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
/
Studies in bivalve aquaculture: metallotoxicity, microbiome manipulations, and genomics & breeding programs with a focus on mutation rate
(USC Thesis Other)
Studies in bivalve aquaculture: metallotoxicity, microbiome manipulations, and genomics & breeding programs with a focus on mutation rate
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Studies in Bivalve Aquaculture: Metallotoxicity, Microbiome Manipulations, and
Genomics & Breeding Programs with a Focus on Mutation Rate.
by
Nathan Daniel Churches
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Department of Molecular Biology)
May 2020
ii
ACKNOWLEDGEMENTS
The research presented here was first and foremost supported by my advisor, Dr. Sergey
Nuzhdin. His appetite for new and exciting science, even before an avenue or structure is in
place in which to perform said science, is the driving force behind my work. Without his support
and confidence in me, this effort would not have been possible. Considering that the idea to
begin a bivalve aquaculture science program on Catalina was little more than a hazy idea when
first presented several years ago, it is amazing to see how far the facilities and lab have come.
Importantly, in my mind, it is also fantastic to have been involved in the spearheading of a new
direction in the Nuzhdin lab that may be pursued for many years to come, specifically in the
aquatic sciences. I am extremely grateful for that opportunity, and look forward to the lab’s
output in that field after my graduate student tenure. In addition to Sergey, there are many people
in his laboratory that have guided me, and I am so lucky to have experienced an environment of
support during my studies. Namely, these people are Brad Foley, Mohammad Abbasi, Josh
Schiffman, Wendy Vu, Peter Chang, Hamdi Kitapci, Gary Molano, Levi Simmons, Maddelyn
Harden, and Anupam Singh. Other colleagues at USC that aided me greatly in my scientific
learning include Dr. Meghan Hall, Dr. Andrew Gracey, Dr. Ian Ehrenreich, Dr. Dean Pentcheff,
Dr. Regina Wetzer, and Dr. Dennis Hedgecock. I would also like to thank Dr. Matt Dean, who
never fails to engage me in interesting scientific discussions, and for his friendship. Thanks for
convincing me to come to USC, Dr. Dean!
Two standouts in my early academic career need to be mentioned, as I truly don’t think I would
be where I am without these key people. Dr. Billie Swalla, thank you for being such an awesome
person, mentor, and friend. I am so lucky to have met you, and consider you among my most
influential mentors. Dr. Mark Ainsworth at Seattle Central Community College, thank you for
making introductory biology so, so interesting for me. Your teaching style and enthusiasm
inspire me as an educator.
In regards to building new hatchery facilities and demonstrating rigorous and thoughtful
approaches to aquaculture management, I cannot thank Dave Anderson enough for mentoring me
iii
(though he would probably decline ever having such a role). I would not be where I am today
without his workmanship as a template for which to strive. Another majorly influential person to
me in this regard is Kelly Stromberg, who showed me everything there is to know about bivalve
animal husbandry, and always with a smile. Dr. Diane Kim has also been a great friend and
colleague, and I look forward to many future projects with her and Kelly. Bernard Friedman,
Thomas Grimm, Gordon King, and Joth Davis are only a few of the industry folks to which I
owe some of my successes. Thank you all for showing me, sometimes literally, the ropes. Many
thanks also to the Waitt Foundation and Paso Pacifico, especially Cherie Prothro and Sarah
Otterstrom, for helping fund my work and invest time into worthy causes.
Each chapter of my graduate work was performed in some part, or in some cases entirely, at the
Wrigley Institute for Environmental Science. There are several people there who supported me
with both facilities help and friendship, including Kellie Spafford, Lorraine Sadler, Chase
Puentes, Vivian Kim, Chad Bertrum, Captain Gord, Maurice Roper, Juan Aguilar, Phil Lopez,
Trevor and Lauren, Josh, Pat, Karl Huggins, Holly Gayler, Karen Erickson, and Sean Connor.
Thank you all for making my work on the island possible!
As a scientific diver, I have a great deal of gratitude for Eric Castillo and Chad Bertrum, for
taking a chance on a total beginner and getting me up to speed in record time. My love for the
ocean has deepened dramatically since learning to dive, thank you! Other dive partners I would
like to thank are Maurice and Ted Sharshan, Riley and Tristan, Andy, and anybody else who
decided to take a ‘boring’ work dive with me (though some were more interesting than planned:
Maurice, the barracuda are still running I hear!). I’d also like to thank the CSUN and CSULB
divers that shared dives and dive lockers with me; their divers, too numerous to name
individually, demonstrated incredible watermanship and poise.
Anybody who knows me well understands that my life revolves around balance, and the work I
do in science and aquaculture is counterweighted by a rich personal life. I don’t think it’s
possible, or healthy, to perform at a high level of academia without friends and interests outside
of your field, and so it is merited to thank people in my personal life here. These are my climbing
partners, surfing buddies, music friends, backpacking companions, and family. Again, too
iv
numerous to name all of them, but I’ll try here…Thank you to my core Olympia friends: Aaron,
Andrew, and Eric Buchelle, Matt Bacon, Nathan Furst-Nichols, Nick Briganti, Brandon Clarno,
Forrest Gates, Andy Dahlen, DJ Jenkins, Joe Wheeler, and Mitch Piper. In Los Angeles, thanks
to Sean & Patricia, Gary & Helen, Chad & Emily, Todd, and the rest of the Hangar 18 climbing
community. To Brad and Alison for being the best roommates ever. Thank you to my
grandparents, Al and Ina Smith, for being two of the most badass people I know. Thank you to
my family Weston and Patty, Jen and Reed, Jewel and Russ, and all my cousins for always
drawing me back to beautiful Utah. For always reminding me that art is important, and for a
connection too deep to describe here, I want to thank my sister Shayla Churches. I also would
like to thank my parents, Dan and Lorinda Churches, for instilling in me a passion for learning
and athletic/healthy pursuits from a young age, for supporting me forever, and for loving me
unconditionally. Finally, a second dissertation would be required to express the depth of my
appreciation, gratitude, and love for Celja Jae Uebel. You make me a better person, thank you!
v
Table of Contents
ACKNOWLEDGEMENTS ............................................................................................................. II
THESIS INTRODUCTION .............................................................................................................. 1
WHY STUDY AQUACULTURE? .......................................................................................................... 1
A FEW IMPORTANT AREAS OF BIVALVE AND BIVALVE AQUACULTURE SCIENCE ......................... 5
CHAPTER 1: METALLOTOXICITY BEHAVIORAL RESPONSE IN BIVALVES ........................................ 5
CHAPTER 2: MICROBIOME MANIPULATIONS .................................................................................... 7
CHAPTER 3: MUTATION RATE ESTIMATES ....................................................................................... 9
THE FACILITIES ............................................................................................................................... 11
CHAPTER 1: METALLOTOXICITY BEHAVIORAL RESPONSE IN BIVALVES ...................................... 15
ABSTRACT ....................................................................................................................................... 16
INTRODUCTION ................................................................................................................................ 17
METHODS ........................................................................................................................................ 19
RESULTS .......................................................................................................................................... 26
DISCUSSION AND FUTURE DIRECTIONS .......................................................................................... 27
CONCLUSIONS ................................................................................................................................. 33
FIGURES ........................................................................................................................................... 34
CHAPTER 2: MICROBIOME INVESTIGATIONS IN THE PACIFIC OYSTER ............. 37
ABSTRACT ....................................................................................................................................... 37
INTRODUCTION ................................................................................................................................ 37
RESULTS .......................................................................................................................................... 41
DISCUSSION ..................................................................................................................................... 44
CONCLUSION ................................................................................................................................... 47
MATERIALS AND METHODS ............................................................................................................ 48
ACKNOWLEDGEMENTS .................................................................................................................... 55
FIGURES ........................................................................................................................................... 55
SUPPLEMENTARY FIGURES .............................................................................................................. 59
CHAPTER 3: MUTATION RATE ESTIMATES IN THE PACIFIC OYSTER* ................ 69
ABSTRACT ....................................................................................................................................... 69
INTRODUCTION ................................................................................................................................ 70
METHODS ........................................................................................................................................ 74
RESULTS .......................................................................................................................................... 77
DISCUSSION ..................................................................................................................................... 80
FIGURES AND TABLES ..................................................................................................................... 90
SUPPLEMENTAL CHAPTERS ................................................................................................. 106
SUPPLEMENTAL CHAPTER 1: GENERATION OF MUSSEL (MYTILUS
GALLOPROVINCIALIS) FAMILY LINES, IN CONCERT WITH RELEVANT
PHENOTYPING SCHEDULE, IN ORDER TO ELUCIDATE VIABILITY- AND
GROWTH-ASSOCIATED BIOMARKERS. ............................................................................ 107
BACKGROUND ............................................................................................................................... 107
METHODS ...................................................................................................................................... 109
RESULTS ........................................................................................................................................ 114
vi
DISCUSSION ................................................................................................................................... 116
FIGURES ......................................................................................................................................... 118
SUPPLEMENTAL CHAPTER 2: SPERM AND EGG CHEMOATTRACTION STUDIES:
A SERIES OF PRELIMINARY EXPERIMENTS SUGGESTS THE CASE FOR EGG
COMPETITION IN A MARINE DIOECIOUS SIMULCASTER ........................................ 124
INTRODUCTION .............................................................................................................................. 124
METHODS ...................................................................................................................................... 127
RESULTS ........................................................................................................................................ 131
DISCUSSION ................................................................................................................................... 135
FIGURES ......................................................................................................................................... 143
REFERENCES ............................................................................................................................... 150
SUPPLEMENTAL DOCUMENTS ............................................................................................. 174
BIVALVE DNA EXTRACTION PROTOCOL ..................................................................................... 174
LIBRARY GENERATION PROTOCOL: BIVALVE SPECIFIC DDRAD ................................................ 177
1
THESIS INTRODUCTION
The word aquaculture simply means to grow organisms, animal or plant, from a body of
water, fresh or salt. Because this term is inherently so large, it is important to narrow the field
when considering anything related. During my Ph.D. studies, I focused primarily on marine
bivalve aquaculture academics. Its techniques, processes, and of course state of science. In this
introductory section I describe firstly why bivalve aquaculture is an important and worthy field
of study, and follow with a more general background for my specific thesis focus areas. It is my
intention that this work demonstrates that combining traditional experimental science and
genomics techniques can provide the necessary insights in order to further develop bivalves as an
ecological model species and a (delicious) sustainable food source.
Why Study Aquaculture?
It is simple: wild fisheries are expected to collapse globally within the next 100 years if
current rates of over-exploitation continue (Worm et al., 2006 (see figure 1 below); Jaenike J,
2007). In the US, fisheries have plateaued in the last 20 years, with ~10 billion pounds landed in
both 1995 and 2015, and fluctuating little in-between (Fisheries of the US, 2015, NMFS). This
plateau is due primarily to the aforementioned declining of seafood wild-stocks and subsequent
regulation of fragile fisheries. Demand for seafood, however, has continued to rise,
approximately 7-10% for some of the fastest growing categories (Ruggles, 2018). The only
solution to balance this as-now unbalanced equation is to meet demand via an increase in global
aquaculture production. In the late aughts and early 2010s, seafood from aquaculture exceeded
that of wild-caught sources for the first time (FAO, 2018) on a global scale, a trend unlikely to
2
change in the future. But not all aquaculture is created equal; some forms can be detrimental to
the very ocean it is seeking to protect.
One of the most sustainable
options to generate proteins from the
ocean via aquaculture are them from
the class Bivalvia, two-shelled varieties
of mollusk. This class includes, clams,
cockles, mussels, oysters, scallops, and
geoducks, among others. Bivalves have
worldwide distribution, which means
that just about any area on the planet
can find a suitable native organism to farm (e.g. Gaitan-Espita et al. 2016). They require zero
fresh water input, which is an increasingly critical resource for many areas as climate change
increases drought periods (Rogers 2018). As they naturally occur in extreme densities, they do
not incur increased disease risk from intensive farming and therefore also do not necessitate the
use of antibiotics, fertilizers, or anti-fouling chemicals. Finally, they alleviate any need to harvest
“feed fish” (which are often unsustainable fisheries in and of themselves), as bivalves filter feed
ambient plankton from the surrounding environment.
On a national scale, however, the US is far behind when it comes to aquaculture
production of most varieties, including bivalve aquaculture. Despite rising demand, the value of
the US-produced Seafood market has remained relatively unchanged for the last two decades –
$5B in 1995 and $4.8B in 2015, after correcting for inflation (Fisheries of the US, 2015,
NOAA). The lack of growth in the seafood and aquaculture sectors contribute to the nation’s
Figure 1. Trajectories of collapsed fish and invertebrate taxa
over the past 50 years (diamonds, collapses by year; triangles,
cumulative collapses). Data are shown for all (black), species-
poor (500 species, red) LMEs. Regression lines are best-fit
power models corrected for temporal autocorrelation (figure and
description from Worm et al. 2006, their figure 3A).
3
increasing seafood deficit, currently estimated to exceed $14B annually (Fisheries of the US,
2015, NOAA). For bivalves, the US contributions to the global shellfish supply remain at ~3% or
less (FAO 2016). For example, the US harvested over 574k tons of mollusks in 2016, which only
represents 3.3% of global 2016 mollusk totals (17.1M tons, net value of $29.2BUSD; FAO
2016). In fact, shellfish remain among the most imported aquaculture products into the US, with
Canada, Chile, and China representing a majority of the import sources (Imports and Exports,
Annual Summary, NOAA 2017). In addition to baseline economic inefficiencies associated with
importing food, there come additional risks in terms of ecological impact. Importing food from
other countries weakens or eliminates the potential for US environmental oversight in areas of
production origin, and imported foods inherently carry a higher greenhouse gas footprint.
Nationally, there is major ground to cover if import/export imbalances are to be
mitigated, but we certainly have the space: the US has the world’s largest coastal and open water
economic development areas (dubbed the Exclusive Economic Zone, EEZ), at 11.4 km
2
(4.4
million mi
2
, 281M acres) (The Economist, 2016). Currently, bivalve aquaculture production
accounts for only a miniscule fraction of acreage-use in the EEZ; California utilizes ~6000 acres
of its 130M available in the EEZ, or roughly 0.0046%. A majority of the acreage for aquaculture
is leased on the East Atlantic and Gulf coasts, which produce 64% of US aquaculture, while the
US West Coast accounts for the remaining 36%, worth $504M (Fisheries of the United States,
2016, NOAA). This equates to $0.44 per acre annually in the US West Coast EEZ from
sustainably produced aquaculture (CA, WA, and OR: 318,746 mi
2
’; AK: 1,445,613 mi
2
;
Combined: 1,764,359 square miles = 1.3B acres; NOAA 2018; 2013 USDA Census of
Aquaculture). Even when all seafood production (wild catch and aquaculture) is considered, we
only produce $4.84 per acre in the US West Coast EEZ, including Alaska (NOAA, 2018). In
4
contrast, in 2017 California produced $50B of agricultural goods across California’s 43M farm-
slated acres, resulting in an average of $1,162 generated per acre annually (California
Department of Food and Agriculture, 2018). The state of California is therefore 240x more
efficient at producing terrestrial crops than the entire US West Coast is at producing seafood of
any variety from the EEZ. But this estimate includes wild caught seafood. From a strictly farmed
perspective, we are over 2500x as efficient in growing food from land in California as we are
from growing food (aquaculture alone) in the entire US West Coast ocean arena.
This is perhaps using the data to paint too dark a picture, because when we consider the
dollar amount generated per acre which is actually utilized by aquaculture, we see the US West
Coast has 24,489 aquaculture-slated acres in total (15,283 acres in WA, 3539 in OR, 5573 in CA,
94 in AK) which produced $504M, therefore averaging $20,580 per acre annually (NOAA,
2018). When the ocean is purposefully allocated for aquaculture use, it is clearly quite profitable!
The US can benefit from increased bivalve aquaculture production and ocean-farmed
crop development. Indeed, various stakeholders are recognizing the need and opportunity in this
sector, and current expansion plans for CA include an increase in acreage at Santa Barbara
Mariculture (from 25 to 72 acres), a second 172-acre farm in Santa Barbara (Cassidy Teuful,
personal communications, September 2018), and 2,000 acres near Ventura Harbor (all projected
within 1-3 years, Ventura Shellfish Enterprises). These expansions will by no means fully
alleviate the $14B seafood deficit, but they are a good start. And because bivalves are essentially
wild organisms, a rigorous science-based approach to their farming will be necessary if we plan
to increase protein production and market phenotypes via selective breeding and domestication.
5
A Few Important Areas of Bivalve and Bivalve Aquaculture Science
As incoming bivalve enthusiasts, we should also consider the rich body of literature that
can inform healthy expansions in the industry and continue to support ecological understanding
of these important organisms. There is not the space to review this extensive literature in this
thesis, but I recommend ‘Bivalve Molluscs’ by Elizabeth Gosling (2003) for a fantastic biology
and physiology reference, ‘The Mollusks’ by Sturm, Pearce, and Valdez (2006) for a more
extensive primer on the general state of molluscan research, and ‘the Hatchery Culture of
Bivalves, Practical Manual’ by Helm and Bourne (2004) for an almost religious text for
scientists needing hatchery-relevant physiology, biological, and rearing systems information.
These are only a few of the building blocks of this field, and there are myriad questions
remaining that will help inform a sustainable expansion of a bivalve seafood industry. With the
above said, I focused on a few main areas in my graduate studies and will briefly expand on
them here. Each sub-topic has a more in-depth and fully cited introduction in the subsequent
chapters. Together, I hope my graduate work adds to the general knowledge in the field of
bivalve biology, such that it might aid in the development of a sustainable aquaculture
expansion, and inform evolutionary and ecological models for these interesting animals.
Chapter 1: Metallotoxicity Behavioral Response in Bivalves
Dissolved bio-available metals can impact marine systems in dramatic ways, especially
when introduced via anthropogenic routes and in excess of natural quantities. Water quality and
dissolved metal content and their interaction with organisms are among the most important
dynamics to understand in terms of choosing appropriate aquaculture sites. Metallotoxicity, in
low-to-moderate-severity, can be detrimental to organisms that are ‘low’ on the trophic scale, or
6
have a life history that includes planktonic phases (such as bivalves). In high-severity scenarios,
metals can be toxic to macro-organisms, including vertebrates (such as humans), as
metallotoxins can accumulate as trophic level increases (bioaccumulation). Thus, even organisms
that are seemingly unaffected by metal pollution may contribute, or be affected by, an
ecosystem-wide metal accumulation issue. An outcome we’re all familiar with is the
implementation of a series of government issued warnings about the appropriate amount of
seafood to consume in order to avoid mercury poisoning.
The effects of metallotoxicity to organisms are seemingly pan-physiological, depending
on the species of study, affecting everything from growth, stress protein (hsp) production,
immune function, and, ultimately, survival (references 1-5, 19, chapter 1). For a Southern
Californian example, the Port of San Diego has implemented a series of efforts to reduce the
amount of copper introduced to the San Diego Bay, due to high levels of recovered ionic copper
compounds. Copper, in particular, has been shown to be damaging to marine organisms, and is
introduced to marine environments primarily via car brakes and subsequent run-off, and anti-
fouling boat hull paint.
Early in my graduate career, I was introduced to Dr. Megan Hall (then a graduate
student), and the work she was undertaking in Dr. Andrew Gracey’s lab concerning copper
metallotoxicity in the mussel Mytilus californianus. She was demonstrating that exposure to the
bio-available ion Cu
2+
was capable of shifting RNA expression profiles in larval cultures of these
mussels, and dramatically reduced survival after exposure. Dr. Hall graciously mentored me in
the scientific application of bivalve hatchery techniques, and so we began having discussions
about the deformity of bivalve larvae when exposed to copper, and the literature concerning
bivalve larval biology and physiology. Around the same time period, Dr. Mohammad Abbasi
7
(also then a graduate student), had developed a motion sensing software in the Nuzhdin lab,
capable of tracking and quantifying movement in fruit flies. Thus, eventually the idea to quantify
swimming behavior in isogenic families exposed to copper toxicity came about. Feeding,
swimming, and growth are critical and highly correlated behaviors for bivalve larvae (pre-
metamorphosis). So, the hypothesis became that swimming behavior would be affected by
copper metallotoxicity, and henceforth so would growth, feeding, and survival be likewise
affected. This chapter presents the work done to investigate this idea, which applied a new
technique to quantifying swimming behavior in bivalves, further supported the scientific
literature regarding reduced larval survival in metallotoxic conditions, and suggested that
familial response may differ in regards to swimming behavior in copper rich environments.
Chapter 2: Microbiome Manipulations
Microbiomes and microbiome-host interaction research are experience a surge in
scientific popularity at the moment, and duly so. Host and microbe interactions have recently
been shown to affect dietary preference in humans, disease resistance in plants, and commercial
trait parameters in livestock, to name just a few exciting phenomenon. This field is sure to
uncover more exciting discoveries in the coming years. Sergey and I became interested in the
notion that bivalves may have interesting characteristics associated with their microbiomes after
reviewing plant literature, and associated crop improvements via microbial manipulations. Some
traits that are desirable to humans in their plant communities are reliably increased by
encouraging specific root-bacteria interaction via soil inoculations. Tomatoes, for example,
grown in poor soil (phosphorous deficient) were shown to experience faster growth when
supplemented with specific bacterial inoculants, which in turn affected the soil and rhizosphere
8
microbiome profile, and resulted in increased nutrient uptake in the tomato. Plants receive
nutrients from the soil and rhizobia, so it makes sense that altering this would affect plant health,
but the pattern of microbiomes affecting phenotype holds for terrestrial farm animals too (e.g.
see chapter 2 references: Mackie, R. I., 2002; Sommer and Backhed, 2013; Mao et al., 2015;
Nakamura et al., 2017; Myer et al., 2017). Could the same phenomenon be observed, and
possible harnessed for commercial gain, in bivalves as well?
Microbiome studies in bivalves are not as abundant in the literature, though we do know
that phenotypes might be influenced by microbiome manipulations, and harmful algal blooms
may be predicted by observing microbiome shifts in sentinel species (chapter 2 references:
Douillet et al., (1994); and Givens et al., (2014), respectively). Among that which remains
unknown is the capacity in which bivalve commercially relevant phenotypes (most importantly
growth and survival) could be affected by inoculation at a young age, and the ability of bivalves
to differentiate their own microbiomes from that of the outside water-world. Considering that a
bivalve can syphon 1.5 L per hour gram
-1
of seawater per day (the equivalent of a human
drinking an Olympic size swimming pool each day), it becomes clear that a sedentary, filter-
feeding, and aquatic life-history may present significant challenges to keeping distinct microbial
communities (Jorgensen 1996). Because this concept remained understudied, we decided to use
the tank system I built on Catalina Island to investigate the stability of bivalve microbiomes (see
facilities, below). This was seen as an interesting question because if bivalve microbiomes are
stable and/or manipulatable, then inoculants may work well for bivalves as a supplement, and
possibly boost (or supersede) family-line oriented breeding programs.
Additionally, humans consume many bivalves raw and in their entirety, so understanding
microbes in our food becomes valuable from a consumption safety standpoint. However, before
9
any implementation of a large-scale inoculant or probiotic (a.k.a. ‘yogurt for oysters’, joke a la
Levi Simons) would ever be undertaken by a commercial hatchery, basic scientific
understanding of bivalve microbiota needed to be generated. Chapter 2 focuses on this work, and
ultimately demonstrated that oyster gut-associated microbiomes are both manipulable and
changeable, yet separated in their community structure from that of the surrounding water. The
fact that oyster microbiomes were differentiating (on the order of days) suggests that bivalves
may employ a strategy in maintaining digestion-associated communities that align with their life-
history, and is a distinctly different approach than that which has been evolved by mobile,
higher-order organisms.
Chapter 3: Mutation Rate Estimates
Bivalves have an interesting phenomenon which has been documented in their natural
populations, which is that their effective population size, Ne, is much lower than would be
expected considering their actual population size, N (see chapter 3 for references). This
observation has been explained to be a result of a theory dubbed the “Sweepstakes Reproductive
Success” hypothesis, which states, essentially, that marine broadcast spawners have such a
difficult task in mating, in terms of coordinating gametogenesis, gamete release, and
oceanographic conditions, that the chance of any two individuals successfully mating is similar
to that of a sweepstakes. Remembering that marine organisms (and Pacific oysters in particular)
can be extremely fecund (many millions of gametes per spawn cycle), SRS posits that lower than
expected Ne is due to a few lucky ‘winners’ having the chance to repopulate a reef in any given
year, while most remain unsuccessful at mating.
10
In another branch of the ‘bivalves are odd’ literature, it has been hypothesized that
oysters might harbor many deleterious recessive alleles (~14-15 per individual), which interact
negatively with individuals during highly physiologically complex larval stages, are lethal for
many genotypes in the zygotic stage (via direct or epistatic interactions), and are responsible for
large deviations from expected segregation ratios in the resulting surviving cohort (chapter 3
references: Launey and Hedgecock, 2001; Plough and Hedgecock, 2011). The prevailing view is
that viability loci are therefore likely influencing the observation of inbreeding depression seen
in F2 Pacific oysters and other oysters (chapter 3 reference: Bierne, et al., 1998), and
furthermore, that the species might harbor a “greater genetic load than any animal species thus
far studied” (quote from Plough and Hedgecock, 2011). This latter statement, regarding genetic
load, had not yet been investigated by means of a trio analysis in order to estimate the
generational mutation rate.
How do these phenomena (SRS hypothesis, high genetic load, and generational mutation
rate) interact, if at all? Is it possible that a high mutation rate plays a larger role than previously
considered for observed low Ne? And what does this mean for farmers, who need to attempt to
create domesticated varieties of bivalves? Bivalve aquaculture will be greatly impacted by the
answers to these questions. If low Ne is found in localized waters where broodstock might be
sourced, and inbreeding depression is typically observed by F2, and this is due in part to an as-yet
unconfirmed high mutation rate, then the industry will need to take steps to safeguard its crop
and breeding strategies accordingly. In this chapter I attempt to shed some light on the
generational mutation rate of the Pacific oyster by sequencing 4 sets of trios in 4 families, for a
total of 16 offspring genotypes. Our results indicate that an extremely high mutation rate (1 x 10
-
5
per site per generation), consistent with observations of low Ne observations and the neutral
11
theory of evolution, may be occurring in Pacific oysters. This work is currently being drafted for
a manuscript and is the only primary chapter (i.e. not supplemental) which has not been peer
reviewed as of the time of the publication of this thesis.
The Facilities
The above projects obviously would not be possible without adequate facilities. Through
the funding of a NOAA SBIR and from a grant from the Waitt Foundation, I was able to build
out a hatchery space at the Wrigley Institute for Environmental Science. This process took a
considerable amount of my graduate school efforts, as I needed to manufacture and engineer the
parts to suit my needs. A brief description of the facilities I built follows here.
The general idea behind the hatchery space was to create a system with multiple tanks
that could be used for rearing bivalve larvae, juveniles, or hold adult broodstock. The impetus
came from the first few mussel families I reared after arriving in graduate school in 2014, which
took an incredible amount of effort by simply using stand-alone clear buckets. If I intended to
scale up family lines and holding tanks to accommodate more than 4 or 5 families, I would need
to streamline the process. Inherent challenges come with this, including the problem of keeping
conditions as similar as possible between tanks (food, temperature, salinity, pH, etc.) while
keeping the very tiny organisms in each individual tank isolated. In order to accomplish this
goal, I consulted with industry colleagues, the Helm and Bourne (2004) bivalve manual, and
organized an industry-to-academia aquaculture conference. After this work, I designed a new
system from the ideas gleaned. The Wrigley Institute was kind enough to let me use space in
their new-as-of-2015 “Blue House” to generate the hatchery, which has temperature and lighting
controls, both programmable.
12
For adequate water volume footprint, I designed a system that had a 220L sump (a.k.a.
reservoir) that was attached to a pump that would push water through the system. Incoming
water was sent through a series of mesh filters to a desired size, most commonly 2 microns, and a
UV sterilizing lamp. This filtered seawater could subsequently be pumped to any of the 50
individual tanks I manufactured, each equipped with a shut-off valve allowing for use of any
number of tanks up to 50. Water flow into any one tank was controlled via 5 gallon/hour
irrigation emitters, and aeration of the tanks was achieved via an air compressor, air valve,
rubberized tubing, and standardized air stone to each unit. The air stone setup was designed such
that the stone fit in the central and lowest part of each tank, so that aeration was uniform
throughout the water column. Water could either be 1) continuously cycled through the system in
a one-way flow, 2) continuously re-cycled by closing input and drains resulting in an
approximately 1200L recirculating system, or 3) set to some intermediate combination of the
two. The pump can be programmed to re-start via a floating ball switch when the sump is either
full or empty, depending on the experimental needs. Populations of organisms as small as ~50
microns were kept isolated in their tanks, while allowing water and food to pass through, via an
upwelling drain and so-called “banjo” screen that effectively blocked their ability to pass through
the drain. Several sizes of these banjo screens were fabricated, each using 3” PVC, mesh, hot
glue, and PVC piping, so that as populations grew in size the mesh would be less likely to clog.
The tanks themselves were made with 20L Nalgene carboys, the bottoms of which I chopped off
before inverting to create a conically bottomed tank, which then had a drain drilled so that they
held 12L each. The screw-caps (now at the bottom) of the Nalgene tanks were manipulated using
PVC sheeting, resin, and retrofitted with a ball valve such that the tanks drained at the lowest
point and there were no nooks or puddles left for microscopic animals to be missed. It may seem
13
obvious to design a tank this way - with the drain at the lowest point - but it is often overlooked
in other systems. Lids were made by routing out Home Depot bucket lids and fitting them to
each inverted Nalgene tank. Through these lids I was able to standardize where the air, water,
and Hobo temperature sensor sat in each tank. I built shelving out of ply-wood and used a router
to carve a negative shape which reflected the carboy shape and leveraged the handles for
stability. This system is the primary experimental setup used for the majority of the work in my
dissertation, and took innumerable hours of trial and error to successfully engineer.
Other systems needed to be generated in order to support this research, described briefly here.
Algae: I built a system that housed 4 ‘sun-cones’, which are tall cylindrical tanks that
hold up to 200L. These were in addition to sun-cones already on the island in other labs’ spaces.
For lighting, I used the ambient light in the WIES “Blue-House” on ¾ sides, and a rack of full
spectrum 6-foot ballast tube-lights, with additional reflection via reflectix paneling on the
remaining side. This system had a separate air compressor from the main hatchery system, and it
worked extremely well for generating several microalgae, including Tetraselmis spp., Isochrysis
galbana, Rhodomonas spp., and Chaetoceris spp.
Gamete Clearance Raceway: Separating gametes among spawning mussels is an often
overlooked and complex issue. Mussels tend to increase their spawn-out when in the presence of
other spawning mussels, but simultaneously inhale gametes into their body cavity. Thus, simply
“group spawning” and isolating an individual after identification of spawn-initiation can result in
‘contaminated’ or uncontrolled spawns. This is an especially important point to consider when
trying to generate isogenic family lines. Indeed, at a particular point in my graduate career, I
discovered that some of the family lines I had generated were actually from unknown parents
14
(using whole genomes, PCA analysis, and STRUCTURE, data not shown). With this experience
in mind, I designed a system that could house up to 12 spawning individuals at a time and had a
one-way direction of water flow through it, such that spawning mussels could purge their cavity
of contaminating gametes inhaled during group-spawning. A trial showed less than 5%
contamination using this method, determined by observing dividing zygotes discovered in a
theoretically pure egg isolation, versus up to 50% in non-raceway individuals (data not shown).
This was dubbed the GCS: Gamete Clearance Raceway.
Longline: Creating bivalve families necessarily generates a need for a place to house
them, and it helps to have broodstock on hand for hatchery science. Unfortunately, most bivalves
will not flourish in on-land systems (juveniles will never exceed a few centimeters, broodstock
will not ripen without extensive & intensive systems), and so we solved this issue by creating a
longline. Initially spearheaded by the Manahan and Hedgecock labs, the project was ultimately
completed with help from the Wrigley office and the Nuzhdin lab, with major construction
coordination from Dave Anderson and permitting help from Jessica Dutton. The line itself is
~150 ft long, secured on both ends with two 40 lb. Danforth anchors and heavy chain, suspended
by two can/nun buoys and a series of mooring buoys, and is located in the southwest corner of
Catalina Harbor. Using this, we have a series of cages that house mussels and oysters that were
generated during my graduate studies. Weekly visual inspections, quarterly dive inspections, and
annual Catalina Harbor surveys are all part of the monitoring process. Together, the hatchery,
algae, gamete raceways, and longline create one of the most unique aquaculture science facilities
on the West Coast, and have enabled the leveraging of significant funds for future/ongoing
programs (see photo of systems, next page).
.
15
Two Harbors
Long Line Site
Hatchery and Longline Facilities. A) The hatchery facilities at the Wrigley Institute on Catalina
Island. The 50-tank family line system is on the left, which connects to the sump, UV bulb, pump,
and filtration system in the center of the photo. The drain can be seen with a wooden foot ramp over
the pipe. The bench system for work is located on the right, alongside a ~500L circular broodstock
holding tank, and the Gamete Clearance Raceway is in the foreground on the right. Not shown in this
photo are micro-algal rearing systems, to the rear of the photographer. B) The longline system in
Catalina Harbor. The scenic photo is for orientation purposes, as “Cat Harbor” is relatively little
known. The inset on bottom left shows the floating buoys from the shore; the inset on the right shows
a hanging ‘aquapen’ typically used to house mussels and oysters in this system.
16
Chapter 1: Metallotoxicity Behavioral Response in Bivalves
Associated Publication*: Hall M, Foley B, Cheung E, Abbasi M, Churches ND (2016). A New
Behavioral Phenotyping Strategy for Pacific Oyster (Crassostrea gigas) Larvae Reveals Cohort-
Level Effects on Copper Toxicity Swimming Response. Ann Aquac Res 3(3): 1025.
Abstract: Copper is among the most studied marine metallotoxins. It is both heavily utilized in
commercial applications (e.g. industrial discharges and antifouling hull coatings) and readily
bioavailable in the water column. In bivalve mollusks, common responses to copper toxicity
include increased mortality rates and disruption of normal development, especially during early
life history stages. Bivalve studies, however, focus primarily on physiological and morphological
changes in conditional or field experiments, while behavior remains relatively unexplored. This
study profiles the larval movement characteristics of 48-hour old larval Pacific oysters
Crassostrea gigas (C. gigas) under increasing concentrations of Cu2+. C. gigas full sibling
families were subjected to a series of increasing Cu2+ concentrations in Filtered Sea Water
(FSW), from 0 ppb to 36 ppb, for n=10 conditions. Across all trials, a negative correlation was
observed with increasing Cu2+ loads and percent survival, normal morphological development,
and average width. MovTrack, in-house developed tracking software, was used to quantitatively
show that Cu2+ concentration and total movement of larvae are not dependently linked. A
familial component of Cu2+ stress reaction was potentially observed, with some genetic lines
showing significant differences in movement metrics, supporting the hypothesis that Cu2+
toxicity response may have a heritable component. This study provides evidence that previously
documented physiological responses to Cu2+ toxicity are fundamentally a cellular response,
17
rather than a synergistic effect of altered behavior and cell-level disruption. Finally, this work
provides a proof of concept for MovTrack software as a reliable phenotyping strategy for
quantitative measurement of marine larvae behavior.
Introduction
Anthropogenic introduction of toxins in oceanic environments can have detrimental effects for
marine life. Heavy metals are among the most studied marine toxins (metallotoxins), and have
been shown to negatively affect survival, growth, stress protein pathways, and immune functions
in several species [1- 5]. Among the metallotoxins, copper (Cu, especially Cu
2+
) has been
extensively researched due to widespread introduction. As an example, San Diego Bay and the
Port of Los Angeles have exceeded the EPA allowable Water Quality Criterion (WQC) of 4.8
parts per billion (ppb) in recent years [1,6-7]. The primary source of Cu contamination is
antifouling hull paint, used by both civilian and naval vessels. Other sources included storm
water runoff, rainfall, and municipal and industrial discharges [6]. This study, and similar Cu
pollutant load projects [4,8-12], clearly demonstrate the need to further understand the sources
and effects of increased Cu load for both ecosystems and individual species.
Bivalves have often been the target of metallotoxicity investigations. In addition to extensive
populations, many bivalves exhibit a broadcast spawning reproduction strategy and a planktonic
life history stage, during which gametes and larval/embryonic stages are particularly vulnerable
to ambient seawater conditions. Moreover, mollusks are commercially important, harvested for a
total of more than 13.2 million metric tons in 2012 [13]. As such, there has been an interest in
funding studies related to increasing wild harvest and aquacultural yields, and understanding
18
toxicity response. Several studies have demonstrated the wide-ranging negative consequences
that bioavailable Cu
2+
has on bivalves and other mollusks [14- 26]. Because the vast majority of
aquaculture farms are in- or near-shore operations, understanding anthropogenic Cu loading
effects to waterways is of great economic value.
The Pacific oyster, Crassostrea gigas, is among the most extensively cultured bivalves. For C.
gigas, the percent normal development and percent survivor ship has been used to approximate
the concentration at which 50% of population experiences lethality (LC ) when exposed to Cu
2+
.
For embryos and larvae, the LC50 has been estimated at 5-20 ppb, and upwards of 500 ppb for
established adults [3,27-31]. Each of the studies on metallotoxicity in bivalves mentioned here,
excluding environmental collection experiments, have similar strategies: to expose organisms to
varying concentrations of agonist and observe physiological and morphological impacts. Reports
on larval swimming behavior in the metallotoxin literature are scant. A single study reported that
swimming behavior in C. gigas larvae showed overall increases when exposed to leachates of
chemically treated timber at days 3 and 7 [32]. While informative, this study did not control for
Cu concentration across experiments, and leachates included other toxins including chromium
and arsenic. A second study mentioned erratic swimming behavior in C. gigas in relation to Cu
load, but this was never quantified [28]. Previous work has hypothesized that swimming strategy
and feeding may be linked in bivalve larvae, demonstrating a direct link to behavior and access
to nutrients [33]. This brings about the hypothesis that a negatively synergistic effect of cellular
responses and altered behavior (and therefore access to nutrients) is accountable for previously
observed physiological changes, rather than cellular responses alone.
19
Here, we address this knowledge gap by using new phenotyping strategy for marine bivalve
larvae. In this study, a series of full sibling C. gigas families were created from a unique
population of Southern California oysters, sourced from Carlsbad Aquafarms, USA. Each family
(or cohort) was exposed to a series of increasing Cu
2+
concentrations in Filtered Sea Water
(FSW), from 0 ppb to 36 ppb, for a total of n=10 conditions. After 48 hours in condition, a
quantitative measure of swimming behavior was recorded using an in-house developed computer
tracking software, MovTrack. In parallel to quantitative behavioral measurements, phenotypic
assays were performed to document survival, developmental pace and abnormalities, and growth
rates. The purpose of this work is to i) further explore estimated LC50 values for an unstudied
population of C. gigas, ii) understand between-and-among family variance in response to Cu
load, and iii) demonstrate a new behavioral assay for bivalve larvae. The timing of this project is
especially pertinent, as the US Environmental Protection Agency (EPA) is set to release new
Ambient Water Quality Criteria for copper documents, currently in the public comment phase.
Methods
Animals: Adult C. gigas samples were kindly donated from Carlsbad Aquafarms (Carlsbad,
California, USA) on May 12th, 2016. Individuals collected were held at Wrigley Marine Science
Center (WMSC, Catalina Island, California, USA) in flow-through tanks with raw seawater until
time of experimentation, which was as late as July 29th, 2016. Animals were fed a mixture of
Isochrysis galbana and/or Tetraselmis spp.ad libitum on a semi-daily basis to keep gravid.
Sea Water and Chemical Preparation, Trial Conditions: Seawater at WMSC laboratories was
sourced from the point at Big Fisherman’s Cove (Catalina Island, California, USA). From this
20
source, salt water is sent through a series of pleated reusable filters, down to 2µm to remove
animals and other debris, and hereafter referred to as Filtered Sea Water (FSW). For Cu2+ trials,
a 100 millimolar stock solution was prepared from solid state CuSO4 crystals (EM Science,
product number CX2203-1) mixed with Milli-Q filtered fresh water. From the stock solution, a 0.1
millimolar working solution was prepared using filtered Milli-Q fresh water. Experimental
concentrations were then obtained by diluting CuSO4 working solution in 1L FSW at appropriate
ratios. A total of 7 trials were performed, each with a control group (0 ppb, i.e. FSW) and 6
experimental conditions. Three replicates per condition were performed, except for trial 7 in which
six replicates were conducted for three conditions at the upper end of the Cu2+ spectrum (Table
1). Experimental conditions were as follows for trials 1-3 in ppb Cu: 0, 3, 6, 9, 12, 15, 18; trials 4-
6 in Cu ppb: 0, 6, 12, 18, 24, 30, and 36; trial 7: 0, 24, 30, and 36 (Table 1). Trials 1-3 were initial
experiments, which clearly did not capture the range in which 48 hpf larvae were viable, so trials
4-6 were performed with a greater range. Because some overlap occurred between trials 1-3 and
4-6, we performed a final trial (trial 7) to increase representation in the full dataset at higher copper
concentrations.
Spawning and Toxicity Assay: Gametes were obtained by following the common aquaculture
technique of strip-spawning gravid adult C. gigas [34]. Eggs and spermatozoa with the greatest
number and maturity were selected for fertilization, which was determined by visual subjective
analysis for roundness and motility, respectively. The creation of full-sib families was done by
mixing mature gametes from a single male and single female in a beaker with approximately 250
mL FSW, such that the sperm: egg ratio was approximately 5:1. After one hour, fertilized eggs
were rinsed through a 20 µm nylon sieve to reduce risk of polyspermy.
21
After approximately 1.5-2 hours, fertilization counts were performed by concentrating
egg/sperm solutions into approximately 20-30 mL of FSW in a 50 mL conical tube, which was
gently bubbled from the lowest point to achieve homogeneity. Four 20 microliter aliquots of
egg/sperm solution were counted on a Sedgewick rafter to obtain an accurate estimate of total
fertilized egg concentration in the population. Evidence of fertilization was noted by observation
of a polar body or cell cleavages, with most organisms achieving a 2-8 cell stage by the 2 hour
mark.
Twenty-one 1L polycarbonate bottles were then filled with FSW and amended with a CuSO4
working solution to achieve test concentrations of CuSO4 ranging from 0ppb to 36ppb, as
described above. Each experimental condition was performed in triplicate to control for batch
effects. Bottles were stirred well and left to sit for up to one hour to allow for equilibration before
stocking with C. gigas. Using the estimated fertilized embryo concentration, tanks were stocked
at a density of 15 fertilized eggs ml-1. Next, stocked trial bottles were placed in a Percival Intellus
incubator at 25˚C on a 12:12 hour light:dark cycle for all trials, except for the first trial performed
in which the temperature was set at 18C. Though the temperature was different for trail 1, the
overall patterns of development and movement were not significantly different from later trials, so
data was included in subsequent analysis (Table 1). The total time from the start of spawning to
placement in an incubated experimental condition was less than three hours. At 48 hours post
fertilization, bottles were randomly selected one at a time, and removed from the incubator to be
processed for phenotyping (described below).
22
Percent Survival Counts: After the 48 hour mark, the contents of each experimental trial were
filtered through a nylon mesh sieve (20 µm) to collect larvae, rinsed with FSW, and then
concentrated into approximately 30 mL FSW in separate 50 mL conical tubes. Concentrated larvae
samples were gently bubbled and four aliquots were counted on a Sedgewick rafter to estimate
remaining larvae count for each experimental condition. Final counts were compared to initial egg
and fertilization stocking concentrations to estimate percent survival.
Movement Phenotyping Assays: At the 48 hour mark, after performing the percent survival
counts, a 2 ml movement assay solution was prepared by mixing room temperature FSW and larvae
isolates at appropriate volumes to achieve a concentration of 250 animals ml-1. (Video assays were
not performed in the cases where significant die-offs made it infeasible to collect a total of 500
animals. These populations were considered ‘crashed’). Movement assay solutions were then
mixed in a single well of a 3x2 Falcon Polystyrene Microplate, which was placed under an
Olympus SZ-PT dissecting microscope with Techniquip 150W Fiber Optic Illuminator at 20x total
magnification for recording. Consistency between trials, and avoidance of capturing ‘petri-edge
movement effects’, was achieved by measuring and marking the center point of the dish and
focusing the camera on that point consistently among all trials. Thus, the edge of the petri dish was
not in view during video assays, and larvae could swim in and out of frame during any given assay.
This magnification strategy was chosen based on MovTrack software feasibility control studies,
described below. Temperature was recorded for the first two trials using an Onset Hobo
thermocoupler K-Type device. Temperature fluctuations within control trials were not found to
fluctuate more than 0.15C, and so were disregarded as a covariate.
23
Videos were recorded using a Canon Rebel Ti5 camera with 1920x1080 pixel resolution at 29
frames per second (fps). A single five-minute video was recorded for each replicate.
Morphology and Percent Normal Development: In addition to movement, we recorded other
phenotypes common to metallotoxicity literature, including morphology and percent normal
development. After video recording, remaining concentrated larval samples were used to
determine average width using an EpiScope Microscope at 40x total magnification. Ethanol was
added to aliquots from larval concentrates to reduce motility, and a series of still images were
taken and measured using the built in Olympus DP2-BSW XV Imaging Processing Software. A
single width measurement was conducted on surviving larvae by using the Arbitrary Line
Measurement Tool, which gives a measurement in micrometers, and spanning the longest possible
axis for each individual larva. A target of 50 total measurements were taken per assay replicate on
the first 50 animals encountered, and was achieved for each replicate unless otherwise noted in
table 1. To calculate proportion normal development, the same photos were used, and at least 50
larvae per experimental trial were noted as either normally or abnormally developed, based on
shape (roundness, presence of velum, D-hinge, etc.). Ratios were calculated by dividing normal
by abnormal counts for each replicate, and were log transformed, for linear regression.
MovTrack Software: To track the movement of bivalve larvae, we used MovTrack, a tool
developed in our lab to allow for high-throughput analysis of behavior from video-recorded
organisms. MovTrack is implemented in Matlab, and produces summaries of animal movement
from video input. MovTrack allows for the adjustment of settings such as luminance thresholding
24
and frame-rate resolution because different experimental setups will produce videos that vary in
such features.
A brief description of optimization of this software for this study follows. Because MovTrack was
originally developed for land based assays to track fruit flies, Drosophila spp, the software had to
be tested for aquatic applications. In order to do this, California mussels, Mytilus californianus,
were spawned out using common “heat shock” methods. Larvae were reared using standard
bivalve hatchery techniques essentially similar to the ones described herein, then concentrated and
counted at the 48-hour mark, to a final concentration of 500 animals ml-1. We optimized for
volume and magnification and found that a volume of 2 ml of this concentrate was sufficient to
cover the sides of the petri well in our 3x2 Falcon plates and provide adequate volume for larvae
to swim and not be ‘edge-kept’ by the meniscus of the water body. We also determined that the
best magnification range for data was between 12.5 and 20x magnification, in terms of minimizing
within-video movement variance across time. During this period, we optimized the cutoff
threshold for object detection by eye, at 40 intensity units (based on the 8-bit grayscale intensity
spectrum 0-255). Finally, we also optimized the frame sub-sampling rate, and found manually that
a rate of one frame per second allowed for efficient detection of real movement (instead of noise).
To test the sensitivity of the software to fluctuations in larval movement activity, the authors
then performed a video assay using M. californianus 48 hour larvae at 12.5x magnification using
a starting (control) concentration of 500 animals ml-1 in 2 ml total FSW, and serially diluting the
assay solution at 5, 10, 15, 25, 50 and 75%. From these control assays, it was clear that MovTrack
was detecting movement profiles declining at the same rate as the dilution scheme, and that a target
of 250 total animal’s ml-1 would be sufficient for assays herein.
25
Statistical Analysis: Statistical analysis was performed using the built in linear model functions
in R i386 3.3.1, using the average of size, percent survival, or movement as response variables
[35]. Cu parts per billion or dummy variables assigned to replicate number was used as explanatory
variables. Size correction for total movement was performed by employing equation 1:
𝑀
"#
=
∑(𝑀
'#
)
𝑀
)#
*
∑(𝑆
'#
)
𝑆
)#
,
-
Equation 1
Where MAi is the adjusted movement for trial ‘i’, Mri is the raw movement per second output
(measured in pixels) from MovTrack software for trial ‘i’, Mni is the total count ‘n’ of Mr for trial
‘i’, Sri is the raw size measurement (width, measured in microns) for trial ‘i’, and Sni is the total
count ‘n’ of size measurements for trial ‘i’. Thus, the output is in units of average pixel change per
square micrometer (pix/uM^2), or in other words how much a given volume of C. gigas had moved
per assay.
In order to estimate LC50 at the 48 hour mark, we standardized survivorship estimates by first
generating a proportion of survival at each concentration compared to the control concentration (0
ppb) for that cohort. Because the survivorship curve was nonlinear, we log transformed these
proportion survival counts, and performed a linear regression of log survival for all replicates
across changes in Cu2+ concentration. The LC50 is the point at which the predicted survivorship
is ln (2)=0.69 lower than the intercept. Our estimated LC50using this method was at 10.66 ppb.
26
Results
Full-Sibling Family Generation: Each of the 7 C. gigas families produced underwent some
scenario of increasing Cu2+ toxicity exposure, as is described in methods. Percentages for families
achieving D-hinge stages, surviving but not reaching D-hinge, and populations ‘crashed’ are
shown in figure (1C).
Survival, LC50 Estimates, Growth, and Percent Normal Development: Cu2+ greatly
influenced the chances of a population ‘crashing’ in experimental conditions, and severely delayed
or impeded normal development to the stages observed here. Figure (1A) shows a common pattern
seen under the microscope. Control populations exhibited a round shape and developed velum
through to the D-hinge stage, often reached by the 48 hour mark. Experimental condition larvae
often showed slower, abnormal development, with many obvious cellular division disruptions.
Higher Cu2+ concentration populations showed many abnormally developed larvae that were
evidently living and swimming, but which rarely or never achieved D-hinge morphology at 48
hours (Figure 1B).
Survival counts across all trials showed a highly significant decline with increasing Cu2+ load
(Figure 2A, est -0.065, t=-6.86, adj r2 = 0.240, P=1.9e-10). Considered individually, all families
displayed similar average survivorship curves to the collective data: increasing Cu2+ negatively
affects survival, sometimes significantly among a family (Table 1, Supplemental Table 1). The
LC50 dose across all populations was determined to be approximately 10.66 ppb (Figure 2B).
Average size at the 48 hour post fertilization mark was negatively correlated with increasing Cu2+
concentration (Figure 2B, est -0.005, t=-5.63, adj r2 = 0.193, P=1.1e-7). Within family, growth
27
rates were sometimes not statistically significant, though trends were always toward decreasing
size (Table 1, Figure 2C).
There was a pronounced decrease in the ratio of normally developed larvae with increasing Cu2+
concentrations (Figure 2C, est -0.095, t=-16.17, adj r2= 0.678, P=2.2e-16).
Larval Swimming Behavior: Larval swimming behavior responded differently depending on
which group was tested. Five out of the seven families tested returned a positive slope when fitted
with a linear model, while the other two returned negative slope values. When all of the movement
data is considered, normalized for size, and fitted with a movement-by-parts per billion Cu2+
linear model, a significant result is found (Linear model: est=0.0186, df=107, t=2.19, P= 0.0307).
However, increasing the complexity of the models to include any combination of the variables of
family, percent normal development, and survival - and their interactions - reduced or eliminated
significant correlations between swimming behavior and increasing Cu2+ concentrations. A
Bayesian Information Criteria (BIC) test across all possible models showed that family by
movement was the simplest model to explain swimming behavior. Put differently, in our tests,
Cu2+ concentration is not a reliable predictor of overall trends in movement direction, either
increasing or decreasing, though 6 of 7 families showed a non-zero slope in their response. Thus,
there may be a familial component of Cu response in larval swimming behavior at 48 hours.
Discussion and Future Directions
28
Copper Effects on Morphology: The data presented here is consistent with previous literature in
regards to copper effects on normal development, percent survival, and average size [1,3,14-30].
Copper likely disrupts cellular pathways, leading to very strange and retarded developmental
patterns, as seen in figure (1). We calculated an LC50 of 10.66 ppb, which is within the range of
previous literature [3,27-31]. This number is, however, very likely higher than a total lifespan
estimate of LC50 as the assays herein terminated at the 48 hour mark, and many families were
considered ‘crashed’ and so were not able to contribute to LC50 calculations (9 total families at or
under 10% survival; 22 total families without sufficient animals to record videos). If an attempt
was made to rear the C. gigas families to full term (i.e. to the juvenile/adult stages), the LC50
would likely shift toward the lower end of the parts per billion spectra tested here. It is likely that
animals that survived to the 48 hour mark in this study in conditions less than our estimated LC50
value (<10.66 ppb) were still significantly developmentally delayed or completely stunted and
would exhibit increased mortality rates at later larval stages than controls. This hypothesis is
supported by figure (1C), which show an obvious impact on development beginning at 3-6 ppb
Cu2+ in terms of stocks ‘crashed’, and by figure (2C), which demonstrates that abnormal
development is more frequently documented than normal development starting at approximately
6-9 ppb. Put succinctly, this study focused on a narrow portion of the C. gigas life stage, which is
the case in most (if not all) Cu2+ toxicity literature, leaving little room for accurately predicting
long term trends for exposed populations. Future studies should therefore consider extending the
rearing process to later time-points to fully understand Cu2+ induced developmental and
environmental defects.
29
Movement Effects: BIC tests showed that cohort alone, a proxy for genetically distinct families,
was the best model to explain effects on movement profiles recorded in this study, and that Cu2+
concentration was not a good predictor of larval movement after considering family effects. One
possible explanation for this result is that families respond differently to Cu2+ toxicity due to
genetic heritage. Families produced in this study were isogenic, and no gametic contamination was
observed or expected for any one cohort. Lineage based responses for this study are
conjecture, because ambient seawater conditions were not tested prior to trials, and it is not
possible to control for even gamete maturity across individual oysters. However, it is clear that
cohorts responded differently to copper exposure, when each of the slopes are observed for
individual family response (Figure 2D). In 5 out of 7 trials, cohorts responded with slightly-to-
heavily increased movement as copper concentration increased, while 2 of 7 showed significantly
negative reduction in movement (Figure 2D, Table 1). This result may also be a function of
differing baseline movement among cohorts, which again points to a possible genetic influence in
general movement. Decoupling of genetic influence on general movement versus that caused by
Cu2+ concentration is not possible with this data set. However, 6/7 trials showed changes in
movement from control trials, suggesting that Cu2+ is the main factor mediating these trends.
As larvae, C. gigas do not have shells until later stages, and so may employ swimming to stressful
environmental stimuli. For example, bivalves are known to exhibit chemically mediated behavioral
responses including predator avoidance and settling strategies [36-38], and respond behaviorally
to other environmental conditions such as wave action during larval stages [39]. The fact that
family, and not Cu2+ concentration, was the simplest model to explain the movement data was
therefore somewhat surprising, especially considering the extensive literature documenting Cu2+
toxicity on physiology. It is possible that Cu2+ does not affect the development of the velum’s
30
ciliation or activity of the cilia, and is therefore not affecting the movement profiles evenly across
all families tested here. One factor that may be different among Cu2+ exposed groups is the
direction of swimming motion, as abnormally developed larvae seem to swim in more erratic
patterns. This may be a function of hydrodynamics due to strange body shape, but direction of
swimming motion remains quantitatively unverified here. More trials across different groups of
Pacific oysters, as well as more granular analysis of movement profiles involving direction and
angular velocity, would perhaps demonstrate differences in swimming behavior at increased Cu2+
concentrations.
MovTrack Software Application: One of the objectives of this work was to apply in-house
developed quantitative imaging software (MovTrack) in a novel aquatic setting. These
experiments were successful in implementing MovTrack, which the authors believe has much
potential for further application movie quantitation experiments. For future users of MovTrack, a
few parameters which need to be considered depending on the experiment are discussed here. First,
the video frames are subsampled at set intervals, by an amount that the user has specified. A short
sampling interval will be appropriate if animals are moving rapidly, while a longer interval risks
reaching saturation (i.e. all animals will move at a greater-than-body-length in each interval). If
animals are moving slowly, however, a longer interval is more appropriate, as it will reduce noise
(for instance from light fluctuation, or minor jitter). In our study, we found that we were able to
detect sufficient movement changes among treatments using a resolution of 1 frame per second.
The difference in pixel intensity between pairs of consecutive sampled frames is then calculated
as the “difference frame”. This difference frame is the measure of change between time points.
The difference frame is then transformed to an 8-bit gray-scale image, and thresholded to filter out
31
minor changes in frames not caused by movement of organisms (such as light fluctuations, or
minor changes in non-focal features). It is best to maximize difference in light intensity between
foreground (target) and background, within the experimental setup, to facilitate threshold
selection. We applied a threshold of 40 (based on the 8-bit grayscale intensity spectrum 0-255).
The threshold is then applied to the pixel intensities of the grayscale image, to convert grayscale
to binary “different” or “not different” bins. The measure of movement between two consecutive
frames is then calculated by summing over all the pixels in the thresholded difference frame. By
repeating this calculation across the video, the user produces a time series estimate of movement.
Considering these parameters, essentially any video that is standardized for camera position and
contains movement at least one pixel large can be quantified. MovTrack has many other
applications, including path tracking of individuals, which will be described in future publications
(Abbasi et al., in prep).
Implications: Bivalves employ a non- or partially-lecithotrophic free swimming trochophore and
veliger life history. These organisms may not be able to decouple swimming and feeding during
these early stages, as the velum functions as both a swimming and feeding organ. If C. gigas larvae
stop swimming, it is unclear if they are able to continue feeding effectively. Adult bivalves are
known to filter particulate matter in their gills and eject unwanted material as pseudofeces, but
whether or not larvae are capable of this behavior is also unknown [37]. The fact that movement
was not shown here to be affected similarly across all trials during Cu2+ stress, but survival and
percent normal development all showed similar trajectories, indicates that feeding behavior (and
therefore increased or decreased access to nutrients) is likely not a factor in the mode of lethality
32
for Cu2+ toxicity. Thus, our hypothesis that a negatively synergistic effect of cellular responses
and altered behavior (and therefore access to nutrients) is accountable for previously observed
physiological changes, rather than cellular responses alone, is not supported here.
Though our original hypothesis was not supported, genetically mediated response to Cu2+ load
may have been observed in this study. Clear differences in growth curves, percent survival, and
developmental delay were evident between the families (though all showed similar directional
trends). Swimming behavior was different among families, with some showing a generally
increased profile with increasing Cu2+ concentrations, and others showing a decrease (Figure 2D).
If cohort level differences in swimming behavior belie genetically mediated cellular level response
to Cu2+ toxicity, there are implications for aquaculture selective breeding efforts and ecological
toxicity responses.
Field studies may be able to confirm genetically based differences in toxicity response; sites with
high Cu loads which continue to support adult C. gigas and other bivalves may reveal metallo-
tolerant genotypes and phenotypes. The C. gigas used for this study were sourced from the Pacific
Northwest and planted at the Agua Hediona site by Carlsbad Aquafarms approximately 25 years
ago, with some subsequent broodstock supplementation since. More research into Cu2+ hydro-
geo-bio dynamics of Agua Hidiona would be required before comparative analysis would be
possible for environmental selection parameters. On the other hand, the Agua Hidiona population
presents a stock that has almost certainly been selected for higher temperature tolerance,
considering this population is one of the southern-most sites at which C. gigas is commercially
harvested in the United States. Comparisons between this population and C. gigas grown in the
US Pacific Northwest and Canadian West coast may provide an interesting comparative group in
both metallotoxicity, warming earth scenarios, and any additive effects of these phenomenon.
33
Conclusions
This work reaffirms the rich body of scientific literature demonstrating negative Cu2+ toxicity
effects in bivalves, and adds a new behavioral phenotyping strategy via use of MovTrack software.
Future work in this area would benefit from a focus on full term rearing of exposed larvae,
understanding the effects of multi-generational exposure to large Cu2+ load (whether reared in a
hatchery or observed in the wild), and testing of more Cu2+ sensitive species.
34
Figures
Table 1: Cu
2+
effect results across seven isogenic C. gigas cohorts (previous page).
Trials/cohorts are indicated in the furthest left column and boldly boxed. Each trial contains four
metrics calculated in this study: average size (in µm), abnormal/normal estimate (ratio: number
of abnormal counted over number of normal counted), percent survival (ratio: count surviving
over initial population estimate), and adjusted average movement (proxy for how much a given
volume of C. gigas larvae had moved per assay, see materials and methods: VIII. Statistical
Analysis). Trial 7 was created to increase representation at higher copper concentrations for the
full dataset.
Trial Number Metric 0 3 6 9 12 15 18 24 30 36
1 Average Size 62.41 55.91 55.48 55.82 55.90 56.59 55.40 - - -
Abnormal/Normal
Estimate 0.17 0.25 0.19 0.39 0.31 0.74 1.98 - - -
Percent Survival 1.91 2.27 1.98 1.79 1.93 1.61 2.01 - - -
Adjusted Average
Movement 3.43 (n=2) 1.92 2.41 2.33 1.84 (n=2) 1.55 1.65 - - -
2 Average Size 74.68 (n=2) 66.98 (n=2) 59.80 (n=2) 51.33 (n=1) 58.45 53.43 (n=2) N/A - - -
Percent Normal
Development 0.11(n=2) 0.29 1.26 (n=2) 1.39 (n=2) 0.98 1.725 (n=2) 12 (n=1)* - - -
Percent Survival 0.53 0.87 0.97 0.54 0.61 0.28 0.27 - - -
Adjusted Average
Movement 2.29 (n=2) 1.47 (n=2) 2.96 (n=2) 3.03 (n=1) 2.48 3.79 (n=2) N/A - - -
3 Average Size 56.60 (n=1) 55.52 50.63 48.96 49.17 47.80 (n=2) 48.18 (n=1) - - -
Percent Normal
Development 0.12 (n=1) 0.52 0.88 (n=2) 3.63 (n=1) 0.94 N/A 34 (n=1)* - - -
Percent Survival 1.98 1.84 1.51 1.00 1.53 0.44 0.47 - - -
Adjusted Average
Movement 1.54 (n=1) 2.66 3.80 (n=2) 2.26 3.43 (n=1) 4.57 (n=2) 1.62 (n=1) - - -
4 Average Size 58.76 - 52.99 - 51.15 - 49.81 49.51 48.08 48.45
Percent Normal
Development 0.45 0.62 1.02 2.17 3.74 4.37 12.97
Percent Survival 0.53 0.66 0.46 0.62 0.65 0.34 0.46
Adjusted Average
Movement 2.99 (n=2) - 3.66 - 3.30 - 4.46 (n=2) 3.90 3.23 3.82 (n=2)
5 Average Size 65.63 - 60.83 - 56.51 - 51.48 49.75 41.83 (n=1) 45.17 (n=2)
Percent Normal
Development 0.43 0.47 1.52 2.04 3.11 7.33 (n=1) 8.31 (n=2)
Percent Survival 0.48 0.80 0.43 0.36 0.27 0.02 0.20
Adjusted Average
Movement 1.30 - 1.07 - 2.41 (n=1) - 1.94 (n=2) 2.98 (n=1) N/A 2.10 (n=2)
6 Average Size 50.61 - 51.77 - 53.71 - 50.65 52.63 50.25 48.38
Percent Normal
Development 0.50 (n=2) 0.65 (n=2) 0.63 1.51 6.02 8.24 14.43 (n=2)
Percent Survival 1.36 0.78 0.85 0.64 0.32 0.14 0.60
Adjusted Average
Movement 3.13 - 3.16 (n=2) - 3.09 - 3.01 (n=2) 4.05 (n=2) 3.69 (n=1) 3.45 (n=2)
7** Average Size 79.12 (n=3) - - - - - - 59.37 56.39 52.48
Percent Normal
Development 0.04 (n=3) - - - - - - 0.85 3.60 5.43
Percent Survival 2.54 (n=3) - - - - - - 0.78 0.61 0.40
Adjusted Average
Movement 3.02 (n=3) - - - - - - 2.96 2.55 3.13
*= population considered 'crashed'beyond normal larval recognition, removed from subsequent PND analysis
** n=6, unless otherwise noted
Parts Per Billion Copper
35
Figure 1 – Cu
2+
Effects on the Development of C. gigas Larvae: A) Representative photos of
Cu
2+
trials for C. gigas larvae. Rows demonstrate an increase in Cu
2+
load, while the columns
give 24 hour time points. The red lines indicate size, which are as follows from left to right: top -
58.13 µm, 47.26 µm, 196.49 um; bottom -51.45 µm, 57.29 µm, 59.96 µm. V = velum organ, N =
Cell Nuclei. B) Representative photo for abnormal development. Compare to furthest right
column in A. Notice strange structures, aberrant cells, and an elongated body frame. Red line
represents size: 60.97 µm. C) Population stages as a function of Cu
2+
ppb, given as a percentage
of total trials. Numbers in white are total replicates (n) in a given category for that particular
stage.
36
Figure 2 - Percent Survival, Lethal LC50 Estimates, Size Effects, and Percent Normal
Development for C. gigas Larvae Exposed to Cu
2+
: All graphs here are data taken at the 48
hour post fertilization mark. A) Percent survival across all trials as a function of increasing Cu
2+
parts per billion. Linear model: est -0.065, t=-6.86, adj r
2
= 0.240, P=1.9e-10. LC50 is estimated
from this linear model at 10.66 ppb. B) Average size effects for C. gigas larvae exposed to Cu
2+
.
Each point represents an average width, in micrometers, taken from n>50 measurements from a
replicate of the condition described on the x-axis in ppb Cu
2+
. Linear model: est -0.005, t=-5.63,
adj r
2
= 0.193, P=1.1e-7 C) Abnormal to normal development ratios for C. gigas larvae exposed
to Cu
2+
. Note that abnormal development begins to skew at values significantly less than the
estimated LC50 of 10.66 ppb estimated in A, at approximately 6 ppb Cu
2+
here. Linear model: est
-0.095, t=-16.17, adj r
2
= 0.678, P=2.2e-16. D) Movement as a function of increased Cu
2+
load.
All movement data together (black circles) with increasing copper concentration was significant
(Linear model: est: 0.0186, t = 2.19, r
2
=0.03396 P=0.0307), however models that considering
family effects and other variables were not significant. Each line represents a linear model for
genetically distinct families (cohorts) generated in these experiments. Note the different
responses (slopes) among the family lines.
●
●
●
●
●
●
●
●
●
●
● ●
●
● ●
●
●
● ● ●
●
● ● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
● ● ●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
● ● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ● ●
●
0 5 10 15 20 25 30 35
−6 −4 −2 0 2 4 6
Larval survival with increased copper
Copper concentration in ppb
Log of scaled surival
r
2
= 0.24
A
●
●
●
●
● ● ● ●
●
●
● ●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
0 5 10 15 20 25 30 35
3.8 3.9 4.0 4.1 4.2 4.3 4.4
Larval size with increased copper
Copper concentration in ppb
Log µm
r
2
= 0.2
B
●
●
●
●
●
●
●
●
●
● ●
●
● ●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
0 5 10 15 20 25 30 35
−3 −2 −1 0 1 2 3
Normal development with increased copper
Copper concentration in ppb
Log µm
r
2
= 0.68
C
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
0 5 10 15 20 25 30 35
0 1 2 3 4 5 6
Size−adjusted movement with increased copper
Copper concentration in ppb
Mean movement (pixels/s)
r
2
= 0.31
D
37
Chapter 2: Microbiome Investigations in the Pacific Oyster
Associated Publication*: Simons AL, Churches ND, Nuzhdin SV (2018). High turnover of
faecal microbiome from algal feedstock experimental manipulations in the Pacific oyster
(Crassostrea gigas). Microb Biotechnol 11:848–858.
Abstract: The composition of digestive microbiomes is known to be a significant factor in the
health of a variety of hosts, including animal livestock. Therefore, it is important to ascertain
how readily the microbiome can be significantly altered. To this end, the role of changing diet on
the digestive microbiome of the Pacific oyster (Crassostrea gigas) was assessed via weekly fecal
sampling. Over the course of 12 weeks, isolated individual oysters were fed either a control diet
of Tetraselmis algae (Tet), or a treatment diet which shifted in composition every 4 weeks.
Weekly fecal samples from all oysters were taken to characterize their digestive bacterial
microbiota. Concurrent weekly sampling of the algal feed cultures was performed to assess the
effect of algal microbiomes, independent of the algal type, on the microbiomes observed in the
oyster samples. Changing the algal feed was found to significantly associated with changes in the
fecal microbiome over a time scale of weeks between control and treatment groups. No
significant differences between individual microbiomes were found within control and treatment
groups. This suggests the digestive microbiome of the Pacific oyster can be quickly and
reproducibly manipulated.
Introduction
The microbiome has increasingly been identified as playing a critical role in a host’s
various physiological functions, ranging from infection response (Wen et al., 2008) to digestion
38
(Bahrndorff et al., 2016). Gaining a better understanding of host-microbiome interactions, then,
is of particular interest in a variety of fields ranging from human health (Belkaid et al., 2015) to
food management techniques (García-Orenes et al., 2013). Modern agriculture has benefited
directly from associative phenotype-microbiome studies. For example, soil microbial
communities, an analog for animal microbiomes, have been shown to: 1) affect the efficiency of
plant nutrient uptake via solubilization of minerals, 2) guard plants against pathogenic
organisms, and 3) regulate growth via synthesis of plant hormones (see review: Hayat et al.
2010). This work has led to the commonplace use of microbial inoculation and ‘microbe-
encouraging’ soil mixtures in terrestrial commercial and hobbyist plant cropping. For livestock
farming, microbiome studies are becoming more common as the link between products, such as
meat and milk, and associated microbiomes becomes clearer (Mackie, R. I., 2002; Sommer and
Backhed, 2013). For the dairy cow, microbiome studies have revealed that each separate stomach
contains distinct microbiome profiles, which are associated with differential genetic regulation in
each organ (Mao et al., 2015). As a whole, the cow microbiome has been associated with
phenotypes of commercial interest such as disease (Nakamura et al., 2017) and food conversion
efficiency (e.g. Myer et al., 2017). Recent work on humans also suggests a degree of vertical
inheritance for the gut microbiome (Bäckhed, 2015). Increased understanding of host-
microbiome phenotypic associations will likely continue to play an increasing role in food
management and breeding strategies, as well as human medicine.
This is especially true in commercial aquaculture, where evidence suggest that host-
microbiome interactions could have a role both in the management of bacterial pathogens (Tan et
al., 2016), as well as improving host growth rates (Douillet et al., 1994; Cruz et al. 2012). Within
aquaculture, a better understanding of these interactions would be of particular interest to the
39
cultivation of Pacific oysters (Crassostrea gigas, (Cgi)), a significant portion of a $19 billion
global aquaculture industry (FAO, 2016). Though they are becoming more common, genetic
improvement strategies for oysters are difficult processes due to their multi-year life cycles, and
the inherent difficulty in retaining pedigreed individuals in oceanic environments. Even a
moderate increase of 5% in growth rate would allow farmers to reduce time-to-market by
between 27 and 54 days (FAO, 2005). Furthermore, genomic resources are still somewhat
lacking for Cgi. While its genome has been recently sequenced (Zhang et al, 2012), and
quantitative trait loci analysis has demonstrated a genetic component to desirable traits such as
growth rate (Guo et al., 2011), it has still proven difficult to apply this knowledge towards
improvements in oyster crop via selective breeding (Dégremont et al., 2015). A better
understanding of how to improve commercial phenotypes of cultured oysters through influencing
their microbiota might provide a shorter route to crop improvement for oysters than genomic
approaches. For example, improvements in larval growth rates of approximately 20% were
reported through the addition of a probiotic bacterial strain to the feed of Cgi (Douillet et al.,
1994).
Human health is also adversely impacted by a general lack of knowledge of bivalve
microbiomes and associated physiological phenomenon because these animals are frequently
consumed raw and in their entirety. Consumers therefore ingest the whole of the bacterial
community also, which may transfer highly toxic pathogenic bacteria (Givens et al., 2014). This
can lead to deadly human diseases such as vibriosis, which is fatal in 15-30% of cases (U.S.
Food and Drug Administration, 2003). On an ecological scale, the monitoring of bivalve
microbiomes could be used as an early warning system for the onset of harmful algal blooms
(McPartlin et al., 2016), which may also cause lethal human health issues such as paralytic
40
shellfish poisoning (Hurley et al., 2014). Many rural communities depend largely on wild-
harvested bivalves as a source of protein, and the only way to completely avoid the risk of PSP is
to eliminate this food staple completely (Paralytic Shellfish Poisoning Fact Sheet, 2002) .
Addressing the question of how readily Cgi can accumulate pathogens from their environment,
as well as assessing the efficacy of current pre-market depuration methods for commercial Cgi
crops, would be quite useful to the field of public safety.
Taken together, it is clear that commercial aquaculture and human food safety
management would benefit from a deeper understanding of how readily the microbiome, and in
particular the gut microbiome, of cultivated Cgi can be altered. However, the literature
concerning bivalve-associated microbiota is still in its infancy, and the transferability of
techniques and approaches between species has not been established. Previous work with
Crassostrea virginica (King et al., 2012) has demonstrated a differentiation in the gut
microbiome of populations from different geographic localities, most likely due to regionally
distinct marine bacterial communities. Similar results have been shown with Cgi and, using a
combination of antibiotics and transplantation, significant shifts in the composition of the gut
microbiome have been observed over the course of a week (Lokmer et al., 2016). This study
aims to establish the first associations between diet and corresponding microbiome profiles in
Cgi. Here, by controlling environmental parameters and adjusting diets between conspecifics of
Cgi, it was shown that dietary variance was shown to correspondingly affect the digestive
microbiome. Additionally, as our group has used the same laboratory setup and similar feeding
schedule to do related work with other bivalves, such as the mussel Mytilus galloprovincialis,
there is the potential to apply this study’s methods to other farmed shellfish species.
41
Results
Sequencing and OTU Visualization
In total, approximately 3.81 million high quality paired-end read sequences, clustered at
97% similarity into 4009 Operational Taxonomic Units (OTUs), were identified across the 128
samples successfully sequenced in this study (see figure 1 for sampling scheme). The distribution
of sequences assigned per OTU is highly skewed with approximately 50% of all counted
sequences found in the 11 most abundant OTUs, and 90% of all sequences found within the most
abundant 144 OTUs.
Of the 128 samples analyzed in this study 17 comprise the bacterial communities found
in the algal feedstock, 103 comprise the bacterial communities found in the oyster fecal pellets,
and 8 comprise the samples directly extracted from the stomach of the oysters during the final
week of sampling. While the read depth for the samples directly extracted from the gut were too
low to use for statistical analysis (under 1000 reads per sample (See Table S1)), some individual
taxa could be identified. For the oysters feeding on Tet at the time of extraction their stomach
communities were dominated by Deinococcus, Shewanella, Marivita, and Vibrio. For those
feeding on Chae at the time of extraction, their gut communities were dominated by Albimonas
and Ruegeria. Figure 2 shows relative sequence abundance of the top 10 most represented
bacterial genera between control and experimental treatment groups in the 120 remaining
samples after filtration. It is visually clear that bacterial community composition diverges as a
function of both algal type and treatment group (control vs. experimental), suggesting a digestive
microbiome that is dynamically responsive to feed type.
42
MDS Plots and b-diversity
Samples were then analyzed using techniques described in Lokmer et al. 2016. The
significance of the experimental factors explaining observed measures of b-diversity were
determined using a permutational multivariate analysis of variance (Permanova) on the Bray-
Curtis dissimilarity values between samples. Both the weighted Unifrac distance and Bray-Curtis
dissimilarities between samples yielded similar results (Figs S1 – S10). Visualizing these
differences was done using multidimensional scaling (MDS) plots, with the two axes describing
the largest amount of total variation in b-diversity, in order to reduce the dimensionality of the
microbiome data.
Algal Microbiomes
It was observed that each algal culture had distinct associated bacterial communities.
MDS plots show that for the control feed, which consisted of a continuous culture of Tet
throughout the 3 month experiment, the bacterial communities remained stable, whereas
experimental feed bacterial communities changed with algae culture type, Isochrysis spp. (Iso) or
Chaetoceros spp. (Chae) (Figures S2 and S7). When comparing control feeds to experimental
feeds within months, algae type was a significant factor in explaining the Bray-Curtis
dissimilarity between each algal associated microbiome sample (F(2,14)=3.0974, p=5e-4) for the
duration of this study (Figure 3). While changes in the algal microbiomes were easy to observe
in all algal samples, even in control feedstocks on a monthly basis, weekly sampling time was
found not to play a significant role in determining the composition of each algal microbiome
(F(1,15)=1.1433, p=0.3246). This supports the idea that algae communities are relatively stable
within feed types, but distinct between them. Time was not a significant factor in explaining the
43
algal Tet microbiome at both the weekly (F(1,9)=1.3948, p=0.2243) and monthly time scales
(F(1,9)=1.663, p=0.1099), which indicates a stable algal microbiome composition over time.
Oyster Fecal Microbiomes
The 103 fecal samples observed were significantly differentiated based on the type of
feed received (F(2,99)=9.9143, p<1e-4). When comparing the control group to experimental
group in month 2 (Tet vs. Iso) and month 3 (Tet vs. Chae), significant differences were found in
fecal sample community composition (Figure 4. F(1,35)=5.9068, p<1e-4) and F(1,29)=7.5761,
p<1e-4, respectively.). For the fecal samples taken from oysters in the control group, the
composition of each sample’s microbiome did not remain stable on a weekly time scale
(F(1,37)=2.7147, p=0.005999), or on a monthly time scale (F(1,37)=3.3479, p= 0.0014).
However, replicates were not reported as a significant factor in determining bacterial community
composition within the fecal control (F(1,37) = 1.2451, p = 0.1242) or fecal treatment (F(1,71) =
1.0083, p = 0.447) groups.
Next, the fecal microbiomes from the control and treatment groups of oysters were
compared as feedstocks were changed. During the first month of this experiment both the control
and treatment groups of oysters received identical Tet feed. It was found that oysters raised in
identical conditions on identical diets will tend to cultivate highly similar bacterial communities
as experimental status was found not to be a significant factor in differentiating fecal
microbiomes (Figure 4. F(1,41)=1.0903, p=0.3208). During the second month, when the
treatment group received Iso algae, the fecal microbiomes diverged with experimental status
becoming a significant factor in explaining differences in fecal microbiome compositions (Figure
4. F(1,35)=5.9068, p<1e-4). During the third month, for fecal samples obtained during weeks 9
44
through 11, experimental status becomes an even more significant factor in describing
differences between fecal microbiomes (Figure 4. F(1,20)= 8.0843, p<1e-4).
Temporal Variability in Microbiomes
The results for both algal and fecal microbiome comparisons are in general agreement
with observations made of changes observed in the relative abundance of the most common
sample taxa over time (Fig 2), as well as an extended Local Similarity Analysis (eLSA) of
patterns of co-occurrence between common taxa across samples (See Table S2). The fecal and
algal microbiomes remained significantly distinct throughout the entire course of the study for
the control group (F(1,48)=8.7385, p<1e-4. See figures S4 & S5, S9 & S10). For the treatment
groups there was significant differentiation between the fecal and algal microbiomes
(F(1,71)=5.0757, p<1e-4) for the duration of the study, similar to the control group. The monthly
time scale was also a significant factor in differentiating both algal (F(1,8)=3.4973, p=0.0043)
and fecal microbiomes (F(1,61)= 10.386, p<1e-4).
Discussion
Algal Microbiomes
How strongly associated are the algal microbiomes to a particular algal culture? While
the water used in culturing algae was taken from a water supply passed through a 1 micron filter
and a UV treatment, the presence of algal microbiomes illustrates that these are not axenic
cultures. Nor, as many marine microalgae species are auxotrophic for many bacterially generated
micronutrients, could these cultures likely even be viable (Kazamia et al., 2012). Auxotrophic
associations are known to exist between the algae and their associated bacterial communities
45
frequently found in this study, such as those between Ruegeria and Tet or Iso (Arora, et al.,
2012) or Maritivita and Chae (Kimura and Tomaru, 2014; Cruz-López and Maske, 2016).
Similarly, a number of bacteria found in the algal communities are known consumers of algae,
and algal exudates, such as members of Tamlana and Ruegeria (Arora et al., 2012; Chauhan and
Saxena, 2016). These relationships are reflected in the composition of our algal culture
communities, although underlying factors were not the focus of this study. It should be noted that
even within a controlled environment the bacterial communities within the Tet culture varied
over the course of the study. This could suggest internal community dynamics within the algal,
variations in the concentrations of trace elements in the water supply, or both (Harrold et al.,
2018).
Oyster Fecal Microbiomes
In general, the composition of the bacterial communities found in the fecal samples are
significantly different, with one trend being a significant rise in the relative abundance of genera
such as Litoribacillus, Shewanella, and Vibrio. The rise in the relative abundance of these
bacterial genera are not unexpected as a number of their member species are known copiotrophs
(Goldberg et al., 2017; Kim, 2017). These observations fall in line with studies showing the
prevalence of the phyla Proteobacteria, and in particular members of Vibrio, and Bacteroidetes
in Cgi fecal communities (Hernández-Zárate and Olmos-Soto, 2006; Fernández et al., 2014;
Wang et al., 2014). These results also suggest that it may be possible to design future
experiments to elucidate methodologies for earlier and faster prediction of Vibriosis promoting
conditions in natural environments.
In observing the most abundant bacterial taxa at monthly intervals (the frequency at
which the treatment group algal feed was changed), a number of the patterns were observed with
46
the analysis of the samples’ bacterial community composition. One, the dominant taxa in the
fecal microbiomes appears distinct from their corresponding algal microbiomes for both the
control and treatment group of oysters. This would suggest a role for the oyster gut environment
in shaping a bacterial community to being significantly different from that of its food. This study
therefore reflects previously observed differences between the bacterial communities in the local
water column and in the guts of both the Pacific and Eastern oysters (King et al., 2012; Lokmer
et al., 2016). Two, the dominant bacterial taxa are distinct between algal cultures. Three, there is
a significant divergence observed in the control and treatment fecal groups as the treatment algal
feed changes after the first month. This suggests a role for algal feedstock in shaping the
digestive bacterial community, and that such shaping can happen on the scale of weeks. Four, the
general trend observed in the fecal bacterial communities were that they contained a high relative
abundance of copiotrophic genera as compared to the algal bacterial communities, independent
of the time or type of algae. For example, the dominant bacteria found in the Tet feedstock varies
over time, but the fecal communities in the control group oysters fed only Tet consistently have a
significant abundance of Vibrio and Shewanella. A consistent shift is also seen in the fecal
samples from the treatment group of oysters, which consistently show a rise in the relative
abundance of Vibrio as compared to the algal feedstock communities.
What remains to be determined is whether the rise in abundance of specific copiotrophic
bacteria in fecal samples is primarily determined by the nutritional profile of the algal culture
used as feed, or by another factor such as associated digestive bacterial communities. There was
observed repeated rise in the relative abundance of genera such as Marivita and Vibrio in the
oysters only fed Tet, even with a varying composition of the Tet culture’s bacterial community.
This points to a role for the algal culture’s nutritional profile in shaping the composition of the
47
gut microbiome, which thereby influences the fecal microbiome community. The fact that fecal
samples were collected at the same post-feeding interval, coupled with no significant difference
between within each time-point’s replicate bacterial communities, suggests that fecal bacterial
communities are indeed changing as a function of dietary influence. A future avenue for research
would be increasing controlled bacterial doses in feed, and observing associated gut and fecal
microbiome response.
Conclusion
In this study, oyster digestive microbiomes were experimentally manipulated via a
change in diet, which provides evidence that a stable, rather than varying, diet will tend to yield a
more stable digestive microbiome as assessed from fecal samples. For example, the oysters in the
control group consistently showed an abundance of Vibrio in their feces, while those in the
treatment had shifts in their fecal bacterial communities at the same time scale as changes in their
feed. This study illustrates also that changes in diet can yield significant changes in the
composition of the digestive microbiome on the scale of weeks. This plasticity suggests that the
digestive microbiome of oysters will be able to respond quite rapidly to perturbations such as the
introduction of probiotic or pathogenic bacteria. This work additionally suggests that oysters
under similar environmental conditions will have fecal microbiomes which respond similarly to
shifts in diet. This may be true regardless of genetic background, as oysters in this study were
from a semi-wild cohort found at an aquaculture farm, though it is conceded that the continuity
between replicates may be as easily explained by potentially high genetic homogeneity (i.e. low
Ne) in bivalves in wild (Hedgecock and Pudovkin, 2011) and aquaculture farm populations
(Hedgecock and Sly, 1990).Furthermore, the fact that shifts in fecal microbiomes were similar
48
between replicates indicates that the experimental condition (i.e. control group or experimental
group) was the stimulating factor, as opposed to some unknown protocol variable. The individual
was not a significant factor in determining the composition of the fecal microbiome for either the
control (F(1,37) = 1.2451, p = 0.1242) or treatment group of oysters (F(1,71) = 1.0083, p =
0.447) throughout the course of this study. The authors conclude that further work on studying
host-microbiome interactions in Pacific oysters could be done with the expectation of a
significant degree of reproducibility between individuals, at least for those reared in a similar
environment.
Materials and Methods
Oyster collection
The 15 oysters used in this study were transported as adults from a semi-enclosed lagoon
at the Carlsbad Aquafarm (Carlsbad, CA) to an aquaculture test facility in the Wrigley Marine
Sciences Center (WMSC) located near the town of Two Harbors on Catalina island. In the three
weeks prior to this study these oysters were held in a common tank and fed a 1:1 cell mix of Tet
and Iso algal cultures at a density of 100,000 cells/mL, on a daily basis.
Experimental setup and sampling
The oysters were separated into 2 groups and kept on a regular feeding and sampling
schedule for the 12 week duration of the study. The control group (n=5) was only fed Tet for the
duration of the study. The experimental group (n=10) was feed Tet for the first 4-week period,
followed by Iso for the second 4-week period, and finishing with Chae for the final 4-week
period. For the duration of the study all 15 of the oysters were kept in separate 12 liter tanks
containing 1 micron filtered seawater and fed algal cultures at a density of 100,000 cells/mL, 12
49
hours per day, 3 times per week (FAO, 1975). Following each feeding the water was changed
using the 1 micron filtered seawater supply. All of the algal cultures were grown on site at
WMSC in 200 liter containers using the same 1 micron and UV filtered seawater supply.
Each week the fecal pellets from the oysters, as well as the bacteria pelleted from the
algal cultures, were collected. A weekly sampling schedule was chosen as both commercial
depuration processes (Lee, 2008), and prior work studying the effects of geographic
transplantation on Cgi microbiota (Wegner et al., 2013; Lokmer et al., 2016), have demonstrated
that significant shifts in Cgi gut communities can regularly occur within 3-5 days. In order to
collect fecal pellets, the oysters were removed from their tanks following a water change and
placed into 2 liter containers filled with 0.2 micron filtered UV sterilized seawater 12 hours prior
to sampling. The fecal pellets were collected using individual disposable pipettes and then frozen
in 1.5mL collection tubes at -20C for later DNA extraction. Concurrent to the oyster fecal pellet
collection, separate 15mL conicals of each feedstock were centrifuged at 13,000g for 5 minutes
and the pellet of algae and bacteria was removed with a disposable pipette and frozen in 1.5mL
collection tubes at -20C for later DNA extraction. On the 12th and final week of the study the
stomach contents of oysters were extracted instead of collection of fecal pellets. To extract the
stomach contents, each oyster was shucked, the stomach surface and surrounding tissues rinsed
using a 1% bleach solution (Provost et al., 2011), and the stomach directly excised and emptied
into a 1.5mL sample collection tube using a sterile razor blade. These samples were then frozen
at -20C for later DNA extraction.
The final number of usable samples consisted of 110 fecal pellet samples, 10 stomach
samples, and 18 pelleted algal samples. The typical sample volume for the fecal pellets and the
50
pelleted algal cultures, was approximately 300uL. For the stomach samples the sample volumes
were approximately 600uL. See figure 1 for feeding and sampling schedule.
DNA extraction
DNA was extracted from all of the samples using the QIAamp PowerFecal DNA Kit
(Qiagen, Carlsbad, CA). The manufacturer’s recommended protocol was followed for all sample
types, and yielded 100uL of a solution containing extracted DNA in a proprietary buffer named
C6. These solutions were stored in 1.5mL Eppendorf tubes and frozen at -20C.
As a first check on the concentration of extracted DNA in the samples, 2uL of each
sample were then quantified using a Qubit 3 fluorometer and a Qubit Quantitation Assay Kits
(Thermo-Fisher). The manufacturer’s quantification protocol for low concentration double
stranded DNA was used to assess the concentration of the extracted sample. Samples which
yielded no detectable concentration of DNA were then omitted from the rest of the analysis
pipeline.
Polymerase chain reaction (PCR)
A 16 S rRNA rRNA region corresponding approximately to the V4–V5 regions with
uniquely barcoded 515f and 926r PCR primers was amplified. The PCR reactions were set up for
each week’s samples using 25uL reaction volumes with the following composition: 1uL of
extracted DNA with a concentration of approximately 0.5 ng/uL, 1.5uL of a 1:1 515f:926r 10uM
primer mix, 12.5uL of PCR water, and 10uL of HotMaster mix (VWR, Visalia, CA). The set of
samples amplified each week varied in number, typically 12, and included one negative control
containing 1uL of PCR water in place of extracted DNA along with a unique forward and reverse
primer allocated for each week’s control sample. During the 4th week an even and staggered
51
mock microbial community was amplified, each with its own unique forward and reverse primer
pair, as a positive control (Parada et al., 2015). These mock communities were generated based
on, with predetermined relative microbial abundances, microbial taxa commonly found in the
waters of the San Pedro channel. The expected and experimentally determined values for the
relative abundance of each mock community member was then used to determine an average
sequencing error rate. This analysis was done using MOTHUR (Schloss et al., 2009). The per-
base pair error rate used was 4.571e-4.
The PCR cycling protocol was as follows: 120 s initial denaturation at 95°C, then 30
cycles: 45 s denaturation at 95°C, 45s annealing at 50°C, 90 s extension at 68°C, 5 min final
extension at 68°C.
Another check of DNA concentration was then made, this time of the PCR product, using
the same protocol as the raw DNA extraction samples. PCR products which did not have a
concentration of at least 1 ng/uL were regenerated from the original extracted sample material.
For PCR product samples with a sufficient concentration of DNA, 2uL of reaction product was
mixed with 3uL of SYBR Safe DNA Gel Stain (Thermo Fisher Scientific) and was analyzed
immediately on a 1.0% agarose gel. To check if the amplicon length matched the expected
amplicon length, the samples were separated with one well containing 5uL of O’GeneRulerTM
100 bp Plus DNA Ladder (Thermo Fisher Scientific). As a negative control, each week’s sample
set was separated with one well containing 2uL of PCR water and 3uL of SYBR Safe DNA Gel
Stain. The separation voltage gradient was set at 100 V per 10cm and run for 60 minutes.
All samples were cleaned and diluted to a uniform concentration of 1 ng/uL of DNA in a
solution of TE using a DNA Clean & Concentrator-25 kit (Zymo Research). A pooled sample
containing 5 uL of each cleaned sample was then made and DNA fragments with a length under
52
200 bp were removed using an Agencourt AMPure XP bead cleaning kit (Beckman-Coulter).
This pooled sample was then cleaned and concentrated into a solution of TE at a concentration of
approximately 10 ng / uL before being sequenced.
Sequencing
The pooled sample was sequenced using 150 bp pair-end Illumina sequencing on the
MiSeq platform at Laragen (Culver City, CA). All Illumina sequencing data is available at
NCBI under the BioProject PRJNA416146.
Sequence quality control and preprocessing
All sequencing libraries were processed together. Quality control, OTU clustering, and
taxonomy assignment were performed in MOTHUR (Schloss et al., 2009), using the MOTHUR
MiSeq SOP (Kozich, Westcott, Baxter, Highlander, & Schloss, 2013). Only overlapping regions
of the contigs and removed any sequences with ambiguous bases and/or homopolymers of with
lengths of at least 8 bp were retained in order to ensure good quality and reduce the number of
spurious OTUs. To further reduce the number of spurious OTUs singletons (i.e. OTUs
containing a single read) were also removed for any downstream analysis. The sequences were
aligned to SILVA 128 reference alignment (Quast et al., 2013) cut to the 515f to 926r region
(Walters, et al., 2015). Taxonomy was assigned with 80% confidence cut-off, using the Silva
v128 taxonomy (Yilmaz, et al., 2014) in conjunction with the Naïve Bayesian Classifier (Wang
et al., 2007) used in MOTHUR. Single-linkage pre-clustering was performed with a cutoff of
two allowed differences (Huse, Welch, Morrison, & Sogin, 2010). Chimeras were then removed
and the remaining sequences used to create 97% OTUs using the average-linkage clustering
method.
53
As the length of the sequenced reads was typically 300 bp, out of an expected amplicon
length of 337 bp, the resulting sequences could only be consistently determined down to the level
of genus.
Statistical analysis
The primary statistical tools used in this project were the Phyloseq (McMurdie and
Holmes, 2013) and “Vegan” R packages (Oksanen et al., 2013), with MOTHUR generated data
as the input. In order to compare b-diversity measures between samples the raw sequence count
data was converted to relative sequence abundances where the scaling factor was the total
number of sequences per sample.
The b-diversity measures used in this study were the Bray–Curtis distances and weighted
UniFrac distances (Hamady et al., 2010), both generated using the Phyloseq package. The results
were further analyzed by MDS, implemented by the ordinate function in Phyloseq, and
Permanova using 10,000 permutations, implemented in the adonis function in the Vegan
package. The significance tests using the Weighted Unifrac distance values yielded similar
results to those using the Bray-Curtis dissimilarity and so only the results using the Bray-Curtis
dissimilarity were displayed.
MDS plots were color coded by various experimental variables, such as the type of algal
feed used. The visual clarity of how samples were clustering was enhanced using the stat_ellipse
function in ggplot2 R package, with a bounding ellipse drawn at the 95% confidence interval. It
should be noted that in groups with less than four samples no unique ellipse could be drawn.
The most abundant OTU bar plots were generated using the plot_bar function in
Phyloseq. The source data for these plots were taken as the relative abundance of the 10 most
abundant OTUs per month across all samples.
54
Analysis of potentially significant co-occurrences between the relative abundance of
common OTUs was carried out using eLSA (Xia et al., 2011). Only OTUs with a relative
abundance greater than 1% were used, in order to manage the data set’s complexity, across all
time points using a replicate value of 5 for control group samples and 10 for the treatment group
samples. Only OTU co-occurrences, defined both for Spearman and Pearson correlations, with a
false discovery rate of under 0.05 were designated as significant (See Table S2).
55
Acknowledgements
We would like to thank David Anderson for help with managing our lab facilities on
Santa Catalina island and running this study. We would also like to thank the Fuhrman lab group
at the University of Southern California for help with choosing primers for our 16 S rRNA
analysis. This study was supported by the Wait Foundation, and the University of Southern
California’s Wrigley Marine Science Center. Our group has no conflict of interest to state with
regards to this project.
Figures
Figure 1, Experimental Sampling Schema: Oyster feeding and sampling schedule with
naming schema. The month in this experiment refers to the particular 4 week window of time
for a given algal culture used in feeding the treatment group of oysters. Fecal and algal
samples were collected weekly. Direct gut extraction samples were collected during week 12,
but not used in analysis due to low read depth. No fecal samples were collected during week
12. Further details described in the materials and methods section.
56
Figure 2, Bacterial Community Composition Changes: The relative abundances of the 10
most prevalent bacterial genera found in all algal and fecal samples. Note that the control feed,
CON_FEED, represent the Tet feedstock and is shared between control and treatment groups
during the first month. The treatment feed, EXP_FEED, represents the Iso feedstock during
month 2 and Chae feedstock during month 3. CON_FECAL represents fecal samples from the
control group of oysters, which are only feed Tet, and EXP_FECAL represents fecal samples
from the oysters receiving the treatment feed.
57
Figure 3, Differentiation in Algal Associated Microbiomes: MDS plot, using the Bray-Curtis
dissimilarity between algal samples during each month of the study. CON_FEED represents the
control Tet feedstock, which is shared between control and treatment groups during the first
month. The treatment feed, EXP_FEED, represents the Iso feedstock during month 2 and Chae
feedstock during month 3. Note that groups with less than four samples have too few points for a
unique ellipse to be drawn using the stat_ellipse function in Phyloseq.
58
Figure 4, Divergence in Digestive Microbiomes by Experimental Status: MDS plot, using the
Bray-Curtis dissimilarity between fecal samples during each month of the study. CON represents
fecal samples from oysters in the control group, which are always feed Tet, and EXP represents
fecal samples from oysters in the treatment group. Both control and treatment groups are fed Tet
during the first month. In months 2 and 3 the EXP group is feed Iso and Chae respectively.
59
Supplementary figures
(supplemental tables can be accessed online associated with this publication).
Figure S1: MDS plot, using the weighted Unifrac distance between samples from all 12 weeks,
showing the differences between algal microbiomes between all three algal cultures. Samples
are colored by algal culture (CHAE = Chaetoceros, ISO = Isochrysis, and TET = Tetraselmis).
Significance of algal culture in differentiating algal microbiomes (F(2,14)=2.6954, p=0.0045).
Significance of weekly sampling basis time in differentiating algal microbiomes
(F(1,15)=0.99099, p=0.4205)
Note: Groups with less than four samples have too few points for a unique ellipse to be drawn
using the stat_ellipse function in Phyloseq.
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
−0.3
0.0
0.3
−0.50 −0.25 0.00 0.25 0.50
Axis.1 [30.2%]
Axis.2 [24.2%]
FeedType
● ●
● ●
● ●
CHAE
ISO
TET
60
Figure S2: MDS plot using the weighted Unifrac distance between all fecal samples.
Differences are shown between fecal microbiomes based on their algal feed. Samples are
colored by algal culture (CHAE = Chaetoceros, ISO = Isochrysis, and TET = Tetraselmis).
Significance of algal culture in differentiating fecal microbiomes (F(2,99)=9.7561, p<1e-4).
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
−0.2
0.0
0.2
0.4
0.6
−0.50 −0.25 0.00 0.25 0.50
Axis.1 [18.2%]
Axis.2 [13.4%]
FeedType
●
●
●
CHAE
ISO
TET
61
Figure S3: MDS plot using the weighted Unifrac distance between fecal samples from the
control group of oysters as well as the Tetraselmis algal microbiomes.
CON_FECAL_MONTH_1 = Fecal samples from control group weeks 1 to 4,
CON_FECAL_MONTH_2 = Fecal samples from control group weeks 5 to 8,
CON_FECAL_MONTH_3 = Fecal samples from control group weeks 9 to 11.
TET_MONTH_1 = Bacterial communities in Tetraselmis feed weeks 1 to 4, TET_MONTH_2
= Bacterial communities in Tetraselmis feed weeks 5 to 8, TET_MONTH_3 = Bacterial
communities in Tetraselmis feed weeks 9 to 11. This figure demonstrates consistency in
control fecal microbiomes and control dietary microbiomes.
Significance weekly sampling basis time in differentiating all fecal samples from the control
group (F(1,37)=2.8668, p=0.0172).
Significance of monthly sampling basis time, phase, in differentiating all fecal samples from
the control group (F(1,37)=3.6169, p=0.0046).
Significance of weekly sampling basis time in differentiating all Tetraselmis microbiomes
(F(1,9)=1.3278, p=0.2514).
Significance of monthly sampling basis time, phase, in differentiating all Tetraselmis
microbiomes (F(1,9)=1.7322, p=0.09699).
Significance of sample type, fecal versus algal microbiome, in differentiating fecal samples
from the control group from the Tetraselmis microbiomes (F(1,48)=9.5276, p<1e-4).
Note: Groups with less than four samples have too few points for a unique ellipse to be drawn
using the stat_ellipse function in Phyloseq.
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
−0.8
−0.6
−0.4
−0.2
0.0
0.2
−0.4 0.0 0.4
Axis.1 [21.8%]
Axis.2 [17.2%]
PhaseAndStatus
●
●
●
●
●
●
CON_FECAL_MONTH_1
CON_FECAL_MONTH_2
CON_FECAL_MONTH_3
TET_MONTH_1
TET_MONTH_2
TET_MONTH_3
62
Figure S4: MDS plot using the weighted Unifrac distance between fecal samples from the
treatment group of oysters as well as all of the algal microbiomes.
EXP_FECAL_MONTH_1 = Fecal samples from treatment group weeks 1 to 4,
EXP_FECAL_MONTH_2 = Fecal samples from treatment group weeks 5 to 8,
EXP_FECAL_MONTH_3 = Fecal samples from treatment group weeks 9 to 11.
TET_MONTH_1 = Bacterial communities in Tetraselmis feed weeks 1 to 4, ISO_MONTH_2
= Bacterial communities in Isochrysis feed weeks 5 to 8, CHAE_MONTH_3 = Bacterial
communities in Chaetoceros feed weeks 9 to 11. Notice consistent cluster migration across
feed types for each experimental group, corresponding to dietary regime change.
Significance of sample type, fecal versus algal microbiome, in differentiating fecal samples
from the control group from the algal microbiomes (F(1,71)=4.7314, p<1e-4).
Significance of monthly sampling basis time, phase, in differentiating the algal microbiomes
(F(1,8)=4.1009, p=0.0031).
Significance of monthly sampling basis time, phase, in differentiating the fecal microbiomes
(F(1,61)= 11.566, p<1e-4).
Note: Groups with less than four samples have too few points for a unique ellipse to be drawn
using the stat_ellipse function in Phyloseq.
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
−0.5
0.0
0.5
−0.4 −0.2 0.0 0.2 0.4
Axis.1 [16.4%]
Axis.2 [12.4%]
PhaseAndStatus
●
●
●
●
●
●
CHAE_MONTH_3
EXP_FECAL_MONTH_1
EXP_FECAL_MONTH_2
EXP_FECAL_MONTH_3
ISO_MONTH_2
TET_MONTH_1
63
Figure S5: MDS plot using the weighted Unifrac distance between control and treatment group
fecal samples.
CON_FECAL_MONTH_1 = Fecal samples from control group weeks 1 to 4,
CON_FECAL_MONTH_2 = Fecal samples from control group weeks 5 to 8,
CON_FECAL_MONTH_3 = Fecal samples from control group weeks 9 to 11.
EXP_FECAL_MONTH_1 = Fecal samples from treatment group weeks 1 to
4,_FECAL_MONTH_2 = Fecal samples from treatment group weeks 5 to 8,
EXP_FECAL_MONTH_3 = Fecal samples from treatment group weeks 9 to 11.
The significance of the differences between fecal samples taken from control and treatment
groups of oysters during phase 1 (F(1,41)=1.0247, p=0.3666).
The significance of the differences between fecal samples taken from control and treatment
groups of oysters during phase 2 (F(1,35)= 6.3725, p<1e-4).
The significance of the differences between fecal samples taken from control and treatment
groups of oysters during phase 3 (F(1,20)= 6.7674, p<1e-4).
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
−0.4
−0.2
0.0
0.2
0.4
0.6
−0.4 0.0 0.4
Axis.1 [18.2%]
Axis.2 [13.4%]
PhaseAndStatus
●
●
●
●
●
●
CON_FECAL_MONTH_1
CON_FECAL_MONTH_2
CON_FECAL_MONTH_3
EXP_FECAL_MONTH_1
EXP_FECAL_MONTH_2
EXP_FECAL_MONTH_3
64
Figure S6: MDS plot, using the Bray-Curtis dissimilarity between samples from all 12 weeks,
showing the differences between algal microbiomes between all three algal cultures. Samples
are colored by algal culture (CHAE = Chaetoceros, ISO = Isochrysis, and TET = Tetraselmis).
Note: Groups with less than four samples have too few points for a unique ellipse to be drawn
using the stat_ellipse function in Phyloseq.
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
−0.50
−0.25
0.00
0.25
0.50
0.75
−0.25 0.00 0.25 0.50
Axis.1 [28.5%]
Axis.2 [20.1%]
FeedType
●
●
●
CHAE
ISO
TET
65
Figure S7: MDS plot using the Bray-Curtis dissimilarity between all fecal samples.
Differences are shown between fecal microbiomes based on their algal feed. Samples are
colored by algal culture (CHAE = Chaetoceros, ISO = Isochrysis, and TET = Tetraselmis).
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
−0.50
−0.25
0.00
0.25
−0.50 −0.25 0.00 0.25 0.50
Axis.1 [16.3%]
Axis.2 [13.1%]
FeedType
●
●
●
CHAE
ISO
TET
66
Figure S8: MDS plot using the Bray-Curtis dissimilarity between fecal samples from the
control group of oysters as well as the Tetraselmis algal microbiomes.
CON_FECAL_MONTH_1 = Fecal samples from control group weeks 1 to 4,
CON_FECAL_MONTH_2 = Fecal samples from control group weeks 5 to 8,
CON_FECAL_MONTH_3 = Fecal samples from control group weeks 9 to 11.
TET_MONTH_1 = Bacterial communities in Tetraselmis feed weeks 1 to 4, TET_MONTH_2
= Bacterial communities in Tetraselmis feed weeks 5 to 8, TET_MONTH_3 = Bacterial
communities in Tetraselmis feed weeks 9 to 11. This figure demonstrates consistency in
control fecal microbiomes and control dietary microbiomes.
Note: Groups with less than four samples have too few points for a unique ellipse to be drawn
using the stat_ellipse function in Phyloseq.
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
−0.6
−0.3
0.0
0.3
0.6
−0.4 0.0 0.4
Axis.1 [23.4%]
Axis.2 [12.2%]
PhaseAndStatus
●
●
●
●
●
●
CON_FECAL_MONTH_1
CON_FECAL_MONTH_2
CON_FECAL_MONTH_3
TET_MONTH_1
TET_MONTH_2
TET_MONTH_3
67
Figure S9: MDS plot using the Bray-Curtis dissimilarity between fecal samples from the
treatment group of oysters as well as all of the algal microbiomes.
EXP_FECAL_MONTH_1 = Fecal samples from treatment group weeks 1 to 4,
EXP_FECAL_MONTH_2 = Fecal samples from treatment group weeks 5 to 8,
EXP_FECAL_MONTH_3 = Fecal samples from treatment group weeks 9 to 11.
TET_MONTH_1 = Bacterial communities in Tetraselmis feed weeks 1 to 4, ISO_MONTH_2
= Bacterial communities in Isochrysis feed weeks 5 to 8, CHAE_MONTH_3 = Bacterial
communities in Chaetoceros feed weeks 9 to 11. Notice consistent cluster migration across
feed types for each experimental group, corresponding to dietary regime change.
Note: Groups with less than four samples have too few points for a unique ellipse to be drawn
using the stat_ellipse function in Phyloseq.
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
−0.4
0.0
0.4
−0.6 −0.3 0.0 0.3
Axis.1 [15.1%]
Axis.2 [12%]
PhaseAndStatus
●
●
●
●
●
●
CHAE_MONTH_3
EXP_FECAL_MONTH_1
EXP_FECAL_MONTH_2
EXP_FECAL_MONTH_3
ISO_MONTH_2
TET_MONTH_1
68
Figure S10: MDS plot using the Bray-Curtis dissimilarities between samples.
CON_FECAL_MONTH_1 = Fecal samples from control group weeks 1 to 4,
CON_FECAL_MONTH_2 = Fecal samples from control group weeks 5 to 8,
CON_FECAL_MONTH_3 = Fecal samples from control group weeks 9 to 11.
EXP_FECAL_MONTH_1 = Fecal samples from treatment group weeks 1 to
4,_FECAL_MONTH_2 = Fecal samples from treatment group weeks 5 to 8,
EXP_FECAL_MONTH_3 = Fecal samples from treatment group weeks 9 to 11.
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
−0.50
−0.25
0.00
0.25
0.50
−0.4 0.0 0.4 0.8
Axis.1 [16.3%]
Axis.2 [13.1%]
PhaseAndStatus
●
●
●
●
●
●
CON_FECAL_MONTH_1
CON_FECAL_MONTH_2
CON_FECAL_MONTH_3
EXP_FECAL_MONTH_1
EXP_FECAL_MONTH_2
EXP_FECAL_MONTH_3
69
Chapter 3: Mutation Rate Estimates in the Pacific Oyster*
Abstract: Terrestrial selective breeding programs for food crops serve as a template for ocean-
based crops, and are significantly more advanced in most cases. One consideration for marine
organisms and their susceptibility to artificial selection is their mutation rate, which is thought to
be among the highest observed for one of aquaculture’s most heavily produced crop: bivalves.
However, many of the current estimates of heritable mutation load have been indirectly
estimated through allozyme or microsatellite inference in the Pacific oyster (Crassostrea gigas).
While these studies do have solid theoretical framework to back up their estimates, and observed
larval survival associated with so-called ‘viability loci’ support a high genetic load hypothesis, to
date, no trios (two parents and their offspring) have been developed and sequenced for the
purposes of gaining a direct estimates of heritable mutation rate. Using Pacific oysters, this study
utilized three single parent pairings, and randomly selected juvenile individuals from each
corresponding family, in order to empirically estimate heritable mutation rate. Our findings rank
among the highest ever recorded for any organism (1x10
-5
per nucleotide). The results herein
inform best breeding practices, and provide the first such study in an aquaculture species with
considerable commercial importance.
*This work is currently being drafted as a manuscript, with myself as the first author, under the
title: “Trio sequencing data provides first direct estimates of heritable mutation load in the
commercially important species the Pacific oyster (Crassostrea gigas)”. Other collaborators
include Scott Applebaum, Francis Pan, Gary Molano, Donal Manahan, Dennis Hedgecock, Peter
Chang, and Sergey Nuzhdin.
70
Introduction
Mutation rates in bivalves: Mutation rates are an important factor to consider when developing
either a new model organism or scientifically managed food crop. Bivalves fill each of these
categories, as they are used for myriad studies: sentinel species in marine toxicity (Beyer et al.
2017), physiological investigations in relation to energy utilization and ocean acidification (Pan
et al. 2018; Liao et al. 2019), associated gene expression for intertidal dynamics and aerobic-to-
anaerobic transitions (Gracey et al. 2008), unique evolutionary phenomenon including Doubly
Uniparental Inheritance of mitochondria (Ghiselli et al. 2012), microbiome and host interactions
(Simons et al. 2018), novel immune system functions (Canesi and Pruzzo, 2016), and genomic
studies to investigate local adaptation and responsiveness to selective breeding programs (She et
al. 2018), to name just a few topic areas. Some of these bivalve-related studies rely on family
lines of their respective species (e.g. Plough and Hedgecock, 2016), and many rely on the larval
stages due to their sensitivity to environmental parameters and sheer biomass at larval stages
(e.g. Waldbusser et al. 2013; more on larval stages later). Bivalves also rank among the most
popular mollusks for human consumption, and much effort is currently underway to investigate
their potential for domestication.
Marine bivalve heritable mutation rate, however, is relatively understudied. The first to
attempt a thoroughly estimate the genetic load (i.e. deleterious mutations inherited per
generation) of the Pacific oyster (Crassostrea gigas, (Cgi)) took place in 2001 (Launey and
Hedgecock, 2001), likely spurred by the then-recent finding that allozyme diversity in
Tasmanian and Japanese Cgi populations were very high – among the most diverse of any
species yet studied (Ward et al. 2000). Utilizing microsatellite data observed among F2 and F3
families, their seminal paper in this field estimates 8-14 highly deleterious recessive mutations in
71
the wild progenitors of their family lines (Launey and Hedgecock, 2001). These early predictions
were based on extrapolation of microsatellite data to the whole genome after noting “widespread
selection against identical-by-descent (IBD) marker homozygotes”, indicating that IBD markers
were likely associated with nearby recessive mutations that proved lethal during development
(Launey and Hedgecock, 2001). A follow up study (Plough and Hedgecock 2011) investigating
inbreeding depression among F2 family lines in Cgi utilized 80 microsatellite markers and
uncovered a similar estimate of approximately 14-15 vQTL harbored per wild adult oyster,
which is much higher than the estimated 3-4 ‘lethal equivalents’ per genome for animals as
estimated by Lynch and Walsh (1998). High genetic load, so follows Plough and Hedgecock’s
(2011) theory, could be the explanatory variable behind the related phenomena of
heterosis/hybrid vigor and inbreeding depression observed in both terrestrial crops (especially
plants), and in bivalve farming operations, as first reported by Singh and Zuorous (1978). Other
studies support the hypothesis of high mutational rate in Cgi. These include observations of
unexpected coding sequence SNPs when developing genotyping assays (Sun, Shin, and
Hedgecock, 2015), high estimates of SNP density across the genome, 0.023 (Zhang et al. 2013),
and a SNP frequency of 1 in every 60 base pairs in coding regions for markers used in a 2007
study, the density of which was found to be negatively correlated to codon usage (Sauvage et al.
2007). For the closely related oyster Ostrea edulis, similarly, a high mutational load was
observed in the form of nonsynonymous mutations in a set of 37 nuclear loci, at 0.3%, and it was
estimated that any single oyster may carry upwards of 4800 slightly deleterious, or non-neutral,
alleles (Harrang et al. 2013). Two species of scallop have likewise shown high rates of SNP
detection in intronic regions, at about 1% (Arias et al. 2009), though this study did not elucidate
exonic SNP prevalence. The hypothesis that bivalves may incur a huge number of novel
72
generational mutations is also supported simply because of the sheer quantity of gametes that
they produce annually, which is on the order of millions or perhaps billions; a high mutational
load may be a direct consequence of such cellular proliferation (Williams, 1975).
Larval stages of bivalves are important to consider in the context of a high mutation rate,
as much of the selection against deleterious alleles occurs in the early stages of development
(e.g. Launy and Hedgecock, 2001; Plough and Hedgecock, 2011, and references therein). Plough
and Hedgecock, 2011, showed convincingly that lethal combinations of mutations were
responsible for early viability selection by implementing a highly granular sampling scheme
across larval and juvenile stages, and demonstrating severe marker segregation occurring 18-30
days post fertilization (see figures 2 and 3, Plough and Hedgecock, 2011), accounting for 90% of
all distorted markers observed. This time period (18-30 d.p.f.) is associated with a dramatic
change in life history as the larvae metamorphose from planktonic and free-swimming to their
final sessile form, and this transition is associated with highly differentially expressed genes
novel to this stage (Li et al. 2016). Each of the above-mentioned studies pertaining to bivalves
utilize low representation methods (e.g. allozyme or microsatellite), and few utilize pedigreed
family lines for trio analysis (Launey and Hedgecock, 2001; Plough and Hedgecock, 2011).
Echoing Harrang et al. (2013): if we are to gain a better understanding of mutation rate in
broadcast spawning marine organisms, we need to describe various categories of mutations
(SNPs, insertions, deletions, and their respective intronic or exonic predictions) in multiple
species, using wild progenitors to reduce bias, and, I would add, at various stages of
development.
The literature concerning high genetic load for Cgi is also supported by observations of
lower-than-expected effective population size (Ne) for this species (first published Hedgecock,
73
1994), which are estimated between 50 and 500 individuals for wild populations. This
observation helped to inform the theory of Sweepstakes Reproductive Success (SRS) as an
explanation for Ne/N ratios (Hedgecock, 1994; SRS review: Hedgecock and Pudovkin, 2011;
more on SRS in the discussion). SRS posits, briefly, that organisms which need to coordinate
mating events in the highly stochastic ocean environment (e.g. marine invertebrates, some fishes)
have a sweepstakes-like chance of reproductive success, which allows for only a few individuals
to repopulate each generation. As mutation rate is expected to scale inversely with Ne under
neutral theory, the observations in the literature seem congruous (Nei, 1987), yet mutation rate
remains unmeasured for Cgi from direct, large empirical data sets.
For this study, we begin to address this gap by utilizing a breeding experiment to directly
estimate mutation rate in the commercially important Cgi. The breeding scheme included two
males and four females sourced from wild populations at the Molluscan Broodstock Program at
the Oregon State University as progenitors, and four offspring (F1) from each combination of
parents reared at USC’s Wrigley Institute for Environmental Science on Catalina Island. From
these trios, we were able to generate Next Generation Sequencing (NGS) libraries using a
commercially available library generation kit, sequenced on the Illumina platform. Using this
data, the first estimations made from deeply sequenced trios for generational mutation rate were
made. Employing highly conservative methods, our results suggest that an estimation of the
generational mutation rate for Cgi might be made at ~1 x 10
-5
per nucleotide, and a minimum of
approximately 200-1000 SNPs inherited per generation
per individual. This is among the highest
reported rates of mutation for any eukaryotic organism, and is 2-3 orders of magnitude larger
than the upper limit of mutation rate made from a recent review of empirically estimated Ne and
mutation rate correlations (Lynch, 2016; and see discussion). We argue that this high
74
generational genetic load is reflective of a life-history which includes an extremely variable
chance at reproduction (SRS, Hedgecock and Pudovkin, 2011).
Methods
Broodstock origin, husbandry, and breeding scheme: Parental broodstock for this experiment
was provided by the Molluscan Broodstock Program at the Oregon State University, and were
deemed ‘wild’ for all intents and purposes. Two males and four females made up the breeding
scheme for this experiment, each of which provided several individuals of F1 offspring
(supplemental figure 1). Their tissues were divided and subsequently stored in a -80C freezer
until DNA extractions were needed. Larval cohorts were raised using common hatchery
techniques as described in Helm and Bourne, 2004. Briefly, adult Cgi were strip spawned, after
which their gametes were isolated and washed using 0.2 micron filtered sea water (FSW), and
subsequently mixed in order to produce appropriate families. Larvae were reared in 200 liter
tanks at 24C, at a density of approximately 10k per liter. All cohorts were fed even ratios of
microalgae throughout the rearing process, and held in the hatchery until they reached
approximately 2 cm in diameter, after which they underwent sacrificing for this study. Oysters
were subsequently named by their F1 family name and given a random letter to assign
individuality (e.g. 51_C = oyster from family 51, individual C). Parents were assigned a male
(M) or female (F) designation and a number (e.g. F100 = female parent oyster, individual 100).
Breeding scheme and naming can be found in supplemental figure 1; supplemental table 1
contains all names, lane well assignments, and barcode indices for each sample (more on library
generation below).
75
DNA Extraction: DNA extraction protocols specific for bivalve DNA were generated by myself
over the course of several years in the lab. The full protocol can be found in the supplemental
documents section in this thesis. Briefly, for both adult and juvenile samples, a tissue digestion
was performed using a diluted solution of pure proteinase K in Tissue and Cell Lysis buffer from
ThermoFisher Scientific (1:15 ratio). After appropriate tissue digestion, an MPC Protein
Precipitation Solution was used to remove excess lipids and proteins, and a column extraction
using a Zymo Research Quick-DNA Microprep kit was performed, followed by a column
cleaning using a Zymo Research DNA Clean and Concentrate (DCC) kit. The extracted DNA
was tested for molecular weight using a 1% agarose gel, and tested for salt and/or other
contaminants using a ThermoScientific NanoDrop machine. DNA concentrations (and all other
concentrations required throughout NGS library generation) were quantified using a Qubit 2.0
Fluorometer from ThermoScientifc.
Library Generation: For this study, we utilized a Kappa Biosystems Hyper Plus Library
Generation Kit, which includes randomly shearing fragmentase and individualized indexes.
Indexing by sample can be found in supplementary table 1. Preliminary quality control for the
libraries was performed at the USC Genome Core, including BioAnalyzer and qPCR analysis,
which ensured that libraries met standards in terms of target fragment size, concentration, and
appropriate adapter ligation. Libraries were then shipped to the NovoGene Corporation, which
performed further quality control analysis and sequencing using four lanes of Illumina Platform
PE150-compatible NovoGene HiSeq.
76
Computational analysis: All Illumina data can be downloaded under the NCBI BioProject
PRJNA566001.
Illumina reads from 25 samples were mapped to the Crassostrea gigas v9 assembly
(Zhang et al. 2012) using BWA MEM 0.7.9a-r78640 (Li & Durbin 2011). Polymorphisms were
called using the GATK pipeline (McKenna et al. 2010), which considers PCR duplicate removal,
indel realignment and base quality score recalibration, and calls variants across all samples
simultaneously through the HaplotypeCaller program. Variants were filtered using standard hard
filtering parameters according to GATK Best Practices recommendation (DePristo et al. 2011,
van der Auwera 2018): MQ > 56, QD > 24, MQRankSum < 8, FS < 12, and ReadPosRankSum <
3. These parameters ensure that the reads are mapped to a unique place in the assembly with high
quality (MQ), that the reads carrying both alleles are comparable in terms of mapping quality
(MQRankSum), that the actual variants are called with high quality (QD), and that the variants
are not biased towards one strand of the genome (FS) or towards the end of the reads
(ReadPosRankSum). Initial analysis of these results confirm that 5 samples were not the F1
progenies of any of the sequenced parents and were subsequently discarded. In addition, 2
samples were identical to other already-sequenced samples and were combined, leaving 18
unique samples for subsequent analysis.
Mutations in each F1 progeny were identified at genomic positions supported by 30, 40,
or 50 reads in all three members of the trio and where the F1 progeny carried an allele not found
in either parent. As a conservative measure, common mutations found between two or more
families were removed. Since VCF files only contain segregating positions across the genome
and not all sequenced positions, SAM file outputs from BWA were used to identify genomic
positions supported by 40 reads in all three members of the trio. Similar to the filtering of
77
segregating sites, genomic calls required scores of MQ > 56, QD > 24, FS < 12, and
ReadPosRankSum < 3. These analyses were repeated for the 30-read and 50-read requirement.
STRUCTURE (Pritchard et al. 2000, Falush et al. 2007, Hubisz et al. 2009) was used to
assess admixture and population structure based on 42,765 WGS polymorphic loci. The analysis
was run under the admixture and correlated allele frequency model. Ten independent runs of
10,000 burn-in Markov Chain Monte Carlo (MCMC) iterations followed by 50,000 iterations
were performed for 2 to 8 clusters (K = 2 to 8). Values of α were reported at < 0.1 in all runs at
50,0000 iterations, which is a measure of convergence. Results were inspected using
STRUCTURE HARVESTER (Earl, 2012) and quantified using the Evanno method (Evanno et
al. 2005). Independent runs at the same K yielded the same clustering with similar log-likelihood
values, and this consistency was taken into account by the Evanno method when assessing
clustering. Principal component analysis was performed using the SNPRelate package in R
(Zheng et al. 2012). Agegenet (Jombart 2008, Jombart et al. 2011) was used to characterize
genetic diversity between samples and produce Nei’s genetic distance matrixes.
Mutation frequency across DNA base-pair type, and figures & tables were all generated
using Microsoft Excel, Microsoft Powerpoint, and/or R statistical package (R Core Team, 2016).
Results
Sequencing Results: Illumina sequencing returned approximately 100-200M reads per
individual, and a retention rate of ~83% was observed after removing duplicate reads
(supplemental table 3). Variant calling of the Illumina data identified 5,093,185 SNPs
segregating among the 18 samples.
78
Nei’s genetic distance, PCA, and STRUCTURE: Nei’s genetic distance estimates
demonstrated that individuals 39_D and 40_A-C were likely not derived from the parental
crosses as assumed, and were not used in subsequent analysis (supplemental figure 2).
STRUCTURE and PCA supported this decision, as these individuals created distinct colors or
clusters, respectively (supplemental figure 2). All F1 members of family 51 were closely related
to each other, but the maternal lineage was not clear, as F116 and F117 both produced a Nei’s
distance of ~0.175 and 0.11, respectively. These values were higher than other parental averages
for Nei’s D (e.g. family 52 average Nei’s distance to predicted maternal lineage F117 =0.048).
Further, F116 and F101 were identical (Nei’s genetic distance = 0.0098), which was a second
line of evidence that F116 had been contaminated. When mutation rates were called for families
(see next section), family 52 had approximately 2- to 3-fold the number of mutation events than
those reported for individuals with confident combinations of Nei’s genetic distance,
STRUCTURE, and PCA results, and were therefore subsequently left out of mutation rate
estimates for this study. The total number of remaining individuals after this quality control was
8 F1 offspring (39_A, 39_C, 39_D, 40_G, 52_A, 52_B, 52_C, 52_G) across three total families
(39, 40, and 52).
Heterozygous sites and mutation rate across three coverages: The total number of no-
mutation, homozygous, and heterozygous mutated F1 states that arose from a fully homozygous
parental state (i.e. no-mut: ♂ A/A x ♀ A/A = F1 A/A; homo: ♂ A/A x ♀ A/A = F1 T/T; het: ♂
A/A x ♀ A/A = F1 A/T;) were tallied across three coverages and are shown in Table 1. For each
coverage threshold, all 6 possible sites among the trio needed to have met or exceeded the
corresponding coverage. Homozygous mutations are expected to be an artefact of library
79
generation and/or sequencing, and so were used as a control for our dataset: as coverage
increases from 30à50X, the number of homo mutations decreases dramatically, while the
number of het mutation calls decreases proportionally to the “total calls in trios” (no-mut + het +
homo). Other data checks included nspecting the distribution of het calls to ensure even coverage
across the genome, and ensuring that there was no positive correlation between mutation rate and
Nei’s genetic distance, which would indicate a bias as a function of relatedness to parental line
(supplemental figure 3). The “heterozygous SNP rate” was then determined by dividing the
total number of het cases into the “total calls in trios”, and was estimated across the series of
coverages. From this table, we observe an average rate of 1.73 x 10
-5
at 30X, 1.86 x 10
-5
at 40X,
and 1.90 x 10
-5
at 50X (table 1; figure 1). This is a conservative method for estimation of
mutation rate per site, because we do not consider examples of heterozygous parental state or
other types of possible mutations which would only add to our count. We hereafter refer to this
simply as the “mutation rate”. Other data relationships, including het versus homo SNP call rate,
ratios of these calls, and their relationship to total calls can be seen in supplemental figure 4.
Base-pair substitution rates: Using the calls provided during computational analysis, base-pair
SNP-type was calculated for A, T, C, and G, and for “*”, which was an indication of a deletion
spanning that particular base pair. This analysis demonstrated relatively even representation of
each possible SNP-type across each individual and each coverage, averaging ~25% for each (see
Table 2.1 and 2.2).
80
Discussion
High nucleotide diversity (p), large genetic load, and low effective population size fit a
model for high mutation rate: As mentioned in the background section above, allozyme studies
in the early 1990’s suggested that among animals, marine invertebrates are perhaps the most
diverse at the nucleotide level (Ward et al., 1992; Solé-Cava and Thorpe, 1991). Launey and
Hedgecock (2001) demonstrated, using 19 microsatellite markers, that the lethal genetic load in
Cgi (they estimate ~8-14 highly deleterious loci per organism) is larger than in mammals, birds,
drosophila, and conifers (Lynch, and Walsh, 1998), and suggested that a high mutation rate
might be sufficient to explain the segregation distortion which is common in bivalve genetic
studies. With current molecular techniques, the suggestion of very large genetic variation has
been supported, and nucleotide diversity (p) in Cgi has been shown to be elevated. Sauvage et al.
(2007) estimated 1 SNP for every 40 base pairs in non-coding regions, and that nonsynonymous
mutations make up a fairly high proportion of polymorphism for Cgi (pn / ps =0.16). Gagnaire et
al. (2018) estimated p = 0.0099 and 0.0101, for two continentally separated wild Cgi populations
using high density RAD sequencing data. In line with these studies, the oyster reference genome
assembly also demonstrated high polymorphism rates among a wild individual (p = 0.013), and
further suggested that high frequencies of indel mutations, and subsequent frameshifts, may also
provide a source for lethal mutations which are not considered in the above references, or this
study (Zhang et al. 2012).
Large effective population (Ne) sizes should theoretically remove deleterious mutations
(Kimura, 1983). However, Cgi and other highly fecund marine invertebrates and fishes are
thought to have very low Ne, and a recent review found many studies which supported the
hypothesis for Ne/N ratios much less than 0.01 for these types of organisms (Hedgecock and
81
Pudovkin, 2011). The first estimations of effective population sizes of hatchery populations for
Cgi were made in 1990, when allozymetric comparison of wild and reproductively isolated
commercial stocks indicated that an effective population size could be on the order of 8.9 and
40.6 for each respective hatchery population, though no wild stock Ne estimates were produced
in this study (Hedgecock and Sly, 1990). As genetic drift is theoretically inversely proportional
to Ne, some of the first wild stock estimates in Cgi were made by taking the inverse of random
genetic drift during temporal studies, where Ne was approximated between 50, 200-400, or 500,
depending on the year blocks between the 1970s-1990s. (Hedgecock, 1994; Hedgecock and
Pudovkin, 2011). These estimations of N e for Cgi are indeed several orders of magnitude lower
than for typical census (N) populations in healthy oyster beds which may have millions of
individuals, and rank among the lowest Ne reported for marine broadcast spawners, comparable
only to the European flat oyster Ostrea edulis (see Hauser and Carvalho, 2008, Table 1). The
Sweepstakes Reproductive Success (SRS, introduced above) hypothesis predicts just such an
observation, namely that animals under SRS pressure will “maintain less molecular genetic
diversity over evolutionary time-scales than expected from their abundance” (Hedgecock and
Pudovkin, 2011).
Under the neutral theory (Watterson 1975; Nei 1987), the level of expected diversity at
equilibrium is expected to be:
[ Q = 4Ne
µ ] Equation 1
where Q = calculated nucleotide diversity per site, and µ = mutation rate. Substituting previously
inferred average nucleotide diversity between genomes for Q (p = 0.013) and existing Ne
estimations, we can calculate that the mutation rate for Cgi might be expected to be between 6.5
x 10
-6
to 6.50 x 10
-5
(see Table 3). The substitution of p for Q makes some assumptions (namely
82
that Tajima’s D for this population is at zero), but here it is an appropriate approach, because Q
can tend to inflate as sample size increases whereas p will remain stable, both should
theoretically be of similar magnitude, and we are using p estimates generated from the entire
genome (Subramanian 2016 and references therein; their figures 3A and 6A). Results from this
exercise closely match the average of the estimated mutation rates previously predicted (~10
-05
;
Plough et al. 2016; see our Table 3 “µ predicted”). The average mutation rate empirically
observed in this study falls precisely within the range of predicted values from this exercise
(1.78x10
-05
, 1.92x10
-05
, and 1.9x10
-05
, at 30, 40 and 50X coverage, respectively. See Table 3 “µ
demonstrated”). Employing our most conservative µ in terms of coverage (50X), we can
calculate Q using Equation 1 for a range of predicted Ne, and show that this calculated Q value
falls very close to previously estimated p values (0.013) for an Ne of approximately 100-200 (see
Table 3 “Q Calculated (50X µ)”).
Frequent Mutation = Functional Adaptation Mechanism?
As mentioned throughout, evidence from the literature suggests a high genetic load and
several vQTL inherited per generation for Cgi, and a lower than expected Ne at small geospatial
scales (Hedgecock 1994). SRS suggests that chaotic oceanographic and physiological parameters
are the driving force behind low Ne, as the chances of two ripe organisms achieving a mating
event is similar to winning a sweepstakes. But how is high mutational load interacting with the
phenomenon of lower than expected Ne? Could a high mutational load be utilized as a within-
cohort adaptive strategy for Cgi and other dioecious simulcasters (organisms which spawn great
quantities of eggs and sperm from separate sexes)?
83
Adding to the picture is a recent study demonstrating the effect of environmental
parameters and a possible mechanism of local adaptation for oysters at larval stages. She et al.
(2018) exposed a combined 10♂x10♀ set of oyster crosses to one of three saline conditions for 8
days at larval stages: control (30 ppt), hyposaline (15 ppt), and hypersaline (45 ppt). DNA
analysis revealed that each of the populations had differentiated SNP profiles (e.g. 438
differentiated SNPs localized to 231 genes for the control:hypersaline comparison), and that the
hypersaline group had significantly enriched expression profiles when compared to controls (see
She et al. 2018: fig. 7). This study is remarkable in the fact that one ‘population’ of parental
oysters can result in demonstrably different end-cohorts, in terms of genetic profiles (SNPs),
based on very short windows (only 8 days) of slightly disparate environmental parameters
experienced during larval stages. She et al. hypothesized that this may be a mechanism of local
adaptation. Other studies have demonstrated a genetic cline associated with geography, and
subsequent origin effects in reciprocal transplant experiments, for the congener Crassostrea
virginica (Buford et al. 2016). Reviews on the topic of local adaptation for marine organisms
have demonstrated that many populations are not as interconnected as once thought (Palumbi,
2004), and that planktonic dispersers seem to make up a majority of species demonstrating high
local adaptation, and often at very small spatial scales (1m or less. See review: Sanford and
Kelly, 2011). Bringing this into the context of commercial applications, selection for bivalves
often occurs at a larval stage, as farmers will size-select large individuals via sieving, assuming a
correlation between fast growing larvae and high production juveniles and adults. Given the
extremely high fecundity of these organisms, new studies demonstrating genetic signatures of
single-generation adaptive responses, and the historic literature describing a high genetic load,
one possible hypothesis emerges that a high mutation rate may influence the ability of wild
84
marine dioecious simulcasters to cope with extreme SRS selective pressures by providing
copious novel genetic material for dynamic localized environmental adaptation.
Perhaps bacteria can provide an analogy that furthers this argument (albeit with
limitations). Bacteria divide rapidly and utilize high mutation rate alleles in order to ‘locally’
adapt to new environments (e.g. novel mouse gut environs), but the advantages of this approach
disappear after primary adaptation (e.g. Giraud et al. 2001). Furthermore, literature suggests that
there is a balance for bacteria between mutation rate and fitness advantages, as extremely high
mutation can be maladaptive in certain novel environments, and these rates quickly evolve to a
reduced level after adaptation (Sprouffske et al. 2018). Of course, some significant caveats exist
when comparing prokaryotes to eukaryotes (see Lynch, 2010). But effective population size (Ne)
has been shown to scale inversely with mutation rate (Nei, 1987; Lynch et al. 2016), and a
sufficiently low Ne may “…compromise the ability of selection to maintain high-fidelity
replication and/or repair mechanisms” (Lynch, 1995, Lynch 2006; Lynch 2007; quote from
Lynch, 2010; Lynch et al. 2016; and Piganeau and Eyre-Walker, 2009). Because Ne and
mutation rate are linked, this necessarily interacts with the so-called ‘cost-of-fidelity’ hypothesis
(Lynch 2010 and references therein), which describes a balance between lower mutation rate and
the costs associated with increased DNA repair fidelity. Bivalves must experience very large
costs to fidelity, considering their gamete count may approach billions per individual in the
oldest and largest organisms. Human germline mutation rates are higher in males than in females
and de novo mutations increase with a father’s age, both of which are most likely explained by
an increased cell proliferation rate (Rahbari et al. 2016 and references therein). Mutation rate in
Cgi and other bivalves therefore might be further pushed, due to their extreme fecundity across
both sexes. Considering the relationship of Ne and mutation rate, as recently reviewed by Lynch
85
(2016) and Krasovek et al. (2017), and high costs of fidelity associated with gametic
proliferation, we might predict a mutation rate of 1 x 10
-7
per nucleotide for Cgi by extension of
pan-organism data (Lynch 2016, figure 3b and their regression model; Krasovec et al., 2017,
figure 2). However, we find in this study a much more elevated mutation rate than the current
literature would allow us to predict based on historic mutation rate reports alone, approximately
two to three orders of magnitude more frequent: solving for the lowest conceivable Ne (x=1) for
Lynch’s 2016 regression versus our regression, which now includes Cgi, we see an upper limit of
the mutation rate at 1 x 10
-7
and 1 x 10
-4
, respectively.
(figure 5. See supplemental table 2 for
data as compiled originally by Lynch, 2016, now with the addition of Cgi).
Bringing the discussion into the realm of experimentation, a recent study in beetles
demonstrated evidence for increased germline mutation rate when exposed to novel challenging
conditions (Berger et al. 2017). This lends an important observational instance of how changing
the cost of fidelity (in this case via heightened resource limitation in warmer environments) can
lead to increased inherited genetic load in a eukaryotic model. Further, new computer simulation
and experimental evidence in Saccharomyces cerevisiae suggests that small deme-deme
migration is sufficient to fix mutator genes in a larger meta-population (Raynes et al., 2019),
despite previous research indicating that positive selection for mutators is only beneficial at
sufficiently large population sizes (Tenaillon et al. 1999; Andre and Godelle 2006; Wylie et al.
2009; Raynes et al. 2014, 2018). Marine invertebrates have been demonstrated to show “chaotic
genetic patchiness [heterogeneity]” at small spatial scales over evolutionarily-relevant time (e.g.
<50m, 2 years; limpets; Johnson and Black, 1982; Johnson and Black, 1984), and has also been
observed in urchins (Watts et al. 1990; Edmands et al. 1996), bivalves (Hedgecock, 1994; David
et al. 1997), and fishes (see review: Hauser and Carvalho, 2008). When coupled with small Ne,
86
as discussed above, these observations may afford us to hypothesize that Cgi and other bivalves
might be experiencing population structures similar to small demes that sometimes interact, as a
consequence of SRS. In this way, the Ranyes et al. (2019) study lend a potential evolutionary
mechanism to keep persistent high mutation load in the wild, namely via low Ne deme-deme
interactions.
As a summary to this section, I hypothesize that bivalves, which are capable of producing
many hundreds of millions of gametes in both sexes per spawn cycle, and who may never
achieve complete adaptation (a.k.a. ‘phenotypic optima’ a la Fischer, 1930) due to chaotic ocean
conditions & high SRS selection pressure (and subsequently low Ne), and who are potentially
under persistent oceanographically mediated small deme-deme migration, may be under
selective pressure to retain 1) high fecundity and 2) high mutation rates, and that 3) this may
allow for increased per-generation local adaptation success.
Common proofreading machinery may demonstrate a possible mechanistic route to higher
mutation rates in oyster gametes.
DNA polymerase δ and ε (cgDP-δ and cgDP-ε) each have 3’à5’ proofreading and
exonuclease capabilities associated with their catalytic subunits, and mutations in these regions
can significantly increase mutation rate, by orders of magnitude in some cases (Prindle and Loeb,
2012, and references therein). If one of these high-fidelity polymerases does make a mistake
(insertion, deletion, or mismatch), the DNA mismatch repair pathway (MMR) provides a second
tier of proofreading and correction mechanisms. The mutS family of proteins play a critical role
during the initial steps of the MMR pathway, as they recognize mismatches, bind to the error
site, and initiate the recruitment of downstream MMR proteins (Acharya et al., 2005; Sutton,
87
2015). Of the four common DNA polymerases among eukaryotes (α, β, δ and ε), the Cgi genome
annotation contains at least one set of entries pertaining to the various subunits associated with
each (NCBI), and contains at least two mutS-like protein annotations, homologs 4 and 5 (cg-
mutS-h4 and cg-mutS-h5, respectively).
Protein alignments using the annotated catalytic subunits of cgDP-δ & cgDP-ε, and cg-
mutS-h4 & cg-mutS-h5, to a variety of organisms – including several that likely are under high
SRS pressure - indicate that conservation of each of these proteins is high in Cgi, with coverage
at >97% as compared to the Saccharomyces cerevisiae reference sequence. Searching for
common mutation-rate increasing mutations in cgDP-δ & cgDP-ε did not reveal any obvious
differences in these critical sites (see supplemental figure 5, FASTA files available upon
request). However, both cgDP-δ and cg-mutS-h5 form separate clades on distance-uncorrected
neighbor joining trees when analyzed against this same set of species, along with their congener,
the Eastern oyster (Crassostrea virginica) (see supplemental figure 6). This may prove a
significant clue as to a mechanism, as DNA polymerase δ has been shown to be much less
accurate for deletions involving repetitive regions (Fortune et al. 2005), of which the oyster
genome is replete (Zhang et al. 2012). Further, conifers, which are known also to have extremely
high rates of mutation and massive per-individual fecundity, also tend to have large loads of
repetitive sequences (Jaramillo Correa et al. 2010; Kovach et al. 2010; Liu et al. 2011; Moritsuka
et al. 2012; Kusumi et al. 2015).
The source of the mutation load uncovered in this study likely stems from gametic
proliferation. Through in-situ hybridization and histological examination of mantle and gonad of
a vasa-like germline determinant, Fabioux et al. (2004) demonstrated that stem cells exist
throughout conjunctive mantle tissues during the ‘rest period’ and quickly proliferate during
88
reproductive-period initiation, indicating that the proliferation of these stem cells are responsible
for seasonal germ cell renewal. Plough et al. (2016) and others referenced throughout this
chapter have hypothesized that the extreme fecundity in Cgi and other marine invertebrates may
incur a higher likelihood of deleterious mutation residence time and initial frequency. Occam’s
razor supports the hypothesis that the genetic load estimated in our dataset is likely a function of
high mutation rate in parental gamete cells, as the only other explanation is an extreme somatic
rate of mutation accumulation in relatively young juvenile oysters. Assuming that captured here
are indeed genetic loads associated with gametic proliferation, perhaps the combination of
observed pol-δ outliers in oysters and highly repetitive genomes presents a mechanistic method
to produce low-fidelity genome replication, and remains a future direction of study in this
species.
Implications for commercial stock & selective breeding
Farmers want bivalves that have better survival rates, grow to market size faster, stay
alive on ice longer, generate consistent shell phenotypes (e.g. shape and color), and retain
preferable meat-to-shell ratios throughout the year. Much of the current research in aquaculture
seeks to develop selectively bred and improved varieties with the aid of some genetics and/or
genomics, as a method to increase the rate of return on desirable traits as compared to traditional
Mass Selection (MS) alone. Our study, and historic literature noting the rapid rate of the
surfacing of inbreeding depression within low- or fully-related family lines (typically within 2-3
generations), indicates that perhaps the most appropriate approach would be through a
combination of MS and Genomic Selection (GS). This would alleviate the need for developing
family lines that are prone to inbreeding depression and are difficult to keep genetically isolated,
89
as MS supplemented or in combination with GS requires the development of large cohorts in
which relatedness among individuals need not be high (Goddard and Hayes, 2007). Other
requirements include developed linkage maps and sufficient markers (SNPs) for analysis, which
the oyster currently enjoys (e.g. Hedgecock et al. 2015). Other bivalves may require more robust
genomics efforts in these arenas before they can begin a GS program. However, a GS breeding
program in conifers, which are also thought to have high mutational loads, has recently been
demonstrated to have the potential to significantly increase genetic gains (for diameter and
density) in approximately half the time of a traditional breeding program, from 17 down to 9
years (Li and Dungey, 2018). In their simulations, Li and Dungey (2018) also showed that a
higher heritability (0.2 vs 0.5) and larger training populations (3000) increased the rate of genetic
gain. This fits nicely into common bivalve breeding operations: heritability for commercial traits
is quite high (shell characteristics range from ~0.36-0.49, wet weight at ~0.35, and a correlation
between shell and meat traits at ~0.79; Kong et al., 2015) and extremely large production runs
are common, numbering in the thousands to hundreds of thousands. Other breeding schemes
might be feasible or profitable with Cgi, such as those which take advantage of hybrid vigor
from high-relatedness family lines, and there remains significant gaps in our understanding of
genotype by environment interactions in marine-farmed populations, but a program which
integrates MS and GS is likely to succeed - and possibly at a reduced effort due to the difficulties
in family line rearing in bivalves - if well managed.
90
Figures and Tables
Table 1: Mutations and Rates across coverage thresholds. Individuals are labelled with their
respective “Total Calls (M)”, which indicate the number of loci that met or exceeded the
coverage threshold at all 6 possible sites among the trio, excluding any between-family duplicate
sites. Combined SNPs indicate the sum of Homo and Het mutation calls, and the “Het SNPs
Rate” is what we deem the conservative generational mutation rate for Cgi.
30X Coverage, All Sites
F
1
Mat Pat
Total
Calls
(M)
Combined
SNPs
Homo
SNPs
Het
SNPs
Combined
SNPs Rate
Homo
SNPs
Rate
Het SNPs
Rate
39_A F100 M20 42.53 953 14 939 2.240E-05 3.291E-07 2.208E-05
39_C F100 M20 36.66 847 3 844 2.310E-05 8.182E-08 2.302E-05
39_D F100 M20 44.47 980 17 963 2.203E-05 3.822E-07 2.166E-05
40_G F101 M20 33.10 720 2 718 2.175E-05 6.041E-08 2.169E-05
52_A F117 MCB26 169.11 2488 309 2179 1.471E-05 1.827E-06 1.288E-05
52_B F117 MCB26 168.86 2187 203 1984 1.295E-05 1.202E-06 1.175E-05
52_C F117 MCB26 167.51 2070 131 1939 1.235E-05 7.821E-07 1.158E-05
52_G F117 MCB26 69.28 971 6 965 1.401E-05 8.661E-08 1.393E-05
Average 1.732E-05
St. Dev 4.851E-06
40X Coverage, All Sites
F
1
Mat Pat
Total
Calls
(M)
Combined
SNPs
Homo
SNPs
Het
SNPs
Combined
SNPs Rate
Homo
SNPs
Rate
Het SNPs
Rate
39_A F100 M20 15.93 382 1 381 2.397E-05 6.276E-08 2.391E-05
39_C F100 M20 13.92 360 0 360 2.586E-05 0 2.586E-05
39_D F100 M20 16.25 404 2 402 2.486E-05 1.230E-07 2.474E-05
40_G F101 M20 13.03 321 0 321 2.463E-05 0 2.463E-05
52_A F117 MCB26 76.12 953 70 883 1.252E-05 9.195E-07 1.160E-05
52_B F117 MCB26 75.96 886 52 834 1.166E-05 6.846E-07 1.098E-05
52_C F117 MCB26 75.00 793 23 770 1.057E-05 3.067E-07 1.027E-05
52_G F117 MCB26 22.74 382 1 381 1.680E-05 4.398E-08 1.676E-05
Average 1.859E-05
St. Dev 6.469E-06
50X Coverage, All Sites
F
1
Mat Pat
Total
Calls
(M)
Combined
SNPs
Homo
SNPs
Het
SNPs
Combined
SNPs Rate
Homo
SNPs
Rate
Het SNPs
Rate
39_A F100 M20 9.12 214 0 214 2.347E-05 0 2.347E-05
39_C F100 M20 8.16 193 0 193 2.366E-05 0 2.366E-05
39_D F100 M20 9.25 219 0 219 2.367E-05 0 2.367E-05
40_G F101 M20 7.84 190 0 190 2.422E-05 0 2.422E-05
52_A F117 MCB26 30.89 421 15 406 1.363E-05 4.856E-07 1.314E-05
52_B F117 MCB26 30.79 417 7 410 1.355E-05 2.274E-07 1.332E-05
52_C F117 MCB26 30.18 374 5 369 1.239E-05 1.657E-07 1.223E-05
52_G F117 MCB26 12.39 235 1 234 1.897E-05 8.073E-08 1.889E-05
Average 1.908E-05
St. Dev 5.040E-06
91
Figure 1: Mutation Rate is Stable Across Coverage. For each individual, mutation rates stayed
in a similar range across coverages thresholds of 30, 40, and 50X. Each individual is colored
across its own coverage data points, and a linear model was estimated from the combined data. A
very small slope (9 x 10
-8
) and R
2
(0.016) suggest that coverage is not a significant factor in
determining mutation rates as estimated from our data. See also supplemental figure 4.
y = 9E-08x + 1E-05
R² = 0.01661
5.00E-06
1.00E-05
1.50E-05
2.00E-05
2.50E-05
3.00E-05
30 40 50
Mutation Rate
Coverage
Mutation Rate Across Coverage Thresholds
Oy_39_A
Oy_39_C
Oy_39_D
Oy_40_G
Oy_52_A
Oy_52_B
Oy_52_C
Oy_52_G
Linear (Het SNPs Rate)
3.00 x 10
-5
2.50 x 10
-5
2.00 x 10
-5
1.50 x 10
-5
1.00 x 10
-5
92
Table 2: Similar distribution across SNP-type for all individuals. Across all individuals and
coverages, SNP-type (A, C, G, T, or “*”, which indicates a deletion event at that nucleotide)
remained at even ratios among all possible types (table 2.1), demonstrating a skew maximum of
31% (T/total, individual 39_C at 50X) and minimum of 18% (C/total, individual 39_C at 50X)
for any individual among all 120 possible categories, and an average of 24% for all SNP types
across individuals. Taking the entire dataset into consideration (table 2.2), rather than by
individual, we see even distribution of SNP type counts and ratios averaging 0.2493 (~25%)
across each SNP-type, excluding deletion events.
COVERAGE A C G T * A C G T * A C G T * A C G T *
30X 242 215 209 268 6 214 186 187 247 10 262 214 203 272 13 209 185 136 184 4
40X 96 96 87 101 1 91 74 84 110 1 108 92 86 116 0 94 78 63 86 0
50X 56 57 43 57 1 49 35 47 61 1 57 53 49 60 0 58 41 39 52 0
COVERAGE A C G T * A C G T * A C G T * A C G T *
30X 591 499 469 620 2 559 454 435 532 7 538 427 439 532 5 281 220 212 252 2
40X 245 213 190 234 1 236 213 184 199 2 189 184 188 207 3 101 89 86 105 1
50X 107 101 93 105 0 117 101 90 102 0 82 94 92 101 1 65 58 53 58 1
51_A SUMMARY 52_B SUMMARY 52_C SUMMARY 52_G SUMMARY
39_A SUMMARY 39_C SUMMARY 39_D SUMMARY 40_G SUMMARY
A C G T * (del)
30X 2896 2400 2290 2907 49
40X 1160 1039 968 1158 9
50X 591 540 506 596 4
A C G T * (del)
30X 0.27 0.23 0.22 0.28 0.005
40X 0.27 0.24 0.22 0.27 0.002
50X 0.26 0.24 0.23 0.27 0.002
TOTAL DATA SUMMARY
TOTAL DATA, RATIO
1
2
93
Neutral Theory This study
Q or p constant Ne µ predicted µ demonstrated Coverage Q Calculated (50X µ)
0.013 4 50 6.50E-05 1.78 x 10
-5
30 0.00395
0.013 4 100 3.25E-05 1.93 x 10
-5
40 0.00791
0.013 4 200 1.63E-05 1.98 x 10
-5
50 0.01581
0.013 4 300 1.08E-05
0.02372
0.013 4 400 8.13E-06
0.03162
0.013 4 500 6.50E-06
0.03953
Table 3: Predicted mutation rates (µ) from literature fall mirror empirically demonstrated
µ: Under Neutral Theory (Watterson 1975; Nei 1987), and following equation 1 as described
above and utilizing previously estimated Ne (Hedgecock, 1994) and nucleotide diversity (Q
substituted as previously estimated p; Zhang et al. 2012), the predicted mutation rate (µ) fell
between [ 6.50 x 10
-5
< µ < 6.50 x 10
-6
] (left side of table). Our study empirically demonstrates a
µ that falls precisely within this range (right side of table), and demonstrates a calculated Q value
for the 50X rate of µ which falls very close to previously estimated p values for an N e of
approximately 100-200.
94
Figure 5: The Pacific oyster (Cgi) is an outlier among eukaryotes thus far described.
Organisms with robust mutation rate estimates, as described in Lynch 2016 (Lynch’s figure 3b
and supplemental table 1 therein), are plotted with the addition of the Pacific oyster mutation
rates described in this study. The x-axis is a log10 scale. The black dotted line represents a fitted
power regression (y=3x10
-7
x
-0.341
) with data sans Cgi, while the blue dotted line refits the power
regression with Cgi estimates included (y = 0.0003x
-0.881
,
see supplemental table 2 for full
dataset).
1.0E-10
1.0E-09
1.0E-08
1.0E-07
1.0E-06
1.0E-05
1.0E-04
1.0E+02 2.0E+03 4.0E+04 8.0E+05
Base-substitution mutation rate
(/ site / generation) (u)
Effective population size (Ne)
Mutation Rate Across Eukaryotes
Pacific oyster (Crassostrea gigas)
Mouse-ear Cress (Arabidopsis thaliana)
Roundworm (Caenorhabditis briggsae)
Roundworm (Caenorhabditis elegans)
Water flea (Daphnia pulex)
Fruit fly (Drosophila melanogaster)
Postman Butterfly (Heliconius melpomene)
Human (Homo sapiens)
Common mouse (Mus musculus)
Rice (Oryza sativa)
Chimpanzee (Pan troglodytes)
Effective Population Size (N
e
) (log scale)
Base-substitution mutation rate
(per site, per generation) (µ)
1.0E-10
1.0E-09
1.0E-08
1.0E-07
1.0E-06
1.0E-05
1.0E-04
1.0E+02 2.0E+03 4.0E+04 8.0E+05
Base-substitution mutation rate
(/ site / generation) (u)
Effective population size (Ne)
Mutation Rate Across Eukaryotes
Pacific oyster (Crassostrea gigas)
Mouse-ear Cress (Arabidopsis thaliana)
Roundworm (Caenorhabditis briggsae)
Roundworm (Caenorhabditis elegans)
Water flea (Daphnia pulex)
Fruit fly (Drosophila melanogaster)
Postman Butterfly (Heliconius melpomene)
Human (Homo sapiens)
Common mouse (Mus musculus)
Rice (Oryza sativa)
Chimpanzee (Pan troglodytes)
1 x 10
2
2 x 10
3
4 x 10
4
8 x 10
5
1 x 10
-4
1 x 10
-5
1 x 10
-6
1 x 10
-7
1 x 10
-8
1 x 10
-9
95
Supplemental Figures
Supplemental Figure 1: Mating Design. Females (vertical column) and males (horizontal
column) were selected and bred in the following pairings. The red and blue highlighting indicate
the females and males used for this study’s parent stock, respectively, and the yellow coloring
indicates the family line naming adapted from these crosses. Four individuals from each family,
and corresponding parental tissues, were sequenced using random amplification libraries using
Kappa Biosystem’s Nextera kit.
96
Supplemental Table 1, Specimen Naming and sequencing information. Table describes the
total number of individuals sequenced (left column), with expected family separated by color
blocks. Well assignments, naming for sequencing at Novogene, and respective P5 and P7 indices
are listed.
Specimen
Barcode
Well
NAMING FOR NOVOGENE P5 Index P7 Index
39-4 A3 Oy_39_4 TATAGCCT CGCTCATT
51-E B3 Oy_51_E ATAGAGGC CGCTCATT
51-B C3 Oy_51_B CCTATCCT CGCTCATT
51-C D3 Oy_51_C GGCTCTGA CGCTCATT
51-F E3 Oy_51_F AGGCGAAG CGCTCATT
51-D F3 Oy_51_D TAATCTTA CGCTCATT
52-C G3 Oy_52_C CAGGACGT CGCTCATT
52-5.2 H3 Oy_52_5.2 GTACTGAC CGCTCATT
52-5.1 A4 Oy_52_5.1 TATAGCCT GAGATTCC
52-6.1 B4 Oy_52_6.1 ATAGAGGC GAGATTCC
52-6.2 C4 Oy_52_6.2 CCTATCCT GAGATTCC
M20 D4 Oy_M20 GGCTCTGA GAGATTCC
M-CB26 E4 Oy_M_CB26 AGGCGAAG GAGATTCC
F100 F4 Oy_F100 TAATCTTA GAGATTCC
F101 G4 Oy_F101 CAGGACGT GAGATTCC
F116 H4 Oy_F116 GTACTGAC GAGATTCC
F117 A5 Oy_F117 TATAGCCT ATTCAGAA
52-G B5 Oy_52_G ATAGAGGC ATTCAGAA
39-C H2 Oy_39_C GTACTGAC TCCGGAGA
39-D G2 Oy_39_D CAGGACGT TCCGGAGA
39-G F2 Oy_39_G TAATCTTA TCCGGAGA
40-A E2 Oy_40_A AGGCGAAG TCCGGAGA
40-B D2 Oy_40_B GGCTCTGA TCCGGAGA
40-C C2 Oy_40_C CCTATCCT TCCGGAGA
40-G B2 Oy_40_G ATAGAGGC TCCGGAGA
97
Supplemental Figure 2: Nei’s genetic distance estimates, STRUCTURE, and PCA
demonstrate family structure. Nei’s genetic distance (A) for all families indicated some
contamination or mixing of tissues for individuals 39_G and 40_A-C (closer to 0 indicates a
closer genetic relationship), and so were omitted from further mutation analysis. Family 51 also
has a closer relationship to F117 than for its predicted maternal line, F116, which would
eventually necessitate this family’s removal from future mutation rate analysis. STRUCTURE
(B) and PCA (C) analysis also demonstrate three strong groupings of families, and indicate a
similar trend to (A) in terms of individuals with known and unknown provenance. The total
number of remaining individuals after this quality control was 8 F1 offspring (39_A, 39_C,
39_D, 40_G, 52_A, 52_B, 52_C, 52_G) across three total families (39, 40, and 52).
Oy_F100 Oy_39_A Oy_39_C Oy_39_D Oy_39_G Oy_M20 Oy_40_A Oy_40_B Oy_40_C Oy_40_G Oy_F101 Oy_F116 Oy_51_A Oy_51_B Oy_51_C Oy_51_D Oy_51_F Oy_M_CB26Oy_52_A Oy_52_B Oy_52_C Oy_52_G Oy_F117
Oy_F100 0 0.0662392 0.0642052 0.0639193 0.1745879 0.176437 0.2139988 0.1758677 0.2122276 0.1976148 0.2279042 0.19703 0.2128612 0.21388 0.2069862 0.2203461 0.2042202 0.1993691 0.2206964 0.220252 0.21958 0.208807 0.154469
Oy_39_A 0.0662392 0 0.0781521 0.0700207 0.1895995 0.0629077 0.2288429 0.1912714 0.2285013 0.1427096 0.2387177 0.2094696 0.2271771 0.2323189 0.2257706 0.2403341 0.2252961 0.2207099 0.2407307 0.2368565 0.2380468 0.2271906 0.1731441
Oy_39_C 0.0642052 0.0781521 0 0.0831209 0.183334 0.0628461 0.2264055 0.1859764 0.222388 0.1378759 0.2331649 0.2040366 0.2277749 0.2268495 0.2204938 0.2351295 0.2225612 0.2168845 0.2334471 0.2337832 0.2327682 0.2209479 0.1660555
Oy_39_D 0.0639193 0.0700207 0.0831209 0 0.1899291 0.0602877 0.2304602 0.1893905 0.2298963 0.1483994 0.2401908 0.2103161 0.2311265 0.2332113 0.2258575 0.2415182 0.2250322 0.2212292 0.2400487 0.2404866 0.2366003 0.2294522 0.1742555
Oy_39_G 0.1745879 0.1895995 0.183334 0.1899291 0 0.1751218 0.18661 0.1520213 0.186575 0.1798481 0.1996299 0.1659743 0.1888422 0.1898231 0.183847 0.1977504 0.1833251 0.1744591 0.1955964 0.1958871 0.1963808 0.1859204 0.1288006
Oy_M20 0.176437 0.0629077 0.0628461 0.0602877 0.1751218 0 0.2137787 0.1764241 0.212867 0.065168 0.2200478 0.1912068 0.217588 0.2178384 0.209648 0.2258417 0.2132757 0.2068429 0.2252661 0.2235809 0.2206885 0.2123936 0.1529758
Oy_40_A 0.2139988 0.2288429 0.2264055 0.2304602 0.18661 0.2137787 0 0.1871043 0.0434766 0.2180712 0.2334493 0.2069258 0.2282842 0.2292806 0.2264009 0.2369175 0.2206724 0.217924 0.2377099 0.2341811 0.2333251 0.2224789 0.1733353
Oy_40_B 0.1758677 0.1912714 0.1859764 0.1893905 0.1520213 0.1764241 0.1871043 0 0.1846478 0.1794437 0.2003274 0.1672801 0.1897368 0.191726 0.1869368 0.1971486 0.1828343 0.1787327 0.2002783 0.1952347 0.1940377 0.1832829 0.1238895
Oy_40_C 0.2122276 0.2285013 0.222388 0.2298963 0.186575 0.212867 0.0434766 0.1846478 0 0.2230141 0.2397653 0.2119926 0.2270215 0.2291978 0.2247636 0.2364246 0.2227375 0.2168504 0.2366609 0.2334865 0.2319164 0.2224276 0.1736611
Oy_40_G 0.1976148 0.1427096 0.1378759 0.1483994 0.1798481 0.065168 0.2180712 0.1794437 0.2230141 0 0.0735651 0.065558 0.2251811 0.225928 0.2172838 0.2326773 0.2189167 0.2120661 0.2317838 0.2297104 0.2271715 0.2193732 0.161167
Oy_F101 0.2279042 0.2387177 0.2331649 0.2401908 0.1996299 0.2200478 0.2334493 0.2003274 0.2397653 0.0735651 0 0.0098269 0.2412456 0.2423669 0.2353766 0.2482339 0.2350366 0.2298926 0.2480315 0.2437324 0.2435581 0.2373901 0.1840783
Oy_F116 0.19703 0.2094696 0.2040366 0.2103161 0.1659743 0.1912068 0.2069258 0.1672801 0.2119926 0.065558 0.0098269 0 0.1800185 0.180964 0.1760394 0.1880346 0.1741021 0.1869168 0.210954 0.2087468 0.2079405 0.19972 0.1340235
Oy_51_A 0.2128612 0.2271771 0.2277749 0.2311265 0.1888422 0.217588 0.2282842 0.1897368 0.2270215 0.2251811 0.2412456 0.1800185 0 0.0332963 0.0501541 0.0427573 0.0639817 0.0503942 0.1518327 0.145589 0.1422064 0.1467092 0.1151156
Oy_51_B 0.21388 0.2323189 0.2268495 0.2332113 0.1898231 0.2178384 0.2292806 0.191726 0.2291978 0.225928 0.2423669 0.180964 0.0332963 0 0.0442231 0.0724281 0.0484636 0.0495176 0.1392334 0.1616353 0.1413987 0.1460046 0.1153997
Oy_51_C 0.2069862 0.2257706 0.2204938 0.2258575 0.183847 0.209648 0.2264009 0.1869368 0.2247636 0.2172838 0.2353766 0.1760394 0.0501541 0.0442231 0 0.0594603 0.0736862 0.0487375 0.0896604 0.1331911 0.1125399 0.1115708 0.1067548
Oy_51_D 0.2203461 0.2403341 0.2351295 0.2415182 0.1977504 0.2258417 0.2369175 0.1971486 0.2364246 0.2326773 0.2482339 0.1880346 0.0427573 0.0724281 0.0594603 0 0.0730126 0.0531555 0.1161315 0.1419077 0.1292832 0.1230726 0.1195331
Oy_51_F 0.2042202 0.2252961 0.2225612 0.2250322 0.1833251 0.2132757 0.2206724 0.1828343 0.2227375 0.2189167 0.2350366 0.1741021 0.0639817 0.0484636 0.0736862 0.0730126 0 0.0486701 0.1288812 0.125487 0.1296619 0.1515656 0.1071341
Oy_M_CB260.1993691 0.2207099 0.2168845 0.2212292 0.1744591 0.2068429 0.217924 0.1787327 0.2168504 0.2120661 0.2298926 0.1869168 0.0503942 0.0495176 0.0487375 0.0531555 0.0486701 0 0.0538026 0.0550532 0.0521944 0.051498 0.1095551
Oy_52_A 0.2206964 0.2407307 0.2334471 0.2400487 0.1955964 0.2252661 0.2377099 0.2002783 0.2366609 0.2317838 0.2480315 0.210954 0.1518327 0.1392334 0.0896604 0.1161315 0.1288812 0.0538026 0 0.0840336 0.0677818 0.0672864 0.0511809
Oy_52_B 0.220252 0.2368565 0.2337832 0.2404866 0.1958871 0.2235809 0.2341811 0.1952347 0.2334865 0.2297104 0.2437324 0.2087468 0.145589 0.1616353 0.1331911 0.1419077 0.125487 0.0550532 0.0840336 0 0.0642179 0.0714097 0.0468932
Oy_52_C 0.21958 0.2380468 0.2327682 0.2366003 0.1963808 0.2206885 0.2333251 0.1940377 0.2319164 0.2271715 0.2435581 0.2079405 0.1422064 0.1413987 0.1125399 0.1292832 0.1296619 0.0521944 0.0677818 0.0642179 0 0.0416606 0.0486484
Oy_52_G 0.208807 0.2271906 0.2209479 0.2294522 0.1859204 0.2123936 0.2224789 0.1832829 0.2224276 0.2193732 0.2373901 0.19972 0.1467092 0.1460046 0.1115708 0.1230726 0.1515656 0.051498 0.0672864 0.0714097 0.0416606 0 0.0464615
Oy_F117 0.154469 0.1731441 0.1660555 0.1742555 0.1288006 0.1529758 0.1733353 0.1238895 0.1736611 0.161167 0.1840783 0.1340235 0.1151156 0.1153997 0.1067548 0.1195331 0.1071341 0.1095551 0.0511809 0.0468932 0.0486484 0.0464615 0
A
B C
F116
F117
MCB26
F100
M20
F101
FAMILY 51 AND 52
FAMILY 39
FAMILY 40
98
0
1000
2000
3000
4000
5000
6000
7000
0 500000 1000000 1500000 2000000 2500000
Chromosome (Scaffold) Number
Position
All Het Calls
0
500
1000
1500
2000
2500
3000
0 50000 100000 150000 200000 250000 300000
Chromosome (Scaffold) Number
Position
All Het Calls (truncated)
-400
600
1600
2600
3600
4600
5600
6600
0 200000 400000 600000 800000 1000000
Chromosome (Scaffold) Number
Position on Chrom
Family 39 Het Call Distribution (full)
Fam39_A
Fam39_C
Fam39_D
0
500
1000
1500
2000
2500
3000
0 50000 100000 150000 200000 250000 300000
Chromosome (Scaffold) Number
Position on Chrom
Family 39 Het Call Distribution (truncated)
Fam39_A
Fam39_C
Fam39_D
0
1000
2000
3000
4000
5000
6000
7000
0 500000 1000000 1500000 2000000 2500000
Chromosome (scaffold) Number
Position on Chromosome
Family 54 Het Call Distribution (full)
Fam52_A
Fam52_B
Fam52_C
Fam52_G
0
200
400
600
800
1000
1200
1400
1600
1800
2000
0 50000 100000 150000 200000
Chromosome (scaffold) Number
Position on Chromosome
Family 54 Het Call Distribution (further
truncation)
Fam52_A
Fam52_B
Fam52_C
Fam52_G
y = -0.0173x + 0.0749
R² = 0.51014
ANOVA P-value = 5.62e-05
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0 0.5 1 1.5 2
Nei's Genetic Distance, Maternal
Mutation Rate * 1M
Mutation Rate x Nei's Mat, 40X
Fam39
Fam40
Fam52
y = -0.013x + 0.0685
R² = 0.49815
ANOVA P-value = 2.076e-05
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0 0.5 1 1.5 2
Nei's Genetic Distance, Paternal
Mutation Rate * 1M
Mutation Rate x Nei's Pat, 40X
Fam39
Fam40
Fam52
A B
C D
E F
G H
99
Supplemental Figure 3 (prev. page): Read Distribution and Quality Control for sequencing
data. Read distribution was determined to be even across the Cgi genome for all data (A and B),
and for all families and individuals inspected (39: C & D; 52: E & F; Family 40 not shown but
similar to others). Each inset (B, D, F) are truncated versions of (A, C, E, respectively) in order
to better visualize the data homogeneity. G and H plot the mutation rates estimated for all data
against the relationship of the offspring to the parent as per Nei’s genetic distance for the
maternal (G) and paternal (H) line, for our estimates at 40X coverage. A negative correlation
indicates that there is no bias in the data which would create higher mutation rates as a function
of increased distance from the parent.
100
Supplemental Figure 4: Relationships between SNP count, total calls, and coverage
thresholds. SNP count, both Homo and Het decreased as a function of coverage from 30à50X
(A), though Homo SNP counts approached or achieved zero by 50X, whereas Het SNP counts
remained at similar orders of magnitude (notice separate y-axis). The ratio of Homo/Het SNP
calls also decreased to <5% as coverage went from 30à50X (B), indicating a reduction of likely
erroneous Homo calls to a nominal amount in relationship to Het calls. The total calls decreased
as a function of increasing coverage (C), but still remained at approximately the same order of
magnitude, averaging 38.6M at 40X, and 17.4M at 50X. The relationship between the number of
Het calls (i.e. trustworthy mutation events) decreases as a function of coverage, and does so
roughly in accordance with total calls (D). In the unlikely scenario that we have only managed to
receive sequences that are very highly prone to mutation events in our sequencing data, we’ve
generated a very conservative mutation rate estimate by taking only the het SNP calls at each
coverage, and dividing that number into the predicted genome size of 560Mb. This is likely so
excessively conservative an approach that these numbers are untrustworthy, in the author’s
opinion.
0
500
1000
1500
2000
2500
0
50
100
150
200
250
300
30 40 50
Het SNP count
Homo SNP count
Coverage
SNP count by coverage
Hom SNPs
Het SNPs
0.000
0.020
0.040
0.060
0.080
0.100
0.120
0.140
0.160
30 40 50
Ratio, Homo/Het SNP Count
Coverage
SNP Ratio by Coverage
0
20
40
60
80
100
120
140
160
180
30 40 50
Number Calls/1M
Coverage
Total Calls by Coverage
0
500
1000
1500
2000
2500
0
20
40
60
80
100
120
140
160
180
30 40 50
Het SNP count
Total calls in Trios (in M)
Coverage
Total calls and Het SNPs across Coverage
Total calls in trios (M)
Het SNPs
0.00E+00
5.00E-07
1.00E-06
1.50E-06
2.00E-06
2.50E-06
3.00E-06
3.50E-06
4.00E-06
4.50E-06
30 40 50
Het calls/560Mb
Coverage
Conservative Het Mut Rate
A D
B E
C
101
Organism Mutation Rate Effective Population Size
Pacific oyster (Crassostrea gigas) 0.0000185925325 225
Mouse-ear Cress (Arabidopsis thaliana) 0.0000000069500 288669
Roundworm (Caenorhabditis briggsae) 0.0000000013277 267380
Roundworm (Caenorhabditis elegans) 0.0000000014500 541379
Water flea (Daphnia pulex) 0.0000000056900 826011
Fruit fly (Drosophila melanogaster) 0.0000000051650 863020
Postman Butterfly (Heliconius melpomene) 0.0000000029000 2068966
Human (Homo sapiens) 0.0000000135126 21091
Common mouse (Mus musculus) 0.0000000054000 177315
Rice (Oryza sativa) 0.0000000071000 52817
Chimpanzee (Pan troglodytes) 0.0000000120000 28750
Roundworm (Pristionchus pacificus) 0.0000000020000 1750000
Supplemental Table 2: Mutation Rate and Effective Population Size Among Eukaryotes.
Original data from Lynch, 2016, supplemental table 1, and references therein. Here, the authors
add the Pacific oyster and associated mutation rate as demonstrated in this study.
102
DNA Polymerase-! ; Catalytic Subunit
No easily discernible mutations found in conserved regions
931
Church et al 2013
612 644
Prindle and Loeb 2012
Saccharomyces cerevisiae
Drosophila melanogaster
Homo sapiens
Mus musculis
Oncorhynchus mykiss
Salmo salar
Strongylocentrotus purpuratus
Crassostrea gigas
Crassostrea virginica
Saccharomyces cerevisiae
Drosophila melanogaster
Homo sapiens
Mus musculis
Oncorhynchus mykiss
Salmo salar
Strongylocentrotus purpuratus
Crassostrea gigas
Crassostrea virginica
Saccharomyces cerevisiae
Drosophila melanogaster
Strongylocentrotus purpuratus
Crassostrea gigas
Crassostrea virginica
Homo sapiens
Mus musculus
Oncorhynchus mykiss
Salmo salar
Salmo trutta
Salvelinus alpinus
Oncorhynchus tshawytscha
Oncorhynchus kisutch
Oncorhynchus nerka
275 286 297
411
446 456
Saccharomyces cerevisiae
Drosophila melanogaster
Strongylocentrotus purpuratus
Crassostrea gigas
Crassostrea virginica
Homo sapiens
Mus musculus
Oncorhynchus mykiss
Salmo salar
Salmo trutta
Salvelinus alpinus
Oncorhynchus tshawytscha
Oncorhynchus kisutch
Oncorhynchus nerka
275 286 297
Saccharomyces cerevisiae
Drosophila melanogaster
Strongylocentrotus purpuratus
Crassostrea gigas
Crassostrea virginica
Homo sapiens
Mus musculus
Oncorhynchus mykiss
Salmo salar
Salmo trutta
Salvelinus alpinus
Oncorhynchus tshawytscha
Oncorhynchus kisutch
Oncorhynchus nerka
275 286 297
DNA Polymerase-ε ; Catalytic
Subunit
Church et al. 2013
A
B
103
Supplemental Figure 5 (previous page): DNA Polymerase Delta and Epsilon do not
common mutator-inducing mutations. Protein alignments for DNA polymerase delta and
DNA polymerase epsilon catalytic sites did not return any immediate clues as toward a
molecular mechanism for such high mutation in Cgi. Alignments (and trees, supplemental figure
6) were done using the Multiple Sequence Alignment (MSA) tool MUSCLE, which stands for
MUltiple Sequence Comparison by Log- Expectation, accessed on the EMBL-EBI web tool
package. Both DNA polymerase catalytic sites tested suggested that all functional domains were
present.
104
Supplemental Figure 6: Neighbor-joining trees suggest that Cgi is possible outgroup for
conserved proofreading enzymes. DNA polymerase Delta and mutS-homolog5 create
outgroups when placed in neighbor-joining trees, as per the default parameters on the EMBL-
EBI MUSCLE output tool. Future work in this field might consider investigating the function of
these enzymes as a possible molecular mechanism to maintain high generational mutation rates.
Saccharomyces cerevisiae
Drosophila melanogaster
Strongylocentrotus purpuratus
Crassostrea gigas
Crassostrea virginica
Homo sapiens
Mus musculus
Oncorhynchus mykiss
Salmo salar
Salmo trutta
Salvelinus alpinus
Oncorhynchus tshawytscha
Oncorhynchus kisutch
Oncorhynchus nerka
DNA Polymerase-ε ; Catalytic Subunit
*No NCBI entries described for Mytilus galloprovincalis, Pinus spp.; Ruditapes spp.; Patella spp.
Neighbor-joining Tree, no distance corrections
Left: Cladogram, Right: Real
Saccharomyces cerevisiae
Drosophila melanogaster
Homo sapiens
Mus musculis
Oncorhynchus mykiss
Salmo salar
Strongylocentrotus purpuratus
Crassostrea gigas
Crassostrea virginica
DNA Polymerase-! ; Catalytic Subunit
*No NCBI entries described for Mytilus galloprovincalis, Pinus spp.; Ruditapes spp.; Patella spp.
Neighbor-joining Tree, no distance corrections
Left: Cladogram, Right: Real
Saccharomyces cerevisiae
Oncorhynchus nerka
Salmo trutta
Salvelinus alpinus
Homo sapiens
Mus musculus
Strongylocentrotus purpuratus
Crassostrea gigas
Crassostrea virginica
muS; Homolog 5
*No NCBI entries described for Mytilus galloprovincalis, Pinus spp.; Ruditapes spp.; Patella spp.
Neighbor-joining Tree, no distance corrections
Left: Cladogram, Right: Real
Saccharomyces cerevisiae
Crassostrea gigas
Crassostrea virginica
Strongylocentrotus purpuratus
Homo sapiens
Mus musculus
Salvelinus alpinus
Salmo trutta
Salmo salar
Oncorhynchus mykiss
Oncorhynchus tshawytscha
muS; Homolog 4
*No NCBI entries described for Mytilus galloprovincalis, Pinus spp.; Ruditapes spp.; Patella spp.
Neighbor-joining Tree, no distance corrections
Left: Cladogram, Right: Real
105
Supplemental Table 3: Read distribution across samples.
Lane ID
Number of
Reads
Sequenced
Individual
Number of Reads
Uniquely Mapped
(Duplicates
Removed)
Rate of
Read
Retention
Oy_39_A.USPD16101799_HK5H2DSXX_L3 196773234 Oy_39_A 159551718 0.810841
Oy_39_C.USPD16101817_HK5H2DSXX_L3 161934668 Oy_39_C 130939477 0.808594
Oy_39_D.USPD16101818_HK5H2DSXX_L3 204217812 Oy_39_D 165618217 0.810988
Oy_39_G.USPD16101819_HK5H2DSXX_L3 199508414 Oy_39_G 159266963 0.798297
Oy_40_A.USPD16101820_HK5H2DSXX_L3 156407454 Oy_40_A 126997441 0.811965
Oy_40_B.USPD16101821_HK5H2DSXX_L3 170118216 Oy_40_B 135730048 0.797857
Oy_40_C.USPD16101822_HK5H2DSXX_L3 182264520 Oy_40_C 132457712 0.726733
Oy_40_G.USPD16101823_HK5H2DSXX_L3 156846530 Oy_40_G 112091322 0.714656
Oy_51_A.USPD16101806_HK5H2DSXX_L3 425047300 Oy_51_A 341701207 0.803913
Oy_51_B.USPD16101801_HK5H2DSXX_L3 211719494 Oy_51_B 176477975 0.833546
Oy_51_C.USPD16101802_HK5H2DSXX_L3 180887504 Oy_51_C 145693343 0.805436
Oy_51_D.USPD16101804_HK5H2DSXX_L3 634936150 Oy_51_D 542695232 0.854724
Oy_51_F.USPD16101803_HK5H2DSXX_L3 252375700 Oy_51_F 204168088 0.808985
Oy_52_A.USPD16101800_HK5H2DSXX_L3 217590408 Oy_52_A 532275371 0.835891
Oy_52_A.USPD16101808_HK5H2DSXX_L3 419185958
Oy_52_B.USPD16101807_HK5H2DSXX_L3 211587558 Oy_52_B 419524375 0.852998
Oy_52_B.USPD16101809_HK5H2DSXX_L3 280235764
Oy_52_C.USPD16101805_HK5H2DSXX_L3 336974954 Oy_52_C 277748109 0.82424
Oy_52_G.USPD16101816_HK5H2DSXX_L3 131410656 Oy_52_G 110676651 0.84222
Oy_F100.USPD16101812_HK5H2DSXX_L3 123146508 Oy_F100 106658358 0.866109
Oy_F101.USPD16101813_HK5H2DSXX_L3 123046990 Oy_F101 104553754 0.849706
Oy_F116.USPD16101814_HK5H2DSXX_L3 171057500 Oy_F116 141626703 0.827948
Oy_F117.USPD16101815_HK5H2DSXX_L3 224793502 Oy_F117 188467797 0.838404
Oy_M20.USPD16101810_HK5H2DSXX_L3 196029298 Oy_M20 169857343 0.86649
Oy_M_CB26.USPD16101811_HK5H2DSXX_L3 196741234 Oy_M_CB26 170989007 0.869106
106
SUPPLEMENTAL CHAPTERS
A Note on Supplemental Chapters
The purpose of including the following two chapters are to document the generation of
interesting datasets that merit further work or analysis. During my graduate studies, I was unable
to dedicate the necessary time to complete these projects in addition to the work already
presented, but I would very much dislike the data to fall into a notebook that sits on a permanent
shelf (be it digital or literal). Therefore, I have decided to write a data report here in hopes that
either I or some future scientist can further these projects.
107
SUPPLEMENTAL CHAPTER 1: Generation of Mussel (Mytilus
galloprovincialis) family lines, in concert with relevant phenotyping schedule,
in order to elucidate viability- and growth-associated biomarkers.
Background: generation of Mediterranean mussel (Mytilus galloprovincialis) family lines.
Terrestrial selective breeding programs for food crops and livestock serve as a template
for ocean-based farmed protein production, and are comparatively significantly more advanced
in most cases. For example, many commonly farmed plants and animals have a published
genome, a critical component for elucidating best-approaches to reduce inbreeding risk,
phenotype causality, population genetics, allelic introgression, and best breeding approaches
(e.g. corn (Aanen et al. 2009), wheat (Appels et al. 2018), rice (Matsumoto et al. 2005), cows
(Zimin et al. 2009), pigs (Groenen et al. 2012), chicken; (Warren et al. 2017)). In addition,
several cultivars and varieties with highly valuable phenotypes exist for popular food items,
including genes associated with flavor in tomato (Gao et al. 2019), alleles responsible for iron
enrichment in rice (Ludwig et al. 2019), and abiotic stress resistance haplotypes for wheat (Miao
et al. 2017). Another of the key advantages that terrestrial crops have in comparison to oceanic
foods is that they have been previously domesticated through (often thousands of) years of
traditional selective breeding (Larson et al. 2014), which have greatly aided in the effort to
connect phenotype to genotype and separate genome by environment phenomenon. Seafoods, on
the other hand, have a severe dearth of domesticated species. Salmon, trout, and other finfish
represent the most developed breeding programs for aquaculture (López et al. 2019; Pulcini et al.
2015; Fabrice 2018), but for the majority of seafood, the item on the table is essentially a wild
animal. Such is the case for bivalves. Pacific oysters (Crassostrea gigas, (Cgi)) represent the
108
most well studied bivalves in terms of potential for selective breeding, with a published genome
(Zhang et al. 2012), established linkage maps (Hedgecock et al. 2007), and pedigree based
breeding programs (see Chapter 3 of this thesis for many references). Other bivalves have some
resources such as transcriptomic profiles (e.g. the Manilla clam, Mun et al. 2017), but the
Mediterranean mussel (Mytilus galloprovincialis (Mgal)) sits in a close second place to Cgi in
terms of genomic resources, having a recently published genome (Murgarella et al. 2016) and
many population genetics studies associated with hybrid zone research (Viard et al. 1994; Braby
and Somero, 2006; Shields, Heath, and Heath 2010; and Crego-Prieto et al. 2015, to name just a
few). The reason that both oysters and mussels tend to have more developed genomic resources
are many, but a considerable driver is the fact that they are in the top four most valuable bivalves
globally (FAO, 2018). Even so, Mgal lacks any scientifically managed familial breeding
programs with publicly disseminated data, to the knowledge of the authors. This is despite the
fact that traditional selective breeding programs for bivalves do demonstrate the potential for
improvement across other mussel species (Symonds et al. 2019) and oysters (Kong et al. 2015).
Returning to the theme of terrestrial crops as a template for aquaculture: it seems only
appropriate to begin replication of varietal development via population and family line
generation in bivalves, as the terrestrial literature has demonstrated this as a key tool for
understanding genomic underpinnings of commercially important traits.
Through a NOAA SBIR grant that I authored in collaboration with Catalina Sea Ranch, and a
subsequent proposal we (Nuzhdin and myself) drafted in collaboration with the Hedgecock and
Manahan labs through the Waitt Foundation, I was able to build out the scientific
accommodations described in the general introduction (pgs. XX-XX). The main thrust of the
109
NOAA SBIR was to generate family lines of Mediterranean mussels, Mytilus galloprovincialis,
with a goal of performing a Genome Wide Association Study (GWAS) for commercially
important traits, namely larval growth rate and family survival. Ideally, this study could begin to
elucidate potential QTLs and/or SNPs that may be used in selective breeding programs, or for
understanding the best wild populations to target for broodstock. Using the hatchery, I
implemented an NCII mating design to begin the process of investigating viability- and survival-
associated loci for Mgal. The primary literature inspiring this work comes from a handful of
papers detailing the observance of specific shifts in genomic structure as broadcast spawning
marine invertebrate larval populations mature into their respective surviving cohort
(Hedgecock’s main works, see chapter 3 of this thesis). In addition to family generation and
phenotyping, I took genomic samples from each family population, similar to the approaches just
referenced, in order to understand how genomic profiles might shift as populations undergo
extreme mortality common to type-III survivorship life-histories. As of the time of writing this
thesis, phenotypes have been recorded and 73 double digest Restriction site Associated DNA
sequencing (ddRAD) libraries relevant to 8 families have been generated. Phenotype to genotype
association analysis is a future goal of this project. This write-up serves to demonstrate the state
of the data.
Methods
Broodstock source: Mussels were obtained via SCUBA diving the oil rig “Beta Offshore
Platform Ellen”, in the Los Angeles channel (33.582968, -118.128579). Mussels were hand
collected and checked for species and gravidity before collection. Collections took place in early
February, 2017, and utilized a broodstock collection permit held by Catalina Sea Ranch.
110
Family generation and animal husbandry: Family line husbandry followed a generally
accepted norm for bivalve hatcheries, outlined in Helm and Bourne, 2004. A brief recounting
follows. Ripe broodstock were induced to spawn using the “heat shock” method in a group
spawn tank. Our lab has had little historic success by individually isolating mussels and
subsequently inducing spawning, as these bivalves seem to prefer conspecific gametes and/or
chemical signals in the water in order for any great number of individuals to spawn, unless
specifically conditioned to be ready to spawn in a separate system, which these were not. This
presents the problem of family line contamination, as bivalves will inherently siphon some sperm
and/or eggs when in a group tank. To alleviate the possibility of family line contamination, I
developed a gamete collection raceway (GCR), which consisted of a 1 meter long 4 inch PVC
pipe cut lengthwise and with a water entry on an elevated side and a water exit on the downhill
end, so as to produce a unidirectional water flow through a trough. Individuals who were
observed to begin the process of spawning in the group tank were isolated and allowed to
continue spawning in the GCR for approximately 10 minutes, thereby expelling any
contaminating gametes inhaled. Tests of this system reduced contamination to undetectable
levels, as quantified by observations of fertilized and/or dividing embryos in an expectedly
unfertilized egg clutch (data not shown). After collection of sperm or eggs, single parent crosses
were generated by titrating sperm into an egg solution until approximately 10 sperm could be
observed surrounding each egg on a subsample slide. After approximately 30 minutes, the eggs
were washed with 1 micron filtered seawater (hereafter FSW) to avoid polyspermy effects. After
fertilization, embryos were quantified using a Sedgewick Rafter and associated recommended
counting protocol (see Helm and Bourne, 2004), and divided into tanks at an even starting
111
density of 10,000 per liter, into 12 liter tanks. Starting family size was therefore 120,000
individuals. Note that the Coefficient of Variation, calculated as the standard deviation divided
by the average, was targeted to be between 0.1 and 0.2 or lower, which ensured confidence in
our counts for all populations mentioned here and below. Families were raised in the family line
system described in the introduction of this thesis, which allowed for even distribution of
microalgal feed and aeration between tanks. Other tank parameters were tested and found to be
quite consistent between tanks (see results). Feeding schedules followed a target of one feeding
episode per 3 days, with a density of approximately 100,000 cells of Isochrysis galbana per
milliliter, or equivalent nutritional count as determined by utilizing cross-species microalgae
tables provided in Helm and Bourne, 2004, for Tetraselmis spp. and Chaetocerous spp., as
multi-species microalgae feed schedules were utilized after the larvae reached an appropriate
size. Families were reared and phenotyped as described below, and at day 80 settled onto 1m
lengths of “fuzzy rope”, eventually being out-planted on day 114 on the Cat Harbor long line,
where they currently are held.
Family Design and Phenotyping Scheme: The mating design for this study was a full factorial
cross (NC II design) using 4 females and 2 males, with three tank replicates for each family, for a
total of 8 families replicated across 24 tanks. Each replicate was standardized in terms of
stocking density, feed allocation and feed type, and water parameters tested (pH, temperature,
and salinity), using an Onset Hobo thermocoupler K-Type device and handheld YSI probe
device. The larval stages of the family lines underwent a phenotyping protocol for survival and
size on days 0, 11, 16, and 23. On day 23 the populations were further divided into two
phenotypic ranges, large and small, using a three-tiered sieve system for size gradation. Size was
112
measured using an Olympus SZ-PT dissecting microscope with Techniquip 150W Fiber Optic
Illuminator, at 20x total magnification and capturing images using built in DP2-BSW XV
Imaging Processing Software. Counting for population survival was performed by capturing and
concentrating the entire cohort by use of an appropriately sized mesh sieve, and placing this
concentrate into 50 mL conical vials. A small bubbler was place at the bottom of the conical vial,
such that the cohort concentrate could be homogenized throughout the water column. At this
point, several aliquots were retrieved and placed on a Sedgwick Rafter counting cell, 5% ethanol
was added to each sample to reduce swimming capacity of larvae, and each aliquot quantified
using a compound microscope at 4x magnification. By ensuring that the coefficient of variation
was as small as possible, it is possible to accurately estimate the total number of surviving larvae
in the population.
Samples were collected on each of the phenotyping days for population genomic
sequencing (n individuals ranged from 48 to 11,625, depending on the population and date),
including for both size cohorts collected on day 23. The end result, after replication of families
and division of sizes, was a total of 73 larval populations for which I created ddRAD libraries
using SphI-HF and Mlu-CI restriction enzymes (see below), which were determined in a separate
assay to be adequate for Mgal genomic digestion (see supplemental documents for full ddRAD
protocol). Libraries were pooled, tested for quality control, and sequenced across two illumina
“NovoSeqS4” lanes by the Novogene Corporation.
DNA Extraction: DNA extraction protocols specific for bivalve DNA were generated by myself
over the course of several years in the lab. The full protocol can be found in the supplemental
documents section in this thesis. Briefly, for both adult and larval samples, a tissue digestion was
113
performed using a diluted solution of pure proteinase K in Tissue and Cell Lysis buffer from
ThermoFisher Scientific (1:15 ratio). After appropriate tissue digestion, an MPC Protein
Precipitation Solution was used to remove excess lipids and proteins, and a column extraction
using a Zymo Research Quick-DNA Microprep kit was performed, followed by a column
cleaning using a Zymo Research DNA Clean and Concentrate (DCC) kit. The extracted DNA
was tested for molecular weight using a 1% agarose gel, and tested for salt and/or other
contaminants using a ThermoScientific NanoDrop machine. DNA concentrations (and all other
concentrations required throughout NGS library generation) were quantified using a Qubit 2.0
Fluorometer from ThermoScientifc.
NGS Library Generation: The full protocol for library preparation can found in the
supplemental documents section of this thesis. In brief, a two-enzyme strategy was developed
after determining that SphI-HF and Mlu-CI had adequate cutting behavior for the Mytilus
genome, targeting fragments between ~100-900bp, as determined by a multi-enzyme assay (data
not shown). Libraries were generated using an adapted protocol provided initially by Wendy Vu
for Chickpea, and further modified by myself to accommodate bivalve extractions and updated
lab reagents and techniques. After library creation, the USC Genome Core ran Quality Control
assays, including BioAnalyzer and qPCR analysis, which ensured that libraries met standards in
terms of target fragment size, concentration, and appropriate adapter ligation. Libraries were
shipped to the NovoGene Corporation, which performed further quality control analysis.
114
Results
System conditions control: In order to ensure that variance in abiotic conditions between tanks
as nominal, I took temperature, pH, salinity, and percent dissolved oxygen measurements.
Temperature was shown to be stable between tanks in the rearing system by taking
measurements every 30 seconds for a 22 day period (2/17/2017-3/10/2017), and observing near-
perfect congruency between tanks throughout the system, regardless of position (e.g. top shelf,
West vs. East end of the room, proximity to door, etc.). This was consistent with several previous
results indicating that temperature was stable between tanks (data not shown). During this time,
measurements taken using a YSI instrument indicated that similar to temperature, pH, salinity,
and percent dissolved oxygen were fluctuating very little across tanks. See figure 1 for system
conditions report.
Survival and Growth: Survival curves seemed to be semi-unstable in terms of consistency of
order among all cohort replicates across time (figure 3), but a stabilization of the hierarchy
among families seems to have occurred somewhere around settlement (figure 2A, Day 16 to Day
23 transition). Growth rates differed between families significantly at various timepoints. For
example, an ANOVA and subsequent Tukey’s test from data taken on day 16 indicate that
certain cohorts may be performing, on average, better than others both within family (e.g. 1.1R1
vs. 1.1R3) and among families and replicates (figure 4 top and middle panels). However, the
general trend was not predictive through multiple time points (see figure 4, bottom panel), or to
the settlement stage and outplanting stage (day 114, see figure 2B). However, by the time of
settlement and out-planting, there were clear better performing families (e.g. 1.3 and 2.1
115
compared to 1.4, figure 2B). There was a likely significant negative correlation between
survival and size (figure 2D; R
2
:0.38).
Library Generation: ddRAD libraries generated and sequenced across two lanes of NovoSeqS4
illumina returned with a total of 5.65B and 6.13B reads, respectively, for a total dataset of
approximately 11B reads identifiable to individual after filtering. Three samples, 1.4_D0,
1.2_D0, and 2.1R1_D23, returned only 2,626; 201,888; and 2.26M known reads, respectively,
while one sample (1.4R2_D11) exceeded the average significantly with 107M reads. Excluding
these outliers, the range was 89.1M, with a maximum of 93.9M reads (2.2R2_D11) and a
minimum of 4.8M reads (1.4R3_D23_Sm). Including all sample except 1.4_D0, the dataset has
an average depth of 53M and 56M per sample for lanes 1 and 2, respectively. The total dataset
consists of ~11B known reads. A general display of reads can be seen in figure 5, with heat
mapping to indicate extreme outliers.
A total of 8 family replicates report complete congruence between presence of a genomic
sample and phenotypic profile for growth and survival, while another 5 cohorts have congruence
through day 23 (D23) but have only a big (B) population recorded on that day. One family
replicate (2.4R1) has all but D16 recorded for genomic profiles. Family 1.4 was the only family
to have phenotyping and genomic data across all possible timepoints and replicates, while all
other families have some portion of data missing through at least one replicate. The above is
summarized in figure 5, middle panel for lane 1. Read count for lane 1 and lane 2 is highly
similar (figure 5, bottom panel).
116
Discussion
As these results still require much more analysis, I will keep this discussion fairly brief.
In general, this dataset has the potential to be helpful in elucidating both viability loci for Mgal
and overall associations with phenotype-to-genotypes relationships for larvae and spat (day 114
in this experiment). It will be interesting to see whether the best family, in terms of survival
(family 1.4), has fewer viability loci associated with D23 survivors than the worst families
(family 1.2), which we can assume (in this case) would be coming from the mother, as these
families share the same father. More interesting will be to compare parental DNA sequencing
results to end-point survivors, to understand which genes were incompatible due to epistatic
interactions.
Figure 2D shows a very interesting negative correlation between size and survival at day
23. This is consistent with the idea that slow growing larval cohorts may underperform in
survival as compared to their fast-growing cohorts, which is a commercially important trait, and
perhaps selectable. But is also potentially consistent with the notion that the experimental setup
influenced this relationship: a few remaining large individuals will receive more feed than a
cohort with many remaining small individuals, and thus the relationship between size and
survival is only due to food rations. This latter explanation is not as likely to be true, as the
feeding schedule for this experiment was designed to be ad libitum, especially at larval and
juvenile stages, but cannot be fully ruled out as contributing to this effect. In fact, this
experimental variable would not be possible to control in this system unless a very rigorous
scheme was devised whereby populations were standardized to have even densities of population
throughout the study, and adjusted as-such at regular intervals. This would be nearly impossible
to complete for family lines, as there is simply not enough biological material, but might be
117
possible with a series of very large multi-parent cohorts that are genetically differentiated from
one another. An experiment like this could tease apart the relationship between genetic influence,
between population density, and eliminate nutritional variables. This study would be applicable
only to hatchery design, as application of any results in an experiment like this would be difficult
to apply to wild systems, and so the likelihood of it ever being executed is minimal.
As a final note, I will add that family line generation for the purposes of teasing apart
genotype-to-phenotype relationships in bivalves is a job for either 1) a very large and
accomplished academic team or 2) a large and accomplished commercial research and
development sector. The amount of work required for me to execute this experiment was
extreme, and it will need to be repeated at a larger scale for any value to be added to commercial
breeding schemes. For this reason, I would recommend that mass selection and genomic
selection be used for breeding purposes in bivalve aquaculture until significant progress has been
made in the industry in terms of R&D power and investment of farming companies. This is not
to place a negative spin on this data, as it will absolutely be informative to marine organismal
evolutionary theory if significant relationships between genetic background and traits are
discovered.
118
Figures
Figure 1, stable between-tank conditions in Blue House. The top panel shows temperature
readings, taken every 15 minutes for 21 days, between 7 tanks relevantly spaced throughout the
BlueHouse hatchery. The x-axis is the datapoint number, the y-axis is the temperature in C. The
bottom panel shows a histogram of this data, and demonstrates that the largest gap between
hottest and coolest temperature between tanks was <1C, and was never above 2.5C. Other
relevant salt water parameters are shown in the inset table.
11
13
15
17
19
21
23
1
15
29
43
57
71
85
99
113
127
141
155
169
183
197
211
225
239
253
267
281
295
309
323
337
351
365
379
393
407
421
435
449
463
477
491
505
519
533
547
561
575
589
603
617
631
645
659
673
687
701
715
729
743
757
771
785
799
813
827
841
855
869
883
897
911
925
939
953
967
981
All Data Top&Bottom Rows 2/17/2017 -3/10/2017
2.8 C 2.4 C 1.7 C 1.22 C 1.12 C 1.1 C 1.18 C
0
100
200
300
400
500
600
0 0.5 1 1.5 2 2.5 3 More
Frequency
MaxDelta Temperature Between Tanks
Histogram Max Temp Change
119
Figure 2, size and survival correlations for full dataset. Larval survival curves (A) are shown,
averaged per family for day zero (D0) through day 23 (D23). The secondary y-axis on the right is
scaled down for better visualization of the data. Growth rates (B) are shown between D16 and
D114, just before out-planting at an open environment site. The sub table demonstrates the
number of remaining replicates per family at these time points. Correlations between average
size per family and survival at D23 are shown in (C). Similar data is shown in (D), but each
individual replicate is separated, and a linear regression and associated R
2
is shown.
160
170
180
190
200
210
220
230
240
D16 D23 D114
Size in µm
Larvae-to-Juvenile Transition Growth
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
4000
4500
5000
5500
6000
6500
7000
7500
8000
0 0.2 0.4 0.6 0.8 1 1.2
Chart Title
4000
4500
5000
5500
6000
6500
7000
7500
8000
0 0.2 0.4 0.6 0.8 1 1.2
Chart Title
160
170
180
190
200
210
220
230
240
D16 D23 D114
Size in µm
Larvae-to-Juvenile Transition Growth
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
3 2 2 3 2 1 n
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larvae Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larval Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larvae Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larval Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
R² = 0.38823
0.00
2000.00
4000.00
6000.00
8000.00
10000.00
12000.00
14000.00
16000.00
210 215 220 225 230 235 240 245
Survival in Population
Average Size (µm)
Size x Survival Across All Replicates, D23
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larvae Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larval Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larvae Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larval Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
R² = 0.38823
0.00
2000.00
4000.00
6000.00
8000.00
10000.00
12000.00
14000.00
16000.00
210 215 220 225 230 235 240 245
Survival in Population
Average Size (µm)
Size x Survival Across All Replicates, D23
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larvae Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larval Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larvae Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larval Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
205
210
215
220
225
230
235
240
245
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
D23 Size
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
205
210
215
220
225
230
235
240
245
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
D23 Size
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
D23 Survival
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
D23 Survival
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
205
210
215
220
225
230
235
240
245
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Day 23 Survival x Size Comparison
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larvae Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
160
170
180
190
200
210
220
230
240
D16 D23 D114
Size in µm
Larvae-to-Juvenile Transition Growth
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
D23 Size x Survival
205
210
215
220
225
230
235
240
245
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
D23 Size
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
205
210
215
220
225
230
235
240
245
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
D23 Size
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
D23 Survival
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
D23 Survival
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
205
210
215
220
225
230
235
240
245
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Day 23 Survival x Size Comparison
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larvae Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
160
170
180
190
200
210
220
230
240
D16 D23 D114
Size in µm
Larvae-to-Juvenile Transition Growth
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
D23 Size x Survival
A C
D B
205
210
215
220
225
230
235
240
245
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
D23 Size
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
205
210
215
220
225
230
235
240
245
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
D23 Size
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
D23 Survival
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
D23 Survival
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
205
210
215
220
225
230
235
240
245
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Day 23 Survival x Size Comparison
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
20000
40000
60000
80000
100000
120000
140000
D0 D11 D16 D23 D23
Remaining Larvae
Larvae Survival Curves by Family
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
0
2000
4000
6000
8000
10000
12000
14000
0 0.2 0.4 0.6 0.8 1 1.2
Axis Title
Chart Title
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
160
170
180
190
200
210
220
230
240
D16 D23 D114
Size in µm
Larvae-to-Juvenile Transition Growth
F1.1 F1.2 F1.3 F1.4 F2.1 F2.4
D23 Size x Survival
120
Figure 3, Survival with replicates separated. This graph demonstrates survival for all
replicates across all timepoints (1 = D0, 2 = D11, 3 = D16, 4 = D23). This chart demonstrates
that for some replicates, the hierarchical position is fairly stable, while for other replicates some
instability is seen. However, when taken on average, between D16 and D23, hierarchical position
seems stable (see figure 2A).
0
20000
40000
60000
80000
100000
120000
140000
1 2 3 4
Survival, All 31817 Families, All Data, All Reps
1.1R1 1.1R2 1.1R3 1.2R1 1.2R2 1.2R3 1.3R1 1.3R2 1.3R3 1.4R1
1.4R2 1.4R3 2.1R1 2.1R2 2.2R1 2.2R2 2.3R1 2.3R2 2.4R1
121
Figure 4, D16 and D23 growth patterns. Growth at D16 is shown in the top panel. Diamonds
indicate between-family significant differences, as determined by an ANOVA and Tukey’s HSD
test. The second panel describes the Tukey’s HSD test, with Xs indicating where significant
differences were found. The total number of significant differences are tallied along the diagonal.
The shading indicates where within family comparisons are made, and the yellow highlighting
indicates between-family significant differences (relating to diamonds in top panel). The bottom
panel describes growth rates between D16 and D23.
160
170
180
190
200
210
220
230
240
250
D16 D23
Pre-settlement Larval Growth Rates
1.1R1 1.1R2 1.1R3 1.2R3 1.3R1 1.3R3
1.4R1 1.4R2 1.4R3 2.1R1 2.2R1 2.4R1
122
0
20000000
40000000
60000000
80000000
100000000
120000000
0 10 20 30 40 50 60 70 80
Read Count
Library, Sorted by Size
Sequencing Read Depth, Sorted
LANE 1
LANE 2
123
Figure 5 (previous page), ddRAD library data spread. The top panel shows a heat-map of
returned reads across all replicates, and demonstrates a fairly even rate of sequencing return,
though some had excessive return (blue) or few reads returned (red). The bottom panel describes
the number of reads associated with each replicate and for each timepoint, for lane 1 (lane 2
returned similar results, data not shown). Yellow columns indicate a replicate with full data
presence, from D0 through D23SM. Green highlighting indicates full data presence except for
D23SM, and un-highlighted columns indicate replicates with missing data. The bottom panel
shows the read depth, sorted by size on the x-axis, between lane 1 and 2, and indicates
congruence between return.
124
Supplemental Chapter 2: Sperm and Egg Chemoattraction Studies: a series of
preliminary experiments suggests the case for Egg Competition in a marine
dioecious simulcaster
Introduction
Sexual Competition in Marine Invertebrates.
One of Darwin’s earliest interests as a graduate student were marine invertebrates, of which he
expressed his theory that in the context of sexual selection (mate choice), such basal trophic
creatures “are not sufficiently advanced to allow of the feelings of love and jealousy, or of the
exertion of choice” (1871, p. 613, “The descent of man, and selection in relation to sex”). Since
Darwin’s time, the field of sexual selection has come to recognize that “choice” extends beyond
the immediate organism’s ability to visually or chemically perceive mates and make estimations
of their fitness (see review by Firman et al. 2017). Indeed, male cues that result in an ‘exertion of
choice’ in the form of a chosen partner from a female may even be rendered irrelevant at a post-
copulatory level (e.g. Fisher et al. (2016) demonstrated that precopulatory competition did not
preclude post-copulatory competition in wild crickets). The idea that competition among males
and their ejaculates for access to female gametes and/or successful offspring may continue long
after any copulatory event is now commonly referred to in terms of sperm competition and as
Cryptic Female Choice (CFC) (Thornhill 1983; Eberhard 1996). The phrase employs the word
“cryptic” because it occurs inside the female’s reproductive tract and out of human sight, at least
in the historic sense of the theory. The mechanism(s) of CFC can take many forms (Eberhard
describes 24 of them in his 2009 paper on the topic, which is surely not an exhaustive list,
ranging from penetration allowance, to intrauterine physiological preparations, to offspring care
125
(Eberhard 2009; p. 10026, Table 2). This form of sexual selection on a population is often
attributed as the causal agent behind high differentiation in gonad morphology between the sexes
and between species (e.g. Birkhead and Pizzari, 2002; Schultz et al., 2016).
Post-copulatory sexual selection theory typically attributes internal physiological and
biochemical processes as the fodder for evolution, but what about broadcast spawning organisms
which release gametes from both sexes (which I term “dioecious simulcasting”) for external
fertilization? By comparison, those that fall into this category have been under the investigative
microscope far fewer times than their internally fertilizing counterparts, despite their mode of
reproduction often being considered the ‘ancestral condition’ (Levitan 2005). The reason for the
dearth of work regarding marine broadcast spawners in the context of post-copulatory sexual
selection is primarily due to the difficulty in studying the phenomenon in water. (Note: the term
“copulation”, for dioecious simulcasters, is here adjusted to ‘post-spawning’ or ‘post-spawn-
event’ for accuracy). In fact, when both male and female gametes are in the same environment,
with multiple contributors from both sexes, the classical CFC and sperm competition archetype
might shift (Levitan 2005). What’s more, CFC can be studied more directly by scientists in these
types of systems given an appropriate experimental approach, now possible in scientific
hatcheries.
Levitan was the first to study the phenomenon in a population of broadcast spawning sea
urchins, where his data supported the idea that variance among male and female reproductive
success was lower and higher, respectively, than that of polygamous internally fertilizing
organisms (Levitan 2004). Only in the case where population density is intermediate do we see
the sexual selection variance expected from the classical Bateman’s principle, namely that males
show higher variance than the females in reproductive success (Bateman 1948; Levitan 2005, see
126
figure 6). Outside of these conditions, Levitan posits that sperm density (limited or
overabundant) can decrease both the variance in, and general success of, both sexes’
reproductive output; a ‘nobody wins’ scenario. These conditions, outside of the Bateman “zone”,
may often be experienced by marine dioecious simulcasters. In fact, the interaction between the
density of population and sperm availability is thought to be able to affect the evolution of
Gamete Recognition Proteins (GRPs), such as female sperm-bindin, when outside of the
Bateman zone (Levitan and Ferrel 2006). Though intraspecific differences in the biochemical
makeup of chemoattractants or GRPs remains an open question (Evans and Sherman, 2013),
chemoattractant production (quantity) has recently shown to mediate gamete interactions in a
broadcast spawner (Hussain et al. 2017). Many familiar with the field conclude that egg
competition is likely occurring at some rate for marine broadcast spawners (e.g. Marshal and
Evans, 2005; and see review: Evans and Sherman 2013, and references therein), and recent
works demonstrate some potential within-clutch mechanisms of egg fertilization-capacity
adjustments (Okamoto 2016).
Evans et al. (2012) began the first investigations of the post-spawn interactions between
sperm and eggs in the blue mussel (Mytilus galloprovincialis), a dioecious simulcaster. In their
paper, they employ an elegant experimental rig, which allowed the deposition of two samples of
eggs at either end of a long tube, with a hole in the center in which sperm can be deposited,
dubbed a dichotomous choice chamber (DCC). Using the DCC, the authors were able to
demonstrate that when presented with a choice between two female egg clutches, sperm swam
toward the clutches with which they had experienced more fertilization success with during no-
choice control trials (d.f. = 15, F = 1.95, p = 0.026) (Evans, et al. 2012). Following on this work,
Oliver and Evans (2014) were able to attribute these ‘preference’ phenomenon to chemical cues
127
alone, by placing separated chemoattractant from eggs in DCCs, and largely replicating sperm
swimming choice results (d.f. = 14, F = 2.261, p = 0.016) from Evans et al. 2012. They further
posit that their data demonstrates a link between ideal mate ‘choice’ in sperm, and embryonic
viability (see their figure 1).
Using their study as inspiration, I sought to create my own dichotomous choice chambers
to determine if I could replicate these results using another dioecious simulcaster, the Pacific
oyster, Crassostrea gigas (Cgi). Interestingly, initially I observed that control fertilization trials
would not necessarily predict the percent success of fertilization in competition trials, and that
predicting the ‘winning’ female during competition trials was not possible based on control
fertilization trials. I hypothesized that eggs were responding to chemosensory cues from other
eggs in the environment, and adjusting their fertilization success rate in accordance by some
mechanism. Here I describe results from a new experimental system, in combination with
preliminary DCC results, which support my hypothesis and suggest that eggs may adjust their
fertilization success rate depending on chemosensory cues from competing eggs. In the
discussion section, I argue that my findings are not in contrast to earlier work, but rather that
there is a congruence of these outcomes in the literature. These results, though preliminary, may
have important implications for sexual selection theory, population genomics, and commercial
hatchery work.
Methods
Experimental Overview
There were four types of experiments that produced the results described in this supplemental
chapter. The first set, the Dichotomous Choice Chamber (DCC) trials, were designed such that
each of two females and one male would first undergo a control fertilization, where gametes
128
were mixed in isolation (e.g. FA x MA and FB x MA) in 50 mL conical vials filled to 40 mL, in
order to determine an expected base-line fertilization success rate. Next, eggs or “egg water”
were placed in either end of a DCC and sperm was placed in the center, which allows sperm to
“choose” which direction it would like to travel: toward FA or FB (design by Evans et al. 2012).
These experimental units were done in some number of replicates for each cross, at least n = 2.
The second type of experiment was similar to the first in terms of crossing design, except
that a series of DCCs were designed at three distances that the sperm were required to travel: 8,
16 or 24 cm. This was done in order to determine the distance of any chemo-effect from the
experiments.
The third type of experiment was designed to determine a direct estimate of sperm travel
capacity. This was done by collecting ripe eggs and sperm, as detailed below, and depositing
eggs in one end of a “Long Chamber”, LC, and left to sit for 1 hour in order to diffuse egg
chemoattractants. The LC had a sperm entry port on the left side, and an egg reservoir on the
right side, and a series of ports from which sperm could be collected from along the length,
which totaled 34cm from sperm port to egg reservoir. Thus, sperm travel distances could be
estimated by calculating a curve from observed sperm densities along the LC.
The fourth type of experiment was designed after the results of the first experiments
showed, surprisingly, that eggs may have some capacity to sense each other and adjust their
fertilization capacity when placed in “competition” (see results and discussion), and this
phenomenon was only attributable to chemoattractant presence, as eggs were physically
separated in DCCs. In order to reduce any chamber or specific combination effects influenced by
initial experimental designs, a fourth experiment was devised. In this experiment, again a two
female and one male design was implemented as described for experiment 1. Trials were
129
conducted using a consistent number of eggs from each individual female (10k eggs), and
subsequently adding either FSW (control), or “Egg Water” (EW) isolated from the same female
or from a ‘competing’ female. After allowing 1 hour for fertilization to occur, fertilization trials
were subsampled (200 uL aliquots), and fertilization events recorded in order to estimate the
number of events per 10k eggs. Fertilization was determined by the presence of a polar body, or
by observing cell divisions.
Dichotomous Choice Chambers
Using two PVC 90
degree elbows and sections of ¾” PVC pipe, I was able to construct
Dichotomous Choice Chambers (DCCs) as described in Evans et al. 2012. These contained a
drilled central port for introduction of sperm, as well as two extraction ports directly above the
‘deposition’ chambers, in which eggs were placed. Various types of DCC chambers were
constructed. For the choice trials, chambers exactly similar to those described in Evans et al.
2012 were constructed. In order to determine distance of effect of chemoattractant and sperm
swimming distance estimations, a series of increasingly long chambers were designed. See figure
1, below. In order to ensure that DCC chambers were effective, a control was designed whereby
sperm were entered without any addition of eggs in either chamber (double blank control), or
eggs were entered only in one chamber (single blank control). Sperm were recovered evenly
above each egg chamber in the double blank control, and in greater quantities toward the
chamber with eggs in the single blank control (data not shown).
Fertilization trials and manufacturing of Egg Water
130
Fertilization trials were conducted using gravid Cgi, and strip spawning techniques common to
hatchery science (see Helm and Bourne, 2004). In short, after shucking, a razor was used to
create several lengthwise slits in the gonad, after which gametes were collected using a spray
bottle containing 0.2 micron filtered sea water (FSW). Gametes were then isolated and further
diluted in FSW for examination. Sperm was used only if motility greater than 80% was estimated
after viewing under microscope, and eggs were used if they were determined to be healthy,
numerous, and rounded. Eggs were then quantified using a Sedgewick Rafter and common
counting techniques, such that eggs could be aliquoted in consistent densities for replication
purposes. In order to produce control crosses, male and female gametes were mixed in 50 mL
conical vials such that approximately 10 sperm were actively swimming around each egg during
microscope observations. Total volume was approximately 40 mL. After 1 hour, the eggs are
washed with FSW and replaced into new FSW reservoirs, and let to sit until fertilization events
could be observed. Fertilization events were counted, along with unfertilized eggs, to give a ratio
and/or percent success for fertilization.
“Egg Water” was created using a similar method to that described in Oliver and Evans
(2014), whereby egg spawn is quantified and left to sit at a known quantity (in our case, 10k
eggs) in aerated 50 mL conical vials filled to 40 mL with FSW. After sitting for 1 hour,
necessary to allow chemoattractants to diffuse into the seawater, eggs and large particulates are
filtered from the water using a 1 micron sieve, after which this raw egg water is spun down in a
microcentrifuge to further isolate large particles; the supernatant of this process is now filtered
“Egg Water”, EW. For introducing EW experimentally, the exact same volume that it would take
to reach 10k eggs was added. For example, if FA required 200 uL in order to retrieve 10k eggs,
then 200 uL of EW would be added for all experiments requiring her EW. Thus, the EW added is
131
actually a conservative estimate of the effects of EW on fertilization trials, as we are only taking
a portion of the total chemoattractant load that would be generated by 10k eggs. EW was added
in this manner in order to reduce the total number of eggs required for any given fertilization
trial.
Statistical Analysis
R statistical packages were used using innate statistical packages (R version 3.3.2 (2016-10-31) -
- "Sincere Pumpkin Patch"). For determining significance between outcomes, we used t-tests,
ANOVA, and General Linear Model (GLM) as appropriate, and Tukey’s HSD for multiple
comparisons.
Results
DCC competition trail, experiment 1.
In the first DCC test, control fertilization rates were generated for two females, dubbed CAF6
and CAF7 (CA = Carlsbad Aquafarms; F = Female; n = individual number). Unfortunately,
CAF7 did not have successful fertilizations during control counts, despite having numerous eggs
that appeared to be of high quality. Despite this setback, we still performed a DCC ‘competition’
test in which eggs were placed from each female in either end of a 10cm DCC, and fertilization
rates were estimated after 1 hour of addition of sperm from CAM4 was added to the center port.
Again, CAF7 did not exhibit any fertilization during DCC trials, but CAF6 did exhibit a slight
non-significant increase in fertilization rate when placed in DCC ‘competition’ (control average
fertilization success rate: 13.61%, n = 2; DCC competition average fertilization success rate:
17.30%, n=2; paired two sample t-test, one-tailed p-value = 0.348). Though not significant, this
132
experiment was the first indication that some difference in fertilization success might be
occurring in “egg competition” environments. The fact that the difference in fertilization success
between controls and DCC competition trials was not significant may have been due to the fact
that CAF7 did not have viable eggs, and thus her chemoattractant profile may have been
immature or altered in some way. See figure 2.
DCC competition trial, Experiment 2.
Because experiment 1 demonstrated a potentially interesting result, and because there were no
physical interactions between the eggs in experiment 1, we repeated the experiment but only
placed “egg water” during each DCC competition replicate, instead of eggs. Similar to
experiment 1, two females were used (CAF8 and CAF9) and mated with one male (CAM5). For
this trial, we observed two phenomenon: 1) that the ‘winner’ in terms of competition was not
predicted by control trials, and that 2) each female exhibited a change in fertilization success
when in competition. Again, results between controls and competition were not statistically
significant for either female (paired two sample t-test, assuming unequal variances one-tailed p-
values = 0.103 (CAF8) and 0.136 (CAF9)). Again, we attributed non-significant results due to
low replication and high variance, but nonetheless thought our results were interesting enough to
continue to experiment 3 (below).
DCC competition trial, Experiment 3.
Given the results of the first two DCC competition trials, which were interesting but not
statistically significant, we increased the number of replicates per segment of the trial to 3, and
also implemented a distance element to the experiment to begin to understand if chemoattractant
133
had any noticeable limit, in terms of distance of effect. In this trial, a similar mating scheme was
devised between two females (CAF10 and CAF11) and one male (CAF6). In addition to a short
chamber (8 cm distance), we repeated the experimental setup in medium (16 cm) and long (24
cm) DCCs. Control trials and the three experimental sets were repeated in triplicate for this
experiment. Results indicated a similar finding to the first two experiments, namely that the
fertilization ratios during controls would not predict the better performing female during
competition trials. This relationship held across the three distances, in effect giving this result 9
replications worth of reinforcement (see figure 4). The change in fertilization success for CAF10
was highly significant when in competition versus control conditions (t-test, unequal variances
assumed, one tail, p = 0.00048), and slightly significant for CAF11(t-test, unequal variances
assumed, one tail, p = 0.0506). This result was consistent with experiments 1 and 2 in terms of
trends in relationships between the two females demonstrating a “flipping” effect when placed in
competition in terms of the better performing female.
Long Chamber (LC) Sperm Travel Experiment, experiment 4.
During experiment 3, sperm counts were also taken in order to quantify their swimming distance
in a chamber in response to a chemical queue. In order to reduce the effects of competition on
sperm travel, we devised a single welled Long Chamber (LC), which had several ports along its
34 cm axis (8, 16, 24, and 34 cm) for assessing sperm recovery along a distance gradient. This
was called experiment 4. The distance of 34 cm was chosen given curves estimated during
experiment 3, and was beyond the hypothesized maximum distance of sperm travel from the
literature of approximately 30 cm. For this experiment, sperm were placed in the entry port, and
eggs from a single female were placed in the far chamber an hour before sperm was added at a
134
volume equivalent to constitute 470k eggs. This increase in eggs was created in order to facilitate
sperm attraction. Further, we placed EW in ports at 300 uL (24cm), 200 uL (16 cm), and 100 uL
(8cm), in order to facilitate the creation of a chemoattractant gradient. Sperm was allowed to sit
in the chamber for 1 hour before recovery. This data was analyzed both separately and in
combination with data collected from experiment 3 to devise two curves in order to estimate
maximum sperm travel along a gradient (See figure 5). Solving both equations for the x intercept
(y= 0.0099, or approximately 0) demonstrates an estimated sperm travel distance of 64.71 cm
and 35.9 cm, respectively, for experiment 4 and experiments 3 and 4 combined.
Egg competition assay, experiment 5.
Given the statistically not significant – but still intriguing - results from experiments 1 and 2, and
a similar and statistically significant set of results from experiment 3, the fourth experiment was
devised to separate any effects that may have been occurring due to quantity of egg water,
quality of egg water, DCC chamber effects, and any other unaccounted-for experimental
variable. As described in the methods section, a series of 50 mL conical vials were set up such
that each would receive eggs from one female, and subsequently either FSW (control), or “Egg
Water” (EW) isolated from the same female or from a ‘competing’ female. This was done in
triplicate, for a similar breeding design of two females tested using a single male (Note a
nomenclature change, F1 = Female 1, F2 = Female 2, and the male is not named). Again, the
relationship between the winning female in “competition”, as approximated by the addition of
egg water, could not be predicted by the control conditions. In other words, a fourth series of
evidence for “flipping” had occurred, and F1 performed better than F2 in competition (condition:
F1+F2EW = 1,150 fertilization events per 10k eggs vs. condition: F2+F1EW 850 fertilization
135
events per 10k eggs) despite a dramatic and statistically different performance in control trials
(condition: F1+F1EW = 756 fertilization events per 10k eggs vs. condition: F2+F2EW = 2,608
fertilization events per 10k eggs, ). An ANOVA of the General Linear Model returned the egg
source, egg water source, and the interaction between the two each as highly significant factors
(p = 4.91e-06, p = 0.000127, and p = 0.00102, respectively). Furthermore, differences between
F1+F1FSW vs. F1+F1EW and F2+F2FSW vs. F2+F2EW were not statistically different
(Tukey’s HSD, p >> 0.05), indicating that it is the interaction of eggs and a foreign set of
chemoattractants that is mediating the change in fertilization capacity.
Discussion
General Findings
We found, in each experiment described here, that control fertilization trials would not
accurately predict the ‘winning’ female (in terms of fertilization success) when placed in a
similar environment with eggs and/or chemoattractant from another female, which we call a
proxy for competition. We dubbed this phenomenon a flipping event, and observed its
occurrence in 6 of the trials described here (total observances n = 30, 21 of which contributed to
a statistically significant result in experiments 3 and 5), and an additional 9 times in trials not
described here, in which we tested the capacity for fertilization phenomenon to be affected by
copper toxicity (data not shown). Additionally, we found that egg water quantity, and hence the
amount of chemoattractant in the water, was not the influencing variable, as females placed in
conditions with their own chemoattractant did not perform any better than when placed in
conditions with FSW alone (see experiment 5, figure 6). This resolves the variable of some form
of sperm “super activation” in response to an increased quantity of chemoattractant. These
136
findings are preliminary, and experimental 5 needs to be repeated several more times in order to
lend robustness to the observed phenomenon. However, these preliminary results lend way to the
following working two-part hypothesis: 1) eggs from female Crassostrea gigas are capable of
sensing chemoattractant from conspecific female eggs in a water column, and 2) may adjust their
capacity for fertilization accordingly. In order to support either part of this hypothesis, in
addition to more replication of our experimental units, data would need to be generated showing
some enzymatic activation and/or activity during a multi-egg ejaculate spawn event. After
preliminary data is more robustly replicated, many other interesting experiments can be thought
of, including more complex mating design, such as three or more females tested in control vs
competition experiments (which may result in some three-way competition and sensory
phenomenon), and multiple males tested on the same set of females (which may show that
females react to ratio of sperm to egg chemo-sensory queues, or total females ‘sensed’ in the
environment). These might begin to answer questions of why this type of phenomenon might be
occurring.
Population genetics and ecological implications
There are population and ecological implications for post-spawn sexual selection, which
include drivers for the evolution of gamete phenotypes and interactions (Evans and Sherman
2013). But there might be a synergy with this branch of the literature and another: in the bivalve
population genomics world, there exists a theory put forward by Dennis Hedgecock dubbed
Sweepstakes Reproduction Success (SRS), which seeks to explain the common observation of
lower-than-expected genetic diversity in large marine populations (aka a Ne/N ratio smaller than
0.01 in marine populations, compared to 0.11 on average in terrestrial species) (review:
137
Hedgecock and Pudovin, 2011; and see chapter 3 of this thesis for further references). SRS
theory states that there is “sweepstakes-like chance[s] of matching reproductive activity with
oceanographic conditions conducive to gamete maturation, fertilization, larval development,
settlement, and recruitment to the adult spawning population”. It is an easy line of reasoning to
posit that post-spawn sexual selection may be playing a role in each of these processes, as has
been described in internally fertilizing organisms (see table 1 of Firman et al. 2017). SRS
employs oceanographic conditions as the hypothesized primary driver for the juxtaposition of
chaotic genetic heterogeneity at small spatial scales in direct contrast with large scale low genetic
diversity. Levitan’s data suggests that variances in population density (and associated
synchronicity of spawning events, or lack thereof), which thereby influence reproductive
variance, may play a larger role than previously considered in explaining Hedgecock’s SRS for
marine organisms. However, there remains little evidence or research for the role and
proportionality of variable effects which drive SRS, and therefore post-spawn sexual selection
may be underappreciated in the SRS literature. In other words, the reason that the ‘sweepstakes’
may be so hard to win might not be due to chance oceanographic conditions alone, but also
significantly due to intersex (and intra-sex?, see next section) competition phenomenon unique to
extremely dense populations (oyster reefs can have millions of individuals) with dioecious
simulcasting life history.
And why not egg competition too?
Typically, studies which seek to understand effects on sexual selection by multi-ejaculate
scenarios assume that eggs themselves do not respond to multiple sperm sources (review: Wedell
2010). This assumption is understandable, given that much of the sexual selection literature
138
focuses on internally fertilizing species which tend to produce millions or billions of sperm and
relatively much fewer numbers of eggs (e.g. Bateman sexual conflict zone), and have a
reproductive tract which can facilitate adaptations for handling multiple ejaculates. The
framework put forward by Levitan (2005) expands on earlier concepts, and posits that there is a
spectrum of fertilization success and variance related to density of a given population, which is a
proxy for mating events and gamete interaction. In this theoretical framework, it is only at
intermediate population densities, where mating events are fairly common, that traditional
Bateman’s forces are applicable. Outside of this central range, where population densities are
very low, sperm limitation - due to fewer mating encounters - may drive higher variances of
female reproduction success but overall lower reproductive success. Similarly, when population
densities are very high, sexual conflict from gametic overload (e.g. polyspermy) might occur,
again creating higher female to female variance in reproductive success and an overall lower
reproductive success rate (figure 7).
So where do bivalves sit on this spectrum? In classic SRS theory, it has been assumed
that only a few individuals ever have the chance to choreograph their mating cycles sufficiently
to overcome oceanographic chaos and successfully mate, placing them effectively on the low
‘density’ spectrum of Levitan’s framework, despite extremely high actual population densities
(there can be millions of oysters in any given reef or population). However, in my opinion, an
equally plausible scenario is that bivalves fall to the right of the Bateman sexual selection
scenario and into sexual conflict, having extremely high real population densities, a tendency to
simultaneously increase fecundity-per-individual across a given population, and perhaps much
lower ratio of sperm-to-egg than traditional models consider (many millions of eggs per female
is common). For example, mussel farmers typically describe mass spawning events where their
139
entire crop will go from extremely high meat to shell ratios (a proxy for fecundity, as the
majority of meat in a bivalve food product is gamete content), to very low, in a matter of a few
days. This would place bivalves in a state of sexual conflict, to the right third of figure 7. And
importantly so for both sexes, because they are dioecious simulcasters with each sex producing
millions of gametes. This may drive the evolution of egg response to both polysperm and
polyegg conditions, which we present the first preliminary data for here in oysters. This leads to
one further generalized hypothesis: species with high population densities, at approximately
equal sex ratios, which are dioecious simulcasters, who coordinate mass spawn events, and have
high fecundity across both sexes, will experience sexual selection pressures across both gamete
types (eggs and sperm), which will thus necessitate the evolution of egg response to conspecific
female-ejaculate load.
Why would egg sensing capabilities be an advantage? Our data saw much more evidence
for high reduction in fertilization capacity of a ‘winning’ female in control conditions, as
opposed to large increases for a ‘losing’ female. Because females must produce chemoattractants
to attract sperm, we can assume that a female egg ejaculate might signal the event of increased
incoming sperm content. If an egg clutch is able to anticipate this increased sperm load, and
subsequently adjust its fertilization capacity downward, the egg may be creating a situation for
itself where only the most robust sperm are capable of fertilization. Because there is no
reproductive tract to filter out ‘underperforming’ sperm in a marine dioecious simulcaster,
perhaps this is the mechanism that females employ to ensure quality gamete matching. My
hypothesis echoes a recent review of post-mating sexual conflict, which points out the fact that
there has been historically much less attention given to post-mating (or post-spawn) adaptations
in female gametes (Firman 2018), and that females do in fact have “greater control over
140
fertilization than has previously been appreciated”. One example from this review includes a
study which demonstrated that mice eggs which evolved with polygamous sperm (i.e. the
females and males had many mating opportunities) were more resistant to fertilization than eggs
which evolved from monogamous lineages (i.e. females and males had only one partner; Figure
3, 2017 review, original data from Firman et al. 2015)). Importantly, this review includes both
traditional terrestrial vertebrate and invertebrate models as well as data concerning marine
organisms, including sea urchins, externally fertilizing fishes, and other bivalves.
(2017) also provides possible genes of interest that might inform future experimental
work in this system, and supports my conclusion that it is possible, considering life history and
population conditions experienced by many bivalves, that dioecious simulcasting may give rise
to selection pressures that produce adaptations in eggs which allow them to sense their ‘social’
situation and hence to manage fertilization capacity. The data I present here is the first to suggest
that there may be a genotype or egg-type dependent response within a given egg ejaculate clutch,
as opposed to an evolved and set ‘egg type’ (e.g. monogamous vs polygamous) that is inherited.
The final piece of evidence I will present which supports my various hypothesis comes
from the original literature which informed my DCC chamber experiments in the first place. In
both Evans, et al. (2014) and Oliver and Evans (2016), they present figures which show the
change in fertilization rates between two females (FA and FB) between controls and choice
experiments. Thus, their figures have a system of four quadrants, and they describe a system in
which sperm will travel more frequently toward (and have more fertilization success with, and
offspring survival with,) eggs as predicted by their controls. However, I would argue that this is a
trend supported by the way in which they present their data, and in fact they show that in
approximately 30% (4/13 events; Evans et al. (2012) figure 3B), 43% (6 of 14; Oliver and Evans
141
(2014) figure 1A), 35% (5 of 14; Oliver and Evans (2014) figure 1B) of their egg choice
experiments, they observed the same “flip events” as I present in my data, whereby the control
trials do not accurately predict performance in experimental trials. In my manipulation of their
figures (figure 8, all colorized elements are my own additions), I have divided the quadrats and
identified situations where the winning female during choice trials, in terms of egg choice by
sperm and fertilization, was not predicted by the control trials (green boxes). Perhaps a larger
assay of spawn choice and/or competition experiments, followed by genotyping and molecular
assays, could elucidate the prevalence of ‘flipping’ phenotypes among a subsampled population.
Anecdotal Hatchery Observations
One last line of reasoning that supports my hypothesis of the occurrence of some form post-
spawn sexual selection occurring in bivalves comes from the anecdotal experiences I’ve had in
the hatchery. Many lab-spawned oyster families, without oceanographic variables to worry
about, perform very poorly at larval stages, including fertilization, without much explanation as
to why. This anecdotal observation also seems to be associated with specific pairings, because
some females will work well with some males and not others, and vice versa. Further, spawn
success seems to be significantly correlated with the number of spawners in a group. Are they
(either the adult organisms or their spawned gametes) communicating something to each other?
Is ‘choice’ involved for sperm, eggs, or both? One elegant description of sexual selection, from
Jennion’s and Kokos 2010 book ‘Sexual Selection’, is as follows: ‘sexual selection favors traits
that improve the likelihood of fertilization given limited access to opposite sex gametes due to
competition with members of the same sex’. Using this description of sexual selection (which
explicitly omits sexual direction, male or female, of competition), and combining the literature,
142
my results, and anecdotal experiences above, it is easy to posit that bivalve eggs may have
increased competition with each other in broadcast spawning events than other systems which
down require dioecious simulcasting, which necessitates the evolution of a within-clutch egg
response mechanism.
143
Figures
Figure 1: Dichotomous Choice Chambers (DCCs) and Long Chamber (LC). The photo
shows three of the DCCs devised for the experiments described in this chapter, and the Long
Chamber at bottom. Each has a sperm entry port (middle, DCCs; right side, LC), and an egg
port above the reservoir at the ends of the chambers. The LC has additional sperm recovery
ports along the length of the apparatus.
144
Figure 2: DCC Competition Assay gives preliminary result. Using a 10 cm DCC, eggs
from CAF6 were placed in competition with eggs from CAF7. As CAF7 did not have any
successful fertilization events, only data from CAF6 is shown. This was the first instance of
a change in the capacity of fertilization when placed in competition, though the change was
not significant (t-test, p-value = 0.348). Error bars represent standard deviation.
0.1
0.12
0.14
0.16
0.18
0.2
0.22
0.24
No Competition (Control) With Competition (Trial)
Percent Fertilization Estimate
Figure 2, DCC Competition Assay, Exp. 1
145
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
CAF9 Control CAF8 Control CAF9 CompetitionCAF8 Competition
Fertilization Rate
Assay type
Figure 3, DCC Egg Water Test, Exp. 2
Figure 3: DCC Egg Water trial demonstrates a possible second experimental shift if
fertilization capacity. Using 10 cm DCCs, two females (CAF8 and CAF9) were placed in
a proxy for competition using eggs or egg water (EW). The ‘control’ factor, named on the
x-axis, is a stand-alone fertilization assay, and the ‘competition’ factor represents eggs
placed in a DCC with competing EW. Again, we see a non-significant shift in direction of
fertilization for both females, and a ‘flipping event’, whereby the control condition does not
predict the winning female in terms of fertilization rate in competition (t-test, p-values =
0.103 (CAF8 control vs. competition); 0.136 (CAF9 control vs. competition). Error bars
represent standard deviation.
146
y = 1.361e-0.137x
R² = 0.43738
0
1
2
3
4
5
6
7
0 8 16 24 32
Percent Recovered Sperm
Distance in cm
Figure 5B, All Sperm Recovery Data,
Experiments 3 & 4
y = 32.183e
-0.125x
R² = 0.67685
0
10
20
30
40
50
60
70
0 8 16 24 32
Sperm Recovered in 100K
Distance in cm
Figure 5A, Sperm Recovery in LC, Exp.4
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
CAF11
Control
CAF10
Control
CAF11
Short
CAF10
Short
CAF11
Medium
CAF10
Medium
CAF11
Long
CAF10
Long
Figure 4, DCC Choice assays, Exp. 3
All A:B
comparisons
statistically
significant
(P<0.05)
A
B
A
B
Figure 4: increasing distance in DCC Egg Water trials demonstrate significant
‘flipping event’. Similar to figure 3, after a separate control trial, two females were placed
in competition with an increasing length of DCC, from 8 cm (short), 16 cm (medium), and
24 cm (long). This is the first statistically significant result showing that a ‘flipping event’
occurs, where the control fertilization percent does not predict the winning female in a
competition trial (CAF10 control vs. short, p = 0.00048; CAF11 control vs. short, p =
0.0506), and this result is maintained at short, medium, and long distances. Each trial was
ran in triplicate, error bars represent standard deviation.
Figure 5: Sperm travel capacity estimates in Long Chambers. 5A) Sperm recovery was
tested at various lengths using a Long Chamber. Figure shows number of sperm recovered
at various ports, in 100k increments. Each distance n = 4 replicates. Maximum sperm travel
distance, as estimated when the y-value approaches zero (y = 0.0099), was 64.71 cm. 5B)
data combined from experiment 3 (see figure 4) and experiment 4 (figure 5A), to add more
data to distance recoveries. Note that experiment 3 did not have a 32 cm port, and this may
contribute to the difference in model prediction when both datasets are combined (y =
0.0099, x= 35.9 cm). Sperm maximum travel observations were less than observed for
Suquet et al 2010, which estimated a maximum distance of 1M.
147
756 725
1,150
2,683
2,608
850
0
500
1000
1500
2000
2500
3000
3500
F1+FSW F1+F1EW F1+F2EW F2+FSW F2+F2EW F2+F1EW
Fertlization Events per 10k Eggs
Female + Condition
Figure 6, Egg Competition Trial
A
A
A
B
B
A
All A:B comparisons are significant
(Tukey's HSD, P<0.001).
No A:A or B:B comparisons are significant.
Figure 6: Egg competition trial, DCC effects removed. Experiment 5. In order to test
explicitly the variable of egg water (EW) on fertilization capacity, we redesigned the
experimental unit such that DCCs were not used (see methods). The result from one
experimental iteration is shown above, where Female 1 (F1) is placed in a control condition
(FSW added), chemoattractant load control (F1EW added), or competition experimental
condition (F2EW added). The same experiment, in the opposite direction, was performed for
Female 2 (F2). This is the second statistically significant ‘flipping event’ described, and the
first which controls for the variable of chemoattractant load. Tukey’s test of the General
Linear Model show that addition of FSW and individual-identical EW were not statistically
different for either female, but the addition of a foreign female’s EW did cause a statistically
significant deviation from control conditions. Error bars represent standard deviation,
replicates n = 3 per trial.
148
Figure 7. From Levitan, 2005: Conceptual illustration of the distribution of male and
female mating success. Three zones with different selective pressures are noted: sperm
limitation, ‘‘Bateman,’’ and sexual conflict. The variances are only unequal in the
Bateman zone. At the extremes, the variances are similar for males and females but the
selective pressures may be quite different. At low densities both males and females may
be under selection for increased mating success, whereas at high densities, males may be
selected for increased mating success while females try to resist matings.
149
*
*
Q1 Q2
Q3
Q4
4/13 = 30% experienced ‘flip’
Figure 8: Flipping Event evidence from the literature (original figures from Evans et al.
2012). All colorized work and quadrant identification are my own additions to the original
figures. Datapoints (black dots) for both x and y values are differences between FemaleA (FA)
and FemaleB (FB). For example, in figure 8A, the bottom left most blue square showed that
fertilization was greater for FB than for FA in controls (the x-value is negative), and that
fertilization was similarly better for FB than for FA during choice experiments (the y-value is
negative). The pink lines indicate the quadrants of the x and y values in terms of sign, positive
or negative. Values in Q2 or Q4 have matching x and y signs (positive or negative), whereas
values in Q1 or Q3 have mismatched signs, which are further highlighted by green boxes.
These green boxes indicate an instance where fertilization value signs are incongruous, and
indicate a situation where the control experiment did not predict the winner, indicating a
‘flipping event’. The red boxes indicate a scenario where more sperm migrated toward FA,
and FA also enjoyed an increase (delta 0.15) in fertilization rate. The asterisk on the red boxes
indicate a situation where the relationship is very close to becoming ‘green’. The blue boxes
are similar to the red boxes: both sperm collections and fertilization percentages were better in
FB than in FA. The blue dotted line is an estimate of the slope of the linear equation presented
by the data.
150
REFERENCES
Introduction References
CDFA (California Department of Food and Agriculture) 2018. “California Agricultural Statistics
Review 2017-2018.” 1–105.
FAO. 2018. “The State of World Fisheries and Aquaculture 2018 - Meeting the sustainable
development goals”. Rome.
FAO. 2016. “The State of World Fisheries and Aquaculture 2016. Contributing to food security
and nutrition for all”. Rome. 200 pp.
Gaitán-Espitia, Juan Diego, Julian F. Quintero-Galvis, Andres Mesas, and Guillermo D’Elía.
2016. “Mitogenomics of Southern Hemisphere Blue Mussels (Bivalvia: Pteriomorphia):
Insights into the Evolutionary Characteristics of the Mytilus Edulis Complex.” Scientific
Reports 6(May).
Gosling, E., 2004. “Bivalve Molluscs: Biology, Ecology, and Culture”. Wiley and Sons
Publishing.
Helm, Michael M. and Neil Bourne. 2004. Hatchery Culture of Bivalves, A Practical Manual.
Jaenike, John. 2007. “Comment on ‘Impacts of Biodiversity Loss on Ocean Ecosystem
Services.’” Science 316(5829):1285 LP – 1285.
Jørgensen, C. Barker. 1996. “Bivalve Filter Feeding Revisited.” Marine Ecology Progress Series
142(1–3):287–302.
National Agricultural Statistics Service (NASS). 2014. “2012 Census of Agriculture - Census of
Aquaculture 2013.” United States Department of Agriculture 3(Special Studies, Part 2):1–
16.
National Marine Fisheries Service (2018) Fisheries of the United States, 2017. U.S. Department
of Commerce, NOAA Current Fishery Statistics No. 2017
NOAA Fisheries. 2017. Imports and Exports of Fishery Products Annual Summary, 2017.
NOAA. 2017. Fisheries of the United States, 2016.
NOAA. 2018. “What is the EEZ?” National Ocean Service.
https://oceanservice.noaa.gov/facts/eez.html
Rogers, Paul. 2018. Data from US Drought Monitor. “Here’s how much recent rains have washed
away California’s drought”. The Mercury News.
151
https://www.mercurynews.com/2019/03/07/heres-how-much-recent-rains-have-washed-
away-californias-drought/
Ross, Wilbur, RDML Timothy Gallaudet, Usn Ret, and Chris Oliver. 2018. “Fisheries of the
United States 2017 Report- NOAA National Marine Fisheries Service.” (2017).
Ruggles, Ron. 2018. “Seaweed Enjoys Rising Tide of Popularity.” Nation’s Restaurant News.
https://www.nrn.com/whats-hot/seaweed-enjoys-rising-tide-popularity.
Sturm, C. F., Pearce, T.A., and Valdes, A. 2006. The Mollusks: A Guide to Their Study,
Collection, and Preservation. Universal Publishers.
The Economist (Online). 2016. “Drops in the ocean: Franceas marine territories: Daily chart.”
The Economist Newspaper NA, Inc. https://www.economist.com/graphic-
detail/2016/01/13/drops-in-the-ocean-frances-marine-territories
Worm, Boris, Edward B. Barbier, Nicola Beaumont, J. Emmett Duffy, Carl Folke, Benjamin S.
Halpern, Jeremy B. C. Jackson, Heike K. Lotze, Fiorenza Micheli, Stephen R. Palumbi,
Enric Sala, Kimberley A. Selkoe, John J. Stachowicz, and Reg Watson. 2006. “Impacts of
Biodiversity Loss on Ocean Ecosystem Services.” Science 314(5800):787 LP – 790.
Chapter 1 References: Metallotoxicity Behavioral Response in Bivalves. In order of
citation.
Rivera-Duarte I, Rosen G, Lapota D, Chadwick DB, Kear-Padilla L, Zirino A. Copper toxicity to
larval stages of three marine invertebrates and copper complexation capacity in San Diego
Bay, California. Environ Sci Technol. 2005; 1542-1546.
Lewis C, Ellis RP, Vernon E, Elliot K, Newbatt S, Wilson RW. Ocean acidification increases
copper toxicity differentially in two key marine invertebrates with distinct acid-base
responses. Nat Publ Gr. 2016; 6: 21554.
Okazaki RK. Copper Toxicity in the Pacific Oyster Crassostrea gigas. Bull. of Environ Contam,
Toxicol. 1976; 658-664.
Eriksen RS, Mackey DJ, Van Dam R, Nowak B. Copper speciation and toxicity in Macquarie
Harbour, Tasmania: An investigation using a copper ion selective electrode. Mar Chem.
2001; 99-113.
Brand LE, Sunda WG, Guillard RRL. Reduction of marine phytoplankton reproduction rates by
copper and cadmium. Mar Biol Ecol. 1986;96:225–50.
Chadwick DB, Zirino A, Rivera-Duarte I, Katz CN, Blake AC. Modeling the mass balance and
fate of copper in San Diego Bay. Limnol. Oceanogr., 49(2), 2004, 355–366.
152
Rainbow PS. Biomonitoring of heavy metal availability in the marine environment. Mar Pollut
Bull. 1995; 183-192.
DelValls TÁ, Forja JM, Gonza È Lez-Mazo E, Go È Mez-Parra A, Blasco J. Determining
contamination sources in marine sediments using multivariate analysis. Trends in Anal.
Chem. 1998; 181-192.
Stuart RK, Brahamsha B, Busby K, Palenik B. Genomic island genes in a coastal marine
Synechococcus strain confer enhanced tolerance to copper and oxidative stress. ISME J.
2013; 1139–1149.
Chen CL, Maki JS, Rittschof D, Teo SLM. Early marine bacterial biofilm on a copper-based
antifouling paint. Int Biodeterior Biodegrad. 2013; 71-76.
Foekema EM, Kaag NHBM, Kramer KJM, Long K. Mesocosm validation of the marine No
Effect Concentration of dissolved copper derived from a species sensitivity distribution. Sci
Total Environ. 2015; 173-182.
Pawiro S. Safe Management of Shellfish and Harvest Waters, Ch 2; Bivalves: Global production
and trade trends. 2010 World Health Organization (WHO).
FAO, 2014. The State of World Fisheries and Aquaculture 2014. Rome. 223 pp.
Vázquez-Boucard C, Anguiano-Vega G, Mercier L, Rojas del Castillo E. Pesticide Residues,
Heavy Metals, and DNA Damage in Sentinel Oysters Crassostrea gigas From Sinaloa and
Sonora, Mexico. J Toxicol and Environ Health, Pt A. 2014; 169-176.
Mai H, Cachot J, Brune J, Geffard O, Belles A, Budzinski H, et al. Embryotoxic and genotoxic
effects of heavy metals and pesticides on early life stages of Pacific oyster (Crassostrea
gigas). Mar Pollut Bull. 2012; 2663–2670.
McPherson CA, Chapman PM. Copper effects on potential sediment test organisms: The
importance of appropriate sensitivity. Marine Pollution Bulletin. 2000; 656-665.
Ivanina A V., Hawkins C, Sokolova IM. Immunomodulation by the interactive effects of
cadmium and hypercapnia in marine bivalves Crassostrea virginica and Mercenaria
mercenaria. Fish Shellfish Immunol. 2014; 54-65.
Allam B, Raftos D. Immune responses to infectious diseases in bivalves. J Invert Pathol. 2015;
121-136.
Metzger DC, Pratt P, Roberts SB. Characterizing the effects of heavy metal and vibrio exposure
on hsp70 expression in Crassostreas gigas gill tissue. J Shell Res. 2012; 627-630.
Xu M, Bijoux H, Gonzalez P, Mounicou S. Investigating the response of cuproproteins from
oysters (Crassostrea gigas) after waterborne copper exposure by metallomic and proteomic
approaches. Metallomics. 2014; 338–346.
Viarengo A, Pertica M, Mancinelli G, Burlando B, Canesi L, Orunesu M. In vivo effects of
copper on the calcium homeostasis mechanisms of mussel gill cell plasma membranes.
Comp Biochem Physiol - C Pharmacol Toxicol Endocrinol. 1996; 421-425.
153
Scott DM, Major CW. The Effect of Copper (II) on Survival, Respiration, and Heart Rate in the
Common Blue mussel, Mytilus edulis. Biol Bull. 1972; 679–88.
Grace AL, Gainey LF. The Effects of Copper on the Heart Rate and Filtration Rate of Mytilus
edulis. Mar Poll Bull. 1987; 87–91.
Gainey LF, Kenyon JR. The effects of reserpine on copper induced cardiac inhibition in Mytilus
edulis. Comp Biochem Physiol Part C, Comp. 1990; 177-179.
Curtis TM, Williamson R, Depledge MH. The initial mode of action of copper on the cardiac
physiology of the blue mussel, Mytilus edulis. Aquat Toxicol. 2001; 29-38.
Rosen G, Rivera-Duarte I, Kear-Padilla L, Chadwick B. Effects of copper on marine invertebrate
larvae in surface water from San Diego Bay, CA. SPAWAR Systems Center San Diego.
Presentation Accessed 2016.
Knezovich JP, Harrison FL, Tucker JS. The Influence of Organic Chelators on the Toxicity of
Copper to Embryos of the Pacific Oyster, Crassostrea gigas. Arch Environm Contam,
Toxicol. 1981; 241–249.
Coglianese M, Martin MP. Individual and ineractive effects of envrionmental stress on the
embryonic development of the Pacific oyster, Crassostrea gigas. I. The toxicity of copper
and silver. Mar Environ Res. 1981; 13–27.
Martin M, Osborn KE, Billig P, Glickstein N. Toxicities of ten metals to Crassostrea gigas and
Mytilus edulis embryos and Cancer magister larvae. Mar Pollut Bull. 1981; 305-308.
Brooks SJ, Bolam T, Tolhurst L, Bassett J, Roche J La, Waldock M, et al. Effects of dissolved
organic carbon on the toxicity of copper to the developing embryos of the Pacific oyster
(Crassostrea gigas). Environ Toxicol Chem. 2007; 1756–1763.
United States Environmental Protection Agency. 2003 Draft update of ambient water quality
criteria for copper. EPA 822-R-03-026. November 2003.
Prael A, Cragg S, Henderson SM. Behavioral responses of veliger larvae of Crassostrea gigas to
leachate from wood treated with copper-chrome-arsenic (CCA): a potential bioassay of
sublethal environmental effects of contaminants. Journal of Shellfish Research. 2001; 20:
267-273.
Raby D, Lagadeuc Y, Dodson JJ, Mingelbier M. Relationship between feeding and vertical
distribution of bivalve larvae in stratified and mixed waters. Mar Ecol Prog Ser. 1994; 275-
284.
Applebaum SL, Pan TCF, Hedgecock D, Manahan DT. Separating the nature and nurture of the
allocation of energy in response to global change. Integr and Compar Biol. 2014; 54(2);
284-295.
R Core Team (2013). R: A language and environment for statistical computing. R Foundation for
Statistical Computing, Vienna, Austria. URL http://www.R-project.org/.
154
Cóte IM, Jelnikar E. Predator-induced clumping behaviour in mussels (Mytilus edulis Linnaeus).
J Exp Mar Bio Ecol. 1999; 201-211.
Griffiths CL, Richardson CA. Chemically induced predator avoidance behaviour in the
burrowing bivalve Macoma balthica. J Exp Mar Bio Ecol. 2006; 91-98.
Gosling, E. Bivalve molluscs: biology, ecology and culture. John Wiley & Sons, 2008.
Fuchs HL, Gerbi GP, Hunter EJ, Christman AJ, Javier Diez F. Hydrodynamic sensing and
behavior by oyster larvae in turbulence and waves. J Exp Biol. 2015; 1419-1432.
Chapter 2: Microbiome Investigations in the Pacific Oyster
Arora, M., Anil, A., Delany, J., Rajarajan, N., Emami, K., & Mesbahi, E. (2012). Carbohydrate-
degrading bacteria closely associated with Tetraselmis indica: influence on algal growth.
Aquatic Biology, 15(1), 61-71.
Bäckhed, F., Roswall, J., Peng, Y., Feng, Q., Jia, H., Kovatcheva-Datchary, P.,et al. (2015).
Dynamics and Stabilization of the Human Gut Microbiome during the First Year of Life.
Cell Host Microbe, 17(5), pp.690-703.
Bahrndorff, S., Alemu, T., Alemneh, T., & Lund Nielsen, J. (2016). The Microbiome of
Animals: Implications for Conservation Biology. Int J Genomics, 2016, 5304028.
http://doi.org/10.1155/2016/5304028
Belkaid, Y., & Hand, T. (2014). Role of the Microbiota in Immunity and inflammation. Cell,
157(1), 121–141. http://doi.org/10.1016/j.cell.2014.03.011
Chauhan, P. and Saxena, A. (2016). Bacterial carrageenases: an overview of production and
biotechnological applications. 3 Biotech, 6(2).
Cruz-López, R. and Maske, H. (2016). The Vitamin B1 and B12 Required by the Marine
Dinoflagellate Lingulodinium polyedrum Can be Provided by its Associated Bacterial
Community in Culture. Front Microbiol, 7.
Dégremont, L., Garcia, C. and Allen, S. (2015). Genetic improvement for disease resistance in
oysters: A review. J Invertebr Pathol, 131, pp.226-241.
Douillet, P. and Langdon, C. (1994). Use of a probiotic for the culture of larvae of the Pacific
oyster (Crassostrea gigas Thunberg). Aquaculture, 119(1), pp.25-40.
FAO (1975). Hatchery manual for the Pacific oyster. Rome: Food and Agricultural Organization
of The United Nations.
FAO (2005). Crassostrea gigas. Cultured Aquatic Species Information Programme. Rome: FAO
Fisheries and Aquaculture Department.
155
FAO. 2016. The State of World Fisheries and Aquaculture 2016. Contributing to food security
and nutrition for all. Rome. 200 pp.
Fernández, N. Trabal, Mazón-Suástegui, J., Vázquez-Juárez, R., Ascencio-Valle, F. and Romero,
J. (2014). Changes in the composition and diversity of the bacterial microbiota associated
with oysters (Crassostrea corteziensis,Crassostrea gigas and Crassostrea sikamea) during
commercial production. FEMS Microbiol Ecol, 88(1), pp.69-83.
García-Orenes, F., Morugán-Coronado, A., Zornoza, R. and Scow, K. (2013). Changes in Soil
Microbial Community Structure Influenced by Agricultural Management Practices in a
Mediterranean Agro-Ecosystem. PLoS ONE, 8(11), p.e80522.
Goldberg, S. J., Nelson, C. E., Viviani, D. A., Shulse, C. N., & Church, M. J. (2017, June 15).
Cascading influence of inorganic nitrogen sources on DOM production, composition,
lability and microbial community structure in the open ocean. Environ Microbiol.
Givens, C., Bowers, J., DePaola, A., Hollibaugh, J. and Jones, J. (2014). Occurrence and
distribution of Vibrio vulnificus and Vibrio parahaemolyticus- potential roles for fish,
oyster, sediment and water. Lett Appl Microbiol, 58(6), pp.503-510.
Guo, X., Li, Q., Wang, Q. and Kong, L. (2011). Genetic Mapping and QTL Analysis of Growth-
Related Traits in the Pacific Oyster. J Mar Biotechnol, 14(2), pp.218-226.
Hamady, M., Lozupone, C., and Knight, R. (2010) Fast UniFrac: facilitating high-throughput
phylogenetic analyses of microbial communities including analysis of pyrosequencing and
PhyloChip data. ISME J 4: 17–27.
Harrold, Z., Hausrath, E., Garcia, A., Murray, A., Tschauner, O., Raymond, J. and Huang, S.
(2018). Bioavailability of mineral-bound iron to a snow algae-bacteria co-culture and
implications for albedo-altering snow algae blooms. Appl Environ Microbiol,
pp.AEM.02322-17.
Hayat, R., Ali, S., Amara, U., Khalid, R. and Ahmed, I. (2010). Soil beneficial bacteria and their
role in plant growth promotion: a review. Ann Microbiol, 60(4), pp.579-598.
Hedgecock, D. and Pudovkin, A. (2011). Sweepstakes Reproductive Success in Highly Fecund
Marine Fish and Shellfish: A Review and Commentary. Bulletin of Marine Science, 87(4),
pp.971-1002.
Hedgecock, D. and Sly, F. (1990). Genetic drift and effective population sizes of hatchery-
propagated stocks of the Pacific oyster, Crassostrea gigas. Aquaculture, 88(1), pp.21-38.
Hernández-Zárate G, Olmos-Soto J. Identification of bacterial diversity in the oyster Crassostrea
gigas by fluorescent in situ hybridization and polymerase chain reaction. J Appl Microbiol.
2006;100:664-672.
156
Hurley, W., Wolterstorff, C., MacDonald, R., & Schultz, D. (2014). Paralytic Shellfish
Poisoning: A Case Series. West J Emerg Med, 15(4), 378–381.
http://doi.org/10.5811/westjem.2014.4.16279
Huse, S., Welch, D., Morrison, H., & Sogin, M. (2010). Ironing out the wrinkles in the rare
biosphere through improved OTU clustering. Environ Microbiol(12), 1889-1898.
Kazamia, E., Czesnick, H., Nguyen, T., Croft, M., Sherwood, E., Sasso, S.,et al. (2012).
Mutualistic interactions between vitamin B12-dependent algae and heterotrophic bacteria
exhibit regulation. Environ Microbiol, 14(6), pp.1466-1476.
Kim, S. K. (2017). Marine omics: Principles and applications. Boca Raton, FL, USA: CRC
Press, Taylor & Francis Group.
Kimura, K., & Tomaru, Y. (2014). Coculture with marine bacteria confers resistance to complete
viral lysis of diatom cultures. Aquat Microb Ecol, 73(1), 69-80.
King, G., Judd, C., Kuske, C. and Smith, C. (2012). Analysis of Stomach and Gut Microbiomes
of the Eastern Oyster (Crassostrea virginica) from Coastal Louisiana, USA. PLoS ONE,
7(12), p.e51475.
Kozich, J., Westcott, S., Baxter, N., Highlander, S., & Schloss, P. (2013). Development of a
dual-index sequencing strategy and curation pipeline for analyzing amplicon sequence data
on the MiSeq Illumina sequencing platform. J Appl Environ Microbiol, 79(17), 5112-5120.
Lee, R. (2008). Bivalve depuration: fundamental and practical aspects. Rome: Food and
Agriculture Organization of the United Nations FAO, Communication Division.
Lokmer, A., Kuenzel, S., Baines, J. and Wegner, K. (2016). The role of tissue-specific
microbiota in initial establishment success of Pacific oysters. Environ Microbiol, 18(3),
pp.970-987.
Mackie, R. I. Mutualistic fermentative digestion in the gastrointestinal tract: diversity and
evolution. Integr Comp Biol 42, 319–326 (2002).
Mao, S., Zhang, M., Liu, J. and Zhu, W. (2015). Characterizing the bacterial microbiota across
the gastrointestinal tracts of dairy cattle: membership and potential function. Scientific
Reports, 5(1).
Martínez Cruz, P., Ibáñez, A. L., Monroy Hermosillo, O. A., & Ramírez Saad, H. C. (2012). Use
of Probiotics in Aquaculture. ISRN Microbiol, 2012, 916845.
http://doi.org/10.5402/2012/916845
McMurdie, P.J., and Holmes, S. (2013) phyloseq: an R package for reproducible interactive
analysis and graphics of microbiome census data. PLoS ONE 8: e61217
157
McPartlin, D. A., Lochhead, M. J., Connell, L. B., Doucette, G. J., & O’Kennedy, R. J. (2016).
Use of biosensors for the detection of marine toxins. Essays Biochem, 60(1), 49–58.
http://doi.org/10.1042/EBC20150006
Myer, P. R., Freetly, H. C., Wells, J. E., Smith, T. P. L., & Kuehn, L. A. (2017). Analysis of the
gut bacterial communities in beef cattle and their association with feed intake, growth, and
efficiency 1,2,3. J Anim Sci, 95(7), 3215-3224.
Nakamura, S., Kim, Y. H., Takashima, K., Kimura, A., Nagai, K., Ichijo, T., & Sato, S. (2017).
Composition of the microbiota in forestomach fluids and feces of Japanese black calves
with white scours 1. J Anim Sci, 95(9), 3949-3960.
Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O’Hara, R.B., et al. (2013)
vegan: community ecology package. URL http://CRAN.R-project.org/package=vegan
Parada, A., Needham, D., & Fuhrman, J. A. (2015). Every base matters: assessing small subunit
rRNA primers for marine microbiomes with mock communities, time series and global
field samples. Environ Microbiol, 18(5), 1403-1414.
Paralytic Shellfish Poisoning Fact Sheet. (2002). Alaska Division of Public Health. Anchorage,
AK: Section of Epidemiology.
Provost, K., Dancho, B., Ozbay, G., Anderson, R., Richards, G. and Kingsley, D. (2011).
Hemocytes Are Sites of Enteric Virus Persistence within Oysters. Appl Environ Microbiol,
77(23), pp.8360-8369.
Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P., Glöckner, F. (2013). The
SILVA ribosomal RNA gene database project: improved data processing and web-based
tools. Nucleic Acids Res, 41(DI), D590-D596.
Sommer, F. & Backhed, F. The gut microbiota–masters of host development and physiology. Nat
Rev Microbiol 11, 227–238 (2013).
Tan, L., Chan, K., Lee, L. and Goh, B. (2016). Streptomyces Bacteria as Potential Probiotics in
Aquaculture. Front Microbiol, 7.
U.S. Food and Drug Administration (2003). Carlos’ Tragic and Mysterious Illness How Carlos
almost died by eating contaminated raw oysters. College Park, MD: U.S. Food and Drug
Administration.
Walters, W., Hyde, E. R., Berg-Lyons, D., Ackermann, G., Humphrey, G., Parada, A., Knight,
R. (2015, December). Improved Bacterial 16S rRNA Gene (V4 and V4-5) and Fungal
Internal Transcribe Spacer Marker Gene Primers for Microbial Community Surveys.
mSystems, 1(1).
158
Wang D, Zhang Q, Cui Y, Shi X. Seasonal dynamics and diversity of bacteria in retail oyster
tissues. Int J Food Microbiol. 2014;173:14-20.
Wen, L., Ley, R., Volchkov, P., Stranges, P., Avanesyan, L., Stonebraker, A., et al. (2008).
Innate immunity and intestinal microbiota in the development of Type 1 diabetes. Nature,
455(7216), pp.1109-1113.
Wegner, K., Volkenborn, N., Peter, H. and Eiler, A. (2013). Disturbance induced decoupling
between host genetics and composition of the associated microbiome. BMC Microbiol,
13(1), p.252.
Li C Xia, Joshua A Steele, Jacob A Cram, Zoe G Cardon, Sheri L Simmons, Joseph J Vallino, et
al. Extended local similarity analysis (eLSA) of microbial community and other time series
data with replicates. BMC Syst Biol. 2011, 5(Suppl 2):S15
http://www.biomedcentral.com/1752-0509/5/S2/S15/
Yilmaz, P., Parfey, L. W., Yarza, P., Gerken, J., Pruesse, E., Quast, C., Glöckner, F. O. (2014,
January). The SILVA and "All-species Living Tree Project (LTP)" taxonomic frameworks.
Nucleic Acids Res, 42(DI), D643-D648.
Zhang, G., Fang, X., Guo, X., Li, L., Luo, R., Xu, F., Wang, J. (2012). The oyster genome
reveals stress adaptation and complexity of shell formation. Nature, 490(7418), 49-54.
Chapter 3 References: Mutation Rate Estimates in the Pacific Oyster
Acharya, Samir, Patricia L. Foster, Peter Brooks, and Richard Fishel. 2003. “The Coordinated
Functions of the E. Coli MutS and MutL Proteins in Mismatch Repair.” Molecular Cell
12:233–46.
Andre, Jean Baptiste and Bernard Godelle. 2006. “The Evolution of Mutation Rate in Finite
Asexual Populations.” Genetics 172(1):611–26.
Arias, A., R. Freire, P. Boudry, S. Heurtebise, J. Méndez, and A. Insua. 2009. “Single Nucleotide
Polymorphism for Population Studies in the Scallops Aequipecten Opercularis and
Mimachlamys Varia.” Conservation Genetics 10(5):1491–95.
Berger, David, Josefine Stångberg, Karl Grieshop, Ivain Martinossi-Allibert, and Göran
Arnqvist. 2017. “Temperature Effects on Life-History Trade-Offs, Germline Maintenance
and Mutation Rate under Simulated Climate Warming.” Proceedings of the Royal Society
B: Biological Sciences 284(1866).
Beyer, Jonny, Norman W. Green, Steven Brooks, Ian J. Allan, Anders Ruus, Tânia Gomes, Inger
Lise N. Bråte, and Merete Schøyen. 2017. “Blue Mussels (Mytilus Edulis Spp.) as Sentinel
Organisms in Coastal Pollution Monitoring: A Review.” Marine Environmental Research
130:338–65.
159
Burford, M. O., J. Scarpa, B. J. Cook, and M. P. Hare. 2014. “Local Adaptation of a Marine
Invertebrate with a High Dispersal Potential: Evidence from a Reciprocal Transplant
Experiment of the Eastern Oyster Crassostrea Virginica.” Marine Ecology Progress Series
505:161–75.
Canesi, Laura and Carla Pruzzo. 2016. “Specificity of Innate Immunity in Bivalves: A Lesson
From Bacteria”. Elsevier Inc.
Church, David N., Sarah E. W. Briggs, Claire Palles, Enric Domingo, Stephen J. Kearsey,
Jonathon M. Grimes, Maggie Gorman, Lynn Martin, Kimberley M. Howarth, Shirley V.
Hodgson, Kulvinder Kaur, Jenny Taylor, and Ian P. M. Tomlinson. 2013. “DNA
Polymerase 1 and d Exonuclease Domain Mutations in Endometrial Cancer.” Human
Molecular Genetics 22(14):2820–28.
David, P., M. Perdieu, A. Pernot, and P. Jarne. 1997. “Fine-Grained Spatial and Temporal
Population Genetic Structure in the Marine Bivalve Spisula Ovalis.” Evolution 51(4):1318–
22.
DePristo, M., Banks, E., Poplin, R. et al. 2011. “A framework for variation discovery and
genotyping using next-generation DNA sequencing data”. Nat Genet 43, 491–498.
Earl, D. A. 2012. “STRUCTURE HARVESTER: a website and program for visualizing
STRUCTURE output and implementing the Evanno method”. Conserv. Genet Resour. 4,
359–361.
Edmands, S., P. E. Moberg, and R. S. Burton. 1996. “Allozyme and Mitochondrial DNA
Evidence of Population Subdivision in the Purple Sea Urchin Strongylocentrotus
purpuratus.” Marine Biology (126):443–50.
Evanno, G. et al. 2005. “Detecting the number of clusters of individuals using the software
STRUCTURE: a simulation study”. Mol. Ecol. 14, 2611–2620.
Fabioux, Caroline, Stéphane Pouvreau, Frédérique Le Roux, and Arnaud Huvet. 2004. “The
Oyster Vasa-like Gene: A Specific Marker of the Germline in Crassostrea Gigas.”
Biochemical and Biophysical Research Communications 315(4):897–904.
Falush, D., Stephens, M. & Pritchard, J. K. 2007. “Inference of population structure using
multilocus genotype data: dominant markers and null alleles”. Mol. Ecol. 7, 895–908.
Fisher, R.A. 1930. “The Genetical Theory of Natural Selection”. Clarendon Press.
Fortune, John M., Youri I. Pavlov, Carrie M. Welch, Erik Johansson, Peter M. J. Burgers, and
Thomas A. Kunkel. 2005. “Saccharomyces Cerevisiae DNA Polymerase δ: High Fidelity
for Base Substitutions but Lower Fidelity for Single-and Multi-Base Deletions.” Journal of
Biological Chemistry 280(33):29980–87.
160
Gagnaire, Pierre Alexandre, Jean Baptiste Lamy, Florence Cornette, Serge Heurtebise, Lionel
Dégremont, Emilie Flahauw, Pierre Boudry, Nicolas Bierne, and Sylvie Lapègue. 2018.
“Analysis of Genome-Wide Differentiation between Native and Introduced Populations of
the Cupped Oysters Crassostrea Gigas and Crassostrea Angulata.” Genome Biology and
Evolution 10(9):2518–34.
Ghiselli, Fabrizio, Liliana Milani, Peter L. Chang, Dennis Hedgecock, Jonathan P. Davis, Sergey
V. Nuzhdin, and Marco Passamonti. 2012. “De Novo Assembly of the Manila Clam
Ruditapes philippinarum Transcriptome Provides New Insights into Expression Bias,
Mitochondrial Doubly Uniparental Inheritance and Sex Determination.” Molecular Biology
and Evolution 29(2):771–86.
Giraud, A., I. Matic, O. Tenaillon, A. Clara, M. Radman, M. Fons, and F. Taddei. 2001. “Costs
and Benefits of High Mutation Rates: Adaptive Evolution of Bacteria in the Mouse Gut.”
Science 291(5513):2606–8.
Goddard, M. E. and B. J. Hayes. 2007. “Genomic Selection.” Journal of Animal Breeding and
Genetics 124:323–30.
Gracey, Andrew Y., Maxine L. Chaney, Judson P. Boomhower, William R. Tyburczy, Kwasi
Connor, and George N. Somero. 2008. “Rhythms of Gene Expression in a Fluctuating
Intertidal Environment.” Current Biology 18(19):1501–7.
Harrang, Estelle, Sylvie Lapègue, Benjamin Morga, and Nicolas Bierne. 2013. “A High Load of
Non-Neutral Amino-Acid Polymorphisms Explains High Protein Diversity despite
Moderate Effective Population Size in a Marine Bivalve with Sweepstakes Reproduction.”
G3: Genes, Genomes, Genetics 3(2):333–41.
Hauser, Lorenz and Gary R. Carvalho. 2008. “Paradigm Shifts in Marine Fisheries Genetics:
Ugly Hypotheses Slain by Beautiful Facts.” Fish and Fisheries 9(4):333–62.
Hedgecock, D. 1994. “Does Variance in Reproductive Success Limit Effective Population Sizes
of Marine Organisms?” (January 1994):122–134.
Hedgecock, Dennis and Alexander I. Pudovkin. 2011. “Sweepstakes Reproductive Success in
Highly Fecund Marine Fish and Shellfish : A Review and Commentary.” Bulletin of Marine
Science 87(4):971–1002.
Hedgecock, Dennis, Grace Shin, Andrew Y. Gracey, David van den Berg, and Manoj P.
Samanta. 2015. “Second-Generation Linkage Maps for the Pacific Oyster Crassostrea Gigas
Reveal Errors in Assembly of Genome Scaffolds.” G3: Genes, Genomes, Genetics
5(10):2007–19.
Hedgecock, Dennis and Fred Sly. 1990. “Genetic Drift and Effective Population Sizes of
Hatchery-Propagated Stocks of the Pacific Oyster, Crassostrea Gigas.” Aquaculture 88:21–
38.
161
Helm, M. M., N. Bourne, and A. Lovatelli. 2004. Hatchery Culture of Bivalves. A Practical
Manual.
Hubisz, M. J., Falush, D., Stephens, M. & Pritchard, J. K. 2009. “Inferring weak population
structure with the assistance of sample group information”. Mol. Ecol. Resour. 9, 1322–
1332.
Jaramillo-Correa, J. P., Verdú, M., and González-Martínez, S. C. 2010.“The contribution of
recombination to heterozygosity differs among plant evolutionary lineages and life-forms”.
BMC Evol. Biol. 10, 22.
Johnson, M. S. and R. Black. 1982. “Chaotic Genetic Patchiness in an Intertidal Limpet,
Siphonaria Sp.” Marine Biology 70(2):157–64.
Johnson, Michael S. and Robert Black. 1984. “Pattern Beneath the Chaos: The Effect of
Recruitment on Genetic Patchiness in an Intertidal Limpet.” Evolution 38(6):1371.
Jombart, T. 2008. “Adegenet: a R package for the multivariate analysis of genetic markers”.
Bioinformatics 24, 1403–1405.
Jombart, T. & Ahmed, I. 2011. “Adegenet 1.3-1: new tools for the analysis of genome-wide SNP
data”. Bioinformatics 27, 3070–3071.
Kimura, M., 1983. The Neutral Theory of Molecular Evolution. Cambridge University Press,
New York.
Kong, Ning, Qi Li, Hong Yu, and Ling Feng Kong. 2015. “Heritability Estimates for Growth-
Related Traits in the Pacific Oyster (Crassostrea Gigas) Using a Molecular Pedigree.”
Aquaculture Research 46(2):499–508.
Kovach, A., Wegrzyn, J. L., Parra, G., Holt, C., Bruening, G. E., Loopstra, C. A., Hartigan, J.,
Yandell, M., Langley, C. H., Korf, I., et al. 2010. “The Pinus taeda genome is characterized
by diverse and highly diverged repetitive sequences”. BMC Genomics 11, 420.
Krasovec, Marc, Adam Eyre-Walker, Sophie Sanchez-Ferandin, and Gwenael Piganeau. 2017.
“Spontaneous Mutation Rate in the Smallest Photosynthetic Eukaryotes.” Molecular
Biology and Evolution 34(7):1770–79.
Kusumi, Junko, Yoshihiko Tsumura, and Hidenori Tachida. 2016. “Evolutionary Rate Variation
in Two Conifer Species, Taxodium Distichum (L.) Rich. Var. Distichum(Baldcypress) and
Cryptomeria Japonica(Thunb. Ex L.F.) D. Don (Sugi, Japanese Cedar).” Genes and Genetic
Systems 90(5):305–15.
Launey, Sophie and Dennis Hedgecock. 2001. “High Genetic Load in the Pacific Oyster
Crassostrea Gigas.” Genetics 159(1):255–65.
162
Li, Haimei, Bo Zhang, Guiju Huang, Baosuo Liu, Sigang Fan, Dongling Zhang, and Dahui Yu.
2016. “Differential Gene Expression during Larval Metamorphic Development in the Pearl
Oyster, Pinctada Fucata, Based on Transcriptome Analysis.” International Journal of
Genomics 2016.
Li, Heng and Richard Durbin. 2009. “Making the Leap: Maq to BWA.” Mass Genomics
25(14):1754–60.
Li, Lei, Kelly M. Murphy, Uliana Kanevets, and Linda J. Reha-Krantz. 2005. “Sensitivity to
Phosphonoacetic Acid: A New Phenotype to Probe DNA Polymerase δ in Saccharomyces
Cerevisiae.” Genetics 170(2):569–80.
Li, Yongjun and Heidi S. Dungey. 2018. “Expected Benefit of Genomic Selection over Forward
Selection in Conifer Breeding and Deployment.” PLoS ONE 13(12):1–21.
Liao, Huan, Zujing Yang, Zheng Dou, Fanhua Sun, Sihua Kou, Zhengrui Zhang, Xiaoting
Huang, and Zhenmin Bao. 2019. “Impact of Ocean Acidification on the Energy Metabolism
and Antioxidant Responses of the Yesso Scallop (Patinopecten Yessoensis).” Frontiers in
Physiology 10(JAN):1–10.
Liu, W., Thummasuwan, S., Sehgal, S. K., Chouvarine, P., and Peterson, D. G. 2011.
“Characterization of the genome of bald cypress”. BMC Genomics 12, 553.
Lynch, M., and B. Walsh, 1998. Genetics and Analysis of Quantitative Traits. Sinauer Associates,
Sunderland, MA.
Lynch, M. 1995. Mutation accumulation and the extinction of small populations. Am. Nat. 146,
489–518.
Lynch M. 2006. “The origins of eukaryotic gene structure.” Mol. Biol. Evol 23:450–468.
Lynch, M. 2007. “The Origins of Genome Architecture”. Sunderland, MA: Sinauer Assocs., Inc.
Lynch, Michael. 2010. “Evolution of the Mutation Rate.” Trends Genet 26(8):345–52.
Lynch, Michael, Matthew S. Ackerman, Jean Francois Gout, Hongan Long, Way Sung, W.
Kelley Thomas, and Patricia L. Foster. 2016. “Genetic Drift, Selection and the Evolution of
the Mutation Rate.” Nature Reviews Genetics 17(11):704–14.
Lynch, Michael, I. John Conery, and Reinhard Burger. 1995. “Mutation Accumulation and the
Extinction of Small Populations.” American Society of Naturalists 146(4):489–518.
McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, Garimella K,
Altshuler D, Gabriel S, Daly M. 2010. “The genome analysis toolkit: a MapReduce
163
framework for analyzing next-generation DNA sequencing data”. Genome Res. 20:1297–
1230. doi: 10.1101/gr.107524.110.
Moritsuka, E., Hisataka, Y., Tamura, M., Uchiyama, K., Watanabe, A., Tsumura, Y., and
Tachida, H. 2012. “Extended linkage disequilibrium in noncoding regions in a conifer,
Cryptomeria japonica”. Genetics 190, 1145–1148.
Nei M, 1987. Molecular Evolutionary Genetics. New York: Columbia University Press.
Palumbi SR. 2004. “Marine reserves and ocean neighborhoods: the spatial scale of marine
populations and their management”. Annu. Rev. Environ. Resour. 29:31–68
Pan, T. C. Franci., Scott L. Applebaum, Christina A. Frieder, and Donal T. Manahan. 2018.
“Biochemical Bases of Growth Variation during Development: A Study of Protein
Turnover in Pedigreed Families of Bivalve Larvae (Crassostrea Gigas).” Journal of
Experimental Biology 221(10).
Pennisi, Elizabeth. 2018. “Human Mutation Rate a Legacy from Our Past.” Science
360(6385):143.
Piganeau, Gwenael and Adam Eyre-Walker. 2009. “Evidence for Variation in the Effective
Population Size of Animal Mitochondrial DNA.” PLoS ONE 4(2):2–9.
Plough, L. V., G. Shin, and D. Hedgecock. 2016. “Genetic Inviability Is a Major Driver of Type
III Survivorship in Experimental Families of a Highly Fecund Marine Bivalve.” Molecular
Ecology 25(4):895–910.
Plough, Louis V. 2016. “Genetic Load in Marine Animals: A Review.” Current Zoology
62(6):567–79.
Plough, Louis V. and Dennis Hedgecock. 2011. “Quantitative Trait Locus Analysis of Stage-
Specific Inbreeding Depression in the Pacific Oyster Crassostrea Gigas.” Genetics
189(4):1473–86.
Prindle, Marc and Lawrence Loeb. 2012. “DNA Polymerase Delta in DNA Replication and
Genome Maintenance.” Environmental Molecular Mutagens 53(9):666–82.
Pritchard, J. K. et al. 2000. “Inference of population structure using multilocus genotype data”.
Genetics 155, 945–959.
R Core Team (2016). R: A language and environment for statistical computing. R Foundation for
Statistical Computing, Vienna, Austria.
Rahbari, Raheleh, Arthur Wuster, Sarah J. Lindsay, Robert J. Hardwick, Ludmil B. Alexandrov,
Saeed Al Turki, Anna Dominiczak, Andrew Morris, David Porteous, Blair Smith, Michael
164
R. Stratton, and Matthew E. Hurles. 2016. “Timing, Rates and Spectra of Human Germline
Mutation.” Nature Genetics 48(2):126–33.
Raynes, Y., A. L. Halstead, and P. D. Sniegowski. 2014. “The Effect of Population Bottlenecks
on Mutation Rate Evolution in Asexual Populations.” Journal of Evolutionary Biology
27(1):161–69.
Raynes, Yevgeniy, Paul D. Sniegowski, and Daniel M. Weinreich. 2019. “Migration Promotes
Mutator Alleles in Subdivided Populations.” Evolution 73(3):600–608.
Raynes, Yevgeniy, C. Scott Wylie, Paul D. Sniegowski, and Daniel M. Weinreich. 2018. “Sign
of Selection on Mutation Rate Modifiers Depends on Population Size.” Proceedings of the
National Academy of Sciences of the United States of America 115(13):3422–27.
Sanford, Eric and Morgan W. Kelly. 2011. “Local Adaptation in Marine Invertebrates.” Annual
Review of Marine Science 3(1):509–35.
Sauvage, C., N. Bierne, S. Lapègue, and P. Boudry. 2007. “Single Nucleotide Polymorphisms
and Their Relationship to Codon Usage Bias in the Pacific Oyster Crassostrea Gigas.” Gene
406(1–2):13–22.
Schmidt, Stephanie. 2009. “The Genome Analysis Toolkit.” Proceedings of the International
Conference on Intellectual Capital, Knowledge Management & Organizational Learning
20:254–60.
Scott Wylie, C., Cheol Min Ghim, David Kessler, and Herbert Levine. 2009. “The Fixation
Probability of Rare Mutators in Finite Asexual Populations.” Genetics 181(4):1595–1612.
She, Zhicai, Li Li, Jie Meng, Zhen Jia, Huayong Que, and Guofan Zhang. 2018. “Population
Resequencing Reveals Candidate Genes Associated with Salinity Adaptation of the Pacific
Oyster Crassostrea Gigas.” Scientific Reports 8(1):8683.
Signh, S. M. and E. Zouros. 1978. “Genetic Variation Associated with Growth Rate in the
American Oyster ( Crassostrea Virginica).” Ev 32(2):342–53.
Simons, Ariel Levi, Nathan Churches, and Sergey Nuzhdin. 2018. “High Turnover of Faecal
Microbiome from Algal Feedstock Experimental Manipulations in the Pacific Oyster
(Crassostrea Gigas).” Microbial Biotechnology 11(5):848–58.
Sole-Cava, A. M. and J. P. Thorpe. 1991. “High Levels of Genetic Variation in Natural
Populations of Marine Lower Invertebrates.” Biological Journal of the Linnean Society
44(1):65–80.
Sprouffske, Kathleen, José Aguílar-Rodríguez, Paul Sniegowski, and Andreas Wagner. 2018.
High Mutation Rates Limit Evolutionary Adaptation in Escherichia Coli. Vol. 14.
165
Subramanian S. 2016. “The effects of sample size on population genomic analyses–implications
for the tests of neutrality". BMC Genomics 17:123. 10.1186/s12864-016-2441-8
Sun, Xiujun, Grace Shin, and Dennis Hedgecock. 2015. “Inheritance of High-Resolution Melting
Profiles in Assays Targeting Single Nucleotide Polymorphisms in Protein-Coding
Sequences of the Pacific Oyster Crassostrea Gigas: Implications for Parentage Assignment
of Experimental and Commercial Broodstocks.” Aquaculture 437:127–39.
Sutton, Mark D. 2015. “How MutS Finds a Needle in a Haystack.” Proceedings of the National
Academy of Sciences of the United States of America 112(50):15265–66.
Svensson, Erik I. and David Berger. 2019. “The Role of Mutation Bias in Adaptive Evolution.”
Trends in Ecology and Evolution 34(5):422–34.
Tenaillon, Olivier, Bruno Toupance, Hervé Le Nagard, François Taddei, and Bernard Godelle.
1999. “Mutators, Population Size, Adaptive Landscape and the Adaptation of Asexual
Populations of Bacteria.” Genetics 152(2):485–93.
Van der Auwera GA, et al. 2013. “From FastQ data to high confidence variant calls: the Genome
Analysis Toolkit best practices pipeline”. Curr. Protoc. Bioinforma. 2013;11:11.10.1–
11.10.33.
Waldbusser, George G., Elizabeth L. Brunner, Brian A. Haley, Burke Hales, Christopher J.
Langdon, and Frederick G. Prahl. 2013. “A Developmental and Energetic Basis Linking
Larval Oyster Shell Formation to Acidification Sensitivity.” Geophysical Research Letters
40(10):2171–76.
Ward, R. D., L. J. English, D. J. McGoldrick, G. B. Maguire, J. A. Nell, and P. A. Thompson.
2000. “Genetic Improvement of the Pacific Oyster Crassostrea Gigas (Thunberg) in
Australia.” Aquaculture Research 31(1):35–44.
Watterson, G.A. 1975. "On the number of segregating sites in genetical models without
recombination.", Theoretical Population Biology, 7 (2): 256–276
Watts, R. J., M. S. Johnson, and R. Black. 1990. “Effects of Recruitment on Genetic Patchiness
in the Urchin Echinometra Mathaei in Western Australia.” Marine Biology 105(1):145–51.
Williams, G. C. 1975. “Sex and Evolution”. Princeton University Press, Princeton, NJ.
Zhang, GuofanWang, Jun, Xiaodong Fang…and Jian Wang. 2012. “The Oyster Genome Reveals
Stress Adaptation and Complexity of Shell Formation.” Nature 490(7418):49–54.
166
Supplemental Chapter 1 References: Generation of Mussel (Mytilus galloprovincialis)
family lines, in concert with relevant phenotyping schedule, in order to elucidate viability-
and growth-associated biomarkers.
Aanen, D. K., H. H. de Fine Licht, A. J. M. Debets, N. A. G. Kerstes, R. F. Hoekstra, and J. J.
Boomsma. 2009. “The B73 Maize Genome Complexity, Diversity, and Dynamics.” Science
326(5956):1103–6.
Appels, Rudi, Kellye Eversole…and Le Wang. 2018. “Shifting the Limits in Wheat Research
and Breeding Using a Fully Annotated Reference Genome.” Science 361(6403).
Braby, Caren E. and George N. Somero. 2006. “Ecological Gradients and Relative Abundance of
Native (Mytilus Trossulus) and Invasive (Mytilus Galloprovincialis) Blue Mussels in the
California Hybrid Zone.” Marine Biology 148(6):1249–62.
Crego-Prieto, V., A. Ardura, F. Juanes, A. Roca, J. S. Taylor, and E. Garcia-Vazquez. 2015.
“Aquaculture and the Spread of Introduced Mussel Genes in British Columbia.” Biological
Invasions 17(7):2011–26.
Fabrice, Teletchea. 2018. “Fish Domestication: An Overview.” Pp. 69–90 in IntechOpen.
FAO. 2018. The State of World Fisheries and Aquaculture 2018 - Meeting the sustainable
development goals. Rome.
Gao, Lei, Itay Gonda, Honghe Sun, Qiyue Ma, Kan Bao, Denise M. Tieman, Elizabeth A.
Burzynski-Chang, Tara L. Fish, Kaitlin A. Stromberg, Gavin L. Sacks, Theodore W.
Thannhauser, Majid R. Foolad, Maria Jose Diez, Jose Blanca, Joaquin Canizares, Yimin
Xu, Esther van der Knaap, Sanwen Huang, Harry J. Klee, James J. Giovannoni, and
Zhangjun Fei. 2019. “The Tomato Pan-Genome Uncovers New Genes and a Rare Allele
Regulating Fruit Flavor.” Nature Genetics 51(6):1044–51.
Groenen, Martien A. M., Alan L…and Lawrence B. Schook. 2012. “Analyses of Pig Genomes
Provide Insight into Porcine Demography and Evolution.” Nature 491(7424):393–98.
Harrang, Estelle, Sylvie Lapègue, Benjamin Morga, and Nicolas Bierne. 2013. “A High Load of
Non-Neutral Amino-Acid Polymorphisms Explains High Protein Diversity despite
Moderate Effective Population Size in a Marine Bivalve with Sweepstakes Reproduction.”
G3: Genes, Genomes, Genetics 3(2):333–41.
Hauser, Lorenz and Gary R. Carvalho. 2008. “Paradigm Shifts in Marine Fisheries Genetics:
Ugly Hypotheses Slain by Beautiful Facts.” Fish and Fisheries 9(4):333–62.
Hedgecock, D. 1994. “Does Variance in Reproductive Success Limit Effective Population Sizes
of Marine Organisms?” (January 1994):122–134.
167
Hedgecock, Dennis and Alexander I. Pudovkin. 2011. “Sweepstakes Reproductive Success in
Highly Fecund Marine Fish and Shellfish: A Review and Commentary.” Bulletin of Marine
Science 87(4):971–1002.
Hedgecock, Dennis, Grace Shin, Andrew Y. Gracey, David van den Berg, and Manoj P.
Samanta. 2015. “Second-Generation Linkage Maps for the Pacific Oyster Crassostrea Gigas
Reveal Errors in Assembly of Genome Scaffolds.” G3: Genes, Genomes, Genetics
5(10):2007–19.
Helm, M. M., N. Bourne, and A. Lovatelli. 2004. Hatchery Culture of Bivalves. A Practical
Manual.
Kong, Ning, Qi Li, Hong Yu, and Ling Feng Kong. 2015. “Heritability Estimates for Growth-
Related Traits in the Pacific Oyster (Crassostrea Gigas) Using a Molecular Pedigree.”
Aquaculture Research 46(2):499–508.
Larson, Greger, Dolores R. Piperno, Robin G. Allaby, Michael D. Purugganan, Leif Andersson,
Manuel Arroyo-Kalin, Loukas Barton, Cynthia Climer Vigueira, Tim Denham, Keith
Dobney, Andrew N. Doust, Paul Gepts, M. Thomas P. Gilbert, Kristen J. Gremillion,
Leilani Lucas, Lewis Lukens, Fiona B. Marshall, Kenneth M. Olsen, J. Chris Pires, Peter J.
Richerson, Rafael Rubio De Casas, Oris I. Sanjur, Mark G. Thomas, and Dorian Q. Fuller.
2014. “Current Perspectives and the Future of Domestication Studies.” Proceedings of the
National Academy of Sciences of the United States of America 111(17):6139–46.
Launey, Sophie and Dennis Hedgecock. 2001. “High Genetic Load in the Pacific Oyster
Crassostrea Gigas.” Genetics 159(1):255–65.
Lockwood, Brent L., Kwasi M. Connor, and Andrew Y. Gracey. 2015. “The Environmentally
Tuned Transcriptomes of Mytilus Mussels.” Journal of Experimental Biology
218(12):1822–33.
López, Maria E., Laura Benestan, Jean Sebastien Moore, Charles Perrier, John Gilbey, Alex Di
Genova, Alejandro Maass, Diego Diaz, Jean Paul Lhorente, Katharina Correa, Roberto
Neira, Louis Bernatchez, and José M. Yáñez. 2019. “Comparing Genomic Signatures of
Domestication in Two Atlantic Salmon (Salmo Salar L.) Populations with Different
Geographical Origins.” Evolutionary Applications 12(1):137–56.
Ludwig, Yvonne and Inez H. Slamet-Loedin. 2019. “Genetic Biofortification to Enrich Rice and
Wheat Grain Iron: From Genes to Product.” Frontiers in Plant Science 10(July):1–10.
Matsumoto, Takashi, Jianzhong Wu…and Benjamin Burr. 2005. “The Map-Based Sequence of
the Rice Genome.” Nature 436(7052):793–800.
Miao, Lili, Xinguo Mao, Jingyi Wang, Zicheng Liu, Bin Zhang, Weiyu Li, Xiaoping Chang,
Matthew Reynolds, Zhenhua Wang, and Ruilian Jing. 2017. “Elite Haplotypes of a Protein
168
Kinase Gene TaSnRK2.3 Associated with Important Agronomic Traits in Common Wheat.”
Frontiers in Plant Science 8(March):1–11.
Mun, Seyoung, Yun Ji Kim, Kesavan Markkandan, Wonseok Shin, Sumin Oh, Jiyoung Woo,
Jongsu Yoo, Hyesuck An, and Kyudong Han. 2017. “The Whole-Genome and
Transcriptome of the Manila Clam (Ruditapes Philippinarum).” Genome Biology and
Evolution 9(6):1487–98.
Murgarella, Maria, Daniela Puiu, Beatriz Novoa, Antonio Figueras, David Posada, and Carlos
Canchaya. 2016. “Correction: A First Insight into the Genome of the Filter-Feeder Mussel
Mytilus Galloprovincialis.” PLoS ONE 11(7):1–22.
Plough, Louis V. 2016. “Genetic Load in Marine Animals: A Review.” Current Zoology
62(6):567–79.
Plough, Louis V. and Dennis Hedgecock. 2011. “Quantitative Trait Locus Analysis of Stage-
Specific Inbreeding Depression in the Pacific Oyster Crassostrea Gigas.” Genetics
189(4):1473–86.
Pulcini, Domitilla, Stefano Cataudella, Clara Boglione, Tommaso Russo, Paul A. Wheeler,
Loredana Prestinicola, and Gary H. Thorgaard. 2015. “Testing the Relationship between
Domestication and Developmental Instability in Rainbow Trout, Oncorhynchus Mykiss
(Teleostei, Salmonidae).” Biological Journal of the Linnean Society 114(3):608–28.
Shields, J. L., J. W. Heath, and D. D. Heath. 2010. “Marine Landscape Shapes Hybrid Zone in a
Broadcast Spawning Bivalve: Introgression and Genetic Structure in Canadian West Coast
Mytilus.” Marine Ecology Progress Series 399:211–23.
Symonds, Jane E., Shannon M. Clarke, Nick King, Seumas P. Walker, Brian Blanchard, David
Sutherland, Rodney Roberts, Mark A. Preece, Mike Tate, Peter Buxton, and Ken G. Dodds.
2019. “Developing Successful Breeding Programs for New Zealand Aquaculture: A
Perspective on Progress and Future Genomic Opportunities.” Frontiers in Genetics
10(FEB):1–7.
Viard, F., B. Delay, C. Coustau, and F. Renaud. 1994. “Evolution of the Genetic Structure of
Bivalve Cohorts at Hybridization Sites of the Mytilus Edulis-M. Galloprovincialis
Complex.” Marine Biology 119(4):535–39.
Warren, Wesley C., La Deana W. Hillier, Chad Tomlinson, Patrick Minx, Milinn Kremitzki,
Tina Graves, Chris Markovic, Nathan Bouk, Kim D. Pruitt, Francoise Thibaud-Nissen,
Valerie Schneider, Tamer A. Mansour, C. Titus Brown, Aleksey Zimin, Rachel Hawken,
Mitch Abrahamsen, Alexis B. Pyrkosz, Mireille Morisson, Valerie Fillon, Alain Vignal,
William Chow, Kerstin Howe, Janet E. Fulton, Marcia M. Miller, Peter Lovell, Claudio V.
Mello, Morgan Wirthlin, Andrew S. Mason, Richard Kuo, David W. Burt, Jerry B.
Dodgson, and Hans H. Cheng. 2017. “A New Chicken Genome Assembly Provides Insight
into Avian Genome Structure.” G3: Genes, Genomes, Genetics 7(1):109–17.
169
Zhang, Guofan, Xiaodong Fang, Ximing Guo, Li Li, Ruibang Luo, Fei Xu, Pengcheng Yang,
Linlin Zhang, Fucun Wu, Yuanxin Chen, Jiafeng Wang, Chunfang Peng, Jie Meng, Lan
Yang, Jun Liu, Bo Wen, Na Zhang, and Zhiyong Huang. 2012. “The Oyster Genome
Reveals Stress Adaptation and Complexity of Shell Formation.” Nature 490:49–54.
Zimin, Aleksey V., Arthur L. Delcher, Liliana Florea, David R. Kelley, Michael C. Schatz,
Daniela Puiu, Finnian Hanrahan, Geo Pertea, Curtis P. Van Tassell, Tad S. Sonstegard,
Guillaume Marçais, Michael Roberts, Poorani Subramanian, James A. Yorke, and Steven L.
Salzberg. 2009. “A Whole-Genome Assembly of the Domestic Cow, Bos Taurus.” Genome
Biology 10(4).
Supplemental Chapter 2 References: Sperm and Egg Chemoattraction Studies: a series of
preliminary experiments suggests the case for Egg Competition in a marine dioecious
simulcaster
Bateman, A. J. 1948. “Intra-sexual selection in Drosophila”. Heredity 2:349–368.
Birkhead, Timothy R. and Tommaso Pizzari. 2002. “Postcopulatory Sexual Selection.” Nature
Reviews Genetics 3(4):262–73.
Darwin, Charles. 1871. “The Descent of Man: And Selection in Relation to Sex”. London: J.
Murray, Print.
Depends, Proteins, Don R. Levitan, and David L. Ferrell. 2006. “Selection on Gamete
Recognition.” Science 312(April):267–69.
Eberhard, W.G. 1996. “Female Control: Sexual Selection by Cryptic Female Choice”. Princeton
University Press.
Eberhard, William G. 2009. “Postcopulatory Sexual Selection: Darwin’s Omission and Its
Consequences.” In the Light of Evolution 3:243–61.
Evans, Jonathan P., Francisco Garcia-Gonzalez, Maria Almbro, Oscar Robinson, and John L.
Fitzpatrick. 2012. “Assessing the Potential for Egg Chemoattractants to Mediate Sexual
Selection in a Broadcast Spawning Marine Invertebrate.” Proceedings of the Royal Society
B: Biological Sciences 279(1739):2855–61.
Evans, Jonathan P. and Craig D. H. Sherman. 2013. “Sexual Selection and the Evolution of Egg-
Sperm Interactions in Broadcast-Spawning Invertebrates.” Biological Bulletin 224(3):166–
83.
Firman, Renée C. 2018. “Postmating Sexual Conflict and Female Control over Fertilization
during Gamete Interaction.” Annals of the New York Academy of Sciences 1422(1):48–64.
170
Firman, Renée C., Clelia Gasparini, Mollie K. Manier, and Tommaso Pizzari. 2017. “Postmating
Female Control: 20 Years of Cryptic Female Choice.” Trends in Ecology and Evolution
32(5):368–82.
Firman, Renée C., and Simmons, L.W. 2015. “Gametic interactions promote inbreeding
avoidance in house mice”. Ecol. Lett. 18, 937–943
Fisher, David N., Rolando Rodríguez-Muñoz, and Tom Tregenza. 2016. “Comparing Pre-and
Post-Copulatory Mate Competition Using Social Network Analysis in Wild Crickets.”
Behavioral Ecology 27(3):912–19.
Helm, M. M., N. Bourne, and A. Lovatelli. 2004. Hatchery Culture of Bivalves. A Practical
Manual.
Hussain, Yasmeen H., Martin Sadilek, Shukri Salad, Richard K. Zimmer, and Jeffrey A. Riffell.
2017. “Individual Female Differences in Chemoattractant Production Change the Scale of
Sea Urchin Gamete Interactions.” Developmental Biology 422(2):186–97.
Jennions MD, Kokko H. 2010. “Sexual selection”. Evolutionary behavioral ecology pp. 343–
364. Oxford, UK: Oxford University Press.
Levitan, D. R. 2004. “Density-dependent sexual selection in external fertilizers: Variances in
male and female fertilization success along the continuum from sperm limitation to sexual
conflict in the sea urchin Strongylocentrotus franciscanus”. Am. Nat. 164:298–309.
Levitan, Don R. 2005. “The Distribution of Male and Female Reproductive Success in a
Broadcast Spawning Marine Invertebrate.” Integrative and Comparative Biology
45(5):848–55.
Marshall, D. J. and J. P. Evans. 2005. “Does Egg Competition Occur in Marine Broadcast-
Spawners?” Journal of Evolutionary Biology 18(5):1244–52.
Okamoto, Daniel K. 2016. “Competition among Eggs Shifts to Cooperation along a Sperm
Supply Gradient in an External Fertilizer.” American Naturalist 187(5):E129–42.
Oliver, Mathew and Jonathan P. Evans. 2014. “Chemically Moderated Gamete Preferences
Predict Offspring Fitness in a Broadcast Spawning Invertebrate.” Proceedings of the Royal
Society B: Biological Sciences 281(1784).
Schultz, Nicholas G., Jesse Ingels, Andrew Hillhouse, Keegan Wardwell, Peter L. Chang, James
M. Cheverud, Cathleen Lutz, Lu Lu, Robert W. Williams, and Matthew D. Dean. 2016.
“The Genetic Basis of Baculum Size and Shape Variation in Mice.” G3: Genes, Genomes,
Genetics 6(5):1141–51.
Suquet, M., C. Labbe, R. Brizard, A. Donval, J. R. Le Coz, C. Quere, and P. Haffray. 2010.
“Changes in Motility, ATP Content, Morphology and Fertilisation Capacity during the
171
Movement Phase of Tetraploid Pacific Oyster (Crassostrea Gigas) Sperm.” Theriogenology
74(1):111–17.
Thornhill, R. 1983. “Cryptic female choice and its implications in the scorpionfly Harpobittacus
nigriceps”. Am. Nat. 122, 765–788
Wedell, N. 2009. “Sperm Competition.” Encyclopedia of Animal Behavior 322–28.
Additional Reading and References
Aguirre, J. David, Mark W. Blows, and Dustin J. Marshall. 2016. “Genetic Compatibility
Underlies Benefits of Mate Choice in an External Fertilizer.” American Naturalist
187(5):647–57.
Alonzo, Suzanne H., Kelly A. Stiver, and Susan E. Marsh-Rollo. 2016. “Ovarian Fluid Allows
Directional Cryptic Female Choice despite External Fertilization.” Nature Communications
7:1–8.
Barros, P., P. Sobral, P. Range, L. Chícharo, and D. Matias. 2013. “Effects of Sea-Water
Acidification on Fertilization and Larval Development of the Oyster Crassostrea Gigas.”
Journal of Experimental Marine Biology and Ecology 440:200–206.
Beekman, Madeleine, Bart Nieuwenhuis, Daniel Ortiz-Barrientos, and Jonathan P. Evans. 2016.
“Sexual Selection in Hermaphrodites, Sperm and Broadcast Spawners, Plants and Fungi.”
Philosophical Transactions of the Royal Society B: Biological Sciences 371(1706).
Boni, Raffaele, Alessandra Gallo, Melania Montanino, Alberto Macina, and Elisabetta Tosti.
2016. “Dynamic Changes in the Sperm Quality of Mytilus Galloprovincialis under
Continuous Thermal Stress.” Molecular Reproduction and Development 83(2):162–73.
Breton, Sophie, Charlotte Capt, Davide Guerra, and Donald Stewart. 2017. “Sex Determining
Mechanisms in Bivalves.” (June):1–23.
Crean, Angela J. and Dustin J. Marshall. 2008. “Gamete Plasticity in a Broadcast Spawning
Marine Invertebrate.” Proceedings of the National Academy of Sciences of the United States
of America 105(36):13508–13.
Eads, Angela R., Jonathan P. Evans, and Winn Jason Kennington. 2016. “Plasticity of
Fertilization Rates under Varying Temperature in the Broadcast Spawning Mussel, Mytilus
Galloprovincialis.” Ecology and Evolution 6(18):6578–85.
Eads, Angela R., W. Jason Kennington, and Jonathan P. Evans. 2016. “Interactive Effects of
Ocean Warming and Acidification on Sperm Motility and Fertilization in the Mussel
Mytilus Galloprovincialis.” Marine Ecology Progress Series 562:101–11.
172
Geffard, O., H. Budzinski, S. Augagneur, M. N. L. Seaman, and E. His. 2001. “Assessment of
Sediment Contamination by Spermiotoxicity and Embryotoxicity Bioassays with Sea
Urchins (Paracentrotus Lividus) and Oysters (Crassostrea Gigas).” Environmental
Toxicology and Chemistry 20(7):1605–11.
Henshaw, Jonathan M., Dustin J. Marshall, Michael D. Jennions, and Hanna Kokko. 2014.
“Local Gamete Competition Explains Sex Allocation and Fertilization Strategies in the
Sea.” American Naturalist 184(2).
Ho, M. A., C. Price, C. K. King, P. Virtue, and M. Byrne. 2013. “Effects of Ocean Warming and
Acidification on Fertilization in the Antarctic Echinoid Sterechinus Neumayeri across a
Range of Sperm Concentrations.” Marine Environmental Research 90:136–41.
Hudspith, M., Amanda Reichelt-Brushett, and Peter L. Harrison. 2017. “Factors Affecting the
Toxicity of Trace Metals to Fertilization Success in Broadcast Spawning Marine
Invertebrates: A Review.” Aquatic Toxicology 184:1–13.
Hussain, Yasmeen H., Jeffrey S. Guasto, Richard K. Zimmer, Roman Stocker, and Jeffrey A.
Riffell. 2016. “Sperm Chemotaxis Promotes Individual Fertilization Success in Sea
Urchins.” Journal of Experimental Biology 219(10):1458–66.
Hussain, Yasmeen H., Martin Sadilek, Shukri Salad, Richard K. Zimmer, and Jeffrey A. Riffell.
2017. “Individual Female Differences in Chemoattractant Production Change the Scale of
Sea Urchin Gamete Interactions.” Developmental Biology 422(2):186–97.
Levitan, Don R. 2005. “The Distribution of Male and Female Reproductive Success in a
Broadcast Spawning Marine Invertebrate.” Integrative and Comparative Biology
45(5):848–55.
Lewis, Ceri and Alex T. Ford. 2012. “Infertility in Male Aquatic Invertebrates: A Review.”
Aquatic Toxicology 120–121:79–89.
Luis Stephano, José and Meredith Gould. 1988. “Avoiding Polyspermy in Oyster (Crassostrea
Gigas).” Aquaculture 73(1–4):295–307.
Okamoto, Daniel K. 2016. “Competition among Eggs Shifts to Cooperation along a Sperm
Supply Gradient in an External Fertilizer.” American Naturalist 187(5):E129–42.
Ritchie, Hannah and Dustin J. Marshall. 2013. “Fertilisation Is Not a New Beginning: Sperm
Environment Affects Offspring Developmental Success.” Journal of Experimental Biology
216(16):3104–9.
Rolton, Anne, Philippe Soudant, Julien Vignier, Richard Pierce, Michael Henry, Sandra E.
Shumway, V. Monica Bricelj, and Aswani K. Volety. 2015. “Susceptibility of Gametes and
Embryos of the Eastern Oyster, Crassostrea Virginica, to Karenia Brevis and Its Toxins.”
Toxicon 99:6–15.
173
Schultz, Nicholas G., Jesse Ingels, Andrew Hillhouse, Keegan Wardwell, Peter L. Chang, James
M. Cheverud, Cathleen Lutz, Lu Lu, Robert W. Williams, and Matthew D. Dean. 2016.
“The Genetic Basis of Baculum Size and Shape Variation in Mice.” G3: Genes, Genomes,
Genetics 6(5):1141–51.
Sherman, Craig D. H., Emi S. Ab Rahim, Mats Olsson, and Vincent Careau. 2015. “The More
Pieces, the Better the Puzzle: Sperm Concentration Increases Gametic Compatibility.”
Ecology and Evolution 5(19):4354–64.
Song, Y. P., M. Suquet, I. Quéau, and L. Lebrun. 2009. “Setting of a Procedure for Experimental
Fertilisation of Pacific Oyster (Crassostrea Gigas) Oocytes.” Aquaculture 287(3–4):311–14.
Suquet, M., J. Cosson, A. Donval, C. Labbé, M. Boulais, P. Haffray, I. Bernard, and C. Fauvel.
2012. “Marathon vs Sprint Racers: An Adaptation of Sperm Characteristics to the
Reproductive Strategy of Pacific Oyster, Turbot and Seabass.” Journal of Applied
Ichthyology 28(6):956–60.
Suquet, M., C. Labbe, R. Brizard, A. Donval, J. R. Le Coz, C. Quere, and P. Haffray. 2010.
“Changes in Motility, ATP Content, Morphology and Fertilisation Capacity during the
Movement Phase of Tetraploid Pacific Oyster (Crassostrea Gigas) Sperm.” Theriogenology
74(1):111–17.
Vihtakari, Mikko, Iris E. Hendriks, Johnna Holding, Paul E. Renaud, Carlos M. Duarte, and Jon
N. Havenhand. 2013. “Effects of Ocean Acidification and Warming on Sperm Activity and
Early Life Stages of the Mediterranean Mussel (Mytilus Galloprovincialis).” Water
(Switzerland) 5(4):1890–1915.
174
Supplemental Documents
Bivalve DNA Extraction Protocol
Equipment Needed:
• Animals/Specimens
• Heat Station
• Liquid Nitrogen
• Zymo GDNA Extraction Kit
• Ice
• Qubit Machine & HS-DNA Reagents
• 4
o
C centrifuge (non 4C seems to work well also)
• Razor blades
• Pestle and Mortar
• Proteinase K
• Tissue and Cell Lysis Solution
• MPC Protein Precipitation Reagent
• Eppendorph pestle and pestle vibrating tool
Protocol:
Set Up:
1. Create Lysis solution by adding 20 uL of Proteinase K into 300 uL of Tissue and
Cell Lysis solution. If you anticipate more than one sample, multiply this equation
175
accordingly. (i.e. two samples = 40 uL Proteinase K and 600 uL Tissue & Cell Lysis
solution).
2. Turn Heat Station to 55 C
3. Get Ice
4. Get Liquid Nitrogen
5. Turn centrifuge on and make sure it is set to 4
o
C if applicable
6. Wipe down lab area with 70% EtOH
Grinding Animal Tissues (Skip this step if processing larval samples)
1. If frozen, get animal from -80C Freezer.
2. Cut an approximately 1 cubic cm (or less!) chunk from the animal. You want enough
animal to fill an Eppendorph tube approx. ¼ way, too much will result in poor quality
DNA. Place animal tissue not being used back into the freezer immediately.
3. Place animal tissue into mortar, fill mortar carefully with liquid Nitrogen, and grind.
Place your hand around the pestle so that you don’t send chunks of bivalve flying all over
the place. Repeat this step as necessary adding liquid Nitrogen to keep tissues frozen.
You must get the tissue ground to a FINE POWDER!
DNA Extraction
4. Place animal tissues into a 1.7 mL Maxximum Recovery tube, approximately ¼ to ½ way
full. DO NOT OVERFILL!
5. Add 300 uL of Cell Lysis solution with Proteinase K to each tube.
6. Use Eppendorph pestle and pestle vibrating tool to homogenize tissue and solution as
best you can. It may get very foamy and gooey, that’s fine. Vortex. NOTE: if processing
larval samples, use hand pestle to homogenize gently.
176
7. Incubate at 55 C for at least 15 minutes, 10 minutes, vortex every 5 2 minutes.
8. Place the samples on ice for 3-5 minutes.
9. Add 150 uL of MPC protein Precipitation Reagent to each of your samples and vortex
vigorously for at least 10 seconds. Let stand for 1 minute.
10. Pellet the debris by centrifuging at 4 C for 10 minutes at max speed.
a. If the resultant pellet is clear (i.e. not there), small, loose, the liquid is extremely
milky, etc., add an additional 25 uL of MPC Protein Precip Reagent, mix, and
pellet again.
11. Transfer the supernatant only (i.e. the clear liquid) to clean 1.7 mL Maxx Recovery tubes
in 200 uL aliquots (discard the protein pellet).
12. Start with step 1 of the Zymo G-DNA extraction kit on the “Cell Suspensions and
Proteinase K Digested Samples” protocol. For the first step, add 800 uL of Genomic
Lysis Buffer to each tube.
a. The protocol says that each filter only holds approx. 800 uL, but really they can
hold approx. 1 mL
13. After completing the DNA Extraction kit, quantify DNA concentration with Qubit.
14. Record concentrations, label tubes, and place DNA in -20 C freezer (or on ice if you are
planning to use immediately).
177
Library Generation Protocol: Bivalve Specific ddRAD
PREPARING ddRAD LIBRARIES
Original Written Protocol: Wendy Vu
Bivalve Specific Protocol: Nathan Churches
NOTES: This protocol is for generating a small number of libraries at a time, and has been optimized for mussel
ddRAD seq using SphI-HF and Mlu-CI Restriction Enzymes. This process can be streamlined for 96 well plates by
referencing Wendy’s original protocol. It is assumed that you know how to quantify DNA using Qubit and clean
using MagBio beads for this protocol.
1) Annealing Protocol for Adapters
Description: In this step, you are taking the single stranded adapters that you purchased (both the
barcoded and common adapter) and annealing them so that they are double stranded. These will
later go through the restriction digestion process, and be bound with digested DNA. Before
starting, you need to adjust the molarity of the single stranded adapters to 200 uM. This is
typically done by adding the appropriate amount of water at the onset into the dry tube that the
single stranded DNA comes in.
Protocol:
In a PCR Tube bring together:
o Top strand DNA oligo (200uM starting concentration) 5ul
o Bottom strand DNA oligo (200uM starting concentration) 5ul
o TE 10ul (Tris-EDTA, pH 8.0, can be purchased via Fisher)
178
Total volume 20ul
Primers are 50uM after dilution.
In Thermocycler:
• 95 degrees for 2min
• Ramp to 25 degrees by 0.1 degree per second
• Hold at 25 degrees for 30 min
• Hold at 4 degrees forever
Plus/Minus Check for Annealing
To check the performance of the annealing, run 1ul of annealed adapters side by side with 1ul of
un-annealed adapter in a 4% agarose gel. Sybrsafe does not bind to single stranded DNA, so the
un-annealed adapter lane will look blank (However, when using Safeview, we observed that the
intensity of the un-annealed adapter band was lower than the annealed adapters and also the size
was smaller). The annealed adapters should show up at ~45bp.
Making 4% Agarose gel with Safeview:
• Weigh the agarose and put it in a bottle with 1X TBE buffer overnight to hydrate the
agarose
• Next morning: heat the agarose in the microwave for 2 min (check during heating to
avoid spilling out). Check while heating to observe any floating debris, which is
undissolved agarose.
• Add 5ul of Safeview to 100 ml agarose gel, stir, and pour immediately very carefully.
179
2) Adapter Working Solutions
Description: In the following steps, you will take the adapters, which are now double stranded
and stable at a concentration of 50 uM, and create dilutions that you can work with in the lab.
*Water can work in place of TE.
First Dilution (5uM)
To make working solutions, bring together:
• 5ul of 50uM stock solution of annealed adapters
• 45ul of 1X TE*
• Vortex and spin down
Total Volume: 50ul
Second Dilution, Working Adapter stock (0.5uM)
Dilute the above stock mix 1:10 stock:water
Use Qubit to quantify your working adapter stock. Concentrations should be approximately 60
ng/uL.
NOTE!: The ending concentration is the most important, not the dilution. Ensure you get to
~60ng/uL.
180
3) First restriction digestion and ligation of sample DNA and Barcoded Adapter
Description: In this step, you will digest your DNA using the first RE: SphI-HF. The restriction
enzyme digestion is immediately followed by ligation, and so the barcoded adaptors (as opposed
to the common adapter) are included into the restriction digestion but won’t be affected since
they do not have the restriction enzyme recognition sites. If you change the REs used in this
protocol, you must go back in and ensure your adapters do not have any sites that match the RE
cut site. You should have high molecular weight DNA before this process. This protocol may not
be as effective if you are using degraded DNA.
Protocol:
• In a PCR tube, add enough template DNA for a concentration of 100 ng of DNA in 10 uL
of water.
• Add 2 uL of the barcoded adapter you intend to use for each sample, which should total
approximately 120 ng of adapter.**
**This ratio of [120ng:100ng adapter:DNA] can be optimized for different
genomes and adapter sets, but works well for Mytilus galloprovincialis.
Digestion 1 with SphI-HF
Mix for First Digestion uL
DNA:Adapter Mix 12
H20 5
CutSmart Buffer 2
SphI-HF (20U/uL)* 1
Vol final 20
Incubate 37C for 1 hour
181
Heat at 65C for 20min (inactivation)
*Consult the card that comes with the RE and adjust this table accordingly. This worked as of 2018 for SphI-HF.
Ligation 1
Mix for Ligation 1 uL
Digested DNA&Adapter, prev step 20
H20 23.4
T4 Buffer 5
T4 ligase (400U/uL) 1.6
Vol final 50
Incubate 22C for 1 hour
Heat at 65C for 10 min (inactivation)
Clean using MagBio beads at 1.6X, to rid mixture of small leftover adapter molecules (1.6X * 50
= 80 uL MagBio). Elute in 40 uL of water, and collect 37 uL for the next step. (Leaving 3 uL
ensures no bead contamination for the next steps).
4) Second restriction digestion and ligation reaction using Common Adapter
Description: In this step, you will repeat the process as in step 3, but with MLU-CI restriction
enzyme and the common adapter instead. After cleaning (step 5), you will have a ‘pre-library’
that is ready for PCR enrichment.
Protocol:
Digestion 2 with Mlu-CI
182
Mix for First Digestion uL
DNA From Step 3 37
Common Adapter 4
CutSmart Buffer 4
Mlu-CI (20U/uL)* 1
Vol final 46
Incubate 37C for 1 hour
Heat at 65C for 20min (inactivation)
*Consult the card that comes with the RE and adjust this table accordingly. This worked as of 2018 for Mlu-CI.
Ligation 2
Mix for Ligation 2 uL
Digested DNA&Adapter, prev step 46
T4 Buffer 5
T4 ligase (400U/uL) 1
Vol final 52
Incubate 22C for 1 hour
Heat at 65C for 10 min (inactivation)
5) Library Size Selection
Description: In this step, you will take your libraries and size select for smaller fragments, which
are required for effective PCR. This is done using the properties of MagBio beads.
Protocol:
• Clean using MagBio beads at 1.6X, to rid mixture of small leftover adapter molecules
(1.6X * 52 = 83.2 uL MagBio). Elute in 40 uL of water, and collect 37 uL for the next
step. (Leaving 3 uL ensures no bead contamination for the next steps).
183
• Add .5X MagBio beads to cleaned DNA (37 * .5 = 18.5 uL), and recover supernatant on
step 1 of MagBio Protocol. This should yield ~ 55 uL of small fragments only (37 + 18.5
= ~55). You may continue with the MagBio Protocol with the larger fragments that are
stuck on the beads, if you want to test for efficacy of size selection at later steps. The
large fragments may be tossed out if unneeded.
• Using the 55 uL, clean again at the full 1.8X MagBio Beads (55* 1.8X = 99 uL MagBio).
Run through entire MagBio Protocol, eluting in 40 uL water and recovering 37 uL.
• This is now your size selected ‘pre-library’, ready for PCR.
• Quantify using Qubit.
6) PCR enrichment
Description: You should have only small fragments in your ‘pre-library’, which are now at fairly
low concentrations. PCR will help enrich the concentration, and this step also adds the illumina
platform sequences via the primers. After PCR, you’ll have fragments of DNA in the size range
of approximately 100-1000 bp with a [barcode+illumina_index] on one end, and a
[common_adapter+illumina_index] on the other. This is your library, ready for the BioA! You
can perform several PCR reactions per library in order to increase the concentration of library.
For example, Wendy had 12 PCR reactions per library, resulting in 600 uL of total amplified
library. This can then be cleaned and concentrated for higher concentration libraries.
Mix for PCR uL
184
DNA Pre-Library See Note
Water Fill to 50
5x HF buffer 10
2.5uM dNTPs 5
P1 primer (5uM starting) 4
P2 primer (5uM starting) 4
MgCl2 2
Phusion HF Taq 0.5
Vol final 50
NOTE: It is CRITICAL that you have a ratio of [25ngDNA:2uMPrimer]. 2uM represents the
final concentration of the primers, which are starting at 5uM. Use the equation C1V1=C2V2 to
determine appropriate amount of DNA Pre-Library to add. For example, if you had a pre-library
at a concentration of 20ng/uL, I’d do the following:
C1*V1=C2*V2
(20ng/uL)*V1=(2uM)*(50 uL)
V1 = [(2uM)*(50 uL)]/(20ng/uL)
V1 = 5 uL of pre-library to add
**NOTE: The above strike-through was wrong somewhere…On 1/28/19 through 2/2/2019 I
underwent a BioA optimization of the appropriate ratio and cycling of primers for PCR. The
result shows that for every 1 uL of primer at 5 uM, you need approx. 2ng of DNA. SO, add up to
8ng of your library to achieve the appropriate balance for 4 uL of 5uM primer. EX:
185
DNA total ng Primer volume to add (at 5uM
start conc.)
2 1
3 1.5
4 2
5 2.5
6 3
7 3.5
8 4
186
Amplify using the following PCR protocol:
98C 30sec
10 cycles of:
98C 10sec
65C 45sec
72C 15sec
72C 2min
Hold at 4C
NOTE: Reducing the number of cycles may reduce amplification of bigger fragments >600bp, which will not be
sequenced efficiently by Illumina sequencing. 10 cycles was found best given the 1/28/19 optimization trial.
Optional: Pool PCR reactions if you have performed multiple reactions.
-Clean PCR reactions using MagBio Beads at 1.6X, then quantify PCR reactions using Qubit.
-Test Libraries using BioAnalyzer and qPCR.
-Sequence!
Abstract (if available)
Abstract
The word aquaculture simply means to grow organisms, animal or plant, from a body of water, fresh or salt. Because this term is inherently so large, it is important to narrow the field when considering anything related. During my Ph.D. studies, I focused primarily on marine bivalve aquaculture academics. Its techniques, processes, and of course state of science. In this thesis, I describe in 5 chapters new work pertaining to the field of aquaculture and marine science. In the introductory section I describe firstly why bivalve aquaculture is an important and worthy field of study, and follow with a more general background for my specific thesis focus areas. The first chapter is a study on the behavioral effects of copper toxicity in the Pacific oyster (Crassostrea gigas), our results indicate that toxicity swimming response may have a genetic or familial component. The second focuses on a study of microbiome response to feed type in the Pacific oyster, where we demonstrate an extreme turnover of gut-associated microbiota during dietary change manipulation in contrast with stable control diets. In chapter three I describe a large genomics dataset used to describe the generational mutation rate for Pacific oyster for the first time, confirming that these organisms indeed have one of the most acute mutation rates ever studied in eukaryotes. The fourth and fifth chapters describe preliminary datasets which warrant further investigation, showing evidence for egg fertilization capacity adjustment during female to female competition for the Pacific oyster, and a large family line genotype to phenotype assay performed using the Mediterranean mussel (Mytilus galloprovincialis), respectively. It is my intention that this work demonstrates that combining traditional experimental science and genomics techniques can provide the necessary insights in order to further develop bivalves as an ecological model species and a (delicious) sustainable food source.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Evolutionary genomic analysis in heterogeneous populations of non-model and model organisms
PDF
Transcriptional and morphological impacts of copper on Mytilus californianus larval development in current and future ocean conditions
PDF
Genome sequencing and transcriptome analysis of the phenotypically plastic spadefoot toads
PDF
Genetic architecture underlying variation in different traits in the Pacific oyster Crassostrea gigas
PDF
Complex mechanisms of cryptic genetic variation
PDF
Diversity and dynamics of giant kelp “seed-bank” microbiomes: Applications for the future of seaweed farming
PDF
Ancestral inference and cancer stem cell dynamics in colorectal tumors
PDF
Investigating the potential roles of three mammalian traits in female reproductive investment
PDF
Application of machine learning methods in genomic data analysis
PDF
Genome-wide analysis of genetic load and larval mortality in a highly fecund marine invertebrate, the Pacific Oyster Crassostrea gigas
PDF
Plant genome wide association studies and improvement of the linear mixed model by applying the weighted relationship matrix
PDF
The evolution of gene regulatory networks
PDF
Molecular and behavioral mechanisms of circatidal biological rhythms in intertidal mollusks
PDF
Dps contributes to typical growth, survival, and genome organization in E. coli
PDF
Topics in selective inference and replicability analysis
PDF
Structural and biochemical studies of large T antigen: the SV40 replicative helicase
PDF
Physiological strategies of resilience to environmental change in larval stages of marine invertebrates
PDF
Mechanism study of SV40 large tumor antigen atpase and helicase functions in viral DNA replication
PDF
Towards an understanding of fault-system mechanics: from single earthquakes on isolated faults to millenial-scale collective plate-boundary fault-system behavior
PDF
Phytoplankton bloom initiation in the Southern California Bight: a multi-year local and regional analysis
Asset Metadata
Creator
Churches, Nathan Daniel
(author)
Core Title
Studies in bivalve aquaculture: metallotoxicity, microbiome manipulations, and genomics & breeding programs with a focus on mutation rate
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Molecular Biology
Publication Date
02/07/2020
Defense Date
11/25/2019
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
aquaculture,copper toxicity,Crassostrea gigas,fertilization,larvae,local adaptation,metal toxicity,metallotoxicity,microbiome,mussel,mutation rate,Mytilus,Mytilus galloprovincialis,OAI-PMH Harvest,oyster,plankton,sexual selection,swimming behavior
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Nuzhdin, Sergey (
committee chair
), Dean, Matthew (
committee member
), Ehrenreich, Ian (
committee member
), Gracey, Andrew (
committee member
)
Creator Email
churches@usc.edu,ndchurches@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-267655
Unique identifier
UC11673356
Identifier
etd-ChurchesNa-8157.pdf (filename),usctheses-c89-267655 (legacy record id)
Legacy Identifier
etd-ChurchesNa-8157.pdf
Dmrecord
267655
Document Type
Dissertation
Rights
Churches, Nathan Daniel
Type
texts
Source
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 a...
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
Tags
copper toxicity
Crassostrea gigas
fertilization
larvae
local adaptation
metal toxicity
metallotoxicity
microbiome
mussel
mutation rate
Mytilus
Mytilus galloprovincialis
oyster
plankton
sexual selection
swimming behavior