TRANSMISSIBILITY AND LOCALIZATION OF TETRODOTOXIN IN THE ROUGHSKINNED NEWT, TARICHA GRANULOSA
By
Sarah Wegener

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Integrative Biology – Master of Science
2017

ABSTRACT
TRANSMISSIBILITY AND LOCALIZATION OF TETRODOTOXIN IN THE ROUGHSKINNED NEWT, TARICHA GRANULOSA
By
Sarah Wegener
Tetrodotoxin (TTX) is a powerful neurotoxin that prevents the propagation of action
potentials, leading to paralysis and sometimes death in nearly all animals. However, a diverse
group of marine and freshwater animals possess TTX, which they use for offense, defense, and
communication. One of most studied TTX-mediated interactions is the predator-prey arms race
between the rough-skinned newt (Taricha granulosa) and common garter snake (Thamnophis
sirtalis). Variation in toxicity among populations of newts matched by TTX-resistance in
predatory snakes has captured the focus of much research centered on the hypothesis that the
arms race is the sole driver of variation. Nevertheless, recent studies suggest a more complex
dynamic. Explanations of the dramatic variation in TTX among different populations of newts
can only be constructed once fundamental questions about the origin, function, and transmission
of TTX in newts have been more thoroughly explored. In this study, I took two approaches to
address the origin, function, and transmission of TTX: 1) a cohabitation experiment in which I
paired toxic and non-toxic newts to test whether toxicity can be acquired through contact, and 2)
an experiment to determine the distribution and concentration of TTX in different tissues. The
cohabitation experiment revealed no detectable change in the toxicity of non-toxic male newts,
suggesting that a physical or physiological impediment prevents non-toxic newts from becoming
toxic. The tissue toxicity experiment demonstrated that TTX is present throughout the body in
structurally and functionally diverse tissues, which has many implications for the involvement of
TTX in communication and reproduction in addition to defense.

ACKNOWLEDGEMENTS
I would like to express my sincerest gratitude to my advisor, Dr. Heather Eisthen, for all
of her encouragement, patience, and insistence on high quality work. I am grateful for her
invaluable feedback, advice, and guidance and the many opportunities she gave me to explore
my interests. Her willingness to flexible with me while I pursued my teaching certification and
the academic, financial, and emotional support she provided was integral to completing my
graduate studies and this thesis.
Thank you to Dr. Fred Dyer, Dr. Jason Gallant, and Dr. Ashlee Rowe for being members
of my guidance committee. Their insightful comments and advice and willingness to help
throughout my graduate career shaped the research projects I pursued. I am especially grateful to
Jason and Ashlee for participating in my defense and being so accommodating with scheduling
it.
I appreciate the gracious support of Ashlee and her husband, Dr. Matt Rowe, when I was
preparing for fieldwork. Their equipment and advice significantly contributed to my successful
field season. Without Matt, I would have really struggled to set up a tent.
I could not have completed the research in this thesis without the help of many
collaborators and lab mates to whom I am very grateful. Chen Zhang at MSU’s Mass Spec Core
spent hours teaching me how to operate the equipment, designing and optimizing an LC-MS
protocol, explaining how to interpret and convert the output, and helping me troubleshoot. My
lab mate, Patric Vaelli, collaborated with me and engaged in stimulating conversation that helped
me develop and refine my ideas. Richard Middlebrook provided immense assistance with sample
collection and processing in the lab and helped me stay organized by labeling countless tubes.
All of the members of the Eisthen lab spent extra time to help maintain the animals in my
iii 
 

experiments and were careful to prevent cross contamination when performing animal care
duties. Discussions during lab meetings were always fruitful and enjoyable. Gary Bucciarelli
taught me how to properly biopsy newts. Emma Coddington and Joanna Hernandez played a
vital role during my time in the field: Emma provided her expertise and hospitality and Joanna
helped me learn how to camp and was an excellent newt-catching partner while I was in Oregon.
The cohabitation experiment would not have been possible without Yannik Roell, who caught
newts in Phillips Farm Pond.
Last, but not least, I would like to thank my friends and family who are always willing to
listen to me and provide me with support and encouragement.

iv 
 

TABLE OF CONTENTS

LIST OF TABLES

vi

LIST OF FIGURES

vii

KEY TO ABBREVIATIONS

viii

INTRODUCTION

1

METHODS
Subjects
Subject Collection
Animal Maintenance
Set-up and Sample Collection for the Cohabitation Experiment
Sample Collection for the Tissue Toxicity Experiment
Skin Sampling and Tissue Processing
Toxin Extraction from Tissue
Solid Phase Extraction of Samples
TTX Quantification
Statistical Analysis

6
6
6
7
7
8
9
9
10
10
11

RESULTS
Cohabitation Experiment
Tissue Toxicity Experiment

12
12
13

DISCUSSION
Non-toxic Newts in the Cohabitation Study
Toxic Newts in the Cohabitation Experiment
Tissue Toxicity Experiment

16
16
19
21

CONCLUSION

30

APPENDICES
APPENDIX A: TABLES
APPENDIX B: FIGURES

31
32
44

REFERENCES

53

v 
 

LIST OF TABLES
Table 1. Toxicity of Skin

33

Table 2. Toxicity of Blood

34

Table 3. Toxicity of Feces

35

Table 4. Toxicity of Female Reproductive Tissues

36

Table 5. Toxicity of Male Reproductive Tissues

37

Table 6. Toxicity of Fat Body

38

Table 7. Toxicity of Digestive Tissues

39

Table 8. Toxicity of Excitable Tissues

40

Table 9. Number of Tissue Samples from Toxic Newts with No Quantifiable TTX

41

Table 10. Mean Toxicity, Body Condition, Mass, and SVL of Toxic Newts

42

Table 11. Regression Analysis of Toxicity on Body Condition, Mass, and SVL

43

vi 
 

LIST OF FIGURES
Figure 1. Aerial and Ground View of Sandpiper Pond.

45

Figure 2. Aerial View of Elk River.

46

Figure 3. Aerial View of Phillips Farm Pond.

47

Figure 4. Relative Locations of Field Sites.

48

Figure 5. Toxicity of Philips Farm Pond Male Newts in the Cohabitation
Experiment.

49

Figure 6. Toxicity of All Sandpiper Pond Male Newts in the Cohabitation
Experiment.

50

Figure 7. Toxicity of Treatment and Control Sandpiper Pond Male Newts in the Cohabitation
Experiment.
51
Figure 8. Correlation between SVL and Toxicity in Toxic Newts.

vii 
 

52

KEY TO ABBREVIATIONS
LC-MS/MS

Liquid chromatography-tandem mass spectrometry

Nav

Voltage-gated sodium channel

PA

Pyrrolizidine alkaloids

SVL

Snout-vent length

TTX

Tetrodotoxin

viii 
 

INTRODUCTION
Tetrodotoxin (TTX) is a highly potent neurotoxin that binds to voltage-gated sodium
channels (Navs), preventing the generation of action potentials (Geffeney et al., 2002; Lee and
Ruben, 2008). Most animals are susceptible to the effects of TTX even at very low
concentrations (e.g., the LD50 in humans is 10.2µg/kg) (Hanifin, 2010); however, a diverse group
of marine and freshwater animals including some Platyhelminthes, gastropods, cephalopods,
annelids, echinoderms, teleosts, and amphibians have evolved adaptations that enable them to
possess TTX and, in some instances, consume toxic prey (Miyazawa and Noguchi, 2001;
Moczydlowski, 2013; Noguchi, et al., 2006). In marine species, TTX appears to originate as a
bacterially-produced toxin that accumulates in animal hosts or is passed up the food chain, but
the origin of TTX in freshwater animals, and newts in particular, is contested. Competing
theories of endogenous origins and bacterial symbionts dominate the literature (Chau et al.,
2011; Hanifin, 2010; Williams, 2010). Some authors assert that endogenous production is
supported by regeneration of TTX after secretion (Cardall et al., 2004) and retention of toxicity
in long-term captivity despite being fed non-toxic food (Gall et al., 2012; Hanifin et al., 2002),
but the complicated potential biosynthetic pathways make endogenous origins unlikely (Chau et
al., 2011; Hague et al., 2016). Production by bacterial symbionts seems more plausible based on
observations in marine species. Bacteria cultured from newt skin in our lab produce TTX in vitro
(PM Vaelli, personal communication), strengthening the case for a bacterial origin in freshwater
animals.
The roles of toxicity in the ecology of TTX-bearing organisms include communication,
offense, and defense (Hanifin, 2010; Williams, 2010; Moczydlowski, 2013). TTX acts as a maleattracting pheromone released by female pufferfish (Takifugu niphobles) (Matsumura, 1995), an

1
 

attractant in multiple species of toxic snails (Hwang et al., 2004), and a warning cue indicating
the presence of cannibalistic adults for juvenile California newts (Taricha torosa) (Zimmer et al.,
2006). Blue-ringed octopus (Hapalochlaena maculosa), as well as many species of arrowworms,
flatworms, ribbon worms, and snails, sequester TTX and use it to envenomate their prey
(Williams, 2010). The most notorious function of TTX is defense, which has been extensively
studied in pufferfish and newts (Brodie, 1968; Kodama et al., 1986; Brodie III et al., 2005;
Williams, 2010; Williams et al., 2010; Itoi et al., 2014). Of particular interest is the role of TTX
in predator-prey arms races between sympatric populations of newts (Taricha spp.) and garter
snakes (Thamnophis sp.) (Brodie III et al., 2005). The most well studied of these pairings is that
of the rough-skinned newt (Taricha granulosa, hereafter newts) and common garter snake
(Thamnophis sirtalis, hereafter snakes). Newt toxicity varies widely throughout their range,
which extends from southern California to the Alaskan panhandle (Hanifin et al., 2008; Hague et
al., 2016). Snakes, which have evolved varying degrees of resistance to TTX, live in sympatry
with newts south of the Alaskan panhandle (Mebs et al., 2016). Newt toxicity and snake TTXresistance appear to scale together in some locations (Brodie, Jr. et al., 2002); however,
resistance in snakes sometimes outpaces newt toxicity (Hanifin et al., 2008). The correspondence
between snake TTX-resistance and newt toxicity is attributed to an antagonistic predation-antipredator dynamic (Brodie, Jr. et al., 2002; Williams et al. 2012). In this arms race, the toxicity of
newts is thought to escalate as a result of predation by snakes with increasing resistance to TTX
enabled by mutations in Navs (Geffeney et al., 2005). In addition to the variably toxic
populations on the west coast, an inland population of newts in Moscow, ID possesses
undetectable levels of TTX (hereafter referred to as non-toxic newts) and are sympatric with
snakes that have low levels of TTX-resistance (Hanifin et al., 2008).

2
 

Recent work presents several challenges to the snake-driven predator-prey arms race
hypothesis (Bucciarelli et al., 2016; Hague et al., 2016; Mebs et al., 2016). First, a study of the
congener T. torosa demonstrates that the toxicity of individuals not only varies geographically,
but also fluctuates considerably throughout the year (Bucciarelli et al., 2016). It is likely that the
toxicity of T. granulosa shows similar patterns. Second, other animals, such as caddisfly larvae, 
Limnephilus flavastellus, (Gall et al., 2012b), great blue herons, Ardea herodias (Fellers et al.,
2007), and otters, Lontra canadensis (Stokes et al., 2015), have been observed eating toxic newts
without showing symptoms of intoxication. Whether these organisms are also resistant to TTX
has not been examined. Additionally, non-toxic newts coexist with multiple potential predators,
including garter snakes that have low levels of resistance (Hanifin et al., 2008), suggesting that
chemical defenses could be beneficial to those populations as well (Hanifin et al., 2008; Hague et
al., 2016). Conversely, snakes are absent in the Alaskan panhandle, but populations of newts
possess TTX (Hague et al., 2016). Toxicity is generally low in Alaskan populations; however, a
few individuals were observed to possess levels of TTX that rivaled the highly toxic populations
found in Oregon where newts and highly resistant snakes are sympatric (Hague et al., 2016). The
red-spotted newt (Notophthalmus viridescens), a TTX-bearing member of a clade that is sister to
Taricha, also shows variability in toxicity throughout its range, but toxicity in red-spotted newts
does not correlate with predation pressure (Yotsu-Yamashita et al., 2012). Taken together,
predation pressure by snakes and other animals likely plays a role in newt toxicity but cannot be
the only explanation.
The predator-prey arms race hypothesis presupposes that newt toxicity is a heritable trait
that responds to selection, but the source of TTX in newts and the pathway for TTX biosynthesis
have not been determined. Studies of newt population genetics using a small number of

3
 

fragments of the cytochrome oxidase I gene and 16s RNA (Mebs et al., 2016), microsatellites
(Ridenhour et al., 2007; Hague et al., 2016), and allozymes and fragments of mitochondrial DNA
(Kuchta and Tan, 2005) show that the degree of genetic differentiation among populations of
newts ranging from northern California to the Alaskan panhandle does not match the amount of
variation seen in toxicity. Low genetic differentiation is consistent with high gene flow among
populations. Bucciarelli et al. (2016) observed that lower-toxicity T. torosa males exhibited
weaker site fidelity than their higher-toxicity counterparts, and similar patterns of site fidelity
could result in the observed genetic homogeneity in populations of T. granulosa. Although
multiple studies suggest low levels of genetic differentiation among populations of newts
throughout their range (Kuchta and Tan, 2005; Ridenhour et al., 2007; Mebs et al., 2016), the
relationship between variation in toxicity and genetic differentiation is currently impossible to
determine. These studies are based on a small number of sequences, which may be insufficient to
capture genetic differentiation among populations. Furthermore, a genetic component of toxicity
has not been established. Without identifying and comparing sequences of genes involved in
toxicity, genetic differentiation associated with variation in toxicity cannot be assessed.
In all of the previously-discussed studies, skin toxicity was used as a proxy for newt
toxicity, as presence of TTX in the skin is often associated with defense (Noguchi et al., 2006);
however, the distribution of TTX goes beyond the skin (Wakely et al., 1966). In newts, skin is
the most toxic tissue, but ovary, testes, blood, viscera, and liver contain small amounts of TTX
(Wakely et al., 1966). The location of TTX can relate to its function and origin. In pufferfish, the
distribution of TTX throughout the body is species dependent; however, in some species,
location of TTX corresponds with its function (Noguchi et al., 2006). For example, ovaries of
female pufferfish (Takifugu niphobles) contain concentrations of TTX that far exceed that of the

4
 

skin (Noguchi et al., 2006), and the oocytes produced by these females are laden with TTX. In
this species, TTX from ovulated oocytes acts as a pheromone that attracts males (Matsumura,
1995), so the location of TTX in the ovaries is tied to its reproductive and communicative
functions. Additionally, studies of the pufferfish Takifugu rubipes showed that TTX is absorbed
from the intestines and is selectively transported to the liver, where it is found in large quantities
(Matsumoto et al., 2007). The presence of TTX in the intestine and liver, paired with studies
demonstrating a gain of toxicity when reared on toxic food, provides evidence that TTX comes
from an exogenous source in this species of pufferfish (Matsumoto et al., 2007). Learning more
about the distribution of TTX in newt tissues could provide insight into its source as well as the
role of TTX in newt biology and factors that impact variation in toxicity.
Understanding the function(s) of TTX in newt biology is an essential step in explaining
the substantial variability in toxicity. To address this need, I conducted two experiments: 1) I
cohoused toxic and non-toxic newts to examine possible links among the origin of TTX, the
transmissibility of toxicity via direct contact, and variation in toxicity between individuals; and
2) I described the distribution and concentration of TTX in a large group of tissues to provide
greater insight into the origin and non-defensive roles TTX may play in newts. In the first
experiment, I tested the hypothesis that TTX can be transferred between cohabitating individuals,
which I assessed by cohousing toxic and non-toxic newts and measuring changes in skin toxicity
at intervals of at least one month. In the second experiment, I measured the concentration of TTX
in tissues involved in diverse processes, focusing on the potential functional significance of
location.

5
 

METHODS
Subjects
Male and female rough-skinned newts (Taricha granulosa) were collected from an
unnamed pond (hereafter Sandpiper Pond) between the Pacific Coast Highway and the Sandpiper
Village subdivision in Waldport, Oregon (44° 26’ 53.10" N, 124° 4' 26.60" W) on July 13, 2016
(Figure 1), a section of the Elk River in Port Orford, Oregon (42° 47’ 14.64” N, 124° 28’ 50.58”
W) on July 16, 2016 (Figure 2), and a pond (hereafter Phillips Farm Pond) at Virgil Phillips
Farm Park in Moscow, Idaho (46° 48’ 49.6” N, 117° 0’ 55.86” W) on November 1, 2016 (Figure
3). Sandpiper Pond is approximately 185 km NE of Elk River and 610 km SW of Phillips Farm
Pond. Phillips Farm Pond is approximately 930 km NE of Elk River (Figure 4).
All collection activities and experiments were conducted in accordance with Protocol
10/12-195-00 authorized by Michigan State University’s Institutional Animal Care and Use
Committee. Collection of newts from Oregon (Sandpiper Pond: n = 10 males, 4 females; Elk
River: n = 16 males, 4 females) was conducted in accordance with Oregon Department of Fish
and Wildlife Scientific Taking Permit #104-15. Collection of newts from Idaho (Phillips Farm
Pond: n = 10 males, 4 females) was conducted in accordance with the Idaho Department of Fish
and Game Wildlife Bureau Wildlife Collection/Banding/Possession Permit #150521.
Subject Collection
Newts were collected with dip nets or in partially submerged minnow traps (Promar
Collapsible Minnow Traps, size small and medium, Gardena, CA) baited with worms and glow
sticks. Minnow traps were placed in the water in the evening and checked the next morning.
Animals were removed from the traps and all males were collected. To minimize the impact on
the breeding population, only a small number of females was removed from each location. All

6
 

juvenile newts and non-target organisms were immediately released. Adult newts used in this
study were packed in moss-filled containers placed inside a cooler with ice packs and shipped to
Michigan State University.
Animal Maintenance
All newts were housed in 10-gallon aquaria in an animal facility at Michigan State
University. Aquaria were filled with Holtfreter’s solution (Armstrong et al., 1989) and included a
platform that allowed newts to climb out of the water. Newts were fed live California black
worms (Lumbriculus variegatus) three days per week ad libitum.
In the cohabitation experiment, each aquarium contained two male newts in one of three
combinations: both from Sandpiper Pond (toxic control), both from Phillips Farm Pond (nontoxic control), or one from Sandpiper Pond and one from Phillips Farm Pond (mixed-toxicity
treatment). For the tissue toxicity experiment, newts from Elk River, Sandpiper Pond, and
Phillips Farm Pond were housed in aquaria containing four to six same-sex individuals from the
same population. All aquaria were cleaned daily using different equipment for newts in different
treatments or from different populations. Both rooms were maintained between 18.3 and 22.2 °C
with a 12:12 light-dark cycle for the newts in the cohabitation experiment and a 13:11 light-dark
cycle for those in the tissue toxicity experiment.
Set-up and Sample Collection for the Cohabitation Experiment
Newts from Sandpiper and Phillips Farm Ponds were randomly assigned to one of three
conditions: the toxic control, comprised of two males from Sandpiper Pond (n = 2 aquaria); the
non-toxic control comprised of two males from Phillips Farm Pond (n = 2 aquaria), and the
mixed-toxicity treatment comprised of one male from Sandpiper Pond and one male from
Phillips Farm Pond (n = 6 aquaria).

7
 

Unfortunately, over the course of this 10-month-long experiment, four newts from
Phillips Farms Pond died. Three were in non-toxic control aquaria and one was from a mixedtoxicity treatment aquarium. One non-toxic control newt died before the March sampling date, so
there was only one non-toxic control aquarium from March until May. Two other non-toxic
control newts died before the June sampling date, so for the final two sampling dates, there were
no non-toxic control aquaria. The Phillips Farm Pond newt from a mixed-toxicity treatment
aquarium died before the June sampling date, so n = 6 mixed-toxicity treatment aquaria from
February to May and n = 5 in June and December. Data presented here only include aquaria with
a pair of newts.
Approximately every four weeks between February and June and once in December of
2016, each newt was anesthetized in methanesulfonic acid (MS222, pH 7.2-7.4) before collecting
skin samples. Following the skin biopsy (described below), newts were placed in individual
containers of Holtfreter’s solution until the effects of anesthesia wore off, which was usually
within 15 min. Newts were then returned to their assigned aquaria.
Sample Collection for the Tissue Toxicity Experiment
Post-mortem tissue samples were collected from newts from Elk River (n = 9 males, 4
females), Sandpiper Pond (n = 2 females), and Phillips Farm Pond (n = 1 male, 3 females) to
determine how TTX is distributed throughout the body. Newts were anesthetized in
methanesulfonic acid (MS222, pH 7.2-7.4) for at least 30 min before I measured the newts’
mass, total length, snout-vent length (SVL), head width from the tip of the left quadrate to the tip
of the right quadrate, and head length from snout to quadrate. Two skin samples were collected
from each individual and combined for further analysis. Newts were then decapitated and a small
sample of blood was immediately collected. Brain, feces, heart, kidney, and pieces of fat, skeletal

8
 

muscle, spleen, and liver were collected from all newts. Additionally, testes and vas deferens
were collected from males and a small sample of unfertilized eggs and piece of oviduct were
collected from females.
Skin Sampling and Tissue Processing
Using a sampling protocol modified from Bucciarelli et al., (2014), one skin sample was
removed from each side of the posterior dorsolateral area approximately 1 cm below the
vertebral column using a 2 mm skin biopsy punch (Acu-punch, Acuderm Inc. Fort Lauderdale,
FL). Skin samples and all other tissues collected from newts were weighed before being placed
in individual microcentrifuge tubes with 300 µl 0.1M aqueous acetic acid and homogenized
using a single-use disposable tissue grinder (Pellet Pestle®, Kimble® Kontes®, Rockwood, TN).
Samples were stored at -80 °C until further analysis.
Toxin Extraction from Tissue
All tissue samples collected in the cohabitation and tissue toxicity experiments were
processed to extract TTX, which was then quantified using liquid chromatography-tandem mass
spectrometry (LC-MS/MS). To do so, I followed the extraction protocol described in Bucciarelli
et al. (2014). Specifically, frozen samples were thawed at room temperature, then the macerated
samples were placed in a boiling water bath for 5 min, followed by 5 min in an ice bath. Samples
were then centrifuged for 20 min at 13,000 x g and the supernatants were transferred to
centrifugal filters (Ultrafree-MC GV 0.22 µm pore size, EMD Millipore Corporation, Billerica,
MA) and centrifuged for 20 min at 13,000 x g. An additional 100 µl of 0.1M aqueous acetic acid
was added to the centrifugal filters before a final 20-min centrifugation at 13,000 x g. The
extracts were then diluted with 600 µl 0.1M aqueous acetic acid and either frozen immediately at
-80 °C or prepared for LC-MS/MS.

9
 

Solid Phase Extraction of Samples
Oasis MCX 1 cc Vac Cartridges (30 µm particle size, Waters, Milford, MA) were washed
with 1 ml methanol and then 1 ml of deionized water. 1 ml of the extracted sample was filtered
through the prepared Vac Cartridge over 30 sec followed by washes with 1 ml acetonitrile, 1 ml
methanol, and 1 ml deionized water. The Vac Cartridges were transferred to a manifold
(VacMasterTM 10 Sample Processing Manifold, Biotage®, Charlotte, NC) and eluted twice with
125 µl 0.2M hydrochloric acid/20% methanol with a 30-sec soak time for each elution. The
processed extracts were then dried in a vacuum concentrator (Savant SpeedVac SC110, Thermo
Fisher Scientific, Waltham, MA), reconstituted with 100 µl 1:1 acetonitrile: 0.1M acetic acid,
and transferred to vials (2 ml Screw Top Autosampler Vial and 100 µl inserts, Agilent, Santa
Clara, CA).
TTX Quantification
The concentration of TTX in tissue samples was measured in MSU’s Mass Spectrometry
and Metabolomics Core using a QTRAP 3200 liquid chromatography-tandem mass spectrometry
(LC-MS/MS) system with electron spray ionization (Sciex, Framingham, MA). 5 µl of each
sample was passed through an Xbridge Amide filter (3.5 µm, Waters, Milford, MA) using a
50:50 mix of 10mM ammonium acetate (pH 3.5-4.0): acetonitrile for the mobile phase at a flow
rate of 0.4 ml/min for 5 min. TTX standards ranging from 0 ng/ml to 3,000 ng/ml were run
alongside the samples and used to calculate the concentration of TTX in tissues. TTX standards
below 5 ng/ml were not consistently detected or they decreased the linearity of the standard
curve when included in the calculations and therefore were excluded from the standard curve.
The ability to quantify the concentration of TTX in a tissue sample is constrained by the highest
and lowest values of the standard curve, which comprise the limit of quantification, a range of

10
 

concentrations in which samples can reliably be calculated using the standard curve. Following
standard practice in the field, any tissue sample with a concentration detected as greater than 0
ng/ml but less than 5 ng/ml was excluded from the data set (Keizer et al., 2015). Removal of
these samples from the dataset did not alter the interpretations of the results. The LC-MS/MS
measurements were reported in ng TTX/ml sample and then converted to µg TTX/g tissue using
the amount of 0.1M acetic acid used to reconstitute the sample after vacuum concentration (100
µl) and the mass of the tissue.
Statistical Analysis
Descriptive statistics, ANOVA, and t-tests were calculated using Microsoft Excel and
regressions were calculated using the Data Analysis ToolPack in Excel (Microsoft Office 365
ProPlus, Microsoft, Redmond, Washington). Post-hoc analyses were performed using
QuickCalcs (GraphPad Software, Inc., La Jolla, California).

11
 

RESULTS
Cohabitation Experiment
The toxicity of Phillips Farm Pond newts remained at 0 µg/g during the 10 months of the
experiment. As illustrated in Figure 5, this result was observed both in the newts cohoused with
toxic Sandpiper Pond newts (mixed-toxicity treatment) and those cohoused with another newt
from Phillips Farm Pond (non-toxic control). Although most skin biopsy samples showed no
TTX, in four instances different Phillips Farm Pond newt samples had amounts of TTX that
ranged from the equivalent of 0 to <5 ng/ml (the units of the standard curve), which falls below
the level where the concentration can be determined with confidence. Because the amount of
TTX present could not be quantified accurately, the samples were excluded from analyses. Three
of the four of these newts were cohoused with toxic newts in treatment aquaria and one was
cohoused with another non-toxic newt in a low-toxicity control aquarium. None of these
individuals possessed trace amounts of TTX more than once during the experiment.
All Sandpiper newts consistently possessed TTX; however, the average amount of TTX
varied significantly over time when the treatment and control newt toxicities were pooled
(ANOVA, F = 4.92, df = 59, p = 0.00088; Figure 6). Specifically, the toxicity of newts in
December was significantly greater than in April, May, and June (Tukey Kramer post-hoc: p =
<0.05 for comparisons between December and April, May, and June). The variation in mixedtoxicity treatment newts showed the same significant trend where toxicity of the December
samples was significantly greater than in samples from April, May, and June (ANOVA, F =
1.98, df = 35, p = 0.00630; Tukey Kramer post-hoc: p = <0.05 for comparisons between
December and April, May, and June; Figure 7), but the trend was not significant when the
control newts were analyzed separately (ANOVA, F = 1.98, df = 23, p = 0.13; Figure 7).

12
 

Tissue Toxicity Experiment
TTX was not detected in the skin or any of the other organs test in newts from Phillips
Farm Pond with one exception: a very small quantity of TTX was detected in a heart sample
from the single Phillips Farm Pond male. The concentration of TTX found in this sample (2.34
ng/ml) was below the level of quantification (5 ng/ml), therefore the sample was excluded from
the data set.
All newts from Elk River and Sandpiper Pond possessed TTX. Of the tissues sampled,
the concentration of TTX in the skin in every toxic individual was significantly greater than the
toxicity of any other tissue (ANOVA, F = 7.81, df = 13, p = 3.07x10-11; Tukey Kramer post-hoc:
p = <0.05 for all tissue compared to skin). Highly variable amounts of TTX were also detected in
brain, heart, skeletal muscle, ova, oviduct, testes, vas deferens, kidney, liver, spleen, blood, fat,
and feces in some female and male newts from Elk River and Sandpiper Pond (Tables 1-8).
However, of the newts with toxic skin, only skin, skeletal muscle, brain, ova, and vas deferens
were toxic in all individuals. For blood, fat, feces, heart, kidney, liver, spleen, and testes, at least
one sample from one Elk River or Sandpiper Pond individual had 0 µg/g TTX. Additionally, a
subset of brain, fat, feces, heart, kidney, liver, skeletal muscle, spleen, testes, and vas deferens
samples contained TTX; however, the concentration was between 0 ng/ml—5 ng/ml TTX, which
is below the level of quantification (Table 9), so these samples were excluded from the analyses.
Toxicity of individual organs did not follow a consistent hierarchy based the amount of TTX
found in each organ: the relative toxicities of different organs varied among individuals with no
patterns discernable based on sex or population.
Body condition, the ratio of mass to SVL, is an index of reproductive condition,
movement, and survival in amphibians (MacCracken and Stebbings, 2012; Bucciarelli et al.,

13
 

2016). Morphological measurements varied between males and females (Table 10). Toxic males
had a significantly greater mean mass (t-test, tcrit = 2.20, df = 11, p = 0.000168) and longer
average SVL than toxic females (t-test, tcrit = 2.20, df = 11, p = 0.0000442), but the mean for
body condition was significantly lower in males than females (tcrit = 2.36, df = 7, p = 0.0113).
Toxicity was not significantly different between males and females (tcrit = 2.57, df = 5, p =
0.173).
To determine if toxicity was correlated with body condition, mass, and SVL and toxicity,
I performed regression analyses for all toxic newts (n = 14), all Elk River newts (n = 12), all
toxic females (n = 6), Elk River males (n = 8), and Elk River females (n = 4) (Table 11). The “all
toxic females” grouping includes both Elk River (n = 4) and Sandpiper Pond females (n = 2).
Due to small sample size, a regression analysis was not performed on only Sandpiper Pond
females.
No significant correlation between toxicity and body condition existed for all toxic newts
(R2 = 0.022, p = 0.61), all Elk River newts (R2 = 0.030, p = 0.59), all toxic females (R2 = 0.15, p
= 0.44), Elk River males (R2 = 0.0010, p = 0.94), and Elk River females (R2 = 0.22, p = 0.53).
Similarly, no significant relationship between toxicity and mass existed for all toxic newts (R2 =
0.12, p = 0.23), all Elk River newts (R2 = 0.11, p = 0.30), all toxic females (R2 = 0.05, p = 0.67),
Elk River males (R2 = 0.026, p = 0.70), and Elk River females (R2 = 0.17, p = 0.59). In contrast,
as illustrated in Figure 8, the relationship between toxicity and SVL was significant when all
toxic newts were included in the analysis (R2 = 0.37, p = 0.022). When the two Sandpiper Pond
females, which were significantly more toxic than Elk River newts (F = 30.0, df = 1, p =
0.000142), were excluded from the analysis, the result was not significant for all Elk River newts
R2 = 0.31, p = 0.059), Elk River males (R2 = 0.25, p = 0.21), and Elk River females (R2 = 0.073,

14
 

p = 0.73). The relationship between SVL and toxicity was also not significant when all females
were considered (R2 = 0.38, p = 0.20).

15
 

DISCUSSION
The data presented here for the cohabitation and tissue toxicity experiments have
important implications for understanding the relationship between variation in toxicity and the
origin, transmissibility, and function of TTX in newts, which I will discuss in turn below.
Non-toxic Newts in the Cohabitation Study
Phillips Farm Pond newts did not become toxic during the cohabitation study; however,
garter snakes were present at the field site (personal observation). It also seems likely that they
face other predators, as their toxic counterparts can be successfully consumed by non-snake
predators (Stokes et al., 2015; Gall et al., 2011; Fellers et al., 2007). This raises questions about
why newts are not toxic if predation is a relevant selection pressure and why they do not become
toxic with exposure to other toxic individuals. Approaching these questions requires
consideration of the trade-offs involved in being toxic, and therefore resistant to the toxin, and
the source of the toxin.
TTX resistance in other organisms is due to mutations in Navs that decrease the binding
affinity for TTX (Lee and Ruben, 2008; Chau et al., 2011). Because being toxic requires being
resistant to the toxin (Hanifin, 2010), newts also possess mutations in Navs (Hanifin and Gilly,
2015). Resistance-conferring mutations occur in portions of Navs to which TTX binds, but these
portions are also essential to ion selectivity and flow; thus, changes that increase resistance also
likely impede channel function (Jost et al., 2008). Therefore, being toxic could be very costly,
and, in some environments, the benefits of being toxic may be outweighed by the costs, favoring
a lack of TTX. If populations of toxic and non-toxic newts were separated before some of the
resistance-conferring mutations occurred, toxic and non-toxic newts may not have followed
parallel evolutionary trajectories. Predation pressure likely accounts for only some of the

16
 

variation in newt toxicity; identifying other selective pressures and understanding the function of
TTX in newt biology could provide further insight into the trade-offs that govern toxicity.
In addition to the uncertain ecological and physiological significance of non-toxic
populations of newts, the source of TTX has not been determined. The lack of a detectable
change in non-toxic newt toxicity does not provide additional support for a particular hypothesis.
However, given the production of TTX by bacteria cultured from the skin of toxic newts (PM
Vaelli, personal communication) in addition to TTX-producing bacteria isolated from marine
organisms (Chau et al., 2011), one possible scenario is that newts acquire TTX from bacterial
symbionts. Because toxic and non-toxic newts are geographically separated, TTX-producing
bacteria may not occur in the environment of non-toxic newts. Cohousing toxic and non-toxic
newts exposes non-toxic newts to TTX-producing bacteria, providing many opportunities for the
exchange of microbes. Exchange of microbes between cohabitating conspecifics and even
heterospecifics is well documented in the literature (Archie and Theis, 2011; Leclaire et al.,
2014; Song et al., 2013, Whittaker et al., 2016). For example, in studies of human families living
together, the skin microbiomes of cohabitating individuals, even those that are not related, show
greater similarity than the microbiomes of individuals from different households (Song et al.,
2013). Additionally, microbiota appear to be shared between humans and their dogs (Song et al.,
2013). Skin microbiomes showed the greatest similarity among individuals, and the convergence
appears to be due to the amount of contact between individuals living together (Song et al.,
2013).
Cohabitating newts interacted regularly and were often seen in amplexus (personal
observation), but even if microbes were transferred between toxic and non-toxic newts, no TTX
was detected in non-toxic newts during the experiment. (We have not analyzed the microbiomes

17
 

of these newts, so the degree of similarity between cohoused newts cannot be assessed.)
Assuming that microbes were shared between cohabitating newts, it is very curious that nontoxic newts did not become toxic. Toxic newts may produce a substrate required by the TTXproducing bacteria that non-toxic newts do not. The absence of such a substrate could inhibit the
growth or survival of TTX-producing bacteria or the production of TTX. If the substrate is a
protein or the product of enzymatic activity, toxicity could have a genetic component, as the
coevolutionary arms race presupposes. Morphological comparisons of toxic Washington and
Oregon populations with the non-toxic Idaho population have led to the suggestion that toxic
newts were introduced to Idaho (Nussbaum and Brodie, 1971); however, the mtDNA of nontoxic Idaho newts suggests that they are genetically distinct from toxic populations (Kuchta and
Tan, 2005). Thus, not only could genetic differences in Navs limit resistance as suggested earlier,
but genetic differences among populations could also limit TTX production by microbes.
Based on these data, endogenous production cannot be ruled out as a possible mechanism
underlying toxicity. To date, biosynthesis of TTX has only been identified in bacteria and the
pathway is likely complex (Chau et al., 2011). However, the lack of evidence supporting
endogenous production does not preclude it as a possibility. Additionally, some marine
organisms acquire TTX from their diet, but this is not a likely source of TTX in newts because
they are able to retain and regenerate TTX in the absence of a TTX-containing food source
(Cardall et al., 2004), unlike pufferfish whose toxicity is dependent on consuming toxic food
(Noguchi et al., 2006).
In all, the results of the cohabitation experiment demonstrated that prolonged contact
between toxic and non-toxic individuals was insufficient to induce detectable levels of toxicity in
non-toxic newts.

18
 

Toxic Newts in the Cohabitation Experiment
Similar to Bucciarelli et al.’s (2016) results with T. torosa, the toxicity measured in the
Sandpiper Pond males used in the cohabitation experiment varied widely over time: average
toxicity of Sandpiper males increased from February to March, decreased between March and
June, and then increased again between June and December (Figure 6). Three possible
mechanisms that could account for these fluctuations are: 1) season or breeding condition; 2)
temporal changes in the microbiome; and 3) repeated injury leading to secretion of TTX with an
insufficient inter-sampling interval for complete TTX recovery.
First, newts are semi-aquatic, moving into ponds and streams during the breeding season.
The timing of the breeding season varies based on location, but it generally starts in late
winter/early spring. In addition to changes in their environment, newts undergo concomitant
morphological and physiological changes (Mccurdy, 1931; Jones et al., 2002a, 2002b; Mills and
Rombough, 2008), and toxicity could vary in response to any of these factors. In T. torosa,
toxicity cycles annually, reaching the highest levels during the peak of the breeding season
(Bucciarelli et al., 2016). The fluctuations in toxicity of Sandpiper Pond males observed during
the cohabitation experiment may relate to their breeding season, and the timing of the breeding
season can be inferred from reports of other Oregon populations at similar elevations. Toxicity of
the Sandpiper Pond males increased between February and March, which corresponds to the
breeding season of a population located approximately 65 km NE of Sandpiper Pond (Jones et
al., 2002a) and then declines through June before increasing again in December. If TTX levels
are tied to breeding, then it is likely that newts have a strong internal annual rhythm because the
newts used in this experiment were kept on a constant 12:12 light-dark cycle and had been
isolated from females since July 2015.

19
 

Second, temporal variability in skin TTX levels may not be independent of changes in the
microbiome. Microbial communities are not stagnant; they change or differ based on
environment, genetics, age, and diet (Archie and Theis, 2011; Song et al., 2013; Leclaire et al.,
2014; Whittaker et al., 2016). Changes in the newt microbiome in response to temporal changes
in environment, morphology, and physiology could all account for variation in toxicity. If newts
are viewed as an environment for bacteria, a newt’s location or physiological state could
influence the resources it provisions for bacteria, making different times of year more or less
conducive to TTX production.
Finally, taking skin biopsy samples, while not lethal, can be traumatic and the newts are
left with two small wounds. Secreting TTX could be a response to the trauma, and while newts
have been shown to regenerate TTX after expelling it in response to trauma, recovery took many
months (Cardall et al., 2004). The newts in this experiment were sampled every four weeks for
five months and then again six months later. Wounds healed completely between sampling
events, but if newts secreted TTX during a previous sampling event, the intervening month may
have been insufficient to allow for regeneration of the expelled TTX. Thus, repeated sampling at
such close intervals may have contributed to a cumulative decline in toxicity from March to
June. TTX levels were much higher after a six-month break from sampling, but the December
sampling event was also close to the start of the next breeding season. The suggestion of a cycle
in toxicity in Sandpiper Pond newts and the apparent annual pattern observed in T. torosa by
Bucciarelli et al. (2016) are intriguing, but establishing the occurrence of a cyclical pattern in
toxicity that might parallel breeding events in T. granulosa requires multi-year repetition of
measurements with greater consideration given to the impact of sampling.

20
 

Temporal variation in toxicity has other implications for our understanding of the role of
TTX in newt biology. Previous studies of toxicity in newts have used the amount of TTX in the
skin measured at a single point in time to compare and characterize the toxicity of newts in
different locations (Hanifin et al., 2008; Hague et al., 2016; Mebs et al., 2016). In these studies,
comparisons were made among samples taken at different times of the year and, in some cases,
with very few animals, which could confound the results. Understanding variation in toxicity
may also impact our interpretation of the newt-snake arms race: parity between toxicity and TTX
resistance may vary depending on the timing of toxicity measurements. Rather than describe
populations as toxic or non-toxic based on a sample at a single time point, more extensive studies
that characterize the toxicity profile throughout the entire geographic range of newts over the
course of many years are necessary to fully understand the potentially complex dynamics
underscoring variation in toxicity and the functional significance of TTX.
Tissue Toxicity Experiment
The distribution of TTX throughout the body of newts found in this experiment
reinforces and expands upon that described in previous studies of toxic newts (Wakely et al.,
1966; Hanifin et al., 2003; Lehman et al., 2004). In addition to skin, oviduct, ova, skeletal
muscle, blood, liver, and testes, which were previously known to be toxic (Wakely et al. 1966), I
detected TTX in fat, feces, kidney, vas deferens, brain, and heart, as shown in Tables 1-8.
Toxicity varied greatly among tissues, which may be related to the tissue’s affinity for TTX
(Mebs et al., 2016) or the ability of TTX to pass through cell membranes (Matsumoto et al.,
2007).
Additionally, my experiment provides the first analysis of internal organs in non-toxic
newts: in all instances, if TTX was absent from the skin, as was the case for all the Phillips Farm

21
 

Pond newts, the internal organs also lacked TTX. Conversely, if TTX was present in the skin, as
observed in all Elk River and Sandpiper Pond newts (Table 1), at least some of the internal
organs also contained TTX in each individual (Tables 2-8). The distribution of TTX in tissues
with diverse functions adds further evidence that TTX is involved in more than just defense in
toxic newts.
As noted in previous studies (e.g., Bucciarelli et al., 2016; Hanifin et al., 1999, 2008) and
confirmed in this study, the amount of TTX in skin varies widely among individuals. Aside from
temporal variation, differences in individual resistance and the trade-offs between costs and
benefits of toxicity could contribute to variation. Resistance to TTX is a function of the number,
location, and type of mutations present in Navs and individual toxicity may be limited by the
mutations that individual possesses (Brodie III and Brodie Jr., 1999; Lee and Ruben, 2008).
Therefore, it seems reasonable to speculate that less toxic individuals can survive in mixedtoxicity populations because being surrounded by more toxic individuals could provide
protection in a manner similar to herd immunity. Newts aggregate in large groups in breeding
ponds (Janzen and Brodie, 1989), so less toxic individuals could be protected from predators by
the presence of other highly toxic individuals without incurring the costs that come with being
toxic (Bucciarelli et al., 2016).
The presence of TTX in the blood (Table 2) suggests a means of regulating internal levels
of TTX, which could play a role in mitigating the physiological costs of being toxic and
preventing self-toxicity. As documented in other tetrodotoxic species, possessing TTX imposes
physiological costs through compromises in Navs function, although specific costs in newts have
not been studied (Bucciarelli et al., 2016). Newts may balance the trade-off between toxicity and

22
 

these costs by transporting TTX through the blood from some tissues to others for sequestration
or excretion.
Newts may also regulate toxicity by eliminating TTX in feces, which varied by an order
of magnitude among individuals (Table 3). The amount of TTX I measured in feces was up to an
order of magnitude greater than TTX found in blood, so perhaps TTX accumulates in the
intestines for incorporation into feces until it can be excreted. Newts may need to excrete TTX
because the amount of TTX an individual can possess is likely constrained by its physiology. As
TTX levels increase during the year, newts may need to excrete more TTX. Further studies
examining temporal variation in fecal TTX relative to skin toxicity, which is the only tissue that
can be collected is non-lethally, are needed to gain insight into whether the presence of TTX in
the feces presents a means of general excretion and toxicity regulation, safeguarding newts
against self-intoxication.
Presence of TTX in feces of toxic newts also indicates that newts release TTX in multiple
ways. Newts secrete TTX from the granular glands in their skin as a defense mechanism (Cardall
et al., 2004), but little consideration has been given to other functions of TTX secretion. In
addition to regulating TTX levels, the TTX excreted in feces could play a role in communication.
In the congener T. torosa, TTX is used by larvae as a cue that cannibalistic adults are nearby, and
the presence of TTX elicits antipredator behavior (Zimmer et al., 2006). Although cannibalism
has not been documented in T. granulosa, TTX may communicate similar or even other
information among newts in toxic populations. TTX secreted from the skin and excreted in feces
could indicate where newts are located, attracting others to the area for mating or feeding. TTX
is a pheromone in pufferfish (Matsumura, 1995) and may also act in newts as an aggregation
pheromone during the breeding season, when levels peak (Itoi et al., 2016). Like pufferfish

23
 

(Okita et al., 2013; Itoi et al., 2016), newts can smell TTX (T. torosa: Zimmer et al., 2006; T.
granulosa: HL Eisthen, unpublished data) and T. granulosa are attracted to it (JD Merkel and HL
Eisthen, unpublished data). Additionally, the amount of TTX in feces could function as an
indicator of overall toxicity and thus play a role in mate selection, as toxicity does in T. torosa
(Bucciarelli et al., 2016).
The presence of TTX in both female and male reproductive tissues in toxic newts (Tables
4-6) provides further evidence that TTX could be involved in reproduction. However, I found
that body condition, a commonly used measure of reproductive status in amphibians given by the
ratio of body mass to SVL, does not correlate with toxicity. The lack of relationship holds true
regardless of whether all toxic newts from this experiment are grouped together, separated by
population, separated by sex, or separated by population and sex. My result is consistent with
Bucciarelli et al.’s (2016) finding of no relationship between body condition and TTX level in T.
torosa. The time of year could explain the lack of correlation: I collected most of the samples
between June and August, which is likely well after the height of the breeding season. Males
cease to court females once they begin ovipositing (Jones et al., 2002a), which likely occurs
earlier in the year, so it is possible that the newts are no longer in peak breeding condition.
Alternatively, body condition may not be a good metric of reproductive condition and other
indices may be more appropriate (MacCracken and Stebbings, 2012). The correlation between
mass and toxicity was also non-significant. However, I found a significant negative correlation
between SVL and toxicity when all male and female newts were combined (Figure 8). The mean
SVL of females was shorter than that of males, but females were more toxic (Table 10). Small
size or ovipositing behavior may make females more vulnerable to predators, offsetting the
potential costs of high levels of toxicity. Although I observed sexually dimorphic trends in size

24
 

and toxicity, I found no correlation between toxicity and SVL when the analysis was performed
separately on males and females (Table 11). Small sample sizes, lack of comparisons among
multiple populations, and measurements taken at a single time point may obscure correlations.
Despite the lack of correlation I found in most of the regressions, the presence of TTX in
reproductive tissues suggests that a relationship between toxicity and reproduction exists;
therefore, the role of TTX in reproduction should not be dismissed.
A previous study has shown that TTX is maternally provisioned in the yolk and
cytoplasm of fertilized eggs and that female and egg toxicity are positively correlated, although
the R2 value (0.48) indicates that the correlation does not fully describe the relationship (Hanifin
et al., 2003). Maternally provisioned TTX most likely protects the eggs from predators (Brodie,
1968; Hanifin et al., 2003). As shown in Table 4, whole ova from Elk River and Sandpiper Pond
females were found to contain substantial amounts of TTX, in agreement with the maternal
endowment hypothesis (Hanifin et al., 2003; Gall et al., 2012a). The jelly coat typically contains
small amounts of TTX (Gall et al., 2012; Hanifin et al., 2003), and my results show that the
oviduct, which contributes the jelly coat, contains little or no TTX (Table 4). All toxic females
had large quantities of mature oocytes, so it is possible that the very low levels of TTX I
observed in the oviduct result from depletion due to deposition, but it is also possible that TTX in
the jelly coat is contributed by another source.
In contrast to maternal endowment, little attention has been paid to potential
contributions of TTX to eggs by males. Sperm is transferred to females via a spermatophore,
which is composed of a gelatinous stalk with branching fibrils and a sperm cap on top (Davis and
Twitty, 1964). During courtship, the male attaches the jelly stalk to the substrate at the end of a
lengthy ritual, and the receptive female positions her cloaca over the spermatophore to pick up

25
 

the sperm cap for internal fertilization of the eggs (Propper, 1991). Although the presence of
TTX in spermatophores has not been studied, my measurements of TTX in the testes and vas
deferens of Elk River newts (Table 5) suggest that spermatophores could contain TTX. The
testes had very little TTX, so the sperm cap may not be toxic, but the vas deferens had large
quantities of TTX and plays a role in the development of the stalk of the spermatophore in
concert with cloacal glands (Uzzell, 1969; Hopkins, 1978).
Understanding the potential contribution of TTX from males to offspring requires
knowledge of the distribution of TTX in the spermatophore, which has not been studied.
However, parental endowment of defensive chemicals in other species, such as the ornate moth
(Utetheisa ornatrix), is well documented (Dussourd et al., 1988; González et al., 1999;
LaMunyon and Eisner, 1993). Male and female Utetheisa possess pyrrolizidine alkaloids (PA)
derived from food (Dussourd et al., 1991), and both sexes contribute PA to the eggs, where they
function in defense (Dussourd et al., 1988). Males contribute PA via the spermatophore, and,
along with nutrients, PA constitute a nuptial gift. Females use the nutrients and PA from the
spermatophore to replenish their own stores and also incorporate some of the male’s PA into the
eggs. The concentration of PA in spermatophores also play a role in the females’ postcopulatory
judgment of male quality (Eisner and Meinwald, 1995). Utetheisa males are not alone in the
provisioning offspring with defensive compounds: male assassin bugs (Apiomerus flaviventris),
beetles in the families Meloidae and Pyrochroidae, and danaine butterflies, such as the queen
butterfly (Danaus gilippus), contribute compounds that are included in eggs, and in many cases
females use the concentrations of these compounds as an indicator of male quality (Eisner and
Meinwald, 1995).

26
 

Female newts use multiple cues to assess mates, such as body size and tailfin height
(Janzen and Brodie, 1989), but because of the costliness of being toxic and defensive benefits of
TTX (Bucciarelli et al., 2016), TTX could be a particularly useful indicator of male quality. In
addition, the presence of TTX in male newt reproductive organs found in this study suggests that
TTX may also be transmitted to females in spermatophores as a nuptial gift that can be
incorporated into the egg. Like Utetheisa, female newts mate multiple times and can store and
use sperm selectively (Utetheisa: Eisner and Meinwald, 1995; LaMunyon and Eisner, 1993;
newts: Jones et al., 2002). Female choice in Utetheisa is important for increasing offspring
fitness because choosing males with large amounts of PA enable females to increase the load of
defensive PA in eggs (Iyengar and Eisner, 1999). Preference for toxic spermatophores has also
been observed in toxic and non-toxic moths (Iyengar et al., 2001). A parallel scenario in which
TTX plays a substantial role in the reproductive biology of newts seems possible.
The existence of a naturally occurring non-toxic population of newts opens the possibility
to elucidate the role of TTX in reproduction. Determining whether males contribute TTX to
offspring could be accomplished by mating non-toxic females with toxic males. Given that TTX
is not transferred from toxic to non-toxic male newts through contact and cohabitation,
reproduction could be a means of transmission among adults although non-toxic females may
prefer not to mate with toxic males. Furthermore, if mating does occur, offspring may not be
viable if TTX is contributed from the male and the offspring lack resistance-conferring
adaptations. In either case, TTX could be a barrier to reproduction between toxic and non-toxic
newts, which could provide an explanation for populations devoid of TTX.
The bright yellow bilateral fat bodies, a reproductive structure overlying the gonads in
both males and females, are toxic (Table 6). Fat bodies synthesize steroid hormones (Lupo di

27
 

Prisco et al., 1971) and store lipids that are essential for reproductive function in many
amphibians (Jorgensen, 1986; Girish and Saidapur, 2000). Energy resources from fat bodies are
involved in gametogenesis and maintaining the reproductive organs (Girish and Saidapur, 2000;
Madelaire and Gomes, 2016). I found that TTX is present in the fat bodies of both males and
females at levels similar to blood (Tables 2 and 6). The fat bodies could function as TTX
repositories that supply TTX to the gonads, resulting in levels of TTX that fluctuate with
reproductive state. Chieffi et al. (1980) described a portal system linking the fat bodies with
testes in male edible frogs (Rana esculenta), and if a similar portal system exists in newts it is
possible that TTX from the fat bodies could be transferred directly to the gonads.
I detected small amounts of TTX in the kidney, liver, and spleen in only about half of the
toxic newts (Table 7). In many toxic marine species, the presence of TTX in digestive tissues
results from accumulation of TTX from dietary sources or bacterial symbionts (Hanifin, 2010;
Lago et al., 2015; Noguchi et al., 2006). For example, non-toxic pufferfish gain toxicity after
ingesting toxic food (Noguchi et al. 2006) and the grey side-gilled sea slug (Pleurobranchaea
maculata) gains and loses toxicity based on the toxicity of its diet (Salvitti et al., 2017). In
contrast, newts do not seem to acquire TTX from their food: newts reared in captivity on nontoxic diets do not lose toxicity, which can even increase over time (Hanifin et al., 2003; Cardall
et al., 2004; Gall et al., 2012a) and captive newts regenerate TTX following secretion (Cardall et
al., 2004). Additionally, the relatively low levels of TTX in digestive tissues observed in this
experiment support the idea that newts do not derive TTX from their food. The low level of TTX
that I found in the liver is particularly notable because the liver is responsible for detoxifying and
metabolizing compounds and is one of the most toxic organs in pufferfish (Noguchi et al., 2006).
If the liver were involved in regulating toxicity in newts, larger amounts of TTX would be

28
 

expected. The average concentrations of TTX I found in digestive tissues are near or below
levels in the blood, which was not drained from the tissues prior to analysis, suggesting that TTX
does not accumulate in digestive tissues in newts.
Finally, my analysis of excitable tissues corroborates earlier accounts of TTX in skeletal
muscle (Wakely et al., 1966), but they also show for the first time that TTX is present in both the
brain and heart muscle (Table 8). This result demonstrates that TTX crosses the blood-brain
barrier and also could affect cells in both heart and skeletal muscle, but to date direct evidence
has been lacking. The average concentration of TTX in the brain and heart muscle exceeds that
found in the blood by one to two orders of magnitude (Tables 2 and 8), demonstrating that TTX
is transported into these tissues. It is entirely unclear whether TTX may serve a function in
excitable tissues. Alternatively, perhaps TTX does not serve a function in these tissues; if Navs in
newts are incompletely resistant to TTX and allow partial binding, the toxin could simply
accumulate in tissues that possess Navs. Given that all isoforms need to be resistant to TTX for a
TTX-possessing organism to survive (Jost et al., 2008), a comparative study of Nav sequences in
toxic and non-toxic newts could provide a physiological explanation for the existence of nontoxic populations if mutations are absent in these animals.

29
 

CONCLUSION
My experiments provide a solid foundation for many future studies examining the
function of TTX in newts extending beyond the popular notion of defense against predators.
Non-toxic newts poses a challenge: if TTX is so effective as a defense, why would an entire
population of individuals that is still vulnerable to predation lose their toxicity? Although my
study does not offer an answer to that question, the demonstration that long-term cohabitation
cannot induce toxicity in non-toxic individuals suggests the existence of a physical or
physiological cost or impediment. In addition, I found that toxicity is not a static trait: the
amount of TTX present in skin varies among populations and temporally within individuals, and
TTX is found in many tissues with dissimilar functions throughout the body. My experiments
join the growing body of evidence that toxicity in newts is more complicated than once
conceived. Although defense is certainly a benefit derived from toxicity, the role that TTX plays
is likely multifaceted. Further studies of newts’ reproduction, behavior, microbiota, and sodium
channels are necessary for the explication of this complex and dynamic trait.

30
 

APPENDICES

31
 

APPENDIX A

TABLES

32
 

Table 1. Toxicity of Skin
Skin
Population

n

Toxicity

Range

ER Male

8

20 ± 6.0

2.7 - 50

ER Female

4

37 ± 11

22 - 69

PF Male

1

0

PF Female

3

0±0

0-0

SP Female

2

240 ± 120

120 - 360

Mean ± standard error (SE) for toxicity (“Toxicity”) and range of toxicities (“Range”) observed
for skin samples collected in the tissue toxicity experiment. Toxicity is measured in µg TTX/g
skin. Newts are separated by sex and population; ER = Elk River, PF = Phillips Farm Pond, and
SP = Sandpiper Pond.

33
 

Table 2. Toxicity of Blood
Blood
Population

n

Toxicity

Range

ER Male

8

0.057 ± 0.027

0 - 0.20

ER Female

3

0.032 ± 0.028

0 - 0.088

PF Female

2

0±0

0-0

SP Female

1

0.78

Mean ± SE and range of toxicities observed for blood samples collected in the tissue toxicity
experiment. Toxicity is measured in µg TTX/g blood. Newts are separated by sex and
population; ER = Elk River, PF = Phillips Farm Pond, and SP = Sandpiper Pond.

34
 

Table 3. Toxicity of Feces
Feces
Population

n

Toxicity

Range

ER Male

4

0.14 ± 0.066

0 - 0.28

ER Female

2

0.025 ± 0.025

0 - 0.050

PF Female

3

0±0

0-0

SP Female

1

0.14

Mean ± SE and range of toxicities observed for feces collected in the tissue toxicity experiment.
Toxicity is measured in µg TTX/g feces. Newts are separated by sex and population; ER = Elk
River, PF = Phillips Farm Pond, and SP = Sandpiper Pond.

35
 

Table 4. Toxicity of Female Reproductive Tissues
Ova

Oviduct

Population

n

Toxicity

Range

n

Toxicity

Range

ER Female

4

1.1 ± 0.63

0.064 - 3.0

3

0.0032 ± 0.0032

0 - 0.0095

PF Female

3

0±0

0-0

3

0±0

0-0

SP Female

2

6.8 ± 0.29

6.5 - 7.1

2

0.10 ± 0.052

0 - 0.16

Mean ± SE and range of toxicities observed for ova and oviduct collected in the tissue toxicity
experiment. Toxicity is measured in µg TTX/g tissue. Newts are separated by population; ER =
Elk River, PF = Phillips Farm Pond, and SP = Sandpiper Pond.

36
 

Table 5. Toxicity of Male Reproductive Tissues
Testes

Vas deferens

Population

n

Toxicity

Range

n

Toxicity

Range

ER Male

8

0.013 ± 0.0053

0 – 0.044

8

0.11 ± 0.044

0.014 – 0.37

PF Male

1

0

1

0

Mean ± SE and range of toxicities observed for testes and vas deferens collected in the tissue
toxicity experiment. Toxicity is measured in µg TTX/g tissue. Newts are separated by
population; ER = Elk River and PF = Phillips Farm Pond. No males from Sandpiper Pond were
included in this experiment.

37
 

Table 6. Toxicity of Fat Body
Fat Body
Population

n

Toxicity

Range

ER Male

6

0.058 ± 0.028

0 - 0.19

ER Female

4

0.044 ± 0.012

0.032 - 0.089

PF Female

3

0±0

0-0

SP Female

1

0.32

Mean ± SE and range of toxicities observed for fat bodies collected in the tissue toxicity
experiment. Toxicity is measured in µg TTX/g tissue. Newts are separated by sex and
population; ER = Elk River, PF = Phillips Farm Pond, and SP = Sandpiper Pond.

38
 

Table 7. Toxicity of Digestive Tissues
Kidney

Liver

Spleen

Population

n

Toxicity

Range

n

Toxicity

Range

n

Toxicity

Range

ER Male

9

0.021 ± 0.011

0 - 0.098

7

0.0097 ± 0.0031

0 - 0.019

6

0.012 ± 0.011

0 - 0.070

ER Female

3

0.0089 ± 0.0089

0 - 0.027

3

0.025 ± 0.025

0 - 0.075

2

0±0

0-0

PF Male

1

0

1

0

0

PF Female

3

0±0

0-0

2

0±0

3

0±0

0-0

SP Female

2

0.064 ± 0.031

0.033 - 0.095

2

0.0079 ± 0.0079

1

0.13

0 - 0.016

Mean ± SE and range of toxicities observed for kidney, liver, and spleen collected in the tissue toxicity experiment. Toxicity is
measured in µg TTX/g tissue. Newts are separated by sex and population; ER = Elk River, PF = Phillips Farm Pond, and SP =
Sandpiper Pond.

39
 

Table 8. Toxicity of Excitable Tissues
Brain

Heart Muscle

Skeletal Muscle

Population

n

Toxicity

Range

n

Toxicity

Range

n

Toxicity

Range

ER Male

6

1.7 ± 0.71

0.045 – 3.7

9

0.36 ± 0.27

0 – 2.4

6

0.23 ± 0.094

0.012 – 0.61

ER Female

4

1.9 ± 0.92

0.13 – 4.5

2

0.11 ± 0.11

0 – 0.21

4

0.94 ± 0.6

0.025 – 2.6

PF Male

1

0

1

0

PF Female

3

0±0

0-0

3

0±0

3

0±0

0-0

SP Female

2

3.5 ± 0.84

2.7 – 4.4

1

0.84

2

2.0 ± 0.57

1.4 – 2.5

0
0–0

Mean ± SE and range of toxicities observed for brain, heart muscle, and skeletal muscle collected in the tissue toxicity experiment.
Toxicity is measured in µg TTX/g tissue. Newts are separated by sex and population; ER = Elk River, PF = Phillips Farm Pond, and
SP = Sandpiper Pond.

40
 

Table 9. Number of Tissue Samples from Toxic Newts with No Quantifiable TTX
Elk River Males

Elk River Females

Sandpiper Females

Tissue

Total

0µg/g

0-5ng/ml

Total

0µg/g

0-5ng/ml

Total

0µg/g

0-5ng/ml

Blood

8

1

0

3

1

0

1

n/a

0

Brain

8

0

2

4

0

0

2

0

0

Fat

6

1

0

3

0

1

1

0

0

Feces

5

0

1

3

1

1

1

0

0

Heart

9

5

0

3

1

1

2

0

1

Kidney

9

4

0

4

2

1

2

0

0

Liver

9

2

2

4

2

1

2

1

0

Muscle

8

0

0

4

0

2

2

0

0

Ova

n/a

n/a

n/a

4

0

0

2

0

0

Oviduct

n/a

n/a

n/a

3

2

0

2

0

0

Skin

8

0

0

4

0

0

2

0

0

Spleen

7

4

1

3

2

1

1

0

0

Testes

9

3

1

n/a

n/a

n/a

n/a

n/a

n/a

Vas deferens

9

0

1

n/a

n/a

n/a

n/a

n/a

n/a

Skin from all Elk River (ER) and Sandpiper Pond (SP) newts contained TTX; however, other
tissues from some individuals with toxic skin contained extremely low levels of TTX or lacked it
altogether. The total number of samples collected (Total), and the number of tissue samples that
had either no TTX (0 ng/ml) or levels below the limit of detection (0-5 ng/ml) is indicated above.
The limit of quantification is reported in ng/ml because it is based on the units of the TTX
standards used to generate the standard curve. The concentration of TTX in newt tissues was
converted using the mass of the tissue sample and reported in µg/g. n/a = not applicable.

41
 

Table 10. Mean Toxicity, Body Condition, Mass, and SVL of Toxic Newts
Toxicity

Body Condition

Mass

SVL

Population

n

Mean ± SE

Range

Mean ± SE

Range

Mean ± SE

Range

Mean ± SE

Range

All Toxic Newts

14

57 ± 25

2.7 - 360

4.2 ± 0.25

3.1 - 6.3

20 ± 1.4

12 - 28

79 ± 1.7

69 - 87

All Elk River Newts

12

26 ± 5.6

2.7 – 69

4.1 ± 0.26

3.1 - 6.3

21 ± 1.4

12 - 28

80 ± 1.6

73 - 87

All Toxic Females

6

110 ± 53

22 – 360

4.9 ± 0.36

4.1 - 6.3

15 ± 1.1

12 - 18

73 ± 1.3

70 -79

Elk River Males

8

20 ± 6.0

2.7 – 50

3.6 ± 0.16

3.1 - 4.3

24 ± 1.0

18 - 28

84 ± 1.1

78 - 87

Elk River Females

4

37 ± 11

22 – 69

5.0 ± 0.47

4.2 - 6.3

15 ± 1.4

12 - 18

74 ± 14

73 - 79

Mean ± SE and range of skin toxicity, body condition, mass, and snout-vent length (SVL) observed for toxic newts in the tissue
toxicity experiment. Toxicity is measured in µg TTX/g tissue, body condition is measured in g/mm, mass is measured in g, and SVL is
measured in mm. All Toxic Newts = male and female Elk River Newts and female Sandpiper Pond newts, All Elk River Newts = male
and female Elk River Newts, All Toxic Females = Elk River and Sandpiper Pond newts.

42
 

Table 11. Regression Analysis of Toxicity on Body Condition, Mass, and SVL
Toxicity on Body
Condition

Toxicity on
Mass

Toxicity on
SVL

Population

n

R2

p

R2

p

R2

p

All Toxic Newts

14

0.022

0.61

0.12

0.23

0.37

0.022

All Elk River Newts

12

0.030

0.59

0.11

0.30

0.17

0.059

All Toxic Females

6

0.15

0.44

0.05

0.67

0.38

0.20

Elk River Males

8

0.001

0.59

0.026

0.70

0.25

0.21

Elk River Females

4

0.22

0.53

0.17

0.59

0.073

0.73

R2 and p-values (α = 0.05) for regressions of toxicity on body condition, mass, and snout-vent
length (SVL) observed for toxic newts in the tissue toxicity experiment. All Toxic Newts = male
and female Elk River Newts and female Sandpiper Pond newts, All Elk River Newts = male and
female Elk River Newts, All Toxic Females = Elk River and Sandpiper Pond newts.

43
 

APPENDIX B
FIGURES

44
 

N

Figure 1. Aerial and Ground View of Sandpiper Pond. Top panel: Aerial view of Sandpiper
Pond. Sandpiper Pond (indicated by arrows) is divided by the Oregon Coast Highway (Highway
101) in Waldport, OR. The part of the pond on the west side of the highway is bordered by
Sandpiper Village subdivision and the east side (partially obscured by tree cover) is bounded by
a wooded area. Newts were caught on both sides of the highway. Lower panel: Photograph of the
west side of Sandpiper Pond. Vegetation buffers the pond on all sides. Floating vegetation is
found at the edge of the pond and fallen trees are floating throughout. Sandpiper Pond has gentle
slope with a soft bottom. Photo credits: Google Earth Pro 2017.

45
 

N

Figure 2. Aerial View of Elk River. Newts were captured in a calm, shallow, soft-bottomed
inlet branching off from the Elk River (indicated by the arrow) located approximately 1 km east
of the Oregon Coast Highway (Highway 101). The field site was flanked by a wooded area on
the west and a sandy surface on the right. Part of the field site had floating vegetation, while
other parts were open. The main part of the Elk River, which has a rocky bottom and moderate
current, is separated from the field site by a sandy area. Multiple garter snakes (Thamnophis
sirtalis) were observed on the shore while sampling. Photo credit: Google Earth Pro 2017.

46
 

N

Figure 3. Aerial View of Phillips Farm Pond. Phillips Farm Pond (indicated by the arrow) is
located in Virgil Phillips Farm Park, a public park in Moscow, ID operated by Latah County
Parks and Recreation, with unpaved hiking trails, picnic areas, wetlands, forest, and open fields.
The western side of the pond is deep and permanent whereas the eastern side is shallower and
was partially dried out when we were collecting newts. The pond is surrounded by vegetation,
with some floating vegetation around the edge, open water in the middle, and a generally soft
bottom with rocks closer to the edge. Photo credit: Google Earth Pro 2017.

47
 

Figure 4. Relative Locations of Field Sites. Sandpiper Pond and Elk River were both located
along the Oregon Coast Highway (Highway 101), approximately 185 km apart. Sandpiper Pond
and Phillips Farm Pond are approximately 610 km apart and Elk River and Phillips Farm Pond
are approximately 930 km apart. Photo credit: Google Earth Pro 2017.

48
 

 
Figure 5. Toxicity of Philips Farm Pond Male Newts in the Cohabitation Experiment. The
toxicity of non-toxic males in mixed treatment tanks and non-toxic control tanks remained at 0
µg TTX/g skin over the course of the experiment. Values shown are means.

49
 

 
 
 
 
 

Figure 6. Toxicity of All Sandpiper Pond Male Newts in the Cohabitation Experiment. Data
were collected February – June and again in December. Values shown are means with standard
errors. 6a. Toxicity of all toxic males varied significantly over the course of the experiment (p =
0.00088, df = 59, F = 4.9).

50
 

Figure 7. Toxicity of Treatment and Control Sandpiper Pond Male Newts in the
Cohabitation Experiment. Data were collected February – June and again in December. Values
shown are means with standard errors. Significant temporal variation in toxicity was observed in
the toxic newts in the mixed-toxicity treatment (ANOVA, p = 0.00630, df = 35, F = 1.98, but not
in the control treatment (ANOVA, p = 0.13, df = 23, F = 1.98).

51
 

Sandpiper Female
Elk River Female
Elk River Male

Figure 8. Correlation between SVL and Toxicity in Toxic Newts. SVL and toxicity were
significantly correlated (R2 = 0.37, p = 0.022) for toxic newts in the tissue toxicity experiment.
Data shown here include Sandpiper Pond females (n = 2), Elk River females (n = 4), and Elk
River males (n = 8).

52
 

REFERENCES

53
 

REFERENCES
Archie EA, Theis KR (2011) Animal behaviour meets microbial ecology. Anim Behav 82:425–
436.
Armstrong JB, Duhon ST, and Malacinski GM (1989) Raising the axolotl in captivity. In JB
Armstrong and GM Malacinski (Ed.), Developmental Biology of the Axolotl (pp. 220-227).
New York: Oxford University Press.
Brodie, Jr. ED, Ridenhour BJ, Brodie III ED (2002) The evolutionary response of predators to
dangerous prey: Hotspots and coldspots in the geographic mosaic of coevolution between
garter snakes and newts. Evolution (N Y) 56:2067–2082.
Brodie ED (1968) Investigations on the Skin Toxin of the Adult Rough- Skinned Newt, Taricha
granulosa. 1968:307–313.
Brodie III ED, Brodie Jr ED (1999) Costs of exploiting poisonous prey: Evolutionary trade-offs
in a predator-prey arms race. Evolution 53:626–631.
Brodie ED III, Feldman CR, Hanifin CT, Motychak JE, Mulcahy DG, Williams BL, Brodie ED
(2005) Parallel arms races between garter snakes and newts involving tetrodotoxin as the
phenotypic interface of coevolution. J Chem Ecol 31:343–356.
Bucciarelli GM, Green DB, Shaffer HB, Kats LB (2016) Individual fluctuations in toxin levels
affect breeding site fidelity in a chemically defended amphibian. Proc R Soc B Biol Sci
283:20160468
Bucciarelli GM, Li A, Kats LB, Green DB (2014) Quantifying tetrodotoxin levels in the
California newt using a non-destructive sampling method. Toxicon 80:87–93
Cardall BL, Brodie ED, Brodie ED, Hanifin CT (2004) Secretion and regeneration of
tetrodotoxin in the rough-skin newt (Taricha granulosa). Toxicon 44:933–938.
Chau R, Kalaitzis JA, Neilan BA (2011) On the origins and biosynthesis of tetrodotoxin. Aquat
Toxicol 104:61–72.
Chieffi G, Rastogi RK, Milone M, Iela L (1980) Amphibian reproduction: reproductive
physiology in the male Rana esculenta L. Ital J Zool 47:63–70.
Davis WC, Twitty VC (1964) Courtship Behavior and Reproductive Isolation in the Species of
Taricha (Amphibia, Caudata). Copeia 1964:601–610.
Di Prisco CL, Delrio G, Chieffi G, Bellini Cardellini L, Polzonetti Magni A (1971) Identification
and biosynthesis of steroid hormones in the ovary and fat bodies of female Triturus cristatus
carnifex. Comp Biochem Physiol Part B Comp Biochem 40:53–60.
Dussourd DE, Harvis C, Meinwald J, Eisner T (1991) Pheromonal advertisement of a nuptial gift
54
 

by a male moth (Utetheisa ornatrix). Proc Natl Acad Sci U S A 88:9224–9227.
Dussourd DE, Ubik K, Harvis C, Resch J, Meinwald J, Eisner T (1988) Biparental defensive
endowment of eggs with acquired plant alkaloid in the moth Utetheisa ornatrix. Proc Natl
Acad Sci 85:5992–5996.
Eisner T, Meinwald J (1995) The Chemistry of Sexual Selection. Proc Natl Acad Sci U S A
92:50–55.
Fellers GM, Reynolds-Taylor J, Farrar S (2007) Taricha granulosa (Rough-skinned newt).
Predation. Herpetol Rev 38:317–318.
Gall BG, Stokes AN, French SS, Brodie III ED, Brodie Jr ED (2012a) Female newts (Taricha
granulosa) produce tetrodotoxin laden eggs after long term captivity. Toxicon 60:1057–
1062.
Gall BG, Stokes AN, French SS, Brodie III ED, Brodie Jr ED (2012b) Predatory Caddisfly
Larvae Sequester Tetrodotoxin from Their Prey, Eggs of the Rough-Skinned Newt (Taricha
granulosa). J Chem Ecol 38:1351–1357.
Geffeney S, Brodie Jr ED, Ruben PC, Brodie III ED (2002) Mechanisms of Adaptation in a
Predator-Prey Arms Race: TTX-Resistant Sodium Channels. Science (80- ) 297:1336–1339.
Geffeney SL, Fujimoto E, Brodie III ED, Brodie Jr ED, Ruben PC (2005) Evolutionary
diversification of TTX-resistant sodium channels in a predator-prey interaction. Nature
434:759–763.
Girish S, Saidapur SK (2000) Interrelationship Between Food Availability, Fat Body, and
Ovarian Cycles in the Frog,. J Exp Zool 493:487–493.
Gonzalez A, Rossini C, Eisner M, Eisner T (1999) Sexually transmitted chemical defense in a
moth (Utetheisa ornatrix). Ecology 96:5570–5574.
Hague MTJ, Avila LA, Hanifin CT, Snedden WA, Stokes AN, Brodie ED (2016) Toxicity and
population structure of the Rough-Skinned Newt (Taricha granulosa) outside the range of an
arms race with resistant predators. Ecol Evol 6:2714–2724.
Hanifin CT (2010) The chemical and evolutionary ecology of tetrodotoxin (TTX) toxicity in
terrestrial vertebrates. Mar Drugs 8:577–593.
Hanifin CT, Brodie III ED, Brodie ED (2002) Tetrodotoxin levels of the rough-skin newt,
Taricha granulosa, increase in long-term captivity. Toxicon 40:1149–1153.
Hanifin CT, Brodie III ED, Brodie Jr ED (2003) Tetrodotoxin levels in eggs of the rough-skin
newt, Taricha granulosa, are correlated with female toxicity. J Chem Ecol 29:1729–1739.
Hanifin CT, Brodie Jr ED, Brodie III ED (2008) Phenotypic mismatches reveal escape from
arms-race coevolution. PLoS Biol 6:0471–0482.
55
 

Hanifin CT, Gilly WF (2015) Evolutionary history of a complex adaptation: Tetrodotoxin
resistance in salamanders. Evolution (N Y) 69:232–244.
Hanifin CT, Yotsu-Yamashita M, Yasumoto T, Brodie III ED, Brodie Jr ED (1999) Toxicity of
dangerous prey: variation of tetrodotoxin levels within and among populations of the newt
Taricha granulosa. J Chem Ecol 25:6–11.
Hopkins CCE (1978) The male genital system, and spermatophore production and function in
Euchaeta norvegica Boeck (Copepoda: Calanoida). J Exp Mar Bio Ecol 35:197–231.
Itoi S, Yoshikawa S, Asahina K, Suzuki M, Ishizuka k (2014) Larval pufferfish protected by
maternal tetrodotoxin. Toxicon 78:35–40.
Itoi S et al. (2016) Seasonal changes in the tetrodotoxin content of the pufferfish Takifugu
niphobles. Toxicon 114:53–58.
Iyengar VK, Eisner T (1999) Female choice increases offspring fitness in an arctiid moth
(Utetheisa ornatrix). Proc Natl Acad Sci U S A 96:15013–15016.
Iyengar VK, Rossini C, Eisner T (2001) Precopulatory assessment of male quality in an arctiid
moth "Utetheisa ornatrix": Hydroxydanaidal is the only criterion of choice. Behav Ecol
Sociobiol 49:283–288.
Janzen FJ, Brodie ED (1989) Tall Tails and Sexy Males: Sexual Behavior of Rough-Skinned
Newts (Taricha granulosa) in a Natural Breeding Pond. Copeia 1989:1068–1071.
Jones AG, Adams EM, Arnold SJ (2002a) Topping off: A mechanism of first-male sperm
precedence in a vertebrate. Proc Natl Acad Sci U S A 99:2078–2081.
Jones AG, Arguello JR, Arnold SJ (2002b) Validation of Bateman’s principles: a genetic study
of sexual selection and mating patterns in the rough-skinned newt. Proc R Soc Biol Sci Ser
B 269:2533–2539.
Jorgensen CB (1986) External and internal controls of patterns of feeding, growth and gonadal
function in a temperate zone anural, the toad Bufo bufo. J Zool 210:211–214.
Jost MC, Hillis DM, Lu Y, Kyle JW, Fozzard HA, Zakon HH (2008) Toxin-resitant sodium
channels: parallel adaptive evolution across a complete gene family. Mol Biol Evol
25:1016–1024.
Keizer RJ, Jansen RS, Rosing H, Thissen B, Beijnen JH, Schellens JH, Huitema AD (2015)
Incorporation of concentration data below the limit of quantification in population
pharmacokinetic analyses. Pharmacol Res Perspect 3:1-15.
Kodama M, Sato S, Ogata T, Suzuki Y, Kaneko T, Aida K (1986) Tetrodotoxin secreting glands
in the skin of puffer fishes. Toxicon 24:819–829.
56
 

Kuchta SR, Tan A-M (2005) Isolation by distance and post-glacial range expansion in the roughskinned newt, Taricha granulosa. Mol Ecol 14:225–244.
Lago J, Rodriguez LP, Blanco L, Vieites JM, Cabado AG (2015) Tetrodotoxin, an extremely
potent marine neurotoxin: Distribution, toxicity, origin and therapeutical uses. Mar Drugs
13:6384–6406.
LaMunyon CW, Eisner T (1993) Postcopulatory sexual selection in an arctiid moth (Utetheisa
ornatrix). Proc Natl Acad Sci U S A 90:4689–4692.
Leclaire S, Nielsen JF, Drea CM (2014) Bacterial communities in meerkat anal scent secretions
vary with host sex, age, and group membership. Behav Ecol 25:996–1004.
Lee CH, Ruben PC (2008) Interaction between voltage-gated sodium channels and the
neurotoxin, tetrodotoxin. Channels 2:407–412.
Lehman EM, Brodie ED, Brodie ED (2004) No evidence for an endosymbiotic bacterial origin of
tetrodotoxin in the newt Taricha granulosa. Toxicon 44:243–249.
MacCracken JG, Stebbings JL (2012) Test of a Body Condition Index with Amphibians. J
Herpetol 46:346–350.
Madelaire CB, Gomes FR (2016) Breeding under unpredictable conditions: Annual variation in
gonadal maturation, energetic reserves and plasma levels of androgens and corticosterone in
anurans from the Brazilian semi-arid. Gen Comp Endocrinol 228:9–16.
Matsumoto T, Nagashima Y, Kusuhara H, Sugiyama Y, Ishizaki S, Shimakura K, Shiomi K
(2007) Involvement of carrier-mediated transport system in uptake of tetrodotoxin into liver
tissue slices of puffer fish Takifugu rubripes. Toxicon 50:173–179.
Matsumura K (1995) Tetrodotoxin as a pheromone. Nature 378:563–564.
Mccurdy HM (1931) Development of the sex organs in Triturus torosus. Dev Dyn 47:367–403.
Mebs D, Yotsu-Yamashita M, Ream J, Zajac BK, Zehner R (2016) Tetrodotoxin concentrations
in rough-skinned newts, Taricha granulosa, from populations of their northern distribution
range. Salamandra 52:255–260.
Mills LC, Rombough CJ (2008) Taricha granulosa (Rough-skinned Newt) reproduction.
Herpetol Rev 39:336–337.
Miyazawa K, Noguchi T (2001) Distribution and Origin of Tetrodotoxin. J Toxicol Toxin Rev
20:11–33.
Moczydlowski EG (2013) The molecular mystique of tetrodotoxin. Toxicon 63:165–183.
Noguchi T, Arakawa O, Takatani T (2006) TTX accumulation in pufferfish. Comp Biochem
57
 

Physiol Part D Genomics Proteomics 1:145–152.
Nussbaum RA, Brodie Jr ED (1971) The Taxonomic Status of the Rough-Skinned Newt, Taricha
granulosa (Skilton), in the Rocky Mountains. Herpetologica 27:260–270.
Okita K, Yamazaki H, Sakiyama K, Yamane H, Niina S, Takatani T, Arakawa O, Sakakura Y
(2013) Puffer smells tetrodotoxin. Icthyological Res 60:386–389.
Propper CR (1991) Courtship in the rough-skinned newt Taricha granulosa. Anim Behav
41:547–554.
Ridenhour BJ, Brodie ED, Brodie ED (2007) Patterns of genetic differentiation in Thamnophis
and Taricha from the Pacific Northwest. J Biogeogr 34:724–735.
Salvitti L, Wood SA, Fairweather R, Culliford D, McNabb P, Cary SC (2017) In situ
accumulation of tetrodotoxin in non-toxic Pleurobranchaea maculata (Opisthobranchia).
Aquat Sci 79:335–344.
Song SJ, Lauber C, Costello EK, Lozupone CA, Humphrey G, Berg-Lyons D, Gregory Caporaso
J, Knights D, Clemente JC, Nakielny S, Gordon JI, Fierer N, Knight R (2013) Cohabiting
family members share microbiota with one another and with their dogs. Elife 2013:1–22.
Stokes AN, Ray AM, Buktenica MW, Gall BG, Paulson E, Paulson D, French SS, Brodie III ED,
Brodie Jr ED (2015) Otter Predation on Taricha granulosa and Variation in Tetrodotoxin
Levels with Elevation. Northwest Nat 96:13–21.
Uzzell T (1969) Notes on Spermatophore Production by Salamanders of the Ambystoma
jeffersonianum Complex. Copeia 1969:602–612.
Wakely JF, Fuhrman GJ, Fuhrman FA, Fischer HG, Mosher HS (1966) The occurrence of
tetrodotoxin (tarichatoxin) in amphibia and the distribution of the toxin in the organs of
newts (taricha). Toxicon 3:195–203.
Whittaker DJ, Gerlach NM, Slowinski SP, Corcoran KP, Winters AD, Soini HA, Novotny M V,
Ketterson ED, Theis KR (2016) Social Environment Has a Primary Influence on the
Microbial and Odor Profiles of a Chemically Signaling Songbird. Front Ecol Evol 4:90.
Williams BL (2010) Behavioral and chemical ecology of marine organisms with respect to
tetrodotoxin. Mar Drugs 8:381–398.
Williams BL, Hanifin CT, Brodie ED, Brodie ED (2010) Tetrodotoxin affects survival
probability of rough-skinned newts (Taricha granulosa) faced with TTX-resistant garter
snake predators (Thamnophis sirtalis). Chemoecology 20:285–290.
Williams BL, Hanifin CT, Brodie ED, Brodie ED (2012) Predators usurp prey defenses?
Toxicokinetics of tetrodotoxin in common garter snakes after consumption of roughskinned newts. Chemoecology 22:179–185.

58
 

Wilson NJ, Stokes AN, Hopkins GR, Brodie, Jr. ED, Williams CR (2014) Functional and
physiological resistance of crayfish to amphibian toxins: tetrodotoxin resistance in the white
river crayfish ( Procambarus acutus ). Can J Zool 92:939–945.
Wu Z, Yang Y, Xie L, Xia G, Hu J, Wang S, Zhang R (2005) Toxicity and distribution of
tetrodotoxin-producing bacteria in puffer fish Fugu rubripes collected from the Bohai Sea of
China. Toxicon 46:471–476.
Yotsu-Yamashita M, Gilhen J, Russell RW, Krysko KL, Melaun C, Kurz A, Kauferstein S,
Kordis D, Mebs D (2012) Variability of tetrodotoxin and of its analogues in the red-spotted
newt, Notophthalmus viridescens (Amphibia: Urodela: Salamandridae). Toxicon 59:257–
264.
Zimmer RK, Schar DW, Ferrer RP, Krug PJ, Kats LB, Michel WC, Monographs E, Ferrer P,
Krug J, Kats B (2006) The Scent of Danger : Tetrodotoxin ( TTX ) as an Olfactory Cue of
Predation Risk. Ecol Monogr 76:585–600.
 

59