
Münkemüller T, Lavergne S, Bzeznik B, Dray S, Jombart T, Schiffers K, Thuiller W. 2012. How
to measure and test phylogenetic signal. Methods Ecol Evol. 3(4):743–756. doi:10.1111/j.2041-
210X.2012.00196.x. https://doi.org/10.1111/j.2041-210X.2012.00196.x.
Novacek MJ, Cleland EE. 2001. The current biodiversity extinction event: Scenarios for
mitigation and recovery. Proc Natl Acad Sci. 98(10):5466 LP – 5470.
doi:10.1073/pnas.091093698. http://www.pnas.org/content/98/10/5466.abstract.
77
Olker JH, Elonen CM, Pilli A, Anderson A, Kinziger B, Erickson S, Skopinski M, Pomplun A,
LaLone CA, Russom CL, et al. 2022. The ECOTOXicology Knowledgebase: A Curated
Database of Ecologically Relevant Toxicity Tests to Support Environmental Research and Risk
Assessment. Environ Toxicol Chem. 41(6):1520–1539. doi:10.1002/etc.5324.
https://doi.org/10.1002/etc.5324.
Orme D, Freckleton R, Thomas G, Petzoldt T, Fritz S, Isaac N, Pearse W. 2018. caper:
Comparative Analyses of Phylogenetics and Evolution in R, R package version 1.0.1.
https://cran.r-project.org/package=caper.
Pagel M. 1998. Inferring evolutionary processes from phylogenies. Zool Scr. 26(4):331–348.
Pagel M. 1999. Inferring the historical patterns of biological evolution. Nature. 401(6756):877–
884. doi:10.1038/44766. https://doi.org/10.1038/44766.
Paradis E, Schliep K. 2019. ape 5.0: an environment for modern phylogenetics and evolutionary
analyses in R. Bioinformatics. 35:526–528.
Pinheiro JPS, Windsor FM, Wilson RW, Tyler CR. 2021. Global variation in freshwater physicochemistry and its influence on chemical toxicity in aquatic wildlife. Biol Rev. 96(4):1528–1546.
doi:https://doi.org/10.1111/brv.12711. https://doi.org/10.1111/brv.12711.
Raimondo S, Jackson CR, Barron MG. 2010. Influence of Taxonomic Relatedness and Chemical
Mode of Action in Acute Interspecies Estimation Models for Aquatic Species. Environ Sci
Technol. 44(19):7711–7716. doi:10.1021/es101630b. https://doi.org/10.1021/es101630b.
Revell LJ. 2010. Phylogenetic signal and linear regression on species data. Methods Ecol Evol.
1(4):319–329. doi:https://doi.org/10.1111/j.2041-210X.2010.00044.x.
https://doi.org/10.1111/j.2041-210X.2010.00044.x.
Revell LJ. 2012. phytools: an R package for phylogenetic comparative biology (and other
things). Methods Ecol Evol. 3(2):217–223. doi:10.1111/j.2041-210X.2011.00169.x.
https://doi.org/10.1111/j.2041-210X.2011.00169.x.
Revell LJ, Harmon LJ, Collar DC. 2008. Phylogenetic Signal, Evolutionary Process, and Rate.
Syst Biol. 57(4):591–601. doi:10.1080/10635150802302427.
https://doi.org/10.1080/10635150802302427.
Schlekat CE, Van Genderen E, De Schamphelaere KAC, Antunes PMC, Rogevich EC,
Stubblefield WA. 2010. Cross-species extrapolation of chronic nickel Biotic Ligand Models. Sci
Total Environ. 408(24):6148–6157. doi:https://doi.org/10.1016/j.scitotenv.2010.09.012.
https://www.sciencedirect.com/science/article/pii/S0048969710009733.
Seegert G, Fava J, Cumbia P. 1985. How Representative Are the Data Sets Used to Derive
National Water Quality Criteria ? Aquat Toxicol Hazard Assess Seventh Symp. 854:527–537.
doi:10.1520/STP36288S.
Stephan C, Mount D, Hansen D, Gentile J, Chapman G, Brungs W. 1985. Guidelines for
Deriving Numerical National Water Quality Criteria for the Protection of Aquatic Organisms and
Their Uses. US Environmental Protection Agency.
78
Stone GN, Nee S, Felsenstein J. 2011. Controlling for non-independence in comparative analysis
of patterns across populations within species. Philos Trans R Soc B Biol Sci. 366(1569):1410–
1424. doi:10.1098/rstb.2010.0311. https://doi.org/10.1098/rstb.2010.0311.
Symonds MRE, Blomberg SP. 2014. A Primer on Phylogenetic Generalised Least Squares BT -
Modern Phylogenetic Comparative Methods and Their Application in Evolutionary Biology:
Concepts and Practice. In: Garamszegi LZ, editor. Modern Phylogenetic Comparative Methods
and Their Application in Evolutionary Biology: Concepts and Practice. Berlin, Heidelberg:
Springer Berlin Heidelberg. p. 105–130. https://doi.org/10.1007/978-3-662-43550-2_5.
Valenti TW, Cherry DS, Currie RJ, Neves RJ, Jones JW, Mair R, Kane CM. 2006. Chlorine
toxicity to early life stages of freshwater mussels (Bivalvia: Unionidae). Environ Toxicol Chem.
25(9):2512–2518. doi:https://doi.org/10.1897/05-527R1.1. https://doi.org/10.1897/05-527R1.1.
Warne MSJ, van Dam R. 2008. NOEC and LOEC Data Should No Longer Be Generated or
Used. Australas J Ecotoxicol. 14(1):1–5. doi:10.3316/informit.665706831945866.
Wheeler JR, Grist EPM, Leung KMY, Morritt D, Crane M. 2002. Species sensitivity
distributions: data and model choice. Mar Pollut Bull. 45(1):192–202. doi:10.1016/S0025-
326X(01)00327-7. https://www.sciencedirect.com/science/article/pii/S0025326X01003277.
Whitehead A, Clark BW, Reid NM, Hahn ME, Nacci D. 2017. When evolution is the solution to
pollution: Key principles, and lessons from rapid repeated adaptation of killifish (Fundulus
heteroclitus) populations. Evol Appl. 10(8):762–783. doi:10.1111/eva.12470.
https://doi.org/10.1111/eva.12470.
Yim JH, Kim KW, Kim SD. 2006. Effect of hardness on acute toxicity of metal mixtures using
Daphnia magna: Prediction of acid mine drainage toxicity. J Hazard Mater. 138(1):16–21.
doi:https://doi.org/10.1016/j.jhazmat.2005.11.107.
https://www.sciencedirect.com/science/article/pii/S0304389406004730.
de Zwart D. 2001. Observed regularities in species sensitivity distributions for aquatic species.
In: Species sensitivity distributions in ecotoxicology. CRC Press. p. 157–178.
Statements and Declarations
Funding
This work was supported by University of Southern California (USC) Sea Grant award
NA18OAR4170075.
Competing Interests
The authors have no relevant financial or non-financial interests to disclose.
79
Author Contributions
A.L.C.: Conceptualization, Methodology, Software, Formal analysis, Investigation, Data
curation, Visualization, Writing – Original Draft, Writing – Review & Editing. S.E.:
Conceptualization, Funding acquisition, Supervision, Writing – Review & Editing.
Data Availability
Data pertaining to this manuscript are available on Zenodo at DOI:
10.5281/zenodo.8222812.
80
Figures and Tables
Figure 3-1: Phylogenetic tree and toxicity data heatmap for the acute chlorine dataset (λ = 0.9).
The colored bar next to each species represents its relative sensitivity to the chemical. A red bar
indicates a high degree of sensitivity (i.e. small amount of chemical causes toxic effect), while a
blue bar indicates low sensitivity (i.e. large amount of chemical causes toxic effect)
81
Figure 3-2: Phylogenetic tree and toxicity data heatmap for the acute guthion dataset (λ = 0.56).
The colored bar next to each species represents its relative sensitivity to the chemical. A red bar
indicates a high degree of sensitivity (i.e. small amount of chemical causes toxic effect), while a
blue bar indicates low sensitivity (i.e. large amount of chemical causes toxic effect)
82
83
Figure 3-3: Phylogenetic tree and toxicity data heatmap for the acute lindane dataset (λ = 1). The
colored bar next to each species represents its relative sensitivity to the chemical. A red bar
indicates a high degree of sensitivity (i.e. small amount of chemical causes toxic effect), while a
blue bar indicates low sensitivity (i.e. large amount of chemical causes toxic effect)
84
85
Figure 3-4: Phylogenetic tree and toxicity data heatmap for the acute pentachlorophenol dataset
(λ = 0.99). The colored bar next to each species represents its relative sensitivity to the chemical.
A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes toxic
effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes toxic
effect)
Figure 3-5: Combined phylogenetic tree and toxicity data heatmap for the chronic cadmium
dataset (λ = 1). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect)
86
Figure 3-6: Combined phylogenetic tree and toxicity data heatmap for the chronic copper dataset
(λ = 1). The colored bar next to each species represents its relative sensitivity to the chemical. A
red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes toxic effect),
while a blue bar indicates low sensitivity (i.e. large amount of chemical causes toxic effect)
87
Figure 3-7: Combined phylogenetic tree and toxicity data heatmap for the chronic diazinon
dataset (λ = 1). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect)
88
Figure 3-8: Combined phylogenetic tree and toxicity data heatmap for the chronic lindane
dataset (λ = 1). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect)
89
Figure 3-9: Combined phylogenetic tree and toxicity data heatmap for the chronic phenol dataset
(λ = 1). The colored bar next to each species represents its relative sensitivity to the chemical. A
red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes toxic effect),
while a blue bar indicates low sensitivity (i.e. large amount of chemical causes toxic effect)
90
Table 3-1: Properties of chemicals evaluated in analyses
Chemical CAS RN Origin Class Mode of Action
Ammonia 7664-41-7 Natural Inorganic Osmoregulatory impairmenta
Cadmium 7440-43-9 Natural Metal Metallic iono/osmoregulatory
impairmenta
Chlorine 7782-50-5 Natural Halogen Osmoregulatory impairmenta
Copper 7440-50-8 Natural Metal Metallic iono/osmoregulatory
impairmenta
Mercury 7439-97-6 Natural Metal Metallic iono/osmoregulatory
impairmenta
Nickel 7440-02-0 Natural Metal Metallic iono/osmoregulatory
impairmenta
Phenol 108-95-2 Natural Phenol Polar narcosisa
Toluene 108-88-3 Natural Aromatic
Hydrocarbon Nonpolar narcosisa
Zinc 7440-66-6 Natural Metal Metallic iono/osmoregulatory
impairmenta
4-Nitrophenol 100-02-7 Synthetic Nitrophenol Polar narcosisa
Atrazine 1912-24-9 Synthetic Triazine Narcosisa
Chlorpyrifos 2921-88-2 Synthetic Organophosphate AChE Inhibitionb
DDT 50-29-3 Synthetic Organochlorine Neurotoxicityb
Diazinon 333-41-5 Synthetic Organophosphate AChE inhibitionb
Dieldrin 60-57-1 Synthetic Organochlorine Neurotoxicityb
Endosulfan 115-29-7 Synthetic Organochlorine Neurotoxicityb
Endrin 72-20-8 Synthetic Organochlorine Neurotoxicityb
Glyphosate 1071-83-6 Synthetic Phosphonate Nonpolar narcosisa
Guthion 86-50-0 Synthetic Organophosphate AChE inhibitionb
Lindane 58-89-9 Synthetic Organochlorine Neurotoxicityb
Malathion 121-75-5 Synthetic Organophosphate AChE inhibitionb
Parathion 56-38-2 Synthetic Organophosphate AChE inhibitionb
PCP 87-86-5 Synthetic Organochlorine Electron transport inhibitionb
TBTO 56-35-9 Synthetic Organotin Electron transport inhibitionb
Abbreviations: CAS RN = Chemical Abstracts Service Registry Number, DDT = 1,1′-(2,2,2-
Trichloroethane-1,1-diyl)bis(4-chlorobenzene), PCP = Pentachlorophenol, TBTO = Tributyltin
oxide, AChE = Acetylcholinesterase
a Generic mode of action
b Precise mode of action
91
Table 3-2: Phylogenetic signal results for all dataset variations
Chemical Exposure
Dataset Variation
Complete Subadult Temperature pH Hardness
λ λ λ λ λ
Ammonia Acute 7.30E-05 7.40E-05 0.95
1.00E06 -
Atrazine Acute 7.30E-05 7.30E-05 1.00E-06
1.00E06 1.00E-06
Atrazine Chronic 7.30E-05 7.30E-05 1.00E-06 0.57 0.92
Cadmium Acute 7.30E-05 0.28 1.00E-06 0.029 -
Cadmium Chronic 1 1 1.00E-06
1.00E06 -
Chlorine Acute 0.9 0.69 1.00E-06
1.00E06 0.99
Chlorpyrifos Acute 7.30E-05 7.30E-05 1.00E-06
1.00E06 -
Chlorpyrifos Chronic 7.30E-05 0.23 0.025 0.029 -
Copper Acute 7.30E-05 0.72 0.27 0.13 -
Copper Chronic 1 7.40E-05 1.00E-06
1.00E06 -
DDT Acute 0.00065 1 1.00E-06
1.00E06 -
Diazinon Acute 0.023 0.025 0.92 0.81 -
Diazinon Chronic 1 1 1.00E-06
1.00E06 -
Dieldrin Acute 7.30E-05 0.0028 1.00E-06
1.00E06 -
Endosulfan Acute 0.03 7.30E-05 1.00E-06 0.077 -
Endosulfan Chronic 0.11 0.18 1.00E-06
1.00E06 -
Endrin Acute 7.30E-05 7.30E-05 0.58 0.59 -
Glyphosate Acute 7.40E-05 0.82 1.00E-06
1.00E06 -
Glyphosate Chronic 7.30E-05 7.30E-05 0.29 0.32 -
Guthion Acute 0.56 0.39 1.00E-06 0.028 0.069
Lindane Acute 1 0.84 0.17 0.99 -
Lindane Chronic 1 7.50E-05 1.00E-06
1.00E06 0.019
Malathion Acute 0.35 0.37 1.00E-06
1.00E06 -
Malathion Chronic 7.30E-05 7.40E-05 0.029 1 -
Mercury Acute 0.16 7.40E-05 1.00E-06 0.27 1
Nickel Acute 7.30E-05 0.16 1.00E-06
1.00E06 -
Nitrophenol Acute 7.40E-05 7.40E-05 1.00E-06 1 1
92
Parathion Acute 0.16 0.5 1.00E-06
1.00E06 -
PCP Acute 0.99 0.58 0.63 0.74 -
PCP Chronic 7.40E-05 7.40E-05 0.1 0.086 -
Phenol Acute 0.33 7.40E-05 1.00E-06
1.00E06 -
Phenol Chronic 1 0.9 1.00E-06
1.00E06 -
TBTO Acute 7.30E-05 7.40E-05 1.00E-06 0.24 -
Toluene Acute 7.40E-05 0.96 1.00E-06
1.00E06 -
Zinc Acute 0.033 7.40E-05 1
1.00E06 -
Zinc Chronic 7.30E-05 0.38 1.00E-06 1 1
Abbreviations: DDT = 1,1′-(2,2,2-Trichloroethane-1,1-diyl)bis(4-chlorobenzene), PCP =
Pentachlorophenol, TBTO = Tributyltin oxide
Values in boldface indicate strong phylogenetic signal (λ > 0.5
93
Table 3-3: Results of PGLS regressions for temperature and pH with toxicity data
Chemical Exposure
Dataset Variation
Temperature pH Hardness
Coef p R2 Coef p R2 Coef p R2
Ammonia Acute 8 0.003 0.19 -53 0.14 0.033 - - -
Atrazine Acute 1.1 0.75 -0.02 1.9 0.82 -0.02 0.03 0.3 0.006
Atrazine Chronic 0.045 0.76 -0.01 -0.05 0.97 -0.02 0.008 0.12 0.04
Cadmium Acute -2.3 0.018 0.075 -1.5 0.95 -0.02 - - -
Cadmium Chronic -0.01 0.47 -0.05 -0.02 0.39 0.006 - - -
Chlorine Acute 0.006 0.63 -0.04 0.52 0.009 0.26 1.8 9.9E-06 0.94
Chlorpyrifos Acute -0.04 0.64 -0.01 0.26 0.83 -0.01 - - -
Chlorpyrifos Chronic 0.034 0.45 -0.01 -0.19 0.55 -0.02 - - -
Copper Acute 0.027 0.87 -0.02 0.44 0.28 0.004 - - -
Copper Chronic 0.012 0.11 0.061 -0.04 0.96 -0.05 - - -
DDT Acute 0.1 0.53 -0.01 0.81 0.76 -0.01 - - -
Diazinon Acute -0.47 0.14 0.014 0.39 0.67 -0.01 - - -
Diazinon Chronic -2.2
1E03 0.47 -4.7 0.55 -0.05 - - -
Dieldrin Acute 0.01 0.17 0.016 -0.07 0.1 0.03 - - -
Endosulfan Acute 0.015 0.69 -0.01 -0.51 0.46 -0.01 - - -
Endosulfan Chronic
1.3E05 0.99 -0.09
9E04 0.95 -0.12 - - -
Endrin Acute 0.009 0.55 -0.01 -0.04 0.66 -0.01 - - -
Glyphosate Acute -9.6 0.84 -0.03 11 0.8 -0.05 - - -
Glyphosate Chronic -0.04 0.85 -0.05 -32 0.15 0.13 - - -
Guthion Acute -0.03 0.47 -0.01 0.81 0.29 0.003 0.004 0.59 -0.07
Lindane Acute 0.55 0.1 0.018 1.3 0.54 -0.01 - - -
Lindane Chronic 0.003 0.74 -0.12 5.6 0.065 0.37 -0.01 0.91 -0.05
Malathion Acute 0.91 0.011 0.04 0.59 0.91 -0.01 - - -
Malathion Chronic -0.21 0.68 -0.04 -2.3 0.097 0.2 - - -
Mercury Acute -0.05 0.49 -0.03 0.13 0.76 -0.08
-1.4E04 0.005 1
Nickel Acute 0.81 0.45 -0.02 -81 0.19 0.067 - - -
Nitrophenol Acute 0.1 0.81 -0.08 -2.7 0.78 -0.1 0.008 0.41 -0.03
Parathion Acute 0.16 0.18 0.017 -0.33 0.21 0.016 - - -
PCP Acute -0.79 0.046 0.034 3.6 0.47 -0.01 - - -
PCP Chronic -0 0.79 -0.05 0.17 0.81 -0.04 - - -
Phenol Acute -9.3 0.014 0.079 -5.9 0.95 -0.02 - - -
Phenol Chronic 0.71 0.28 0.043 -16 0.26 0.13 - - -
TBTO Acute -4 0.39 -0.01 -25 0.78 -0.04 - - -
Toluene Acute 14 0.2 0.034 51 0.87 -0.07 - - -
Zinc Acute 0.52 0.28 0.005 0.002 1 -0.03 - - -
Zinc Chronic
-
0.0035 0.78 -0.1 -0.2 0.012 0.79 0.002 0.16 0.55
94
Abbreviations: A-R
2= Adjusted R2
, DDT = 1,1′-(2,2,2-Trichloroethane-1,1-diyl)bis(4-
chlorobenzene), PCP = Pentachlorophenol, TBTO = Tributyltin oxide
Values in boldface indicate a significant (p ≤ 0.05) positive relationship
Values in boldface italics indicate a significant (p ≤ 0.05) negative relationship
95
Supplementary Materials
Table S3-1: Toxicity dataset sample sizes before and after filtering for availability of
experimental life stage, temperature, pH and hardness information
Chemical Exposure
Dataset Variation
Complete Subadult Temperature pH Hardness
Tox
Values Species Tox
Values Species Tox
Values Species Tox
Values Species Tox
Values Species
Ammonia Acute 95 50 43 19 76 41 74 39 - -
Atrazine Acute 144 60 59 30 124 57 88 44 - -
Atrazine Chronic 742 71 258 41 673 66 492 45 - -
Cadmium Acute 233 80 111 31 183 62 126 45 68 23
Cadmium Chronic 164 14 12 5 150 11 104 5 99 3
Chlorine Acute 116 28 24 6 102 23 100 22 - -
Chlorpyrifos Acute 993 112 129 56 821 100 194 77 - -
Chlorpyrifos Chronic 512 57 172 30 390 41 281 30 - -
Copper Acute 435 100 176 37 288 64 306 52 238 37
Copper Chronic 523 34 185 19 341 28 296 24 189 12
DDT Acute 413 115 78 37 381 101 224 77 - -
Diazinon Acute 274 91 95 47 216 82 144 61 - -
Diazinon Chronic 198 21 51 15 139 18 121 14 - -
Dieldrin Acute 280 69 69 27 247 61 140 58 - -
Endosulfan Acute 489 109 110 47 333 88 248 73 - -
Endosulfan Chronic 56 16 50 12 25 13 21 10 - -
Endrin Acute 227 76 58 32 214 70 152 60 - -
Glyphosate Acute 115 35 41 17 99 30 65 20 - -
Glyphosate Chronic 238 26 130 19 207 22 97 11 - -
Guthion Acute 274 55 54 23 242 53 197 49 - -
Lindane Acute 253 102 56 32 218 94 188 85 - -
Lindane Chronic 33 13 7 6 19 9 18 8 - -
Malathion Acute 422 148 127 68 368 138 294 114 - -
Malathion Chronic 223 21 85 13 201 21 180 11 - -
Mercury Acute 75 32 37 10 57 22 33 13 11 11
Nickel Acute 41 20 16 5 28 18 33 14 9 9
Nitrophenol Acute 52 16 21 6 45 15 28 12 - -
Parathion Acute 184 54 66 27 159 52 108 38 - -
PCP Acute 490 116 172 45 382 89 356 84 - -
PCP Chronic 121 27 72 20 95 21 80 24 - -
Phenol Acute 262 79 78 21 205 64 179 49 - -
Phenol Chronic 47 10 22 7 32 9 26 6 - -
TBTO Acute 103 47 25 15 64 33 52 28 - -
Toluene Acute 85 28 40 12 74 22 61 16 - -
Zinc Acute 149 62 64 20 111 43 86 33 24 24
96
Zinc Chronic 77 14 14 5 64 11 24 6 8 4
Abbreviations: DDT = 1,1′-(2,2,2-Trichloroethane-1,1-diyl)bis(4-chlorobenzene), PCP =
Pentachlorophenol, TBTO = Tributyltin oxide
Table S3-2: Summary of NOEC effects information in chronic toxicity datasets
Chemical Effects Species in
Dataset
Average Species
per Effect
Ammonia 7 9 1.29
Atrazine 20 71 3.55
Cadmium 10 14 1.4
Chlorpyrifos 19 60 3.16
Copper 16 34 2.12
Diazinon 11 21 1.91
Dieldrin 4 9 2.25
Endosulfan 8 16 2
Glyphosate 15 26 1.73
Lindane 12 13 1.08
Malathion 15 21 1.4
PCP 13 27 2.08
Phenol 5 10 2
Zinc 8 14 1.75
97
Figure S3-1: Phylogenetic tree and toxicity data heatmap for the complete acute ammonia
dataset (λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to
the chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical
causes toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical
causes toxic effect).
98
Figure S3-2: Phylogenetic tree and toxicity data heatmap for the complete acute atrazine dataset
(λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect).
99
100
Figure S3-3: Phylogenetic tree and toxicity data heatmap for the complete chronic atrazine
dataset (λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to
the chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical
causes toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical
causes toxic effect).
101
102
Figure S3-4: Phylogenetic tree and toxicity data heatmap for the complete acute cadmium
dataset (λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to
the chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical
causes toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical
causes toxic effect).
103
104
Figure S3-5: Phylogenetic tree and toxicity data heatmap for the complete acute chlorpyrifos
dataset (λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to
the chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical
causes toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical
causes toxic effect).
105
Figure S3-6: Phylogenetic tree and toxicity data heatmap for the complete chronic chlorpyrifos
dataset (λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to
the chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical
causes toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical
causes toxic effect).
106
107
Figure S3-7: Phylogenetic tree and toxicity data heatmap for the complete acute copper dataset
(λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect).
108
109
Figure S3-8: Phylogenetic tree and toxicity data heatmap for the complete acute DDT dataset (λ
= 0.00065). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect). (DDT = 1,1′-(2,2,2-Trichloroethane-1,1-diyl)bis(4-chlorobenzene)
110
111
Figure S3-9: Phylogenetic tree and toxicity data heatmap for the complete acute diazinon dataset
(λ = 0.023). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect).
112
113
Figure S3-10: Phylogenetic tree and toxicity data heatmap for the complete acute dieldrin
dataset (λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to
the chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical
causes toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical
causes toxic effect).
114
115
Figure S3-11: Phylogenetic tree and toxicity data heatmap for the complete acute endosulfan
dataset (λ = 0.03). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect).
Figure S3-12: Phylogenetic tree and toxicity data heatmap for the complete chronic endosulfan
dataset (λ = 0.11). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect).
116
117
Figure S3-13: Phylogenetic tree and toxicity data heatmap for the complete acute endrin dataset
(λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect).
118
Figure S3-14: Phylogenetic tree and toxicity data heatmap for the complete acute glyphosate
dataset (λ = 7.4E-05). The colored bar next to each species represents its relative sensitivity to
the chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical
causes toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical
causes toxic effect).
Figure S3-15: Phylogenetic tree and toxicity data heatmap for the complete chronic glyphosate
dataset (λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to
the chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical
causes toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical
causes toxic effect).
119
120
Figure S3-16: Phylogenetic tree and toxicity data heatmap for the complete acute malathion
dataset (λ = 0.35). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect).
Figure S3-17: Phylogenetic tree and toxicity data heatmap for the complete chronic malathion
dataset (λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to
the chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical
causes toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical
causes toxic effect).
121
Figure S3-18: Phylogenetic tree and toxicity data heatmap for the complete acute mercury
dataset (λ = 0.16). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect).
122
Figure S3-19: Phylogenetic tree and toxicity data heatmap for the complete acute nickel dataset
(λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect).
123
Figure S3-20: Phylogenetic tree and toxicity data heatmap for the complete acute 4-nitrophenol
dataset (λ = 7.4E-05). The colored bar next to each species represents its relative sensitivity to
the chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical
causes toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical
causes toxic effect).
124
Figure S3-21: Phylogenetic tree and toxicity data heatmap for the complete acute parathion
dataset (λ = 0.16). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
125
toxic effect).
Figure S3-22: Phylogenetic tree and toxicity data heatmap for the complete chronic
pentachlorophenol dataset (λ = 7.4E-05). The colored bar next to each species represents its
relative sensitivity to the chemical. A red bar indicates a high degree of sensitivity (i.e. small
amount of chemical causes toxic effect), while a blue bar indicates low sensitivity (i.e. large
amount of chemical causes toxic effect).
126
127
Figure S3-23: Phylogenetic tree and toxicity data heatmap for the complete acute phenol dataset
(λ = 0.33). The colored bar next to each species represents its relative sensitivity to the chemical.
A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes toxic
effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes toxic
effect).
128
Figure S3-24: Phylogenetic tree and toxicity data heatmap for the complete acute TBTO dataset
(λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect). (TBTO = Tributyltin oxide)
129
Figure S3-25: Phylogenetic tree and toxicity data heatmap for the complete acute toluene dataset
(λ = 7.4E-05). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect).
130
131
Figure S3-26: Phylogenetic tree and toxicity data heatmap for the complete acute zinc dataset (λ
= 0.033). The colored bar next to each species represents its relative sensitivity to the chemical.
A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes toxic
effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes toxic
effect).
Figure S3-27: Phylogenetic tree and toxicity data heatmap for the complete chronic zinc dataset
(λ = 7.3E-05). The colored bar next to each species represents its relative sensitivity to the
chemical. A red bar indicates a high degree of sensitivity (i.e. small amount of chemical causes
toxic effect), while a blue bar indicates low sensitivity (i.e. large amount of chemical causes
toxic effect).
132
Chapter 4: Heat exposure improves copper tolerance in the intertidal copepod Tigriopus
californicus
Alice L. Coleman and Suzanne Edmands
Abstract
Cross-protection, the phenomenon of exposure to one stressor conferring increased
resistance to a second stressor, is a physiological process that may help organisms cope with
multiple stressor exposure. Here, we investigated a possible case of cross-protection between
copper and heat in the intertidal copepod Tigriopus californicus, which demonstrates a strong
pattern of local adaptation to the latitudinal thermal gradient in the Northern Hemisphere.
Curiously, copper tolerance in this species also broadly matches the thermal gradient, although
there is little evidence for a corresponding latitudinal distribution of copper pollution in its
habitat. Because both heat and copper induce oxidative stress, there is potential for overlap in the
two response mechanisms that might explain the co-occurring latitudinal patterns and potentially
trigger a cross-protection response during sequential stressor exposure. To test this hypothesis,
we measured the tolerances of allopatric T. californicus populations exposed to each stressor
individually, and again after they received a preliminary exposure from the other stressor. This
work was paired with transcriptomic analyses that compared the gene expression responses to
each acute stressor. We observed unidirectional cross-protection between copper and heat, where
copper tolerance increased when copepods were exposed to heat prior to toxicity testing, but not
when the stressor order was reversed. Gene expression was found to be highly populationspecific, although there was evidence to suggest that the two stressors similarly modulated the
antioxidant defense system, energy-yielding catabolism and other processes that could contribute
to the cross-protection observed here.
133
Introduction
Most organisms experience natural fluctuation in local environmental factors on a regular
or stochastic basis throughout the course of their lives. At times, these fluctuations may cause
conditions to reach a level of intensity such that they compromise an organism’s performance
and fitness, becoming stressors (Schulte 2014). Given that natural stressors rarely act in isolation
and anthropogenic disturbances to the environment are forecasted to intensify with climate
change (Tilman et al. 2017), the ability to cope with multiple stressors is increasingly crucial to
fitness. Thus, improving our understanding of the molecular mechanisms that respond to the
interactive and cumulative effects of multiple stressors is imperative to conservation biology.
Classic theory predicts that exposure to multiple stressors will result in either an additive,
synergistic or antagonistic response (Folt et al. 1999). Stressors are considered additive when the
cumulative effect of a group of stressors equals the sum of the effects of each individual
component. Synergistic stressors amplify each other’s negative effects such that the group
exceeds the predicted additive effect, while antagonistic stressors dampen each other, resulting in
a smaller than expected cumulative response. The identification of synergistic stressors has been
highly emphasized in multiple stressor literature (Côté et al. 2016), but understanding of when all
interaction types occur is useful when designing and implementing conservation actions.
Knowledge of antagonistic stressor combinations is particularly important in this context as
management plans might be ineffective or even detrimental if a set of stressors assumed to be
synergistic are actually antagonistic (Brown et al. 2013). Antagonisms can arise in multiple
ways, such as by one dominant stressor removing the most sensitive individuals from a
population, one stressor mitigating the effects of others, or the sharing of protective mechanisms
or pathways between stressors (Breitburg and Riedel 2005b; Rodgers and Gomez Isaza 2021).
134
Here, we are specifically interested in the third scenario, which can lead to the phenomenon
known as cross-protection.
Cross-protection occurs when exposure to one stressor temporarily confers increased
resistance to a second stressor of a different nature (Rodgers and Gomez Isaza 2023). For
example, Todgham et al. (2005) demonstrated that the survival of tidepool sculpins in severe
osmotic and hypoxic conditions increased when the fish underwent a preliminary heat shock.
This phenomenon is not limited to a select group of taxa or certain stressor combinations, having
been documented in over 50 animal species and for more than 16 heterologous stressors
(Rodgers and Gomez Isaza 2023). Cross-protection arises when stressors share either protective
mechanisms (referred to as ‘cross-tolerance’) or cellular signaling pathways that activate
independent protective mechanisms (referred to as ‘cross-talk’) (Rodgers and Gomez Isaza
2023). Because individuals that exhibit cross-protection may possess fitness advantages during
extreme climactic events or exposure to novel combinations of stressors, it has been
hypothesized to serve as a possible “pre-adaptation” that can buffer species from future threats
(Ramegowda et al. 2020; Rodgers and Gomez Isaza 2023). In this study, we explore the
possibility of cross-protection arising in the marine copepod Tigriopus californicus during
exposure to copper (Cu) and high heat.
T. californicus is a small crustacean that inhabits splash pools in the high intertidal zone
along the Pacific coast of North America from Baja California to Alaska (Ganz and Burton 1995;
Edmands 2001). Substantial genetic divergence occurs between Tigriopus populations at various
spatial scales (Edmands 2001; Willett and Ladner 2009), and there is evidence for local
adaptation to multiple environmental factors by this species. The most prominent example of
local adaptation is for temperature, for which several studies have found greater heat tolerance in
135
populations from lower latitudes (Willett 2010; Kelly et al. 2011; Leong et al. 2018). Similar
latitudinal clines in tolerance have also been identified for salinity, which frequently covaries
with temperature, and more surprisingly, for copper (Sun et al. 2015; Leong et al. 2018; Lee et
al. 2021). Specifically, Sun et al. (2015) found that the maximum temperature at a Tigriopus
population’s collection site was positively correlated with its acute copper tolerance, meaning
that populations from warmer, southern areas are more tolerant of copper than northern
populations. A survey by the California Department of Pesticide Regulations of dissolved copper
concentrations in marinas across the state indicated that copper pollution was negatively
correlated with latitude (Burant et al. 2018); however, marinas are not representative of the high
intertidal splash pools that Tigriopus copepods inhabit. Sun et al. (2015) found that T.
californicus populations tolerate copper concentrations several orders of magnitude higher than
those found in their pools, and therefore hypothesized that the variation in copper tolerance in
this species may be an exaptation derived from related response mechanisms for copper and heat
stress. We posit that the association between copper tolerance and latitude in T. californicus
arises from the mutual activation of the oxidative stress defense system by both copper and heat.
Copper is a ubiquitous pollutant in the marine environment resulting from anthropogenic
sources such as antifouling paints, mining and industrial discharge (Brooks and Waldock 2009).
In living systems, copper plays paradoxical dual roles as both an essential trace metal and potent
toxicant. Essential for routine physiology, copper is utilized as a cofactor by a number of
enzymes that function as redox catalysts and dioxygen carriers, but free copper ions are virtually
nonexistent in cells under normal conditions (Harrison et al. 1999; Tapia et al. 2004). Coppercontaining molecules are instead sequestered in different proteins, with toxicity only occurring
when the intracellular copper concentration overwhelms the chaperone system (Harrison et al.
136
1999; Tapia et al. 2004). When excess copper ions are available, they participate in a Fenton-like
reaction with hydrogen peroxide that produces hydroxyl radicals (•OH), a form of reactive
oxygen species (ROS; Pham et al. 2013). Like copper, ROS perform important physiological
functions at low concentrations, but become hazardous at high concentrations by oxidatively
damaging integral cellular structures (Dröge 2002; Lesser 2006; Kong and Lin 2010). The
concentration of ROS in a cell is a dynamic parameter held in check by a balance between their
production and elimination, and when this dynamic equilibrium is disrupted in favor of ROS
production, a cell is said to be experiencing oxidative stress. A large number of environmental
stressors, including both copper and heat, are known to induce oxidative stress in aquatic
organisms through a variety of mechanisms (Lushchak 2011). Heat generates oxidative stress by
increasing the rates of all chemical reactions in a cell, including respiration. Reactive oxygen
species such as superoxide (02
−•
) and hydrogen peroxide (H2O2) are natural byproducts of
oxidative phosphorylation (Belhadj Slimen et al. 2014), meaning that heat-driven increases in
respiration also increase ROS production.
Aerobic organisms possess a robust defense system to combat the deleterious effects of
ROS. This system is composed of both enzymatic and non-enzymatic molecules that act to
neutralize ROS and prevent or repair damage to proteins, lipids and nucleic acids. Major
antioxidant enzymes include superoxide dismutase (SOD), which catalyzes the breakdown of
superoxide into molecular oxygen (O2) and hydrogen peroxide, as well as catalase (CAT) and
glutathione peroxidase (GPx) which both convert hydrogen peroxide into water (Sies 1993).
Non-enzymatic antioxidants such as the tripeptide glutathione (GSH), carotenoids and αtocopherol (Vitamin E) contribute to oxidative defenses by neutralizing ROS via reduction. Heat
shock proteins (HSPs) are also known to respond to oxidative stress by assisting in the correct
137
folding of polypeptide chains and preventing the aggregation of misfolded proteins affected by
ROS (Ghosh et al. 2018). Both heat and copper significantly modify the expression and activity
of HSPs and antioxidants in Tigriopus (Rhee et al. 2009; Schoville et al. 2012; Kim et al. 2014;
Li et al. 2023), suggesting that there are shared features between the two protective mechanisms
that respond to these heterologous stressors. Such similarities might help explain the latitudinal
pattern of copper tolerance in T. californicus and generate cross-protection if preliminary
exposure to one stressor primes the oxidative defense system ahead of exposure to the second.
Here, we expand upon studies of the latitudinal patterns of local adaptation in T.
californicus and investigate whether this species exhibits cross-protection between copper and
heat. Using seven allopatric populations from California, we performed a series of acute stress
assays to measure the tolerance of each population to both stressors in isolation and after a
priming exposure from the other stressor. We additionally evaluated the survival of the
populations during longer term copper and heat experiments to explore whether the latitudinal
patterns in acute tolerance also manifest during chronic exposure. These in vivo experiments
were paired with transcriptomic sequencing of two populations under each stressor to test for
shared features in their gene expression responses that could contribute to cross-protection.
Methods
Copepod collection and culture maintenance
T. californicus were collected from seven sites in California (Fig. 4-1; Table S4-1)
between November 2022 and February 2024: Bodega Bay (BB), Santa Cruz (SC), Pismo Beach
(PB), Leo Carrillo State Beach (LC), Abalone Cove (AB), Santa Catalina Island (SCI) and La
Jolla (LJ). Sampling sites were concentrated in California because there is greater genetic and
phenotypic variation among T. californicus populations found south of Oregon than those in
more northern areas (Edmands 2001). Animals were transported from the field to the University
138
of Southern California (USC) and transferred into 1 L beakers containing 37 μm triple-filtered
natural seawater sourced from the Wrigley Marine Science Center on Santa Catalina Island.
Cultures were periodically fed a mixture of TetraMin fish food (Tetra Holding Inc., USA) and
the blue-green algae Spirulina (Nutraceutical Science Institute, USA) at a concentration of 0.1 g
of each food per liter seawater. Cultures were incubated at 20°C with a 12-hour light:dark cycle
and allowed to acclimate to laboratory conditions for a minimum of four weeks before testing.
Acute copper toxicity tests
Acute copper tolerance was measured as median lethal concentration (LC50), which
represents the concentration of a chemical required to kill half of a test population. In these tests,
twelve adult males were individually distributed into wells of polystyrene 24-well plates (BD
Falcon) containing 2 mL of Cu solution. Animals were checked every 24 hours for mortality
until a total of 96 hours lapsed, with mortality assigned when individuals did not respond to
gentle prodding from a probe. No feeding or solution renewal took place during these assays.
Male T. californicus were utilized in all acute experiments because they are known to be less
stress tolerant than females (Foley et al. 2019). Cu solutions were made with solid CuSO4•5H20
(Sigma) dissolved in 37 μm triple-filtered autoclaved seawater (FASW) in twelve concentrations:
0, 5, 10, 20, 40, 60, 80, 100, 120, 180, 250 and 500 mg/L Cu. All glassware and plastic items
utilized in toxicity tests were acid-washed overnight in 0.1M HCl between uses.
We also performed a “heat shock” variation of this procedure to check for crossprotection in which copepods were first exposed to a period of high heat and then transferred into
a copper toxicity test. In these experiments, males were individually distributed into the wells of
24 well plates that contained 2 mL of triple-filtered FASW. Plates were incubated at 35°C for
one hour, followed by a one-hour recovery period at 20°C. At the conclusion of the recovery
139
period, animals were then transferred into 24-well plates for a 96-hour LC50 test that was
conducted as described above. The same range of copper concentrations was used in both
variations of the acute copper toxicity tests.
Acute heat stress tests
Acute heat tolerance was measured as median lethal temperature (LT50), which
represents the temperature that kills half of a test population. In these experiments, adult males
were individually distributed into the wells of a polyvinyl chloride 96-well plate (BD Falcon)
containing 50 μL of triple-filtered FASW. Heat stress was initiated by a PTC-200 thermal cycler
(MJ Research) which was programmed to ramp from 20°C to the target temperature at a rate of
0.1°C per minute. Plates were then held at the target temperature for one hour, followed by a
one-hour recovery period at 20°C. Assays were repeated in increments of 1°C with fresh animals
until complete mortality occurred. Experimental stress temperatures ranged between 35 and
42°C.
We also performed a “copper shock” variation of this procedure to check for cross
protection in which copepods were exposed to copper prior to an acute heat stress test. To do
this, males were distributed into a 96-well plate containing 50 μL of 60 mg/L Cu solution and
then placed in a 20°C incubator for one hour. Animals were then transferred into a new 96 well
plate containing 50 μL of clean FASW and allowed to recover for 1 hour at 20°C before being
placed into the thermocycler for a heat test conducted as described above. The same range of
temperatures was used in both variations of the acute heat stress experiments.
Chronic stress tests
Chronic tolerance of copper and heat was estimated by tracking the survival of mixed sex
copepods that hatched and were raised in stressful conditions over a 21-day period. Five different
140
treatments were used in these experiments: control (incubation at 20°C, standard FASW), heat
(30°C, standard FASW), low copper (20°C, 3 mg/L Cu in FASW), moderate copper (20°C, 5
mg/L Cu in FASW) and high copper (20°C, 8 mg/L Cu in FASW). Pilot testing indicated that a
combined heat/copper treatment was inviable. The Pismo Beach and Leo Carrillo populations
were not included in these experiments because they were collected (January – February 2024)
after the conclusion of chronic testing (November 2023).
On Day 1 of experiments, gravid females from each population were individually
transferred into acid-washed 25 mL Petri dishes containing 8 mL of treatment media mixed with
0.1 g/L TetraMin and Spirulina and moved into the appropriate temperature incubator. On Day 4,
mothers were removed and all live offspring were counted by manually transferring animals into
new Petri dishes containing fresh media with a glass 5 mL Pasteur pipette (VWR). Five
additional censuses were completed over the following 17 days until Day 21. T. californicus
copepods have a generation time of 3-4 weeks at 20°C (Edmands & Harrison 2003), so 21-day
experiments captured the majority of this species’ reproductive cycle.
RNA isolation, library preparation and sequencing
RNA-sequencing was performed from pools of adult male copepods from Bodega Bay
and La Jolla that were exposed to either 60 mg/L Cu or 35°C for 24 hours. Pool size was 24
animals for the control and heat treatments and 36 animals for the copper treatment. Immediately
after the conclusion of the exposure period animals were transferred into ZR BashingBead Lysis
Tubes (Zymo Research) containing TRI reagent (ThermoFisher Scientific) and homogenized
using a TissueLyser II (Qiagen). RNA was then extracted with a Direct-zol RNA Miniprep kit
(Zymo Research) according to the manufacturer’s instructions. The RNA concentration in each
sample was quantified using an Invitrogen Qubit 4 Fluorometer (ThermoFisher Scientifc) and a
141
Qubit RNA High Sensitivity Assay Kit (ThermoFisher Scientific). Libraries were prepared using
a NEB Next Ultra II Directional Library Prep Kit (New England Biolabs) and sequenced in 150
bp paired-end reads on an Illumina platform at Admera Health (New Jersey, USA).
Stress data analyses
Population LC50s and LT50s were calculated in R with a generalized linear model using
a binomial distribution and the dose.p() function from the package MASS (Venables and Ripley
2002; R Core Team 2022). Significant differences between acute tolerance values were
determined using the ratio_test() function obtained from the package ecotox (Wheeler et al.
2006). Linear regression analyses were performed to evaluate the relationships of LC50 and
LT50 data with latitude, as well as to compare LC50s and LT50s with each other.
Results from the chronic stress tests were assessed via survival analysis. Kaplan-Meier
survival curves were fit to the census data recorded between Days 4 and 21 using the function
survfit2() from the package survival (Therneau 2023). Pairwise comparisons using log-rank tests
of survival curves were performed using pairwise_survdiff() from the package survminer
(Kassambara et al. 2021). Acute and chronic tolerances were compared using linear regressions
of the standard LT50 and LC50 data with the adjusted survival of each population in the heat and
low copper treatments, respectively (Fig. 4-6). Adjusted survival was calculated as the difference
between a population’s survival probability at the final census in the control and each of the heat
and low copper treatments.
RNA-seq data analysis
Raw read quality was evaluated with FastQC (Andrews 2010). Reads were then trimmed
with Trimmomatic (Bolger et al. 2014) using the default parameters and evaluated again with
FastQC. Because the trimming did not appear to improve read quality, we proceeded to map our
142
raw reads to the T. californicus reference genome for the San Diego population (Barreto et al.
2018; accession number GCA_007210705.1) using HISAT2 (Putri et al. 2022). Read counts
after alignment were obtained using HT-Seq (Putri et al. 2022) and then imported into R for
differential expression analysis with DESeq2 (Love et al. 2014). Differential expression analyses
of these data were conducted using the following design: ~population + treatment +
population:treatment. Significance testing was performed using a Wald test and multiple test
corrections were performed following Benjamini and Hochberg (1995). Adjusted p-values that
were less than 0.1 were treated as significant. Specific contrasts were used to evaluate the effects
of population on treatment response as well as to compare the responses of the populations to
each treatment, which was created by using a grouping variable for population and treatment
(BB-control, BB-heat, etc.). Principal component analysis (PCA) of the top 500 genes with the
highest variance after an r-log transformation in each sample was performed using plotPCA()
from DESeq2 to visualize the relationships between all twelve samples as well as when they
were grouped by population. Names of differentially expressed genes (DEGs) were extracted
from the T. californicus organism database found in the package AnnotationHub (Morgan and
Sheperd 2024). Following Li et al. (2020), we searched our DEG lists for antioxidant genes, heat
shock proteins and genes with “stress” or “stress-induced” in their name to determine which
components of the oxidative defense system responded to copper and heat exposure. We
additionally searched for genes related to chitin metabolism and cuticle formation which are
known to respond to heat stress in T. californicus (Schoville et al. 2012; Harada and Burton
2020).
Rank-based enrichment analysis of biological process (BP) gene ontology (GO) terms
was performed using the GO_MWU package (Wright et al. 2015). In this analysis, GO terms
143
with fewer than ten associated genes are filtered out, and the remaining terms are evaluated for
significant enrichment by either up or downregulated genes using the Mann-Whitney U test.
Redundant and highly similar GO terms are then merged according to complete linkage
clustering based on the fraction of shared genes, generating a hierarchical clustering of GO
categories based on the number of genes shared between them.
Results
Acute tolerance
Acute heat tolerance in the seven populations decreased significantly with increasing
latitude (Fig. 4-2A; R
2 = 0.89, p = 8.4E-04), consistent with the previously established latitudinal
gradient in heat tolerance in this species (Willett 2010; Kelly et al. 2011; Leong et al. 2018).
Acute copper tolerance also decreased with increasing latitude, although the linear model we fit
to the data did not meet the significance threshold of p < 0.05 (Fig. 4-2B; R
2 = 0.46, p = 0.056).
Linear regression of the LT50s with the LC50 data indicated that the acute tolerances of both
stressors are positively correlated (Fig 4-2C; R
2 = 0.57, p = 0.031). In the sequential stressor
experiments, thermal tolerance in all populations significantly decreased when they were
exposed to copper prior to thermal testing (Fig 4-3). The opposite effect was observed in the
reverse exposure experiments, in which acute copper tolerance significantly increased in all
populations when the copepods experienced a heat shock prior to copper toxicity testing (Fig.4-
4).
Chronic tolerance
We performed ten replicates of the chronic stress protocol that utilized 20 BB, 16 SC, 18
AB, 18 SCI and 16 LJ females. These females produced a total of 3,121 offspring across all
trials, with the offspring split between populations as follows: 693 BB, 799 SC, 400 AB, 580 SCI
and 649 LJ. In general, all four stress treatments had fewer surviving larvae than the control at
144
the first census on Day 4 (Fig. S4-1). Bodega Bay and Abalone Cove populations had the highest
survival probabilities in the control treatment, followed by Santa Catalina Island, Santa Cruz and
La Jolla (Fig. 4-5A). La Jolla and Santa Catalina Island, the southernmost populations,
demonstrated the highest survival in the heat treatment, while Bodega Bay and Santa Cruz, the
northernmost populations had the lowest survival probabilities (Fig. 4-5B). In contrast, Bodega
Bay had the highest survival in the low copper treatment and tied with Santa Catalina Island for
highest in the moderate copper scenario (Fig. 4-5C, 4-5D). There were no significant differences
in survival between the five populations in the high copper treatment (Fig. 4-5E). When the
survival data were grouped by population, we found that each population exhibited their highest
survival probability in the control treatment (Fig. S4-2), with the exception of La Jolla, which
had equivalent survival in the control and heat treatments. Survival in all populations was most
affected by the high copper treatment (Fig. 4-6, S4-2).
Differential gene expression
RNA-seq was conducted for a total of twelve samples (2 replicates * (2 treatments + 1
control) * 2 populations) of pooled male copepods, which yielded approximately 334.9 million
150 bp reads. On average, 51.91% of Bodega Bay reads and 71.17% of La Jolla reads aligned to
the San Diego reference genome (Table S4-2). Principal components analysis indicated that
population was largely associated with PC1, which explained 89% of the gene expression
variation between all twelve samples, while treatment primarily sorted on PC2, which explained
just 4% of the variance (Fig. 4-7A). La Jolla’s gene expression in the heat treatment appeared to
be more similar to the pattern seen in the control than in the copper treatment (Fig. 4-7C), while
there was more distinct separation between three groups for Bodega Bay (Fig. 4-7B). Relative to
its controls, Bodega Bay differentially expressed 3,114 genes in the heat treatment and 1,333
145
during copper exposure, with 772 DEGs shared by both stressors in this population (Fig. 4-8). La
Jolla differentially expressed more genes (1,825) during copper exposure than heat (1, 782), and
there were 642 DEGs shared in the two treatments (Fig. 4-8). There were 137 DEGs shared
between both populations in both treatments (Fig. 4-8; Table S4-3).
The antioxidant, heat shock protein, stress-induced, chitin and cuticle-associated genes
that were differentially expressed in response to each treatment by both populations are listed in
Table 1. Antioxidant genes were primarily upregulated in both treatments, while heat shock
proteins and stress-induced genes were predominantly downregulated. Genes related to the
cuticle and chitin metabolism were roughly equally distributed between up and down-regulation.
The Bodega Bay heat treatment had the most DEGs (53) in these categories of all four groups,
while the Bodega Bay copper treatment had only 29 DEGs. In contrast, the La Jolla copper
treatment (47) had more responsive DEGs in these categories in the copper treatment than in heat
(35). Of the 137 genes that were differentially expressed by both populations in both treatments,
none were related to chitin or the cuticle, one was an antioxidant, five were heat shock proteins
and one was a stress-induced gene and all genes in these categories were downregulated (Table
S4-3).
Gene Ontology enrichment
During heat exposure, Bodega Bay animals largely upregulated biological processes
related to amino acid metabolism and catabolism, as well as the transport of metals and ions (Fig.
4-9A). Copper exposure also induced upregulation of categories related to the metabolism of
amino acids, carbohydrates and lipids as well as ion transport (Fig 4-9B). GO categories
downregulated by Bodega Bay during both treatments included processing of mRNA, protein
folding and organization as well as cellular responses to stimuli (Fig. 4-9). La Jolla upregulated
146
the metabolism and catabolism of amino acids and macromolecules in both treatments, while
downregulating categories centered on the processing of nucleic acids and regulation of the cell
cycle (Fig. 4-10).
Discussion
In the present study, we provide further support for the previously established latitudinal
trends in acute thermal and copper tolerance in the intertidal copepod T. californicus, although
the evidence for patterns in chronic data varies by stressor. Cross-protection was evident when
copepods were exposed first exposed to heat and then copper, but not in the reverse scenario
which points to the importance of exposure sequence and intensity of the initial stressor in
determining whether a cross-protective benefit will be conferred. Gene expression responses to
stress were highly population-specific, although there was some evidence to suggest that both
copper and heat similarly modulated the antioxidant defense system and other biological
processes in ways that could generate cross-protection.
Latitudinal trends in acute thermal and copper tolerance
The results of our acute single stressor experiments are consistent with the previously
established latitudinal cline in thermal tolerance among T. californicus populations and provide
support for the existence of a concurrent latitudinal cline in copper tolerance. For both copper
and heat, tolerance increases with decreasing latitude, although this relationship was only
significant in the heat data. However, the significant positive association we identified between
LT50 and LC50 values (Fig. 4-2C) suggests that the variation in copper tolerance among
populations can be explained in part by latitude. While the relationship between temperature,
latitude and thermal tolerance is well-established in ectotherms (Sunday et al. 2010), the cause of
the corresponding latitudinal pattern in copper tolerance in T. californicus is less certain.
147
Local adaptation to pollution has been identified in populations inhabiting contaminated
sites for a wide range of species (Brown 1978; Klerks and Levinton 1989; Nacci et al. 2010).
In the case of T. californicus, however, we have very little data on the distribution of dissolved
copper concentrations across its broad habitat range, although Sun et al. (2015) notably found no
significant difference in copper concentrations between pools in Santa Cruz and San Diego. This
work also showed that populations can cope with copper concentrations that far exceed those
found in their pools and the adjacent surface waters (Sun et al. 2015). Together, these findings
and the lack of specific data on environmental pollution levels make it unclear whether T.
californicus populations are experiencing selective pressure from copper. Thus, our results
support the hypothesis that the latitudinal trend in copper tolerance is not a product of natural
selection for improved copper tolerance.
Mismatches between acute and chronic tolerance
Given the support from the results of the acute experiments for the association between
latitude and both acute copper and heat tolerance, we expected to uncover similar trends in the
chronic data. Instead, we found that a population’s survival in chronic stress conditions was not
significantly correlated with its acute tolerance of either stressor (Fig. 4-6), which is indicative of
major mismatches between tolerance types. There were some similarities between acute and
chronic results in the heat experiment, as La Jolla, the southernmost population, had the highest
survival out of the five populations, while Santa Cruz, a northern population, had the lowest (Fig.
4-5B). However, Bodega Bay demonstrated the highest survival out of all populations in both the
low copper treatment and tied for highest in moderate copper (Figs. 4-5C, 4-5D), which
represents a major deviation from the latitudinal pattern in the acute data.
148
Inconsistencies between acute and chronic stress tolerance have been identified
previously. For example, Sgrò et al. (2010) found that Drosophila melanogaster populations
from cold regions exhibited higher tolerance during an acute heat exposure than their warm
region counterparts, but that their advantage collapsed during longer experiments.
Transcriptomic and epigenetic studies have shown that acute and chronic length exposures
induce distinct molecular responses in other organisms (Kovalchuk et al. 2007; Uren Webster et
al. 2018; Kirsten et al. 2021). Although many of the molecular mechanisms that make up the
cellular stress response are highly conserved across the diversity of life (Kültz 2020), the specific
components of this system that respond during a stress event depends both on the characteristics
of the organism experiencing the stress and the exposure event itself. We hypothesize that
distinct physiological demands and response mechanisms for acute and chronic length copper
exposures at least partially underly the tolerance mismatches observed here.
In addition, differential population fitness during long-term culturing in the laboratory
could also contribute to the deviation from the latitudinal pattern in the chronic data. For
example, we anecdotally observed that the Bodega Bay culture was composed of larger animals
and consistently remained densely populated over time. In contrast, the copepods found in the La
Jolla culture were smaller and had to be supplemented with new animals from the field on one
occasion. Larger body size is associated with greater fitness in ectotherms (Kingsolver and Huey
2008), so these size differences could mean that Bodega Bay animals were more fit in our
standard laboratory conditions than La Jolla. However, given that heat tolerance scales
allometrically (Peralta-Maraver and Rezende 2021), wherein small animals are more tolerant
than large ones, smaller body size could explain La Jolla’s superior performance in both the
acute and chronic heat experiments. While chemical sensitivity also exhibits a degree of size-
149
dependency in marine invertebrates (Kang et al. 2019), it is also heavily influenced by an
network of other important physiological factors like pollutant uptake, metabolism and excretion
that could collectively overshadow the effects of body size.
Unidirectional cross-protection between heat and copper
The results of our sequential stressor experiments indicate that the order of exposure and
stressor intensity both contribute to the likelihood of a cross-protection outcome. The heat shock
experiments induced a significant increase in copper tolerance across all populations, which is
indicative of a cross-protection benefit. Studies in other organisms have shown that a heat shock
can increase tolerance of a number of other stressors aside from copper (Neumann et al. 1994;
Todgham et al. 2005; Bilyk et al. 2012), suggesting that heat primes the cellular stress response
to respond to subsequent stressors. Cross-protection is hypothesized to be driven in part by
transcriptional frontloading (Collins et al. 2020), which consists of changes to constitutive gene
expression that enable organisms to maintain physiological tolerance mechanisms that respond to
common environmental stress events (Barshis et al. 2013). Frontloading is associated with
reduced transcriptional plasticity during stress exposure, owed to the higher baseline expression
of genes that respond to the stressor (Barshis et al. 2013). In a sequential exposure scenario,
acclimation to an initial stressor can similarly modify gene expression such that the secondary
stressor requires less of an inducible response, thereby conserving energy (Collins et al. 2020).
In contrast to the heat shock experiments, we found that exposing the copepods to copper
prior to thermal testing resulted sharply reduced heat tolerance in all populations (Fig 4-3). This
negative response is consistent with cross-susceptibility (Todgham and Stillman 2013), the
phenomenon in which exposure to one stressor increases an organism’s vulnerability to a
subsequent stressor, rather than cross-protection. Cross-susceptibility occurs when the
150
physiological and energetic demands imposed by the initial stressor limit an organism’s ability to
mount a response to the secondary stressor, typically by either damaging physiological systems
or depleting energy reserves (Rodgers and Gomez Isaza 2022). The likelihood of crosssusceptibility occurring depends in part on the magnitude and duration of the priming stressor,
which if too severe is more likely to negatively affect an organism’s future tolerance. We
hypothesize that an overly-intense priming stressor is the cause of the cross-susceptibility result
in our copper shock experiments.
While we did not observe any mortalities among the animals exposed to the copper shock
(1 hour at 60 mg/L Cu), data from the single stressor experiments suggests that the copper shock
concentration was a more severe stress event than the heat shock (1 hour at 35°C). None of the
populations recorded any mortalities after a thermal test at 35°C during the standard heat stress
experiments, whereas after 24 hours in 60 mg/L Cu, populations experienced a 27% mean
decrease in sample size (Table S4-3). It is possible that a cross-protection effect could manifest if
the shock concentration used were lower, or if we had given the copepods a longer recovery
period prior to heat testing. Alternatively, the cross-susceptibility induced by this sequence of
stressors may be indicative that there is less robust overlap between the two response
mechanisms than we initially hypothesized.
Population predominantly determines gene expression
The results of the differential expression analyses of the twelve RNA-seq samples (Fig.
4-8A; Fig. S4-3) strongly indicate that population, rather than treatment, is the most important
factor in determining expression in all of the groups we tested. Previous work with T.
californicus has identified substantial genetic divergence between natural populations (Burton
1998; Rawson and Burton 2006; Willett and Ladner 2009), with greater differentiation occurring
151
between southern populations (south of Oregon to Baja California) than northern (Oregon to
Alaska) (Edmands 2001). According to this classification system, Bodega Bay and La Jolla are
both southern populations, meaning that significant genetic differences between them are
expected.
Notably, Bodega Bay differentially expressed substantially more genes (3,114) in the
heat treatment than La Jolla (1,782). This finding is consistent with previous comparative
transcriptomics work in Tigriopus (Li et al. 2020) and other systems (Barshis et al. 2013;
Gleason and Burton 2015; Chen et al. 2019) that identified greater induction of stress DEGs in
the less tolerant group, suggesting that the more tolerant type either possesses a more efficient
response mechanism or experiences a lower level of stress. However, this was not the case in the
copper treatment, in which La Jolla, the more tolerant population (Fig. 4-2B), differentially
expressed more genes (1,825) than Bodega Bay (1,333). The duration of copper exposure for
copepods utilized for RNA-seq was shorter than that of the acute toxicity tests, meaning that the
acute-chronic tolerance mismatches in the populations most likely are not contributing to the
pattern reversal between treatments.
Similar modulation of oxidative defenses and biological processes by both stressors
Although both populations exhibited distinct transcriptional profiles under copper and
heat (Fig. 4-8), there were broad similarities between the two stress responses that could
contribute to cross-protection. For instance, copper and heat both altered the expression of
antioxidant genes (Table 4-1), which are crucial to protecting against oxidative stress. Genes
related to the glutathione s-transferase (GST) enzyme family, which catalyze the conjugation of
the reduced form of glutathione with xenobiotic and hydrophilic molecules (Hayes and Strange
1995), were particularly active (Table 4-1), pointing to major involvement of the glutathione
152
system in the responses to both stressors. High glutathione activity confers protection from
oxidative damage (Kidd 1997), so the priming of glutathione metabolism by heat shock could
provide a major boost to defenses during secondary copper exposure. Both stressors also caused
the copepods to downregulate cell cycle processes and transcription, which are responses that
protect DNA from additional damage (Fuse et al. 1996; Shackelford et al. 2000; Niskanen et al.
2015), and upregulate the energy-yielding catabolism of amino acids and large macromolecules
(Fig 4-9, 4-10). Early initiation of these processes may help generate cross-protection by
increasing energetic resources for use during secondary stressor exposure and reducing the need
to spend more of those resources on controlling on the cell cycle. The upregulation of ion
transport during heat stress (Figs. 4-9A, 14-0A), could also play a role in generating the crossprotection we observed in the heat shock experiments by priming cells to more efficiently
translocate excess copper ions during toxicity. In addition, genes associated with cuticle
formation responded to both stressors, however, the possible contributions of these genes to
cross-protection are uncertain because the specific function of the cuticle during heat stress has
not been determined (Schoville et al. 2012; Harada and Burton 2020).
Surprisingly, we also found that expression of heat shock proteins and protein folding
processes were largely downregulated (Table 4-1; Fig. 4-9, Fig. 4-10) by both copper heat. In
contrast, most other Tigriopus studies identified upregulation of the HSP families during
chemical or oxidative stress exposure (Kim et al. 2014; Li et al. 2020; Li et al. 2023). Given that
we extracted RNA at a relatively late time point compared to other studies, it is possible that our
sampling occurred after the peak of HSP production, and that our results are instead
representative of reestablished HSP homeostasis. If a very rapid increase in HSPs does occur
during stress exposure, it could be a major contributor to cross-protection by reducing the need
153
for additional protein synthesis during exposure to the secondary stressor and allowing resources
to be diverted elsewhere.
Acknowledgements
This work was supported by a Student Research Grant from the Southern California
Academy of Sciences and a Graduate Student Research Award from the Southern California
regional chapter of the Society of Environmental Toxicology and Chemistry awarded to A.L.C.
The authors would also like to thank the UC Davis Bodega Marine Laboratory and the Wrigley
Marine Science Center for providing us with field site access in addition to C. Kenkel and K.
Schoenberger for their contributions to the conceptualization and execution of the RNA-seq
component of this work.
References
Andrews S. 2010. FastQC: A Quality Control Tool for High Throughput Sequence Data
[Online]. http://www.bioinformatics.babraham.ac.uk/projects/fastqc/.
Barreto FS, Watson ET, Lima TG, Willett CS, Edmands S, Li W, Burton RS. 2018. Genomic
signatures of mitonuclear coevolution across populations of Tigriopus californicus. Nat Ecol
Evol. 2(8):1250–1257. doi:10.1038/s41559-018-0588-1. https://doi.org/10.1038/s41559-018-
0588-1.
Barshis DJ, Ladner JT, Oliver TA, Seneca FO, Traylor-Knowles N, Palumbi SR. 2013. Genomic
basis for coral resilience to climate change. PNAS. 110(4):1387–1392.
Belhadj Slimen I, Najar T, Ghram A, Dabbebi H, Ben Mrad M, Abdrabbah M. 2014. Reactive
oxygen species, heat stress and oxidative-induced mitochondrial damage. A review. Int J
Hyperth. 30(7):513–523. doi:10.3109/02656736.2014.971446.
https://doi.org/10.3109/02656736.2014.971446.
Benjamini Y, Hochberg Y. 1995. Controlling the False Discovery Rate: A Practical and
Powerful Approach to Multiple Testing. J R Stat Soc Ser B. 57(1):289–300.
doi:https://doi.org/10.1111/j.2517-6161.1995.tb02031.x. https://doi.org/10.1111/j.2517-
6161.1995.tb02031.x.
Bilyk KT, Evans CW, DeVries AL. 2012. Heat hardening in Antarctic notothenioid fishes. Polar
Biol. 35(9):1447–1451. doi:10.1007/s00300-012-1189-0. https://doi.org/10.1007/s00300-012-
1189-0.
Bolger AM, Lohse M, Usadel B. 2014. Trimmomatic: a flexible trimmer for Illumina sequence
data. Bioinformatics. 30(15):2114–2120. doi:10.1093/bioinformatics/btu170.
154
https://doi.org/10.1093/bioinformatics/btu170.
Breitburg DL, Riedel GF. 2005. Multiple stressors in marine systems. In: Marine conservation
biology: the science of maintaining the sea’s biodiversity. p. 167–182.
Brooks SJ, Waldock M. 2009. Copper Biocides in the Marine Environment. In: Arai T, Harino
H, Ohji M, Langston WJ, editors. Ecotoxicology of Antifouling Biocides. Tokyo: Springer
Japan. p. 413–428. https://doi.org/10.1007/978-4-431-85709-9_24.
Brown BE. 1978. Lead detoxification by a copper-tolerant isopod. Nature. 276(5686):388–390.
doi:10.1038/276388a0. https://doi.org/10.1038/276388a0.
Brown CJ, Saunders MI, Possingham HP, Richardson AJ. 2013. Managing for Interactions
between Local and Global Stressors of Ecosystems. PLoS One. 8(6):e65765.
https://doi.org/10.1371/journal.pone.0065765.
Burant A, Zhang X, Singhasemanon N, Teerlink J. 2018. Study 319 Report: Monitoring of
Dissolved Copper in California Coastal Waterbodies. Sacramento.
Burton RS. 1998. INTRASPECIFIC PHYLOGEOGRAPHY ACROSS THE POINT
CONCEPTION BIOGEOGRAPHIC BOUNDARY. Evolution (N Y). 52(3):734–745.
doi:10.1111/j.1558-5646.1998.tb03698.x. https://doi.org/10.1111/j.1558-5646.1998.tb03698.x.
Chen N, Huang Z, Lu C, Shen Y, Luo X, Ke C, You W. 2019. Different Transcriptomic
Responses to Thermal Stress in Heat-Tolerant and Heat-Sensitive Pacific Abalones Indicated by
Cardiac Performance. Front Physiol. 9.
https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2018.01895.
Collins M, Clark MS, Spicer JI, Truebano M. 2020. Transcriptional frontloading contributes to
cross-tolerance between stressors. Evol Appl. 14(2):577–587. doi:10.1111/eva.13142.
https://doi.org/10.1111/eva.13142.
Côté IM, Darling ES, Brown CJ. 2016. Interactions among ecosystem stressors and their
importance in conservation. Proc R Soc B Biol Sci. 283(1824):20152592.
doi:10.1098/rspb.2015.2592. https://doi.org/10.1098/rspb.2015.2592.
Dröge W. 2002. Free radicals in the physiological control of cell function. Physiol Rev.
82(1):47–95. doi:10.1152/physrev.00018.2001.
Edmands S. 2001. Phylogeography of the intertidal copepod Tigriopus californicus reveals
substantially reduced population differentiation at northern latitudes. Mol Ecol. 10(7):1743–
1750. doi:https://doi.org/10.1046/j.0962-1083.2001.01306.x. https://doi.org/10.1046/j.0962-
1083.2001.01306.x.
Foley HB, Sun PY, Ramirez R, So BK, Venkataraman YR, Nixon EN, Davies KJA, Edmands S.
2019. Sex-specific stress tolerance, proteolysis, and lifespan in the invertebrate Tigriopus
californicus. Exp Gerontol. 119:146–156. doi:10.1016/j.exger.2019.02.006.
https://www.sciencedirect.com/science/article/pii/S0531556518305734.
Folt CL, Chen CY, Moore M V, Burnaford J. 1999. Synergism and antagonism among multiple
stressors. Limnol Oceanogr. 44(3part2):864–877. doi:10.4319/lo.1999.44.3_part_2.0864.
https://doi.org/10.4319/lo.1999.44.3_part_2.0864.
155
Fuse T, Yamada K, Asai K, Kato T, Nakanishi M. 1996. Heat Shock-Mediated Cell Cycle Arrest
Is Accompanied by Induction of p21 CKI. Biochem Biophys Res Commun. 225(3):759–763.
doi:https://doi.org/10.1006/bbrc.1996.1247.
https://www.sciencedirect.com/science/article/pii/S0006291X96912474.
Ganz HH, Burton RS. 1995. Genetic differentiation and reproductive incompatibility among
Baja California populations of the copepod Tigriopus californicus. Mar Biol. 123(4):821–827.
doi:10.1007/BF00349126. https://doi.org/10.1007/BF00349126.
Ghosh S, Sarkar P, Basak P, Mahalanobish S, Sil PC. 2018. Role of Heat Shock Proteins in
Oxidative Stress and Stress Tolerance BT - Heat Shock Proteins and Stress. In: Asea AAA,
Kaur P, editors. Cham: Springer International Publishing. p. 109–126.
https://doi.org/10.1007/978-3-319-90725-3_6.
Gleason LU, Burton RS. 2015. RNA-seq reveals regional differences in transcriptome response
to heat stress in the marine snail Chlorostoma funebralis. Mol Ecol. 24(3):610–627.
doi:https://doi.org/10.1111/mec.13047. https://doi.org/10.1111/mec.13047.
Harada AE, Burton RS. 2020. Consequences of HSF knockdown on gene expression during the
heat shock response in Tigriopus californicus. J Exp Biol. 223(3):jeb208611.
doi:10.1242/jeb.208611. https://doi.org/10.1242/jeb.208611.
Harrison MD, Jones CE, Dameron CT. 1999. Copper chaperones: function, structure and copperbinding properties. JBIC J Biol Inorg Chem. 4(2):145–153. doi:10.1007/s007750050297.
https://doi.org/10.1007/s007750050297.
Hayes JD, Strange RC. 1995. Invited Commentary Potential Contribution of the Glutathione STransferase Supergene Family to Resistance to Oxidative Stress. Free Radic Res. 22(3):193–207.
doi:10.3109/10715769509147539. https://doi.org/10.3109/10715769509147539.
Kang H-M, Lee Jin-Sol, Lee YH, Kim M-S, Park HG, Jeong C-B, Lee Jae-Seong. 2019. Body
size-dependent interspecific tolerance to cadmium and their molecular responses in the marine
rotifer Brachionus spp. Aquat Toxicol. 206:195–202. doi:10.1016/j.aquatox.2018.10.020.
https://www.sciencedirect.com/science/article/pii/S0166445X18308609.
Kelly MW, Sanford E, Grosberg RK. 2011. Limited potential for adaptation to climate change in
a broadly distributed marine crustacean. Proc R Soc B Biol Sci. 279(1727):349–356.
doi:10.1098/rspb.2011.0542. https://doi.org/10.1098/rspb.2011.0542.
Kidd PM. 1997. Glutathione: Systemic Protectant Aginst Oxidative and Free Radical Damage.
Altern Med Rev. 2(3):155–176.
Kim B-M, Rhee J-S, Jeong C-B, Seo JS, Park GS, Lee Y-M, Lee J-S. 2014. Heavy metals induce
oxidative stress and trigger oxidative stress-mediated heat shock protein (hsp) modulation in the
intertidal copepod Tigriopus japonicus. Comp Biochem Physiol Part C Toxicol Pharmacol.
166:65–74. doi:https://doi.org/10.1016/j.cbpc.2014.07.005.
https://www.sciencedirect.com/science/article/pii/S153204561400091X.
Kingsolver JG, Huey R. 2008. Size, temperature, and fitness: Three rules. Evol Ecol Res.
10:251–268.
156
Kirsten K, Pompermaier A, Koakoski G, Mendonça-Soares S, da Costa RA, Maffi VC, Kreutz
LC, Barcellos LJG. 2021. Acute and chronic stress differently alter the expression of cytokine
and neuronal markers genes in zebrafish brain. Stress. 24(1):107–112.
doi:10.1080/10253890.2020.1724947. https://doi.org/10.1080/10253890.2020.1724947.
Klerks PL, Levinton JS. 1989. Rapid Evolution of Metal Resistance in a Benthic Oligochaete
Inhabiting a Metal-polluted Site. Biol Bull. 176(2):135–141. doi:10.2307/1541580.
https://doi.org/10.2307/1541580.
Kong Q, Lin CG. 2010. Oxidative damage to RNA: mechanisms, consequences, and diseases.
Cell Mol Life Sci. 67(11):1817–1829. doi:10.1007/s00018-010-0277-y.
https://doi.org/10.1007/s00018-010-0277-y.
Kovalchuk I, Molinier J, Yao Y, Arkhipov A, Kovalchuk O. 2007. Transcriptome analysis
reveals fundamental differences in plant response to acute and chronic exposure to ionizing
radiation. Mutat Res Mol Mech Mutagen. 624(1):101–113. doi:10.1016/j.mrfmmm.2007.04.009.
https://www.sciencedirect.com/science/article/pii/S0027510707002011.
Kültz D. 2020. Evolution of cellular stress response mechanisms. J Exp Zool Part A Ecol Integr
Physiol. 333(6):359–378. doi:10.1002/jez.2347. https://doi.org/10.1002/jez.2347.
Lee J, Phillips MC, Lobo M, Willett CS. 2021. Tolerance Patterns and Transcriptomic Response
to Extreme and Fluctuating Salinities across Populations of the Intertidal Copepod Tigriopus
californicus. Physiol Biochem Zool. 94(1):50–69. doi:10.1086/712031.
https://doi.org/10.1086/712031.
Leong W, Sun PY, Edmands S. 2018. Latitudinal Clines in Temperature and Salinity Tolerance
in Tidepool Copepods. J Hered. 109(1):71–77. doi:10.1093/jhered/esx061.
https://doi.org/10.1093/jhered/esx061.
Lesser MP. 2006. OXIDATIVE STRESS IN MARINE ENVIRONMENTS: Biochemistry and
Physiological Ecology. Annu Rev Physiol. 68(1):253–278.
doi:10.1146/annurev.physiol.68.040104.110001.
https://doi.org/10.1146/annurev.physiol.68.040104.110001.
Li AJ, Lai RWS, Zhou G-J, Leung PTY, Zeng EY, Leung KMY. 2023. Joint effects of
temperature and copper exposure on developmental and gene-expression responses of the marine
copepod Tigriopus japonicus. Ecotoxicology. 32(3):336–343. doi:10.1007/s10646-023-02643-w.
https://doi.org/10.1007/s10646-023-02643-w.
Li N, Flanagan BA, Partridge M, Huang EJ, Edmands S. 2020. Sex differences in early
transcriptomic responses to oxidative stress in the copepod Tigriopus californicus. BMC
Genomics. 21(1):759. doi:10.1186/s12864-020-07179-5. https://doi.org/10.1186/s12864-020-
07179-5.
Love MI, Huber W, Anders S. 2014. Moderated estimation of fold change and dispersion for
RNA-seq data with DESeq2. Genome Biol. 15(12):550. doi:10.1186/s13059-014-0550-8.
https://doi.org/10.1186/s13059-014-0550-8.
157
Lushchak VI. 2011. Environmentally induced oxidative stress in aquatic animals. Aquat Toxicol.
101(1):13–30. doi:10.1016/j.aquatox.2010.10.006.
https://www.sciencedirect.com/science/article/pii/S0166445X1000411X.
Morgan M, Sheperd L. 2024. AnnotationHub: Client to access AnnotationHub resources.
doi:10.18129/B9.bioc.AnnotationHub.
Nacci DE, Champlin D, Jayaraman S. 2010. Adaptation of the Estuarine Fish Fundulus
heteroclitus (Atlantic Killifish) to Polychlorinated Biphenyls (PCBs). Estuaries and Coasts.
33(4):853–864. doi:10.1007/s12237-009-9257-6. https://doi.org/10.1007/s12237-009-9257-6.
Neumann D, Lichtenberger O, Günther D, Tschiersch K, Nover L. 1994. Heat-shock proteins
induce heavy-metal tolerance in higher plants. Planta. 194(3):360–367.
doi:10.1007/BF00197536. https://doi.org/10.1007/BF00197536.
Niskanen EA, Malinen M, Sutinen P, Toropainen S, Paakinaho V, Vihervaara A, Joutsen J,
Kaikkonen MU, Sistonen L, Palvimo JJ. 2015. Global SUMOylation on active chromatin is an
acute heat stress response restricting transcription. Genome Biol. 16(1):153. doi:10.1186/s13059-
015-0717-y. https://doi.org/10.1186/s13059-015-0717-y.
Peralta-Maraver I, Rezende EL. 2021. Heat tolerance in ectotherms scales predictably with body
size. Nat Clim Chang. 11(1):58–63. doi:10.1038/s41558-020-00938-y.
https://doi.org/10.1038/s41558-020-00938-y.
Pham AN, Xing G, Miller CJ, Waite TD. 2013. Fenton-like copper redox chemistry revisited:
Hydrogen peroxide and superoxide mediation of copper-catalyzed oxidant production. J Catal.
301:54–64. doi:10.1016/j.jcat.2013.01.025.
https://www.sciencedirect.com/science/article/pii/S0021951713000377.
Putri GH, Anders S, Pyl PT, Pimanda JE, Zanini F. 2022. Analysing high-throughput sequencing
data in Python with HTSeq 2.0. Bioinformatics. 38(10):2943–2945.
doi:10.1093/bioinformatics/btac166. https://doi.org/10.1093/bioinformatics/btac166.
Ramegowda V, Da Costa MVJ, Harihar S, Karaba NN, Sreeman SM. 2020. Chapter 17 - Abiotic
and biotic stress interactions in plants: A cross-tolerance perspective. In: Hossain MA, Liu F,
Burritt DJ, Fujita M, Huang BBT-P-MS and C-ST in CP, editors. Priming-mediated stress and
cross-tolerance in crop plants. Academic Press. p. 267–302.
https://www.sciencedirect.com/science/article/pii/B9780128178928000179.
Rawson PD, Burton RS. 2006. Molecular Evolution at the Cytochrome Oxidase Subunit 2 Gene
Among Divergent Populations of the Intertidal Copepod, Tigriopus californicus. J Mol Evol.
62(6):753–764. doi:10.1007/s00239-005-0074-7. https://doi.org/10.1007/s00239-005-0074-7.
Rhee J-S, Raisuddin S, Lee K-W, Seo JS, Ki J-S, Kim I-C, Park HG, Lee J-S. 2009. Heat shock
protein (Hsp) gene responses of the intertidal copepod Tigriopus japonicus to environmental
toxicants. Comp Biochem Physiol Part C Toxicol Pharmacol. 149(1):104–112.
doi:10.1016/j.cbpc.2008.07.009.
https://www.sciencedirect.com/science/article/pii/S1532045608001415.
158
Rodgers EM, Gomez Isaza DF. 2021. Harnessing the potential of cross-protection stressor
interactions for conservation: a review. Conserv Physiol. 9(1). doi:10.1093/conphys/coab037.
https://doi.org/10.1093/conphys/coab037.
Rodgers EM, Gomez Isaza DF. 2022. Stress history affects heat tolerance in an aquatic
ectotherm (Chinook salmon, Oncorhynchus tshawytscha). J Therm Biol. 106:103252.
doi:https://doi.org/10.1016/j.jtherbio.2022.103252.
https://www.sciencedirect.com/science/article/pii/S0306456522000675.
Rodgers EM, Gomez Isaza DF. 2023. The mechanistic basis and adaptive significance of crosstolerance: a ‘pre-adaptation’ to a changing world? J Exp Biol. 226(11):jeb245644.
doi:10.1242/jeb.245644. https://doi.org/10.1242/jeb.245644.
Schoville SD, Barreto FS, Moy GW, Wolff A, Burton RS. 2012. Investigating the molecular
basis of local adaptation to thermal stress: population differences in gene expression across the
transcriptome of the copepod Tigriopus californicus. BMC Evol Biol. 12(1):170.
doi:10.1186/1471-2148-12-170. https://doi.org/10.1186/1471-2148-12-170.
Schulte PM. 2014. What is environmental stress? Insights from fish living in a variable
environment. Davies SA, Dow JAT, Lukowiak K, editors. J Exp Biol. 217(1):23–34.
doi:10.1242/jeb.089722. https://doi.org/10.1242/jeb.089722.
SgrÒ CM, Overgaard J, Kristensen TN, Mitchell KA, Cockerell FE, Hoffmann AA. 2010. A
comprehensive assessment of geographic variation in heat tolerance and hardening capacity in
populations of Drosophila melanogaster from eastern Australia. J Evol Biol. 23(11):2484–2493.
doi:10.1111/j.1420-9101.2010.02110.x. https://doi.org/10.1111/j.1420-9101.2010.02110.x.
Shackelford RE, Kaufmann WK, Paules RS. 2000. Oxidative stress and cell cycle checkpoint
function11Both Drs. Paules and Kaufmann received their doctoral degrees in Experimental
Pathology from the University of North Carolina at Chapel Hill School of Medicine (in 1984 and
1979, respectively). Free Radic Biol Med. 28(9):1387–1404. doi:https://doi.org/10.1016/S0891-
5849(00)00224-0. https://www.sciencedirect.com/science/article/pii/S0891584900002240.
Sun PY, Foley HB, Bao VWW, Leung KMY, Edmands S. 2015. Variation in tolerance to
common marine pollutants among different populations in two species of the marine copepod
Tigriopus. Environ Sci Pollut Res.:16143–16152. doi:10.1007/s11356-015-4846-3.
http://dx.doi.org/10.1007/s11356-015-4846-3.
Sunday JM, Bates AE, Dulvy NK. 2010. Global analysis of thermal tolerance and latitude in
ectotherms. Proc R Soc B Biol Sci. 278(1713):1823–1830. doi:10.1098/rspb.2010.1295.
https://doi.org/10.1098/rspb.2010.1295.
Tapia L, González-Agüero M, Cisternas MF, Suazo M, Cambiazo V, Uauy R, González M.
2004. Metallothionein is crucial for safe intracellular copper storage and cell survival at normal
and supra-physiological exposure levels. Biochem J. 378(Pt 2):617–624.
doi:10.1042/BJ20031174.
Team RC. 2024. R: A language and environment for statistical computing. https://www.rproject.org/.
159
Therneau TM. 2023. A Package for Survival Analysis in R.
Tilman D, Clark M, Williams DR, Kimmel K, Polasky S, Packer C. 2017. Future threats to
biodiversity and pathways to their prevention. Nature. 546(7656):73–81.
doi:10.1038/nature22900. https://doi.org/10.1038/nature22900.
Todgham AE, Schulte PM, Iwama GK. 2005. Cross‐Tolerance in the Tidepool Sculpin: The
Role of Heat Shock Proteins. Physiol Biochem Zool. 78(2):133–144. doi:10.1086/425205.
https://doi.org/10.1086/425205.
Todgham AE, Stillman JH. 2013. Physiological Responses to Shifts in Multiple Environmental
Stressors: Relevance in a Changing World. Integr Comp Biol. 53(4):539–544.
doi:10.1093/icb/ict086. https://doi.org/10.1093/icb/ict086.
Uren Webster TM, Rodriguez-Barreto D, Martin SAM, Van Oosterhout C, Orozco-terWengel P,
Cable J, Hamilton A, Garcia De Leaniz C, Consuegra S. 2018. Contrasting effects of acute and
chronic stress on the transcriptome, epigenome, and immune response of Atlantic salmon.
Epigenetics. 13(12):1191–1207. doi:10.1080/15592294.2018.1554520.
https://doi.org/10.1080/15592294.2018.1554520.
Venables WN, Ripley BD. 2002. Modern Applied Statistics with S. Fourth. New York: Springer.
https://www.stats.ox.ac.uk/pub/MASS4/.
Wheeler MW, Park RM, Bailer AJ. 2006. Comparing median lethal concentration values using
confidence interval overlap or ratio tests. Environ Toxicol Chem. 25(5):1441–1444.
doi:https://doi.org/10.1897/05-320R.1. https://doi.org/10.1897/05-320R.1.
Willett CS. 2010. POTENTIAL FITNESS TRADE-OFFS FOR THERMAL TOLERANCE IN
THE INTERTIDAL COPEPOD TIGRIOPUS CALIFORNICUS. Evolution (N Y). 64(9):2521–
2534. doi:10.1111/j.1558-5646.2010.01008.x. https://doi.org/10.1111/j.1558-5646.2010.01008.x.
Willett CS, Ladner JT. 2009. Investigations of fine-scale phylogeography in Tigriopus
californicus reveal historical patterns of population divergence. BMC Evol Biol. 9(1):139.
doi:10.1186/1471-2148-9-139. https://doi.org/10.1186/1471-2148-9-139.
Wright RM, Aglyamova G V, Meyer E, Matz M V. 2015. Gene expression associated with white
syndromes in a reef building coral, Acropora hyacinthus. BMC Genomics. 16(1):371.
doi:10.1186/s12864-015-1540-2. https://doi.org/10.1186/s12864-015-1540-2.
160
Figures and Tables
Figure 4-1: Map of the seven T. californicus populations collected from the California coastline.
BB = Bodega Bay, SC = Santa Cruz, PB = Pismo Beach, LC = Leo Carrillo, AB = Abalone
Cove, SCI = Santa Catalina Island, LJ = La Jolla.
161
Figure 4-2: Linear regressions between population median lethal temperatures (A) and median
lethal concentrations (B) with latitude and as well as with each other (C). Error bars represent the
standard error of each LT50 or LC50 value. Linear model fit in A: R2 = 0.89, p = 8.4E-04. Linear
model fit in B: R2 = 0.46, p = 0.056. Linear model fit in C: R2 = 0.57, p = 0.031. BB = Bodega
Bay, SC = Santa Cruz, PB = Pismo Beach, LC = Leo Carrillo, AB = Abalone Cove, SCI = Santa
Catalina Island, LJ = La Jolla.
162
Figure 4-3: Median lethal temperature of seven T. californicus populations from standard
thermal tests (circles) and from thermal tests that were preceeded by copper exposure (squares).
Error bars represent the standard error of each LT50 value. BB = Bodega Bay, SC = Santa Cruz,
PB = Pismo Beach, LC = Leo Carrillo, AB = Abalone Cove, SCI = Santa Catalina Island, LJ =
La Jolla.
163
Figure 4-4: Median lethal concentration of seven T. californicus populations from standard
copper toxicity tests (circles) and from toicity tests that were preceeded by heat exposure
(triangles). Error bars represent the standard error of each LC50 value. BB = Bodega Bay, SC =
Santa Cruz, PB = Pismo Beach, LC = Leo Carrillo, AB = Abalone Cove, SCI = Santa Catalina
Island, LJ = La Jolla.
164
Figure 4-5: Survival probability of five populations over a 17 day exposure in the five chronic
stress exposure experiments, grouped by treatment. Curves in the same plot with different letter
labels indicate populations with significantly different survival times. BB = Bodega Bay, SC =
Santa Cruz, AB = Abalone Cove, SCI = Santa Catalina Island, LJ = La Jolla. Control = 20°C and
no copper, Heat = 30°C and no copper, Low Copper = 20°C and 3 mg/L Cu, Moderate Copper =
20°C and 5 mg/L Cu, High Copper = 20°C and 8 mg/L Cu.
165
Figure 4-6: Linear regressions of acute and chronic tolerance for heat (A) and copper (B).
Chronic survival is measured as the difference between survival probability at the final census
date between the control and the heat and low copper stress treatments (Heat = 30°C and no
copper, Low Copper = 20°C and 3 mg/L Cu) Linear model fit in A: R2 = 0.22 p = 0.43. Linear
model fit in B: R2 = 0.6.7E-03, p = 0.89. BB = Bodega Bay, SC = Santa Cruz, AB = Abalone
Cove, SCI = Santa Catalina Island, LJ = La Jolla.
166
Figure 4-7: Principal components analysis (PCA) of the top 500 most variable genes from all
twelve samples (A), and for the samples from each population considered separately (B: Bodega
Bay, C: La Jolla). Expression values were transformed by the rld function in DESeq2. Point
shape refers to a sample’s treatment, and point color refers to population. BB= Bodega Bay, LJ =
La Jolla. Control = 24 hrs in 20°C with no copper, Heat = 24 hrs in 35°C with no copper, Copper
= 24 hrs in 20°C with 60 mg/L Cu.
167
Figure 4-8: Venn diagram of the number of differentially expressed genes (Benjamini-Hochberg
p < 0.1) shared between all population-treatment combinations. BB = Bodega Bay, LJ = La Jolla.
Heat = 24 hrs in 35°C with no copper, Copper = 24 hrs in 20°C with 60 mg/L Cu.
168
Figure 4-9: Hierarchical clustering of biological process (BP) gene ontology (GO) categories in
Bodega Bay copepods exposed to heat (A) and copper (B). Font style indicates level of
statistical significance and colors indicate enrichment of categories with either upregulated (red)
or downregulated (blue) genes. The fraction preceeding each category name indicates the
number of “good candidate” genes exceeding an absolute log2-fold change value of 1 relative to
the total number of genes in the category.
169
Figure 4-10: Hierarchical clustering of biological process (BP) gene ontology (GO) categories in
La Jolla copepods exposed to heat (A) and copper (B). Font style indicates level of statistical
significance and colors indicate enrichment of categories with either upregulated (red) or
downregulated (blue) genes. The fraction preceeding each category name indicates the number
of “good candidate” genes exceeding an absolute log2-fold change value of 1 relative to the total
number of genes in the category.
170
Table 4-1: The number of differentially expressed antioxidant, heat shock protein, stressinduced, chitin and cuticle genes in each treatment.
Category Gene Name
Bodega Bay La Jolla
Heat Copper Heat Copper
Up Down Up Down Up Down Up Down
Antioxidant Ascorbate peroxidase 0 1 1 0 1 0 0 0
Catalase 1 0 0 0 1 0 0 0
Glutathione peroxidase 1 0 0 0 1 0 0 0
Glutathione s-transferase 11 0 0 0 4 0 4 2
Mitogen activated protein kinases 2 0 0 0 0 0 0 0
Oxidoreductase 3 0 1 0 1 1 0 1
Peroxiredoxin 1 0 1 0 3 0 4 0
Superoxide dismutase 2 0 0 0 0 0 1 0
Thioredoxin 0 3 0 3 1 1 1 3
Thioredoxin domain containing protein 0 1 0 1 0 1 1 1
Heat Shock Protein
(HSP)
HSP beta-1 1 0 0 0 0 0 0 0
HSP10 0 1 0 0 0 0 0 0
HSP16.48/16.49 0 1 0 1 0 0 0 1
HSP60 0 1 0 1 0 1 0 1
HSP70 0 4 0 5 0 2 0 5
HSP83 0 1 0 1 0 0 0 1
HSP90 0 1 0 1 0 1 0 0
Stress-Induced Stress-70 protein 0 1 0 1 0 0 0 1
Stress-induced-phosphoprotein 1-like 0 1 0 1 0 1 0 1
Oxidative stressinduced growth inhibitor 1 1 0 0 0 0 0 0 1
Chitin Chitinase 2 4 1 5 1 0 1 5
Chitin deacetylase 4 2 2 2 1 0 0 2
Chitin synthase 0 0 1 0 1 0 0 0
Cuticle Cuticle protein 21 1 0 0 0 1 0 0 0
Cuticle protein 27 0 0 0 0 1 0 0 0
Cuticle protein 38 1 0 0 0 1 0 0 0
Total Responding Genes 31 22 7 22 18 8 12 25
Up = Gene was upregulated relative to control
Down = Gene was downregulated relative to control
171
Supplementary Materials
Figure S4-1: Distribution of the number of offspring at first count from five populations in the
five chronic stress exposure experiments. An asterisk (*) above a bar indicates that the mean
number of offspring in a treatment was significantly less than the mean number of offspring in
the control treatment. Control = 20°C and no copper, Heat = 30°C and no copper, Low Copper =
20°C and 3 mg/L Cu, Moderate Copper = 20°C and 5 mg/L Cu, High Copper = 20°C and 8 mg/L
Cu. BB = Bodega Bay, SC = Santa Cruz, AB = Abalone Cove, SCI = Santa Catalina Island, LJ =
La Jolla.
172
Figure S4-2: Survival probability of five populations over a 17 day exposure in the five chronic
stress exposure experiments, grouped by population. Curves in the same plot with different letter
labels indicate populations with significantly different survival probabilities. Control = 20°C and
no copper, Heat = 30°C and no copper, Low Copper = 20°C and 3 mg/L Cu, Moderate Copper =
20°C and 5 mg/L Cu, High Copper = 20°C and 8 mg/L Cu.
173
Figure S4-3: Heatmap of the rlog transformed expression values of the top 250 most variable
genes across all groups. BB= Bodega Bay, LJ = La Jolla. Control = 24 hrs in 20°C with no
copper, Heat = 24 hrs in 35°C with no copper, Copper = 24 hrs in 20°C with 60 mg/L Cu.
Table S4-1: Approximate coordinates for each of the sampled Tigriopus californicus
populations
Population Latitude
(°N)
Longitude
(°W)
Bodega Bay (BB) 38.32 -123.1
Santa Cruz (SC) 36.95 -122.1
Pismo Beach (PB) 35.09 -120.7
Leo Carrillo (LC) 34.04 -118.9
Abalone Cove (AB) 33.73 -118.4
Santa Catalina Island (SCI) 33.45 -118.5
La Jolla (LJ) 32.83 -117.3
174
Table S4-2: Number of raw reads and percentage of reads mapped to the San Diego reference
genome in each sample
Sample Name Population Treatment Total Reads Alignment
Rate
BA1 Bodega Bay Control 21,115,893 52.93%
BA2 Bodega Bay Control 27,883,234 50.36%
BA3 Bodega Bay Heat 30,539,955 53.43%
BA4 Bodega Bay Heat 30,291,769 53.66%
BA5 Bodega Bay Copper 33,581,001 49.88%
BA6 Bodega Bay Copper 29,273,615 51.21%
LA1 La Jolla Control 25,311,298 62.68%
LA2 La Jolla Control 32,840,962 71.33%
LA3 La Jolla Heat 33,848,509 73.17%
LA4 La Jolla Heat 29,258,125 72.07%
LA5 La Jolla Copper 23,944,229 69.05%
LA6 La Jolla Copper 26,983,968 78.72%
Table S4-3: Copepod mortalities after a thermal stress test at 35°C and after 24 hours in 60 mg/L
Cu
Population Treatment Sample Size Mortalities
Bodega Bay 35°C 12 0
Santa Cruz 35°C 24 0
Pismo Beach 35°C 36 0
Leo Carrillo 35°C 36 0
Abalone Cove 35°C 12 0
Santa Catalina Island 35°C 48 0
La Jolla 35°C 12 0
Bodega Bay 60 mg/L Cu 72 12
Santa Cruz 60 mg/L Cu 114 28
Pismo Beach 60 mg/L Cu 47 13
Leo Carrillo 60 mg/L Cu 48 18
Abalone Cove 60 mg/L Cu 96 35
Santa Catalina Island 60 mg/L Cu 90 29
La Jolla 60 mg/L Cu 113 24
175
Table S4-4: Differentially expressed genes shared by all four treatment groups.
Gene Symbol Gene Name Bodega Bay La Jolla
Heat Copper Heat Copper
LOC131876887 hepatic lectin-like up up up up
LOC131876909 trimethylamine monooxygenase-like up up up down
LOC131877019 uncharacterized LOC131877019 up up up up
LOC131877154 26S proteasome non-ATPase regulatory subunit 8-like down down down down
LOC131877195 heat shock 70 kDa protein cognate 4-like down down down down
LOC131877444 uncharacterized LOC131877444 up up up up
LOC131877456 phytanoyl-CoA dioxygenase, peroxisomal-like up up up up
LOC131877494 proton channel OtopLc-like up up up up
LOC131877497
lysosomal proton-coupled steroid conjugate and bile acid
symporter SLC46A3-like up up up up
LOC131877542 uncharacterized LOC131877542 up up up up
LOC131877579 UDP-glycosyltransferase UGT5-like up up up up
LOC131877758 E3 ubiquitin-protein ligase HUWE1-like down down down down
LOC131877879 von Willebrand factor A domain-containing protein 2-like up up up up
LOC131877888 uncharacterized LOC131877888 up up up up
LOC131877935 GTP cyclohydrolase 1-like up up up up
LOC131878120 cyanophycinase-like up up up up
LOC131878146 zinc finger protein 345-like down down down down
LOC131878271 forkhead box protein D1-like up up up up
LOC131878619 uncharacterized LOC131878619 up up up up
LOC131878931 mannan endo-1,4-beta-mannosidase-like down up up up
LOC131879070 uncharacterized LOC131879070 up up up up
LOC131879170 alkaline phosphatase 4-like down up down down
LOC131879250 uncharacterized LOC131879250 up up up up
LOC131879386 uncharacterized LOC131879386 up up up up
LOC131879440 m7GpppN-mRNA hydrolase-like down down down down
LOC131879525 glutamate decarboxylase-like up up up up
LOC131879562 sodium/potassium-transporting ATPase subunit alpha-B-like up up up up
LOC131879648 perlucin-like up up up up
LOC131879822 protein disulfide-isomerase-like up up up up
LOC131879932 poly(3-hydroxybutyrate) depolymerase-like down up up up
LOC131880156 N-acetylated-alpha-linked acidic dipeptidase 2-like up up up up
LOC131880287 aldo-keto reductase family 1 member B1-like up down up down
LOC131880393 uncharacterized LOC131880393 down down down down
LOC131880395 uncharacterized LOC131880395 down down up up
LOC131880542 NF-X1-type zinc finger protein NFXL1-like down down down down
LOC131881090 zwei Ig domain protein zig-4-like up up up up
LOC131881189 tubulin beta chain-like up up up up
LOC131881262 brachyurin-like down up up up
LOC131881263 brachyurin-like down up up up
LOC131881304 uncharacterized LOC131881304 up up down down
176
LOC131881410 ecdysone receptor-like up down up down
LOC131881557
pre-mRNA-splicing factor ATP-dependent RNA helicase
PRP16-like down down down down
LOC131881669 cartilage oligomeric matrix protein-like up up up up
LOC131881683 innexin inx2-like up up up up
LOC131881773 sucrase-isomaltase, intestinal-like down up up up
LOC131882171 elongation of very long chain fatty acids protein 7-like up up up up
LOC131882396 sphingosine kinase 1-like up up up up
LOC131882573 cullin-2-like down down down down
LOC131882656 nuclear receptor subfamily 1 group I member 3-like up down up down
LOC131882734 uncharacterized LOC131882734 up up up up
LOC131882756 uncharacterized LOC131882756 up up up up
LOC131882851 heat shock protein 60A-like down down down down
LOC131883101 electrogenic sodium bicarbonate cotransporter 1-like up up up up
LOC131883102 uncharacterized LOC131883102 up up up up
LOC131883190 heat shock 70 kDa protein 1-like down down down down
LOC131883226 anti-sigma-I factor RsgI6-like down up up up
LOC131883257
phosphatidylethanolamine-binding protein homolog
F40A3.3-like up up up up
LOC131883348 uncharacterized LOC131883348 up up up up
LOC131883421 soma ferritin-like down down down down
LOC131883585 maltase 1-like down up down up
LOC131884124 long-chain-fatty-acid--CoA ligase ACSBG2-like up up up up
LOC131884219 erythroid differentiation-related factor 1-like down down down down
LOC131884450
hemocyte protein-glutamine gamma-glutamyltransferaselike up up up up
LOC131884767 sodium-dependent glucose transporter 1-like up up up up
LOC131884796 exosome complex exonuclease RRP44-like down down down down
LOC131884862 pancreatic triacylglycerol lipase-like down up up up
LOC131884964 uncharacterized LOC131884964 up up up up
LOC131884990 facilitated trehalose transporter Tret1-like up up up down
LOC131885169 purine nucleoside phosphorylase-like up up up up
LOC131885266 protein c-Fos-like up down up down
LOC131885392 aquaporin-like up up up up
LOC131885413 uncharacterized LOC131885413 up up up up
LOC131885448 proteasome subunit beta type-2-like down down down down
LOC131885609 protein ultraspiracle homolog up up up up
LOC131885767 peptidyl-prolyl cis-trans isomerase E-like down down down down
LOC131885782 uncharacterized LOC131885782 up up up up
LOC131885938
activator of 90 kDa heat shock protein ATPase homolog 1-
like down down down down
LOC131885956 uncharacterized LOC131885956 down down down down
LOC131886042 rhomboid-related protein 1-like down down up down
LOC131886100 uncharacterized LOC131886100 down down up down
LOC131886208 phenoloxidase-activating factor 2-like up up up up
LOC131886565 F-box/LRR-repeat protein 14-like up up up up
177
LOC131886705 acetylcholine receptor subunit alpha-like down up down up
LOC131887236 uncharacterized LOC131887236 down up down up
LOC131887246 serine protease inhibitor 42Dd-like up up up up
LOC131887366 endoglucanase E-4-like down up up up
LOC131887389 uncharacterized LOC131887389 down down down down
LOC131887398 switch-associated protein 70-like up up up up
LOC131887516 phenoloxidase-activating factor 1-like up up up up
LOC131887784 uncharacterized LOC131887784 up up up up
LOC131888374 beta-1,3-galactosyltransferase 1-like up up up up
LOC131888540 myosin-11-like down down down down
LOC131888543 uncharacterized LOC131888543 down down up down
LOC131888698 heat shock protein Hsp-16.48/Hsp-16.49-like down down down down
LOC131889100 legumain-like up up up up
LOC131889587 26S proteasome regulatory subunit 10B down down down down
LOC131889658 T-complex protein 1 subunit eta-like down down down down
LOC131889700 sarcoplasmic calcium-binding proteins I, III, and IV-like up up up up
LOC131889779 uncharacterized LOC131889779 up up up up
LOC131889812 uncharacterized LOC131889812 up up up up
LOC131889939 uncharacterized LOC131889939 up up up up
LOC131889974 uncharacterized LOC131889974 up up up up
LOC131889976 E3 ubiquitin-protein ligase UBR5-like down down down down
LOC131890012 dehydrogenase/reductase SDR family member 7-like up up up up
LOC131890038 uncharacterized LOC131890038 up up up up
LOC131890082 glutaminase liver isoform, mitochondrial-like up up up up
LOC131890125 cleavage stimulation factor subunit 3-like down down down down
LOC131890141 sodium-dependent proline transporter-like up up up up
LOC131890204 uncharacterized LOC131890204 up up up up
LOC131890381 procathepsin L-like up up up up
LOC131890401 carbonic anhydrase-like up up up up
LOC131890772 protein Skeletor up up up up
LOC131891097 stress-induced-phosphoprotein 1-like down down down down
LOC131891304 neprilysin-2-like up up up up
LOC131891407 uncharacterized LOC131891407 up up up up
LOC131891578 uncharacterized LOC131891578 up down up down
LOC131891581 leukocyte elastase inhibitor-like up up up up
LOC131891602 aldo-keto reductase family 1 member B1-like up up up up
LOC131891872
eukaryotic translation initiation factor 4E-binding protein 1-
like up up up up
LOC131892192
uncharacterized oxidoreductase MexAM1_META1p0182-
like down down down down
LOC131892305 uncharacterized LOC131892305 up up up up
LOC131892439 uncharacterized LOC131892439 down up up up
LOC131892594 uncharacterized LOC131892594 up up up up
LOC131892771 DNA topoisomerase 2-alpha-like down down down down
LOC131892816 protein 5NUC-like up up up up
178
LOC131892901 proteasome subunit alpha type-4-like down down down down
LOC131892917 endoplasmic reticulum chaperone BiP-like down down down down
LOC131893007 uncharacterized LOC131893007 down down down down
LOC131893080 uncharacterized LOC131893080 down down down down
LOC131893202 DNA replication licensing factor mcm4-A-like down down down down
LOC131893218 uncharacterized LOC131893218 up up up up
LOC131893225 uncharacterized LOC131893225 up up up up
LOC131893370 uncharacterized LOC131893370 down down down down
LOC131893372 uncharacterized LOC131893372 down down down down
LOC131893447 mitochondrial glycine transporter A-like up up up up
LOC131893449 uncharacterized LOC131893449 down down down down
LOC131893645 probable cytochrome P450 6a14 up up up up
Up = Gene was upregulated relative to control
Down = Gene was downregulated relative to control
179
Chapter 5: Conclusions
Chemical sensitivity is a complex trait, shaped both by exposure to pollution, which may
impose a selective force or cause mutations (Nacci et al. 2010; Bickham 2011), and by deeper
evolutionary relationships that help determine the presence and efficiency of physiological traits
related to chemical metabolism in an organism (Guénard et al. 2011; Guénard et al. 2014). This
dissertation integrates these ideas into two distinct areas of research in aquatic ecotoxicology: the
development of pollution regulations and the study of multiple stressors. Key findings include
the identification of significant taxonomic patterns in toxicity data, but limited potential for
broad phylogenetic relationships to reliably predict chemical sensitivity. Additionally, the marine
copepod Tigriopus californicus exhibited similarities between its responses to copper and heat
that could confer fitness benefits during multiple stressor exposure.
Taxonomic differences in toxicity data
Toxicity data are the backbone of the chemical regulations, yet there are concerns in
ecotoxicology about the availability, composition and usage of these data across all chemicals
(Richard et al. 2009; David B. Buchwalter et al. 2017; DeForest et al. 2017). Chapter 2 uses a
meta-analysis of published toxicity data to assess the shift in their abundance and taxonomic
composition over time and their effects on water quality criteria development, as well as to
explore broad taxonomic patterns of sensitivity. This work unveiled major gaps in the existing
toxicity database, suggesting that redefining the meaning of acceptable data is necessary if the
pace of water quality criteria development is to increase. The amount of biological diversity of a
toxicity dataset did not appear to quantitatively influence the value of a criterion, however,
significant taxonomic differences in sensitivity were evident in the data at the phylum level
which may be necessary to consider during future criteria development.
180
Chapter 3 aims to address the gaps in the toxicity database by evaluating whether
evolutionary relationships can be used to reliably extrapolate new data for untested combinations
of chemicals and species. This was done with a second, broader meta-analysis that quantified the
phylogenetic influence on sensitivity for a large group of chemicals and investigated whether
various chemical properties and experimental conditions such as temperature and pH affected
phylogenetic signal. Strong phylogenetic signal was apparent at high taxonomic levels for only a
small subset of these chemicals and there were no clear shared properties among those datasets
with strong signal, meaning that the phylogenetic methods should only be used to address data
gaps in certain instances. However, when signal was strong, distinct patterns of sensitivity were
evident in the data, building on the broad-scale taxonomic trends identified in Chapter 2.
Similarities between copper and heat tolerance in Tigriopus californicus
Chemical pollution is just one among many environmental stressors that threaten aquatic
life. The effects of multiple stressors vary (Folt et al. 1999), but cross-protection represents one
possible outcome that could confer considerable fitness advantages to organisms in a changing
environment (Rodgers and Gomez Isaza 2023). In Chapter 4, I investigate the potential for crosstolerance in T. californicus between copper and heat on the basis of a shared response
mechanism between the two stressors. Using in vivo stress experiments, I found that the
tolerances of both stressors have similar latitudinal distributions among T. californicus
populations and that cross-protection occurred when animals were exposed to heat prior to a
copper toxicity test. Transcriptomic sequencing indicated that copper and heat induced distinct
gene expression responses in the copepods, although both similarly modulated the antioxidant
defense system and metabolic processes that could help generate the cross-protection response
that I observed.
181
Final remarks
Evolutionary concepts can benefit both fundamental and applied research in aquatic
ecotoxicology and beyond by providing a framework for understanding the processes that control
responses to toxic chemicals. Macroevolutionary relationships can help explain a portion of the
variation in chemical sensitivity that occurs across the diversity of life, while a contemporary
perspective of evolution can improve the realism of toxicity measurements by acknowledging the
potential for populations to develop divergent sensitivities as a result of their local environmental
conditions (Brady et al. 2017). The results presented here illustrate these concepts and open new
questions regarding the impact of taxonomic patterns of sensitivity among closely related groups
as well as the potential for heightened chemical tolerance to arise from means other than direct
selection by chemical exposure.
References
Bickham JW. 2011. The four cornerstones of Evolutionary Toxicology. Ecotoxicology.
20(3):497–502. doi:10.1007/s10646-011-0636-y. https://doi.org/10.1007/s10646-011-0636-y.
Brady SP, Richardson JL, Kunz BK. 2017. Incorporating evolutionary insights to improve
ecotoxicology for freshwater species. Evol Appl. 10(8):829–838.
doi:https://doi.org/10.1111/eva.12507. https://doi.org/10.1111/eva.12507.
Buchwalter DB, Clements WH, Luoma SN. 2017. Modernizing Water Quality Criteria in the
United States: A Need to Expand the Definition of Acceptable Data. Environ Toxicol Chem.
36(2):285–291. doi:10.1002/etc.3654.
DeForest D, Croteau K, Ryan A, Santore B, Toll J. 2017. Retrospective on The United States
Environmental Protection Agency’s guidelines for deriving ambient water quality criteria. Integr
Environ Assess Manag. 13(6):1125–1126. doi:10.1002/ieam.1948.
Folt CL, Chen CY, Moore M V, Burnaford J. 1999. Synergism and antagonism among multiple
stressors. Limnol Oceanogr. 44(3part2):864–877. doi:10.4319/lo.1999.44.3_part_2.0864.
https://doi.org/10.4319/lo.1999.44.3_part_2.0864.
Guénard G, Carsten von der Ohe P, Carlisle Walker S, Lek S, Legendre P. 2014. Using
phylogenetic information and chemical properties to predict species tolerances to pesticides. Proc
R Soc B Biol Sci. 281(1789):20133239. doi:10.1098/rspb.2013.3239.
https://doi.org/10.1098/rspb.2013.3239.
182
Guénard G, von der Ohe PC, de Zwart D, Legendre P, Lek S. 2011. Using phylogenetic
information to predict species tolerances to toxic chemicals. Ecol Appl. 21(8):3178–3190.
doi:10.1890/10-2242.1. https://doi.org/10.1890/10-2242.1.
Nacci DE, Champlin D, Jayaraman S. 2010. Adaptation of the Estuarine Fish Fundulus
heteroclitus (Atlantic Killifish) to Polychlorinated Biphenyls (PCBs). Estuaries and Coasts.
33(4):853–864. doi:10.1007/s12237-009-9257-6. https://doi.org/10.1007/s12237-009-9257-6.
Richard J, Ann R, J. DD, Keith H, Matthew M, Robert K, Vicki D, Tala H, Todd H, Philip S, et
al. 2009. The Toxicity Data Landscape for Environmental Chemicals. Environ Health Perspect.
117(5):685–695. doi:10.1289/ehp.0800168. https://doi.org/10.1289/ehp.0800168.
Rodgers EM, Gomez Isaza DF. 2023. The mechanistic basis and adaptive significance of crosstolerance: a ‘pre-adaptation’ to a changing world? J Exp Biol. 226(11):jeb245644.
doi:10.1242/jeb.245644. https://doi.org/10.1242/jeb.245644.
183
Comprehensive References
Amiard-Triquet C. 2015. Chapter 1 - Introduction. In: Amiard-Triquet C, Amiard J-C,
Mouneyrac CBT-AE, editors. Academic Press. p. 1–23.
https://www.sciencedirect.com/science/article/pii/B9780128009499000012.
Anderson B, Phillips B. 2016. Chapter 6 - Saltwater Toxicity Tests. In: Blasco J, Chapman PM,
Campana O, Hampel M, editors. Marine Ecotoxicology. Academic Press. p. 167–197.
https://www.sciencedirect.com/science/article/pii/B9780128033715000060.
Andrews S. 2010. FastQC: A Quality Control Tool for High Throughput Sequence Data
[Online]. http://www.bioinformatics.babraham.ac.uk/projects/fastqc/.
Ankley G, Barron M, Bergeron C, Burgess R, Cormier S, Etterson M, Hoff D, Lalone C, Mount
D, Suter G, et al. 2017. Update on US EPA’s Revision to the 1985 Guidelines for Deriving
Aquatic Life Criteria. :1–3.
Arthur JW, West CW, Allen KN, Hedtke SF. 1987. Seasonal Toxicity of Ammonia to Five Fish
and Nine Invertebrate Species. Bull Environ Contam. 38:324–331.
Augspurger T, Keller A, Black M, Cope G, Dwyer J. 2003. Water Quality Guidance for
Protection of Freshwater Mussels (Unionidae) from Ammonia Exposure. Environ Sci Technol.
22(11):2569–2575.
Awkerman J, Raimondo S, Jackson C, Barron M. 2014. Augmenting Aquatic Species Sensitivity
Distributions with Interspecies Toxicity Estimation Models. Environ Toxicol Chem. 33(3):688–
695. doi:10.1002/etc.2456.
Barata C, Baird DJ, Mitchell SE, Soares AMVM. 2002. Among- and within-population
variability in tolerance to cadmium stress in natural populations of Daphnia magna: Implications
for ecological risk assessment. Environ Toxicol Chem. 21(5):1058–1064.
doi:10.1002/etc.5620210523. https://doi.org/10.1002/etc.5620210523.
Barreto FS, Watson ET, Lima TG, Willett CS, Edmands S, Li W, Burton RS. 2018. Genomic
signatures of mitonuclear coevolution across populations of Tigriopus californicus. Nat Ecol
Evol. 2(8):1250–1257. doi:10.1038/s41559-018-0588-1. https://doi.org/10.1038/s41559-018-
0588-1.
Barron MG. 2002. Bioaccumulation and Bioconcentration in Aquatic Organisms. In: Hoffman
DJ, Rattner BA, Burton Jr, G, Cairns Jr. J, editors. Handbook of Ecotoxicology. 2nd ed. Boca
Raton: CRC Press. p. 877–892.
Barron MG, Lilavois CR, Martin TM. 2015. MOAtox: A comprehensive mode of action and
acute aquatic toxicity database for predictive model development. Aquat Toxicol. 161:102–107.
doi:10.1016/j.aquatox.2015.02.001.
https://www.sciencedirect.com/science/article/pii/S0166445X15000521.
Barshis DJ, Ladner JT, Oliver TA, Seneca FO, Traylor-Knowles N, Palumbi SR. 2013. Genomic
basis for coral resilience to climate change. PNAS. 110(4):1387–1392.
Beiras R. 2018. Marine Pollution: Sources, Fate and Effects of Pollutants in Coastal Ecosystems.
Elsevier Science. https://books.google.com/books?id=sOtlDwAAQBAJ.
184
Belhadj Slimen I, Najar T, Ghram A, Dabbebi H, Ben Mrad M, Abdrabbah M. 2014. Reactive
oxygen species, heat stress and oxidative-induced mitochondrial damage. A review. Int J
Hyperth. 30(7):513–523. doi:10.3109/02656736.2014.971446.
https://doi.org/10.3109/02656736.2014.971446.
Benjamini Y, Hochberg Y. 1995. Controlling the False Discovery Rate: A Practical and
Powerful Approach to Multiple Testing. J R Stat Soc Ser B. 57(1):289–300. doi:10.1111/j.2517-
6161.1995.tb02031.x. https://doi.org/10.1111/j.2517-6161.1995.tb02031.x.
van den Berg SJP, Maltby L, Sinclair T, Liang R, van den Brink PJ. 2021. Cross-species
extrapolation of chemical sensitivity. Sci Total Environ. 753:141800.
doi:10.1016/j.scitotenv.2020.141800.
https://www.sciencedirect.com/science/article/pii/S0048969720353298.
Bickham JW. 2011. The four cornerstones of Evolutionary Toxicology. Ecotoxicology.
20(3):497–502. doi:10.1007/s10646-011-0636-y. https://doi.org/10.1007/s10646-011-0636-y.
Bickham JW, Sandhu S, Hebert PDN, Chikhi L, Athwal R. 2000. Effects of chemical
contaminants on genetic diversity in natural populations: implications for biomonitoring and
ecotoxicology. Mutat Res Mutat Res. 463(1):33–51. doi:10.1016/S1383-5742(00)00004-1.
https://www.sciencedirect.com/science/article/pii/S1383574200000041.
Bilyk KT, Evans CW, DeVries AL. 2012. Heat hardening in Antarctic notothenioid fishes. Polar
Biol. 35(9):1447–1451. doi:10.1007/s00300-012-1189-0. https://doi.org/10.1007/s00300-012-
1189-0.
Board WE. 2021. World Register of Marine Species. doi:10.14284/170.
Bolger AM, Lohse M, Usadel B. 2014. Trimmomatic: a flexible trimmer for Illumina sequence
data. Bioinformatics. 30(15):2114–2120. doi:10.1093/bioinformatics/btu170.
https://doi.org/10.1093/bioinformatics/btu170.
Brady SP, Richardson JL, Kunz BK. 2017. Incorporating evolutionary insights to improve
ecotoxicology for freshwater species. Evol Appl. 10(8):829–838. doi:10.1111/eva.12507.
https://doi.org/10.1111/eva.12507.
Breitburg DL, Riedel GF. 2005. Multiple Stressors in Marine Systems. In: Marine Conservation
Biology: The Science of Maintaining the Sea’s Biodiversity. p. 167–182.
Brooks SJ, Waldock M. 2009. Copper Biocides in the Marine Environment. In: Arai T, Harino
H, Ohji M, Langston WJ, editors. Ecotoxicology of Antifouling Biocides. Tokyo: Springer
Japan. p. 413–428. https://doi.org/10.1007/978-4-431-85709-9_24.
Brown BE. 1978. Lead detoxification by a copper-tolerant isopod. Nature. 276(5686):388–390.
doi:10.1038/276388a0. https://doi.org/10.1038/276388a0.
Brown CJ, Saunders MI, Possingham HP, Richardson AJ. 2013. Managing for Interactions
between Local and Global Stressors of Ecosystems. PLoS One. 8(6):e65765.
https://doi.org/10.1371/journal.pone.0065765.
185
Buchwalter DB, Cain DJ, Martin CA, Xie L, Luoma SN, Garland T. 2008. Aquatic insect
ecophysiological traits reveal phylogenetically based differences in dissolved cadmium
susceptibility. Proc Natl Acad Sci. 105(24):8321–8326. doi:10.1073/pnas.0801686105.
https://doi.org/10.1073/pnas.0801686105.
Buchwalter David B, Clements WH, Luoma SN. 2017. Modernizing Water Quality Criteria in
the United States: A Need to Expand the Definition of Acceptable Data. Environ Toxicol Chem.
36(2):285–291. doi:10.1002/etc.3654. https://doi.org/10.1002/etc.3654.
Burant A, Zhang X, Singhasemanon N, Teerlink J. 2018. Study 319 Report: Monitoring of
Dissolved Copper in California Coastal Waterbodies. Sacramento.
Burnham K, Anderson D. 2002. Model Selection and Multimodel Inference.
Burton RS. 1998. INTRASPECIFIC PHYLOGEOGRAPHY ACROSS THE POINT
CONCEPTION BIOGEOGRAPHIC BOUNDARY. Evolution (N Y). 52(3):734–745.
doi:10.1111/j.1558-5646.1998.tb03698.x. https://doi.org/10.1111/j.1558-5646.1998.tb03698.x.
Cairns JJ, Heath AG, Parker BC. 1975. Temperature influence on chemical toxicity to aquatic
organisms. J Water Pollut Control Fed. 47(2):267–280.
Canadian Council of Ministers of the Environment. 2007. A Protocol for the Derivation of Water
Quality Guidelines for the Protection of Aquatic Life 2007.
Carew ME, Miller AD, Hoffmann AA. 2011. Phylogenetic signals and ecotoxicological
responses: potential implications for aquatic biomonitoring. Ecotoxicology. 20(3):595–606.
doi:10.1007/s10646-011-0615-3. https://doi.org/10.1007/s10646-011-0615-3.
Carr GJ, Belanger SE. 2018. SSDs Revisited: Part I—A Framework for Sample Size Guidance
on Species Sensitivity Distribution Analysis. Environ Toxicol Chem. 38(7):1514–1525.
doi:10.1002/etc.4445. http://dx.doi.org/10.1002/etc.4445.
Chen N, Huang Z, Lu C, Shen Y, Luo X, Ke C, You W. 2019. Different Transcriptomic
Responses to Thermal Stress in Heat-Tolerant and Heat-Sensitive Pacific Abalones Indicated by
Cardiac Performance. Front Physiol. 9.
https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2018.01895.
Chiari Y, Glaberman S, Serén N, Carretero MA, Capellini I. 2015. Phylogenetic signal in
amphibian sensitivity to copper sulfate relative to experimental temperature. Ecol Appl.
25(3):596–602. doi:10.1890/14-0439.1. https://doi.org/10.1890/14-0439.1.
Clarke KR, Warwick RM. 1998. A taxonomic distinctness index and its statistical properties. J
Appl Ecol. 35(4):523–531. doi:https://doi.org/10.1046/j.1365-2664.1998.3540523.x.
https://doi.org/10.1046/j.1365-2664.1998.3540523.x.
Coleman AL, Edmands S. 2022. Data and Diversity in the Development of Acute Water Quality
Criteria in the United States. Environ Toxicol Chem. 41(5):1333–1343. doi:10.1002/etc.5302.
https://doi.org/10.1002/etc.5302.
Collins M, Clark MS, Spicer JI, Truebano M. 2020. Transcriptional frontloading contributes to
cross-tolerance between stressors. Evol Appl. 14(2):577–587. doi:10.1111/eva.13142.
https://doi.org/10.1111/eva.13142.
186
Cormier SM, Paul JF, Spehar RL, Shaw-Allen P, Berry WJ, Suter GW. 2008. Using field data
and weight of evidence to develop water quality criteria. Integr Environ Assess Manag.
4(4):490–504. doi:https://doi.org/10.1897/IEAM_2008-018.1.
https://doi.org/10.1897/IEAM_2008-018.1.
Côté IM, Darling ES, Brown CJ. 2016. Interactions among ecosystem stressors and their
importance in conservation. Proc R Soc B Biol Sci. 283(1824):20152592.
doi:10.1098/rspb.2015.2592. https://doi.org/10.1098/rspb.2015.2592.
Crain CM, Kroeker K, Halpern BS. 2008. Interactive and cumulative effects of multiple human
stressors in marine systems. Ecol Lett. 11:1304–1315. doi:10.1111/j.1461-0248.2008.01253.x.
Crane M, Newman MC. 2000. What level of effect is a no observed effect? Environ Toxicol
Chem. 19(2):516–519. doi:10.1002/etc.5620190234. https://doi.org/10.1002/etc.5620190234.
DeForest D, Croteau K, Ryan A, Santore B, Toll J. 2017. Retrospective on The United States
Environmental Protection Agency’s guidelines for deriving ambient water quality criteria. Integr
Environ Assess Manag. 13(6):1125–1126. doi:10.1002/ieam.1948.
Ding C, Ma Y, Li X, Zhang T, Wang X. 2016. Derivation of soil thresholds for lead applying
species sensitivity distribution: A case study for root vegetables. J Hazard Mater. 303:21–27.
doi:https://doi.org/10.1016/j.jhazmat.2015.10.027.
https://www.sciencedirect.com/science/article/pii/S0304389415301552.
Dowse R, Tang D, Palmer CG, Kefford BJ. 2013. Risk assessment using the species sensitivity
distribution method: Data quality versus data quantity. Environ Toxicol Chem. 32(6):1360–1369.
doi:10.1002/etc.2190. https://doi.org/10.1002/etc.2190.
Dröge W. 2002. Free radicals in the physiological control of cell function. Physiol Rev.
82(1):47–95. doi:10.1152/physrev.00018.2001.
Duboudin C, Ciffroy P, Magaud H. 2004. Acute-to-chronic species sensitivity distribution
extrapolation. Environ Toxicol Chem. 23(7):1774–1785. doi:10.1002/etc.5620230723.
https://doi.org/10.1002/etc.5620230723.
Duque T, Chowdhury S, Isaia M, Pekár S, Riess K, Scherf G, Schäfer RB, Entling MH. 2023.
Sensitivity of spiders from different ecosystems to lambda-cyhalothrin: effects of phylogeny and
climate. Pest Manag Sci. 80(2):857–865. doi:https://doi.org/10.1002/ps.7818.
https://doi.org/10.1002/ps.7818.
Edmands S. 2001. Phylogeography of the intertidal copepod Tigriopus californicus reveals
substantially reduced population differentiation at northern latitudes. Mol Ecol. 10(7):1743–
1750. doi:10.1046/j.0962-1083.2001.01306.x. https://doi.org/10.1046/j.0962-1083.2001.01306.x.
Elias M, Ankley G, Barron M, Eignor D, Erickson R, Hoff D, Jarvis L, Justice JR, Lalone C,
Mount D, et al. 2019. Developing Protective Aquatic Life Values for Data-Limited PFAS
Chemicals Criteria Development and Minimum Data Requirements ( MDRs ). :1–18.
Felsenstein J. 1985. Phylogenies and the comparative method. Am Nat. 125(1):1–15.
187
Foley HB, Sun PY, Ramirez R, So BK, Venkataraman YR, Nixon EN, Davies KJA, Edmands S.
2019. Sex-specific stress tolerance, proteolysis, and lifespan in the invertebrate Tigriopus
californicus. Exp Gerontol. 119:146–156. doi:10.1016/j.exger.2019.02.006.
https://www.sciencedirect.com/science/article/pii/S0531556518305734.
Folt CL, Chen CY, Moore M V, Burnaford J. 1999. Synergism and antagonism among multiple
stressors. Limnol Oceanogr. 44(3part2):864–877. doi:10.4319/lo.1999.44.3_part_2.0864.
https://doi.org/10.4319/lo.1999.44.3_part_2.0864.
Forbes VE, Calow P. 2002. Species Sensitivity Distributions Revisited: A Critical Appraisal.
Hum Ecol Risk Assess An Int J. 8(3):473–492. doi:10.1080/10807030290879781.
https://doi.org/10.1080/10807030290879781.
Fuse T, Yamada K, Asai K, Kato T, Nakanishi M. 1996. Heat Shock-Mediated Cell Cycle Arrest
Is Accompanied by Induction of p21 CKI. Biochem Biophys Res Commun. 225(3):759–763.
doi:10.1006/bbrc.1996.1247.
https://www.sciencedirect.com/science/article/pii/S0006291X96912474.
Ganz HH, Burton RS. 1995. Genetic differentiation and reproductive incompatibility among
Baja California populations of the copepod Tigriopus californicus. Mar Biol. 123(4):821–827.
doi:10.1007/BF00349126. https://doi.org/10.1007/BF00349126.
Ghosh S, Sarkar P, Basak P, Mahalanobish S, Sil PC. 2018. Role of Heat Shock Proteins in
Oxidative Stress and Stress Tolerance BT - Heat Shock Proteins and Stress. In: Asea AAA,
Kaur P, editors. Cham: Springer International Publishing. p. 109–126.
https://doi.org/10.1007/978-3-319-90725-3_6.
Giddings JM, Wirtz J, Campana D, Dobbs M. 2019. Derivation of combined species sensitivity
distributions for acute toxicity of pyrethroids to aquatic animals. Ecotoxicology. 28:242–250.
doi:10.1007/s10646-019-02018-0. http://dx.doi.org/10.1007/s10646-019-02018-0.
Gleason LU, Burton RS. 2015. RNA-seq reveals regional differences in transcriptome response
to heat stress in the marine snail Chlorostoma funebralis. Mol Ecol. 24(3):610–627.
doi:10.1111/mec.13047. https://doi.org/10.1111/mec.13047.
Grafen A, Hamilton WD. 1989. The phylogenetic regression. Philos Trans R Soc London B, Biol
Sci. 326(1233):119–157. doi:10.1098/rstb.1989.0106. https://doi.org/10.1098/rstb.1989.0106.
Guénard G, Carsten von der Ohe P, Carlisle Walker S, Lek S, Legendre P. 2014. Using
phylogenetic information and chemical properties to predict species tolerances to pesticides. Proc
R Soc B Biol Sci. 281(1789):20133239. doi:10.1098/rspb.2013.3239.
https://doi.org/10.1098/rspb.2013.3239.
Guénard G, Legendre P, Peres-Neto P. 2013. Phylogenetic eigenvector maps: a framework to
model and predict species traits. Methods Ecol Evol. 4(12):1120–1131. doi:10.1111/2041-
210X.12111. https://doi.org/10.1111/2041-210X.12111.
Guénard G, von der Ohe PC, de Zwart D, Legendre P, Lek S. 2011. Using phylogenetic
information to predict species tolerances to toxic chemicals. Ecol Appl. 21(8):3178–3190.
doi:10.1890/10-2242.1. https://doi.org/10.1890/10-2242.1.
188
Guidelines and Methodology Used in the Preparation of Health Effect Assessment Chapters of
the Consent Decree Water Criteria Documents. 1980. Fed Regist. 45(231).
Gunderson AR, Armstrong EJ, Stillman JH. 2016. Multiple Stressors in a Changing World: The
Need for an Improved Perspective on Physiological Responses to the Dynamic Marine
Environment. Ann Rev Mar Sci. 8(1):357–378. doi:10.1146/annurev-marine-122414-033953.
https://doi.org/10.1146/annurev-marine-122414-033953.
Gupta PK. 2018. Chapter 3 - Mechanism of toxicity. In: Gupta PKBT-IT, editor. Illustrated
Toxicology. Academic Press. p. 107–129.
https://www.sciencedirect.com/science/article/pii/B9780128132135000031.
Halpern BS, Frazier M, Potapenko J, Casey KS, Koenig K, Longo C, Lowndes JS, Rockwood
RC, Selig ER, Selkoe KA, et al. 2015. Spatial and temporal changes in cumulative human
impacts on the world’s ocean. Nat Commun. 6(1):7615. doi:10.1038/ncomms8615.
https://doi.org/10.1038/ncomms8615.
Hammond JI, Jones DK, Stephens PR, Relyea RA. 2012. Phylogeny meets ecotoxicology:
evolutionary patterns of sensitivity to a common insecticide. Evol Appl. 5(6):593–606.
doi:10.1111/j.1752-4571.2011.00237.x. https://doi.org/10.1111/j.1752-4571.2011.00237.x.
Harada AE, Burton RS. 2020. Consequences of HSF knockdown on gene expression during the
heat shock response in Tigriopus californicus. J Exp Biol. 223(3):jeb208611.
doi:10.1242/jeb.208611. https://doi.org/10.1242/jeb.208611.
Harrison MD, Jones CE, Dameron CT. 1999. Copper chaperones: function, structure and copperbinding properties. JBIC J Biol Inorg Chem. 4(2):145–153. doi:10.1007/s007750050297.
https://doi.org/10.1007/s007750050297.
Hayes JD, Strange RC. 1995. Invited Commentary Potential Contribution of the Glutathione STransferase Supergene Family to Resistance to Oxidative Stress. Free Radic Res. 22(3):193–207.
doi:10.3109/10715769509147539. https://doi.org/10.3109/10715769509147539.
Hiki K, Iwasaki Y. 2020. Can We Reasonably Predict Chronic Species Sensitivity Distributions
from Acute Species Sensitivity Distributions? Environ Sci Technol. 54(20):13131–13136.
doi:10.1021/acs.est.0c03108. https://doi.org/10.1021/acs.est.0c03108.
Hylton A, Chiari Y, Capellini I, Barron MG, Glaberman S. 2018. Mixed phylogenetic signal in
fish toxicity data across chemical classes. Ecol Appl. 28(3):605–611. doi:10.1002/eap.1698.
https://doi.org/10.1002/eap.1698.
Kamilar JM, Cooper N. 2013. Phylogenetic signal in primate behaviour, ecology and life history.
Philos Trans R Soc B Biol Sci. 368(1618):20120341. doi:10.1098/rstb.2012.0341.
https://doi.org/10.1098/rstb.2012.0341.
Kamilar JM, Muldoon KM. 2010. The Climatic Niche Diversity of Malagasy Primates: A
Phylogenetic Perspective. PLoS One. 5(6):e11073.
https://doi.org/10.1371/journal.pone.0011073.
189
Kang H-M, Lee Jin-Sol, Lee YH, Kim M-S, Park HG, Jeong C-B, Lee Jae-Seong. 2019. Body
size-dependent interspecific tolerance to cadmium and their molecular responses in the marine
rotifer Brachionus spp. Aquat Toxicol. 206:195–202. doi:10.1016/j.aquatox.2018.10.020.
https://www.sciencedirect.com/science/article/pii/S0166445X18308609.
Kazmi SSUH, Wang YYL, Cai Y-E, Wang Z. 2022. Temperature effects in single or combined
with chemicals to the aquatic organisms: An overview of thermo-chemical stress. Ecol Indic.
143:109354. doi:10.1016/j.ecolind.2022.109354.
https://www.sciencedirect.com/science/article/pii/S1470160X22008275.
Kelly MW, Sanford E, Grosberg RK. 2011. Limited potential for adaptation to climate change in
a broadly distributed marine crustacean. Proc R Soc B Biol Sci. 279(1727):349–356.
doi:10.1098/rspb.2011.0542. https://doi.org/10.1098/rspb.2011.0542.
Kidd PM. 1997. Glutathione: Systemic Protectant Aginst Oxidative and Free Radical Damage.
Altern Med Rev. 2(3):155–176.
Kim B-M, Rhee J-S, Jeong C-B, Seo JS, Park GS, Lee Y-M, Lee J-S. 2014. Heavy metals induce
oxidative stress and trigger oxidative stress-mediated heat shock protein (hsp) modulation in the
intertidal copepod Tigriopus japonicus. Comp Biochem Physiol Part C Toxicol Pharmacol.
166:65–74. doi:10.1016/j.cbpc.2014.07.005.
https://www.sciencedirect.com/science/article/pii/S153204561400091X.
Kingsolver JG, Huey R. 2008. Size, temperature, and fitness: Three rules. Evol Ecol Res.
10:251–268.
Kirsten K, Pompermaier A, Koakoski G, Mendonça-Soares S, da Costa RA, Maffi VC, Kreutz
LC, Barcellos LJG. 2021. Acute and chronic stress differently alter the expression of cytokine
and neuronal markers genes in zebrafish brain. Stress. 24(1):107–112.
doi:10.1080/10253890.2020.1724947. https://doi.org/10.1080/10253890.2020.1724947.
Klerks PL, Levinton JS. 1989. Rapid Evolution of Metal Resistance in a Benthic Oligochaete
Inhabiting a Metal-polluted Site. Biol Bull. 176(2):135–141. doi:10.2307/1541580.
https://doi.org/10.2307/1541580.
Kong Q, Lin CG. 2010. Oxidative damage to RNA: mechanisms, consequences, and diseases.
Cell Mol Life Sci. 67(11):1817–1829. doi:10.1007/s00018-010-0277-y.
https://doi.org/10.1007/s00018-010-0277-y.
Kooijman SALM. 1996. An Alternative for NOEC Exists, but the Standard Model Has to Be
Abandoned First. Oikos. 75(2):310–316. doi:10.2307/3546255.
http://www.jstor.org/stable/3546255.
Kovalchuk I, Molinier J, Yao Y, Arkhipov A, Kovalchuk O. 2007. Transcriptome analysis
reveals fundamental differences in plant response to acute and chronic exposure to ionizing
radiation. Mutat Res Mol Mech Mutagen. 624(1):101–113. doi:10.1016/j.mrfmmm.2007.04.009.
https://www.sciencedirect.com/science/article/pii/S0027510707002011.
Kültz D. 2020. Evolution of cellular stress response mechanisms. J Exp Zool Part A Ecol Integr
Physiol. 333(6):359–378. doi:10.1002/jez.2347. https://doi.org/10.1002/jez.2347.
190
LaLone CA, Basu N, Browne P, Edwards SW, Embry M, Sewell F, Hodges G. 2021.
International Consortium to Advance Cross-Species Extrapolation of the Effects of Chemicals in
Regulatory Toxicology. Environ Toxicol Chem. 40(12):3226–3233. doi:10.1002/etc.5214.
https://doi.org/10.1002/etc.5214.
Laskowski R. 1995. Some Good Reasons to Ban the Use of NOEC, LOEC and Related Concepts
in Ecotoxicology. Oikos. 73(1):140–144. doi:10.2307/3545738.
http://www.jstor.org/stable/3545738.
Lee J, Phillips MC, Lobo M, Willett CS. 2021. Tolerance Patterns and Transcriptomic Response
to Extreme and Fluctuating Salinities across Populations of the Intertidal Copepod Tigriopus
californicus. Physiol Biochem Zool. 94(1):50–69. doi:10.1086/712031.
https://doi.org/10.1086/712031.
Legradi JB, Di Paolo C, Kraak MHS, van der Geest HG, Schymanski EL, Williams AJ,
Dingemans MML, Massei R, Brack W, Cousin X, et al. 2018. An ecotoxicological view on
neurotoxicity assessment. Environ Sci Eur. 30(1):46. doi:10.1186/s12302-018-0173-x.
https://doi.org/10.1186/s12302-018-0173-x.
Leong W, Sun PY, Edmands S. 2018. Latitudinal Clines in Temperature and Salinity Tolerance
in Tidepool Copepods. J Hered. 109(1):71–77. doi:10.1093/jhered/esx061.
https://doi.org/10.1093/jhered/esx061.
Lesser MP. 2006. OXIDATIVE STRESS IN MARINE ENVIRONMENTS: Biochemistry and
Physiological Ecology. Annu Rev Physiol. 68(1):253–278.
doi:10.1146/annurev.physiol.68.040104.110001.
https://doi.org/10.1146/annurev.physiol.68.040104.110001.
Letunica I. 2022. phyloT: A phylogenetic tree generator, based on NCBI or GTD taxonomy.
https://phylot.biobyte.de/index.cgi.
Leung KMY, Morritt D, Wheeler JR, Whitehouse P, Sorokin N, Toy R, Holt M, Crane M. 2001.
Can saltwater toxicity be predicted from freshwater data? Mar Pollut Bull. 42(11):1007–1013.
doi:10.1016/S0025-326X(01)00135-7.
Lewis M, Thursby G. 2018. Aquatic plants: Test species sensitivity and minimum data
requirement evaluations for chemical risk assessments and aquatic life criteria development for
the USA. Environ Pollut. 238:270–280. doi:10.1016/j.envpol.2018.03.003.
https://doi.org/10.1016/j.envpol.2018.03.003.
Li AJ, Lai RWS, Zhou G-J, Leung PTY, Zeng EY, Leung KMY. 2023. Joint effects of
temperature and copper exposure on developmental and gene-expression responses of the marine
copepod Tigriopus japonicus. Ecotoxicology. 32(3):336–343. doi:10.1007/s10646-023-02643-w.
https://doi.org/10.1007/s10646-023-02643-w.
Li N, Flanagan BA, Partridge M, Huang EJ, Edmands S. 2020. Sex differences in early
transcriptomic responses to oxidative stress in the copepod Tigriopus californicus. BMC
Genomics. 21(1):759. doi:10.1186/s12864-020-07179-5. https://doi.org/10.1186/s12864-020-
07179-5.
191
Liu Q, Xu Xiaoqun, Zeng J, Huang W, Xu Xudan, Shou L, Chen Q. 2019. Development of
marine water quality criteria for inorganic mercury in China based on the retrievable toxicity
data and a comparison with relevant criteria or guidelines. Ecotoxicology. 28(4):412–421.
doi:10.1007/s10646-019-02032-2. http://dx.doi.org/10.1007/s10646-019-02032-2.
Love MI, Huber W, Anders S. 2014. Moderated estimation of fold change and dispersion for
RNA-seq data with DESeq2. Genome Biol. 15(12):550. doi:10.1186/s13059-014-0550-8.
https://doi.org/10.1186/s13059-014-0550-8.
Lushchak VI. 2011. Environmentally induced oxidative stress in aquatic animals. Aquat Toxicol.
101(1):13–30. doi:10.1016/j.aquatox.2010.10.006.
https://www.sciencedirect.com/science/article/pii/S0166445X1000411X.
McKinlay R, Plant JA, Bell JNB, Voulvoulis N. 2008. Endocrine disrupting pesticides:
Implications for risk assessment. Environ Int. 34(2):168–183. doi:10.1016/j.envint.2007.07.013.
https://www.sciencedirect.com/science/article/pii/S0160412007001444.
McRae CJ, Keshavmurthy S, Meng P-J, Rosset SL, Huang W-B, Chen CA, Fan T-Y, Côté IM.
2022. Variable responses to chronic and acute elevated temperature of three coral species from
reefs with distinct thermal regimes. Mar Biol. 169(7):97. doi:10.1007/s00227-022-04071-6.
https://doi.org/10.1007/s00227-022-04071-6.
Merlot C. 2010. Computational toxicology—a tool for early safety evaluation. Drug Discov
Today. 15(1):16–22. doi:https://doi.org/10.1016/j.drudis.2009.09.010.
https://www.sciencedirect.com/science/article/pii/S1359644609003353.
Moore DRJ, Priest CD, Galic N, Brain RA, Rodney SI. 2020. Correcting for Phylogenetic
Autocorrelation in Species Sensitivity Distributions. Integr Environ Assess Manag. 16(1):53–65.
doi:10.1002/ieam.4207. https://doi.org/10.1002/ieam.4207.
Morgan M, Sheperd L. 2024. AnnotationHub: Client to access AnnotationHub resources.
doi:10.18129/B9.bioc.AnnotationHub.
Münkemüller T, Lavergne S, Bzeznik B, Dray S, Jombart T, Schiffers K, Thuiller W. 2012. How
to measure and test phylogenetic signal. Methods Ecol Evol. 3(4):743–756. doi:10.1111/j.2041-
210X.2012.00196.x. https://doi.org/10.1111/j.2041-210X.2012.00196.x.
Nacci DE, Champlin D, Jayaraman S. 2010. Adaptation of the Estuarine Fish Fundulus
heteroclitus (Atlantic Killifish) to Polychlorinated Biphenyls (PCBs). Estuaries and Coasts.
33(4):853–864. doi:10.1007/s12237-009-9257-6. https://doi.org/10.1007/s12237-009-9257-6.
Neumann D, Lichtenberger O, Günther D, Tschiersch K, Nover L. 1994. Heat-shock proteins
induce heavy-metal tolerance in higher plants. Planta. 194(3):360–367.
doi:10.1007/BF00197536. https://doi.org/10.1007/BF00197536.
Niskanen EA, Malinen M, Sutinen P, Toropainen S, Paakinaho V, Vihervaara A, Joutsen J,
Kaikkonen MU, Sistonen L, Palvimo JJ. 2015. Global SUMOylation on active chromatin is an
acute heat stress response restricting transcription. Genome Biol. 16(1):153. doi:10.1186/s13059-
015-0717-y. https://doi.org/10.1186/s13059-015-0717-y.
192
Novacek MJ, Cleland EE. 2001. The current biodiversity extinction event: Scenarios for
mitigation and recovery. Proc Natl Acad Sci. 98(10):5466 LP – 5470.
doi:10.1073/pnas.091093698. http://www.pnas.org/content/98/10/5466.abstract.
Oksanen AJ, Kindt R, Legendre P, Hara BO, Simpson GL, Stevens MHH, Wagner H. 2007. The
vegan Package. Community Ecol Packag.(May 2014):631–637.
Olker JH, Elonen CM, Pilli A, Anderson A, Kinziger B, Erickson S, Skopinski M, Pomplun A,
LaLone CA, Russom CL, et al. 2022. The ECOTOXicology Knowledgebase: A Curated
Database of Ecologically Relevant Toxicity Tests to Support Environmental Research and Risk
Assessment. Environ Toxicol Chem. 41(6):1520–1539. doi:10.1002/etc.5324.
https://doi.org/10.1002/etc.5324.
Orme D, Freckleton R, Thomas G, Petzoldt T, Fritz S, Isaac N, Pearse W. 2018. caper:
Comparative Analyses of Phylogenetics and Evolution in R, R package version 1.0.1.
https://cran.r-project.org/package=caper.
Ownby DR, Newman MC, Mulvey M, Vogelbein WK, Unger MA, Arzayus LF. 2002. Fish
(Fundulus heteroclitus) populations with different exposure histories differ in tolerance of
creosote-contaminated sediments. Environ Toxicol Chem. 21(9):1897–1902.
doi:https://doi.org/10.1002/etc.5620210917. 10.1002/etc.5620210917.
Pagel M. 1998. Inferring evolutionary processes from phylogenies. Zool Scr. 26(4):331–348.
Pagel M. 1999. Inferring the historical patterns of biological evolution. Nature. 401(6756):877–
884. doi:10.1038/44766. https://doi.org/10.1038/44766.
Paradis E, Schliep K. 2019. ape 5.0: an environment for modern phylogenetics and evolutionary
analyses in R. Bioinformatics. 35:526–528.
Pavlaki MD, Araújo MJ, Cardoso DN, Silva ARR, Cruz A, Mendo S, Soares AMVM, Calado R,
Loureiro S. 2016. Ecotoxicity and genotoxicity of cadmium in different marine trophic levels.
Environ Pollut. 215:203–212. doi:10.1016/j.envpol.2016.05.010.
Peralta-Maraver I, Rezende EL. 2021. Heat tolerance in ectotherms scales predictably with body
size. Nat Clim Chang. 11(1):58–63. doi:10.1038/s41558-020-00938-y.
https://doi.org/10.1038/s41558-020-00938-y.
Pham AN, Xing G, Miller CJ, Waite TD. 2013. Fenton-like copper redox chemistry revisited:
Hydrogen peroxide and superoxide mediation of copper-catalyzed oxidant production. J Catal.
301:54–64. doi:10.1016/j.jcat.2013.01.025.
https://www.sciencedirect.com/science/article/pii/S0021951713000377.
Pinheiro JPS, Windsor FM, Wilson RW, Tyler CR. 2021. Global variation in freshwater physicochemistry and its influence on chemical toxicity in aquatic wildlife. Biol Rev. 96(4):1528–1546.
doi:https://doi.org/10.1111/brv.12711. https://doi.org/10.1111/brv.12711.
Posthuma L, Suter GW, Traas TP. 2001. Species Sensitivity Distributions in Ecotoxicology.
Lewis Publishers.
193
Putri GH, Anders S, Pyl PT, Pimanda JE, Zanini F. 2022. Analysing high-throughput sequencing
data in Python with HTSeq 2.0. Bioinformatics. 38(10):2943–2945.
doi:10.1093/bioinformatics/btac166. https://doi.org/10.1093/bioinformatics/btac166.
Raimondo S, Jackson CR, Barron MG. 2010. Influence of Taxonomic Relatedness and Chemical
Mode of Action in Acute Interspecies Estimation Models for Aquatic Species. Environ Sci
Technol. 44(19):7711–7716. doi:10.1021/es101630b. https://doi.org/10.1021/es101630b.
Raimondo S, Vivian DN, Barron MG. 2010. Web-based Interspecies Correlation Estimation
(Web-ICE) for Acute Toxicity : User Manual. (January).
Ramegowda V, Da Costa MVJ, Harihar S, Karaba NN, Sreeman SM. 2020. Chapter 17 - Abiotic
and biotic stress interactions in plants: A cross-tolerance perspective. In: Hossain MA, Liu F,
Burritt DJ, Fujita M, Huang BBT-P-MS and C-ST in CP, editors. Priming-mediated stress and
cross-tolerance in crop plants. Academic Press. p. 267–302.
https://www.sciencedirect.com/science/article/pii/B9780128178928000179.
Rawson PD, Burton RS. 2006. Molecular Evolution at the Cytochrome Oxidase Subunit 2 Gene
Among Divergent Populations of the Intertidal Copepod, Tigriopus californicus. J Mol Evol.
62(6):753–764. doi:10.1007/s00239-005-0074-7. https://doi.org/10.1007/s00239-005-0074-7.
Revell LJ. 2010. Phylogenetic signal and linear regression on species data. Methods Ecol Evol.
1(4):319–329. doi:https://doi.org/10.1111/j.2041-210X.2010.00044.x.
https://doi.org/10.1111/j.2041-210X.2010.00044.x.
Revell LJ. 2012. phytools: an R package for phylogenetic comparative biology (and other
things). Methods Ecol Evol. 3(2):217–223. doi:10.1111/j.2041-210X.2011.00169.x.
https://doi.org/10.1111/j.2041-210X.2011.00169.x.
Revell LJ, Harmon LJ, Collar DC. 2008. Phylogenetic Signal, Evolutionary Process, and Rate.
Syst Biol. 57(4):591–601. doi:10.1080/10635150802302427.
https://doi.org/10.1080/10635150802302427.
Rhee J-S, Raisuddin S, Lee K-W, Seo JS, Ki J-S, Kim I-C, Park HG, Lee J-S. 2009. Heat shock
protein (Hsp) gene responses of the intertidal copepod Tigriopus japonicus to environmental
toxicants. Comp Biochem Physiol Part C Toxicol Pharmacol. 149(1):104–112.
doi:10.1016/j.cbpc.2008.07.009.
https://www.sciencedirect.com/science/article/pii/S1532045608001415.
Richard J, Ann R, J. DD, Keith H, Matthew M, Robert K, Vicki D, Tala H, Todd H, Philip S, et
al. 2009. The Toxicity Data Landscape for Environmental Chemicals. Environ Health Perspect.
117(5):685–695. doi:10.1289/ehp.0800168. https://doi.org/10.1289/ehp.0800168.
Rodgers EM, Gomez Isaza DF. 2021. Harnessing the potential of cross-protection stressor
interactions for conservation: a review. Conserv Physiol. 9(1). doi:10.1093/conphys/coab037.
https://doi.org/10.1093/conphys/coab037.
Rodgers EM, Gomez Isaza DF. 2022. Stress history affects heat tolerance in an aquatic
ectotherm (Chinook salmon, Oncorhynchus tshawytscha). J Therm Biol. 106:103252.
doi:10.1016/j.jtherbio.2022.103252.
https://www.sciencedirect.com/science/article/pii/S0306456522000675.
194
Rodgers EM, Gomez Isaza DF. 2023. The mechanistic basis and adaptive significance of crosstolerance: a ‘pre-adaptation’ to a changing world? J Exp Biol. 226(11):jeb245644.
doi:10.1242/jeb.245644. https://doi.org/10.1242/jeb.245644.
Schäfer RB, Jackson M, Juvigny-Khenafou N, Osakpolor SE, Posthuma L, Schneeweiss A,
Spaak J, Vinebrooke R. 2023. Chemical Mixtures and Multiple Stressors: Same but Different?
Environ Toxicol Chem. 42(9):1915–1936. doi:10.1002/etc.5629.
https://doi.org/10.1002/etc.5629.
Schlekat CE, Van Genderen E, De Schamphelaere KAC, Antunes PMC, Rogevich EC,
Stubblefield WA. 2010. Cross-species extrapolation of chronic nickel Biotic Ligand Models. Sci
Total Environ. 408(24):6148–6157. doi:https://doi.org/10.1016/j.scitotenv.2010.09.012.
https://www.sciencedirect.com/science/article/pii/S0048969710009733.
Schoville SD, Barreto FS, Moy GW, Wolff A, Burton RS. 2012. Investigating the molecular
basis of local adaptation to thermal stress: population differences in gene expression across the
transcriptome of the copepod Tigriopus californicus. BMC Evol Biol. 12(1):170.
doi:10.1186/1471-2148-12-170. https://doi.org/10.1186/1471-2148-12-170.
Schulte PM. 2014. What is environmental stress? Insights from fish living in a variable
environment. Davies SA, Dow JAT, Lukowiak K, editors. J Exp Biol. 217(1):23–34.
doi:10.1242/jeb.089722. https://doi.org/10.1242/jeb.089722.
Seegert G, Fava J, Cumbia P. 1985. How Representative Are the Data Sets Used to Derive
National Water Quality Criteria ? Aquat Toxicol Hazard Assess Seventh Symp. 854:527–537.
doi:10.1520/STP36288S.
SgrÒ CM, Overgaard J, Kristensen TN, Mitchell KA, Cockerell FE, Hoffmann AA. 2010. A
comprehensive assessment of geographic variation in heat tolerance and hardening capacity in
populations of Drosophila melanogaster from eastern Australia. J Evol Biol. 23(11):2484–2493.
doi:10.1111/j.1420-9101.2010.02110.x. https://doi.org/10.1111/j.1420-9101.2010.02110.x.
Shackelford RE, Kaufmann WK, Paules RS. 2000. Oxidative stress and cell cycle checkpoint
function11Both Drs. Paules and Kaufmann received their doctoral degrees in Experimental
Pathology from the University of North Carolina at Chapel Hill School of Medicine (in 1984 and
1979, respectively). Free Radic Biol Med. 28(9):1387–1404. doi:10.1016/S0891-5849(00)00224-
0. https://www.sciencedirect.com/science/article/pii/S0891584900002240.
Shahidul Islam M, Tanaka M. 2004. Impacts of pollution on coastal and marine ecosystems
including coastal and marine fisheries and approach for management: a review and synthesis.
Mar Pollut Bull. 48(7):624–649. doi:10.1016/j.marpolbul.2003.12.004.
https://www.sciencedirect.com/science/article/pii/S0025326X03005459.
Shannon C. 1948. A Mathematical Theory of Communication. Bell Syst Tech J. 27:379–423.
Sokolova IM. 2013. Energy-Limited Tolerance to Stress as a Conceptual Framework to Integrate
the Effects of Multiple Stressors. Integr Comp Biol. 53(4):597–608. doi:10.1093/icb/ict028.
https://doi.org/10.1093/icb/ict028.
195
Solomon KR, Baker DB, Richards RP, Dixon KR, Klaine SJ, La Point TW, Kendall RJ,
Weisskopf CP, Giddings JM, Giesy JP, et al. 1996. Ecological risk assessment of atrazine in
North American surface waters. Environ Toxicol Chem. 15(1):31–76.
doi:https://doi.org/10.1002/etc.5620150105. https://doi.org/10.1002/etc.5620150105.
Song W, Chen S, Liu J, Chen L, Song N, Li N, Liu B. 2015. Variation of Cd concentration in
various rice cultivars and derivation of cadmium toxicity thresholds for paddy soil by speciessensitivity distribution. J Integr Agric. 14(9):1845–1854. doi:https://doi.org/10.1016/S2095-
3119(14)60926-6. https://www.sciencedirect.com/science/article/pii/S2095311914609266.
Stephan C, Mount D, Hansen D, Gentile J, Chapman G, Brungs W. 1985. Guidelines for
Deriving Numerical National Water Quality Criteria for the Protection of Aquatic Organisms and
Their Uses. US Environmental Protection Agency.
Stephan CE. 1984. Are the “National Guidelines” Based on Sound Judgements? Aquat Toxicol
Hazard Assess Seventh Symp.
Stone GN, Nee S, Felsenstein J. 2011. Controlling for non-independence in comparative analysis
of patterns across populations within species. Philos Trans R Soc B Biol Sci. 366(1569):1410–
1424. doi:10.1098/rstb.2010.0311. https://doi.org/10.1098/rstb.2010.0311.
Sun H, Xia M, Austin CP, Huang R. 2012. Paradigm Shift in Toxicity Testing and Modeling.
AAPS J. 14(3):473–480. doi:10.1208/s12248-012-9358-1. https://doi.org/10.1208/s12248-012-
9358-1.
Sun PY, Foley HB, Bao VWW, Leung KMY, Edmands S. 2015. Variation in tolerance to
common marine pollutants among different populations in two species of the marine copepod
Tigriopus. Environ Sci Pollut Res.:16143–16152. doi:10.1007/s11356-015-4846-3.
http://dx.doi.org/10.1007/s11356-015-4846-3.
Sunday JM, Bates AE, Dulvy NK. 2010. Global analysis of thermal tolerance and latitude in
ectotherms. Proc R Soc B Biol Sci. 278(1713):1823–1830. doi:10.1098/rspb.2010.1295.
https://doi.org/10.1098/rspb.2010.1295.
Symonds MRE, Blomberg SP. 2014. A Primer on Phylogenetic Generalised Least Squares BT -
Modern Phylogenetic Comparative Methods and Their Application in Evolutionary Biology:
Concepts and Practice. In: Garamszegi LZ, editor. Modern Phylogenetic Comparative Methods
and Their Application in Evolutionary Biology: Concepts and Practice. Berlin, Heidelberg:
Springer Berlin Heidelberg. p. 105–130. https://doi.org/10.1007/978-3-662-43550-2_5.
Tapia L, González-Agüero M, Cisternas MF, Suazo M, Cambiazo V, Uauy R, González M.
2004. Metallothionein is crucial for safe intracellular copper storage and cell survival at normal
and supra-physiological exposure levels. Biochem J. 378(Pt 2):617–624.
doi:10.1042/BJ20031174.
R Core Team. 2019. R: A language and environment for statistical computing. https://www.rproject.org/.
R Core Team. 2024. R: A language and environment for statistical computing. https://www.rproject.org/.
196
Tenbrook PL, Palumbo AJ, Fojut TL, Hann P, Karkoski J, Tjeerdema RS. 2010. The University
of California-Davis methodology for deriving aquatic life pesticide water quality criteria. Rev
Environ Contam Toxicol. 209(August):1–172. doi:10.1007/978-1-4419-6883-8_1.
Therneau TM. 2023. A Package for Survival Analysis in R.
Thorley J, Schwarz C. 2018. ssdtools : An R package to fit Species Sensitivity Distributions. J
Open Source Softw. 3(31):2–3. doi:10.1007/b97636.
Tilman D, Clark M, Williams DR, Kimmel K, Polasky S, Packer C. 2017. Future threats to
biodiversity and pathways to their prevention. Nature. 546(7656):73–81.
doi:10.1038/nature22900. https://doi.org/10.1038/nature22900.
Todgham AE, Schulte PM, Iwama GK. 2005. Cross‐Tolerance in the Tidepool Sculpin: The
Role of Heat Shock Proteins. Physiol Biochem Zool. 78(2):133–144. doi:10.1086/425205.
https://doi.org/10.1086/425205.
Todgham AE, Stillman JH. 2013. Physiological Responses to Shifts in Multiple Environmental
Stressors: Relevance in a Changing World. Integr Comp Biol. 53(4):539–544.
doi:10.1093/icb/ict086. https://doi.org/10.1093/icb/ict086.
U.S. Environmental Protection Agency. 2003. Ambient Aquatic Life Water Quality Criteria for
Atrazine - Revised Draft.
U.S. Environmental Protection Agency. 2013. Aquatic Life Ambient Water Quality Criteria for
Ammonia - Freshwater (2013) Fact Sheet.
U.S. Environmental Protection Agency. 2021. National Recommended Water Quality Criteria -
Aquatic Life Criteria Table. https://www.epa.gov/wqc/national-recommended-water-qualitycriteria-aquatic-life-criteria-table.
Uren Webster TM, Rodriguez-Barreto D, Martin SAM, Van Oosterhout C, Orozco-terWengel P,
Cable J, Hamilton A, Garcia De Leaniz C, Consuegra S. 2018. Contrasting effects of acute and
chronic stress on the transcriptome, epigenome, and immune response of Atlantic salmon.
Epigenetics. 13(12):1191–1207. doi:10.1080/15592294.2018.1554520.
https://doi.org/10.1080/15592294.2018.1554520.
Valenti TW, Cherry DS, Currie RJ, Neves RJ, Jones JW, Mair R, Kane CM. 2006. Chlorine
toxicity to early life stages of freshwater mussels (Bivalvia: Unionidae). Environ Toxicol Chem.
25(9):2512–2518. doi:https://doi.org/10.1897/05-527R1.1. https://doi.org/10.1897/05-527R1.1.
Vandegehuchte MB, Janssen CR. 2014. Epigenetics in an ecotoxicological context. Mutat Res
Toxicol Environ Mutagen. 764–765:36–45. doi:10.1016/j.mrgentox.2013.08.008.
https://www.sciencedirect.com/science/article/pii/S1383571813002325.
Venables WN, Ripley BD. 2002. Modern Applied Statistics with S. Fourth. New York: Springer.
https://www.stats.ox.ac.uk/pub/MASS4/.
197
Wang Z, Kwok KWH, Lui GCS, Zhou G-J, Lee J-S, Lam MHW, Leung KMY. 2014. The
difference between temperate and tropical saltwater species’ acute sensitivity to chemicals is
relatively small. Chemosphere. 105:31–43.
doi:https://doi.org/10.1016/j.chemosphere.2013.10.066.
https://www.sciencedirect.com/science/article/pii/S0045653513015002.
Warne Ms, Batley G, van Dam R, Chapman J, Fox D, Hickey C, Stauber J. 2015. Revised
Method for Deriving Australian and New Zealand Water Quality Guideline Values for
Toxicants. (August).
Warne MSJ, van Dam R. 2008. NOEC and LOEC Data Should No Longer Be Generated or
Used. Australas J Ecotoxicol. 14(1):1–5. doi:10.3316/informit.665706831945866.
Wheeler J R, Grist EPM, Leung KMY, Morritt D, Crane M. 2002. Species sensitivity
distributions: data and model choice. Mar Pollut Bull. 45(1):192–202. doi:10.1016/S0025-
326X(01)00327-7. https://www.sciencedirect.com/science/article/pii/S0025326X01003277.
Wheeler James R, Leung KMY, Morritt D, Sorokin N, Rogers H, Toy R, Holt M, Whitehouse P,
Crane M. 2002. Freshwater to saltwater toxicity extrapolation using species sensitivity
distributions. Environ Toxicol Chem. 21(11):2459–2467.
doi:https://doi.org/10.1002/etc.5620211127. https://doi.org/10.1002/etc.5620211127.
Wheeler MW, Park RM, Bailer AJ. 2006. Comparing median lethal concentration values using
confidence interval overlap or ratio tests. Environ Toxicol Chem. 25(5):1441–1444.
doi:10.1897/05-320R.1. https://doi.org/10.1897/05-320R.1.
Whitehead A, Clark BW, Reid NM, Hahn ME, Nacci D. 2017. When evolution is the solution to
pollution: Key principles, and lessons from rapid repeated adaptation of killifish (Fundulus
heteroclitus) populations. Evol Appl. 10(8):762–783. doi:10.1111/eva.12470.
https://doi.org/10.1111/eva.12470.
Wilcut L, Lehmann W, Elias M. 2015. Updating EPA’s Guidelines for Deriving National
Recommended Water Quality Criteria. Invit Expert Meet Revising US EPA’s Guidel Deriving
Aquat Life Criteria.
Willett CS. 2010. POTENTIAL FITNESS TRADE-OFFS FOR THERMAL TOLERANCE IN
THE INTERTIDAL COPEPOD TIGRIOPUS CALIFORNICUS. Evolution (N Y). 64(9):2521–
2534. doi:10.1111/j.1558-5646.2010.01008.x. https://doi.org/10.1111/j.1558-5646.2010.01008.x.
Willett CS, Ladner JT. 2009. Investigations of fine-scale phylogeography in Tigriopus
californicus reveal historical patterns of population divergence. BMC Evol Biol. 9(1):139.
doi:10.1186/1471-2148-9-139. https://doi.org/10.1186/1471-2148-9-139.
Wright RM, Aglyamova G V, Meyer E, Matz M V. 2015. Gene expression associated with white
syndromes in a reef building coral, Acropora hyacinthus. BMC Genomics. 16(1):371.
doi:10.1186/s12864-015-1540-2. https://doi.org/10.1186/s12864-015-1540-2.
Wu J, Yan Z, Liu Z, Liu J, Liang F, Wang X, Wang W. 2015. Development of water quality
criteria for phenanthrene and comparison of the sensitivity between native and non-native
species. Environ Pollut. 196:141–146. doi:10.1016/j.envpol.2014.09.024.
http://dx.doi.org/10.1016/j.envpol.2014.09.024.
198
Yim JH, Kim KW, Kim SD. 2006. Effect of hardness on acute toxicity of metal mixtures using
Daphnia magna: Prediction of acid mine drainage toxicity. J Hazard Mater. 138(1):16–21.
doi:https://doi.org/10.1016/j.jhazmat.2005.11.107.
https://www.sciencedirect.com/science/article/pii/S0304389406004730.
de Zwart D. 2001. Observed regularities in species sensitivity distributions for aquatic species.
In: Species sensitivity distributions in ecotoxicology. CRC Press. p. 157–178.
Abstract (if available)
Abstract
Chemical pollution represents one of the most potent anthropogenic forces of change in the aquatic environment. Precise understanding of the toxic effects of chemicals on aquatic organisms is therefore crucial to pollution mitigation efforts, however, the role of evolutionary processes in shaping chemical sensitivity has been understudied. This dissertation seeks to address this gap in knowledge by integrating theory and methods from evolutionary biology into two distinct areas of research in aquatic ecotoxicology: the development of pollution regulations and the study of multiple stressors. Firstly, I performed two broad meta-analyses to evaluate the taxonomic composition of the toxicity data used to derive water quality criteria and to quantify the extent to which phylogeny determines chemical sensitivity. I then utilized in vivo stress experiments and transcriptomic sequencing to evaluate the interpopulation variation in copper tolerance in the intertidal copepod Tigriopus californicus and to investigate whether the copper-tolerant phenotype is an exaptation derived from the heat stress response mechanism. Toxicity data were found to be lacking in biological diversity, although the taxonomic composition of datasets did not quantitatively affect water quality criteria in most cases. Broad taxonomic patterns of sensitivity were also evident in these data, although phylogeny was a significant predictor of sensitivity for only a minority of chemicals. Additionally, I found that copper tolerance in T. californicus populations was positively correlated with heat tolerance, and that there was overlap between the gene expression responses for both stressors that could confer fitness benefits during multiple stressor exposure. Overall, this work advances the integration of evolutionary biology and ecotoxicology and highlights the importance of evolutionary processes in both a regulatory and ecological context.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Transgenerational inheritance of thermal tolerance in two coral species in the Florida Keys
PDF
Transcriptional and morphological impacts of copper on Mytilus californianus larval development in current and future ocean conditions
PDF
The evolution of pollution tolerance in the marine copepod Tigriopus
Asset Metadata
Creator
Coleman, Alice Louise
(author)
Core Title
Application of evolutionary theory and methods to aquatic ecotoxicology
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Biology (Marine Biology and Biological Oceanography)
Degree Conferral Date
2024-12
Publication Date
01/10/2025
Defense Date
11/22/2024
Publisher
Los Angeles, California
(original),
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
chemical pollution,chemical tolerance,ecotoxicology,evolutionary biology,Marine biology,OAI-PMH Harvest,phylogenetics,thermal tolerance,Tigriopus californicus,toxicity data,water quality criteria
Format
theses
(aat)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Edmands, Suzanne (
committee chair
), Ehrenreich, Ian (
committee member
), Guzman, Laura Melissa (
committee member
), Kenkel, Carly Danielle (
committee member
)
Creator Email
a.coleman1914@gmail.com,alicecol@usc.edu
Unique identifier
UC11399F8ND
Identifier
etd-ColemanAli-13737.pdf (filename)
Legacy Identifier
etd-ColemanAli-13737
Document Type
Dissertation
Format
theses (aat)
Rights
Coleman, Alice Louise
Internet Media Type
application/pdf
Type
texts
Source
20250112-usctheses-batch-1233
(batch),
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Repository Email
cisadmin@lib.usc.edu
Tags
chemical pollution
chemical tolerance
ecotoxicology
evolutionary biology
phylogenetics
thermal tolerance
Tigriopus californicus
toxicity data
water quality criteria