Kobe University Repository : Thesis

Kobe University Repository : Thesis

学位論文題目Tit le

Study on the characterist ics and dynamics of fish environmental DNA(魚類環境DNAの性質および動態に関する研究)

氏名Author Jo, Toshiaki

専攻分野Degree 博士（理学）

学位授与の日付Date of Degree 2021-03-25

公開日Date of Publicat ion 2022-03-01

資源タイプResource Type Thesis or Dissertat ion / 学位論文

報告番号Report Number 甲第7973号

権利Rights

JaLCDOI

URL http://www.lib.kobe-u.ac.jp/handle_kernel/D1007973※当コンテンツは神戸大学の学術成果です。無断複製・不正使用等を禁じます。著作権法で認められている範囲内で、適切にご利用ください。

PDF issue: 2022-07-07

博⼠論⽂

Study on the characteristics and dynamics of

fish environmental DNA

⿂類環境 DNAの性質および動態に関する研究

2021年 1⽉

神⼾⼤学⼤学院⼈間発達環境学研究科

Toshiaki Jo / 徐寿明

Outline

Chapter 1. General Introduction. ............................................................................... 1

1.1. Figures ........................................................................................................................ 11

Chapter 2. Effect of water temperature and fish biomass on environmental DNA

shedding, degradation, and size distribution. .......................................................... 13

2.1. Introduction ................................................................................................................. 13

2.2. Materials and methods ................................................................................................. 16

2.2.1. Tank experiment ....................................................................................................... 16

2.2.1.1. Experimental design .......................................................................................... 16

2.2.1.2. eDNA sampling ................................................................................................. 18

2.2.1.3. DNA extraction ................................................................................................. 20

2.2.1.4. Quantification of eDNA using qPCR .................................................................... 22

2.2.2. Data analysis .......................................................................................................... 23

2.2.2.1. Environmental DNA shedding and decay rates ...................................................... 23

2.2.2.2. Environmental DNA size distribution ................................................................... 25

2.3. Results ........................................................................................................................ 26

2.3.1. Effect of water temperature and fish biomass on eDNA shedding and decay rates ............. 27

2.3.2. Effect of water temperature and fish biomass on eDNA size distribution .......................... 28

2.3.3. Temporal dynamics of eDNA size distribution .............................................................. 28

2.4. Discussion ................................................................................................................... 29

2.4.1. Factors affecting the degradation of eDNA .................................................................. 29

2.4.2. Factors affecting the shedding of eDNA ...................................................................... 30

2.4.3. Environmental DNA size distribution .......................................................................... 31

2.5. Conclusions ............................................................................................................... 33

2.6. Tables .......................................................................................................................... 35

2.7. Figures ........................................................................................................................ 40

Chapter 3. Estimating shedding and decay rates of environmental nuclear DNA

with relation to water temperature and biomass. .................................................... 45

3.1. Introduction ................................................................................................................. 45


3.2.1. Experimental design ................................................................................................. 47

3.2.2. eDNA sampling and extraction .................................................................................. 48

3.2.3. Primers and probe development ................................................................................. 49

3.2.4. Quantification of eDNA samples ................................................................................ 51

3.2.5. Statistical analyses ................................................................................................... 52

3.2.6. Additional experiment for the relationship between eDNA decay rates and its fragment size 55

3.3. Results ........................................................................................................................ 56

3.4. Discussion ................................................................................................................... 58

3.5. Tables .......................................................................................................................... 64

3.6. Figures ........................................................................................................................ 67

Chapter 4. Particle size distribution of environmental DNA from the nuclei of

marine fish. ................................................................................................................ 73

4.1. Introduction ................................................................................................................. 73


4.2.1. Experimental protocol .............................................................................................. 76

4.2.2. Statistical analyses ................................................................................................... 78

4.3. Results and Discussion ................................................................................................ 79

4.3.1. The relationships of eDNA PSD with temperature, fish biomass, and DNA markers ........... 80

4.3.2. Temporal changes of eDNA PSD ................................................................................ 82

4.3.3. Implications and Perspectives .................................................................................... 84

4.4. Tables .......................................................................................................................... 87

4.5. Figures ........................................................................................................................ 90

Chapter 5. Rapid degradation of longer DNA fragments enables the improved

estimation of distribution and biomass using environmental DNA. ........................ 97

5.1. Introduction ................................................................................................................. 97

5.2. Materials and Methods ................................................................................................ 99

5.2.1. Primers and probe development ................................................................................. 99

5.2.2. Tank experiment ..................................................................................................... 100

5.2.2.1. Experimental set-up and water sampling ............................................................ 100

5.2.2.2. DNA extraction ............................................................................................... 102

5.2.2.3. Quantification of eDNA using qPCR .................................................................. 102

5.2.3. Application to field samples ..................................................................................... 104

5.3. Results ...................................................................................................................... 106

5.3.1. Primers and probe development ............................................................................... 106

5.3.2. Degradation curves for long and short amplicons ....................................................... 106

5.3.3. Comparison of eDNA and echo intensity in the field survey .......................................... 107

5.4. Discussion ................................................................................................................. 108

5.5. Tables ........................................................................................................................ 114

5.6. Figures ...................................................................................................................... 118

Chapter 6. Selective collection of environmental DNA with long fragment using

larger filter pore size. .............................................................................................. 121

6.1. Introduction ............................................................................................................... 121

6.2. Materials and Methods .............................................................................................. 123

6.2.1. Water sampling ...................................................................................................... 123

6.2.2. DNA extraction and quantitative real-time PCR ......................................................... 124

6.2.3. Statistical analyses ................................................................................................. 125

6.3. Results and Discussion .............................................................................................. 126

6.3.1. The ratio of long to short mitochondrial eDNA ........................................................... 126

6.3.2. The ratio of nuclear to mitochondrial eDNA .............................................................. 127

6.3.3. The difference of eDNA capture efficiencies between filters .......................................... 128

6.4. Conclusions ............................................................................................................... 130

6.5. Tables ........................................................................................................................ 132

6.6. Figures ...................................................................................................................... 136

Chapter 7. Complex interactions between environmental DNA (eDNA) state and

water chemistries on eDNA persistence suggested by meta-analyses. ................... 140

7.1. Introduction ............................................................................................................... 140

7.2. Materials and Methods .............................................................................................. 143

7.2.1. Literature search and data extraction ....................................................................... 143

7.2.2. Statistical analyses ................................................................................................. 145

7.2.3. Re-analysis of the time-series changes in eDNA particle size distribution ....................... 146

7.3. Results ...................................................................................................................... 147

7.3.1. Literature review .................................................................................................... 147

7.3.2. Model selection ..................................................................................................... 148


7.4. Discussion ................................................................................................................. 149

7.4.1. Meta-analyses of eDNA literature............................................................................. 150


7.4.3. Limitations and perspectives .................................................................................... 155

7.5. Tables ........................................................................................................................ 158

7.6. Figures ...................................................................................................................... 166

Chapter 8. General Discussion................................................................................ 172

8.1. Nuclear and mitochondrial eDNA .............................................................................. 173

8.2. Long and short eDNA fragments ................................................................................ 179

8.3. Linking eDNA characteristics to its dynamics ............................................................ 184

8.4. Further perspectives for the innovation of eDNA applications .................................... 189

8.5. Figure ........................................................................................................................ 196

References throughout the thesis ............................................................................ 197

Appendix ................................................................................................................. 229

Acknowledgements .................................................................................................. 230

1

Chapter 1. General Introduction.

DNA molecules are released as dead individuals, cells, secretions, feces, and pollens

and present in every terrestrial and aquatic environment (Levy-Booth et al., 2007;

Nielsen et al., 2007; Torti et al., 2015). Since the first paper reporting the successful

extraction and purification of microbial DNA from lake sediments (Ogram et al., 1987),

DNA molecules in environment (i.e., environmental DNA [eDNA]) has primarily been

utilized by microbiologists and paleontologists. In the former, by using polymerase

chain reaction (PCR) and in situ hybridization (ISH), researchers achieved to directly

evaluate microbial communities in environmental samples without isolating and

culturing, which often requires multiple tests of biochemical conditions of cultivation

but most of microbes are yet to be unculturable (Amann et al., 1995). These novel

molecular approaches revealed the hitherto unknown diversity of microbes and

revolutionary advanced the understanding of microbial ecology (Alfreider et al., 1996;

Belgrader et al., 1999; Matsui et al., 2001; Amann & Fuchs, 2008; Uchii et al., 2011;

Okazaki et al., 2013; Carini et al., 2016). In the latter, by analyzing DNA in core

samples from frozen or temperate sediments, researchers achieved to obtain the

implications on long-term temporal transition of fauna, flora, and human activities from

the late Pleistocene (~10,000 years ago) to Holocene (past 10,000 years) (Willerslev et

al., 2003; D’Anjou et al., 2012; Giguet-Covex et al., 2014; Zobel et al., 2018).

In addition, eDNA analysis has recently been developed to estimate the

current distribution and abundance of macro-organisms such as fish and amphibians

(Ficetola et al., 2008; Darling & Mahon, 2011; Taberlet et al., 2012; Bohmann et al.,

2

2014; Thomsen & Willerslev, 2015; Takahara et al., 2016; Deiner et al., 2017a). Under

the recent crisis of biodiversity all over the world (Dudgeon et al., 2006; Rockström et

al., 2009; Butchart et al., 2010; Ceballos et al., 2015), the first step against the loss is to

obtain precise information on species distribution and abundance on relevant

spatiotemporal scales. Relative to traditional monitoring methods (e.g., visual census,

video, fishing, trap, acoustic tagging, echo sounder, etc.), owing to the analysis of

genetic materials in environmental samples such as water and soil without capturing nor

observing individuals, eDNA analysis (i) has no or little damage to individuals and their

habitats, (ii) substantially reduces the effort and cost in the field, (iii) enables the species

identification based on nucleotide sequence information without high morphological

expertise, and (iv) produces less variable results among researchers (Darling & Mahon,

2011; Takahara et al., 2016). Since Ficetola et al. (2008) reported the successful

detection of eDNA from American bullfrog (Lithobates catesbeianus) tadpole in ponds,

the non-invasiveness, cost-efficiency, and high sensitivity of eDNA-based biological

monitoring has been reported in various taxa and natural environments (Minamoto et

al., 2012; Thomsen et al., 2012; Tréguier et al., 2014; Fukumoto et al., 2015; Yamamoto

et al., 2016; Bista et al., 2017; Boussarie et al., 2018; Sengupta et al., 2019; Djurhuus et

al., 2020).

However, there are some challenges in biological monitoring via eDNA

analysis. First, although eDNA analysis has a higher detection sensitivity than

traditional methods, eDNA detection is not necessarily perfect (i.e., target eDNA is not

necessarily detected in a site where the individual is present). Thus, even if such a false-

3

negative detection can partly be coped with by occupancy modeling framework, which

can statistically take account of multiple observation and process errors (MacKenzie et

al., 2003; Dorazio & Erickson, 2017; Chen & Ficetola, 2019), the detection/non-

detection of target eDNA can sometimes contradict the presence/absence of target

species. Second, regardless of positive correlations between eDNA concentrations and

biomass/abundance/body size of individuals (Takahara et al., 2012; Klymus et al., 2015;

Yamamoto et al., 2016; Doi et al., 2017; Wu et al., 2018; Yates et al., 2019), it is highly

challenging to establish the method quantifying species biomass/abundance via eDNA

analysis with high level of accuracy and reliability (Hansen et al., 2018; Yates et al.,

2019) except for a few trials which combined quantitative analysis of eDNA with

hydrodynamic modelling to take into account the processes of eDNA production,

transport, and/or degradation (Carraro et al., 2018; Fukaya et al., 2020). Third, the

spatiotemporal range of eDNA signal at a given sampling location and time cannot be

fully understood (i.e., how much time have passed since the eDNA was shed, and how

far away is eDNA transported from?) (Roussel et al., 2015). Why do the uncertainties

relating to eDNA detection and quantification arise, and what should we do to mitigate

and eliminate such uncertainties? This is a big question for all eDNA researchers, and

should be solved for the establishment of eDNA analysis as a more refined tool to

monitor biodiversity and fishery resources (Thomsen & Willerslev, 2015; Evans &

Lamberti, 2018).

Ultimately, these uncertainties can originate from the lack of information on

the characteristics and dynamics of eDNA, which is termed as ‘the ecology of eDNA’ in

4

Barnes & Turner (2016), as follows (Figure 1-1); (a) eDNA characteristics:

physiological (excretion, secretion, exfoliation, decomposition, etc.) and ecological

(reproduction, predator-prey relationship, etc.) sources of eDNA production, and its

physiochemical and molecular states (intra-/extra-cellular, dissolved/adsorbed, particle

size, genetic region, electric charge, etc.), and (b) eDNA dynamics: the processes of

eDNA production, transport, and persistence, and environmental biotic/abiotic factors

affecting such eDNA dynamics. These factors are not independent; eDNA

characteristics can multifacetedly influence its vertical/horizontal transport and

persistence, which eventually determines the spatiotemporal scale of eDNA signal.

Larger and heavier eDNA particles in water can be less dispersed and settle more

rapidly (Robinson & Bailey, 1981; Wotton & Malmqvist, 2001), and DNA molecules

within a cell membrane (intra-cellular DNA) and adsorbed to organic matters and/or

substrates can be less frequently attacked by environmental microbes and extra-cellular

enzymes than extra-cellular, dissolved, and free DNA (Nielsen et al., 2007; Arnosti,

2014). Therefore, understanding of eDNA characteristics can assist to understand eDNA

dynamics, which will refine the knowledge on spatiotemporal scale of eDNA signal,

improve the performance of eDNA detection and quantification, and fill a gap between

eDNA detection/quantification and species presence/abundance in the field.

During this decade, characteristics and dynamics of eDNA from macro-

organisms has been studied to some extent. Among 535 of original papers targeting

eDNA from macro-organisms published in peer-reviewed, international journals during

2008 to 2019, which is based on my search by Google Scholar, 78, 16, 31, and 54

5

papers were assigned to keywords ‘production’, ‘state’, ‘transport’, and ‘persistence’

(Figure 1-2a; Appendix S1). Much of studies corresponding to ‘production’ have

reported the positive effect of species biomass/abundance on eDNA detectability and

concentration in laboratory and natural environments using fish, amphibian, and other

invertebrates (e.g., Takahara et al., 2012; Pilliod et al., 2013; Dougherty et al., 2016;

Yamamoto et al., 2016; Wu et al., 2018; Iwai et al., 2019). However, there are few

studies implying physiological and ecological sources of eDNA except Merkes et al.

(2014) and Dunker et al. (2016), which suggested the risk of false-positive detection of

eDNA derived from carcasses and predator feces in the field. Studies corresponding

‘transport’ have reported eDNA downstream transport distances (e.g., Deiner &

Altermatt, 2014; Jane et al., 2015; Sansom & Sassoubre, 2017), horizontal diffusion

distances (Andruszkiewicz et al., 2019; Murakami et al., 2019), and retention rates to

substrates (Fremier et al., 2019; Shogren et al., 2019). Particularly, eDNA downstream

transport distances greatly varied among studies (from tens of meters to tens of

kilometers), which can be substantially affected by hydrologic and geographic

conditions such as flow velocity, slope, type of substrate, and biofilm (Jane et al., 2015;

Shogren et al., 2018; Fremier et al., 2019). Studies corresponding ‘persistence’ have

reported eDNA decay rate constants based on a first-order exponential model in various

environmental conditions using various taxa (Barnes et al., 2014; Strickler et al., 2015;

Lance et al., 2017; Collins et al., 2018; Seymour et al., 2018). According to these

estimates, eDNA in water seemed to be detectable for days to weeks.

In contrast, the study focusing on the state of eDNA is notably limited; my

6

literature search showed that studies corresponding ‘state’ is only 3.0 % in eDNA

studies (16 out of 535 papers; Figure 1-2b). According to limited number of studies

reporting eDNA states, it is reported that (i) eDNA from macro-organisms is present in

various sizes and forms (<0.2 to >180 µm in diameter), much of which is detected in 1

to 10 µm size fraction (Turner et al., 2014; Wilcox et al., 2015), while it may depend on

experimental conditions and target taxa (Sassoubre et al., 2016; Moushomi et al., 2019),

(ii) eDNA is distributed heterogeneously in water and soil (Shogren et al., 2016; Song et

al., 2017; Chen & Ficetola, 2019), and (iii) eDNA concentration was higher in shorter

DNA fragment sizes (Bylemans et al., 2018a; Wei et al., 2018), whereas not all eDNA is

necessarily highly degraded and almost all length of mitogenomes (>16,000 bp) can be

retrieved from aquatic environment (Deiner et al., 2017b).

These findings implied that not all eDNA is present as extra-membrane free

DNA in environment but some can be as intra-membrane DNA such as cell and tissue

fragments, nuclei, and mitochondria, which can protect DNA molecules from enzymatic

degradation due to environmental microbial activities. However, these inferences on the

relationship between eDNA physiochemical state and persistence have not so far been

examined (e.g., how does the state of eDNA influence the persistence of eDNA, and

how does the quality of genomic information obtained from eDNA differ depending on

its state?). Majority of studies describing eDNA decay rate constants used a first-order

exponential model, while some studies inferred a biphasic or multiphasic degradation of

eDNA (Eichmiller et al., 2016; Bylemans et al., 2018a; Wei et al., 2018). Biphasic

degradation has also been reported in DNA and RNA from microbes, virus, and leaves

7

(Ding & Wu 1999; Poté et al., 2005; Rogers et al., 2011), where degradation processes

were considered not to be necessarily monophasic, being classified into an early phase

with rapid degradation and a remaining phase with slow degradation. With regards to

microbes and viruses, this can be explained by their physiological characteristics such

as living/dead cell, response to environmental carrying capacity, and antibiotic-

resistance (Easton et al., 2005; You et al., 2006; Rogers et al., 2011). Similarly, with

regards to macrobial eDNA, degradation processes may be different depending on its

state (such as intra-/extra-membrane, living/dead cell, particulate/dissolved), which may

determine the persistence of eDNA and its fragment size amplifiable by PCR.

Moreover, contrary to hydrogeographic factors mentioned above, it remains

uninvestigated how the state of eDNA influences the transport of eDNA; different

particle sizes and structures of eDNA can result in different dynamics of

horizontal/vertical transports. Some studies pointed out that eDNA transport did not

follow the same dynamics as the conservative tracer such as ion tracer which assumes

the homogenous distribution of uniform particles (Jerde et al., 2016; Fremier et al.,

2019). These implications are reasonable given various particle sizes and forms of

macrobial eDNA described above. It would be required in the future to develop the

mathematical statistical approach taking various particle sizes and heterogenous

distribution of eDNA into account, for which it is necessary to accumulate the

knowledge of eDNA states such as its particle size distribution and structure.

The aim of my doctoral thesis is to comprehensively refine the relationship

between the characteristics and dynamics of eDNA from macro-organisms, and to

8

obtain the clue to mitigate and eliminate the uncertainties relating to eDNA detection

and quantification. First, in Chapters 2, 3, and 4, by performing tank experiments using

Japanese jack mackerel (Trachurus japonicus) as a model species, I comprehensively

analyzed the effects of biotic/abiotic and molecular factors on eDNA shedding and

degradation. Especially, with regard to a molecular factor, I focused on the

characteristics and dynamics of eDNA derived from mitochondria and nuclei (mt-eDNA

and nu-eDNA, respectively). Most eDNA studies have targeted mt-eDNA, while some

studies have examined the applicability of nu-eDNA, targeting multi-copy ribosomal

RNA (rRNA) gene, and reported its high detection sensitivity and potential usefulness

in eDNA analyses (Minamoto et al., 2017b; Dysthe et al., 2018). However, contrary to

mt-eDNA, the study focusing on characteristics and dynamics of nu-eDNA is very

limited. In my literature search, 47 out of 535 papers targeted nu-eDNA, while only 8

papers were assigned to any of keywords relating to eDNA characteristics and

dynamics. I compared eDNA shedding and degradation rates, its particle size

distributions, and the effects of various factors on them between mt- and nu-eDNA. In

Chapter 8, I discussed the factors influencing the difference in production and

degradation of eDNA between nuclear and mitochondrial eDNA, and the perspectives

and limitations of nuclear eDNA analysis when compared to mitochondrial eDNA

analysis.

Second, in Chapter 5, I analyzed the effect of DNA fragment size on eDNA

degradation and quantification. Given negative relationships between PCR

amplification length and detected DNA copy number/detection rate in fecal samples

9

(Deagle et al., 2006; Kamenova et al., 2018) and the number of eDNA reads in water

samples (Hänfling et al., 2016; Bista et al., 2017), eDNA degradation (that is, the

decrease in its copy number) can be caused by the decrease in DNA fragment length

owing to base cutting and deletion. I thus conducted a tank experiment using Japanese

jack mackerels, verifying a hypothesis that eDNA with longer DNA fragment degrades

faster (i.e., decrement in eDNA copy number with time is larger in longer DNA

fragment size). In addition, if the hypothesis is true, longer DNA fragments in

environmental samples might represent more recent biological information, despite

lower copy number, contrary to that of shorter DNA fragment studied in most eDNA

studies (<200 bp). Therefore, I compared correlations between fish biomass based on

echo intensity and eDNA concentration between different eDNA fragment sizes. In

Chapter 8, I discussed the perspectives and limitations to use longer DNA fragments in

eDNA analyses for ecological monitoring.

Third, in Chapters 6 and 7, I integrated the understanding of eDNA

characteristics and dynamics obtained above. In the former chapter, I tested the

applicability to selectively collect the eDNA with specific particle size. As mentioned

above, eDNA can exist in water with various sizes and states. Among them, relative to

extra-cellular DNA, intra-cellular DNA such as cell and tissue fragments can mainly be

detected at larger size fractions, and may be protected from enzymatic DNA degradation

processes. I investigated the relationship between filter pore size and DNA fragment

size, and verified whether selective collection of such large-sized eDNA increased the

collection efficiency of longer DNA fragments from water samples. Moreover, in the

10

latter chapter (Chapter 7), I conducted meta-analyses targeting previous eDNA studies

to assess how the factors relating to eDNA characteristics such as filter pore size, DNA

fragment size, and target genetic region influenced the persistence and degradation of

aqueous eDNA. Throughout the thesis, I studied the characteristics and dynamics of

eDNA released from macro-organisms, unveiled ‘the ecology of eDNA’ (Barnes &

Turner, 2016) based on complex interactions between eDNA characteristics and

dynamics, and provided the perspectives for the innovation of eDNA analysis based on

these eDNA basic information.

11

1.1. Figures

Figure 1-1. Schematic depiction of the importance and significance of studying the

characteristics and dynamics of eDNA. Characteristics of eDNA such as its

physiological and ecological sources and physiochemical states multifacetedly influence

dynamics of eDNA such as its vertical and horizontal dispersion, downstream transport,

retention, and persistence. Comprehensive understanding of eDNA characteristics and

dynamics allows to refine the spatiotemporal range of eDNA signals, and to fill a gap

between eDNA detection/quantification and species presence/abundance in the field.

Moreover, such a basic information on eDNA can provide us with a groundwork to

develop and update current eDNA analyses for more variety of research area and

interest.

eDNA characteristics

eDNA dynamics

・What is eDNA derived from?・What factors affect eDNA production?

Physiological) excretion, secretion, Physiological) exfoliation, decomposition

Ecological) reproduction, predator-preyEcological) living/dead

Production State・What cellular and molecular structure is

eDNA present with?

Cellular) intra-/extra-membrane, particle size, Cellular) dissolved/adsorbed

Molecular) DNA structure, genetic region, Molecular) fragment size, electric charge

Transport・How is eDNA dispersed vertically and ・horizontally?・How is eDNA retained from substrates?・What factors affect eDNA transport, ・diffusion, and retention?

Persistence・How is eDNA degraded biologically, ・chemically, and physically?・How long is eDNA detectable?・What factors affect eDNA degradation?

By comprehensively understanding

Multifacetedly influencing

Refining the spatiotemporal range of eDNA signal at given location and time

Conquering uncertainties relating to eDNA detection/quantification

Filling a technical gap between eDNA quantification and species presence/abundance in the field

Updating eDNA analysis for more variety of research area and interest

eDNA quantification

Biom

ass/

Abun

danc

e

cycle

Rn

12

Figure 1-2. (a) The number of publications for macro-organisms eDNA analysis from

2008 to 2019 (not including any review papers, news, views, introductions, opinions,

and perspectives), and (b) the overall proportion of publications for eDNA

characteristics and dynamics. Colors of each bar plot show the publication

corresponding each keyword (red: production, yellow: state, green: transport, blue:

persistence, and gray: other). Numerals above each bar plot in (a) represent the number

of eDNA publications on each year. Numerals in a bar plot in (b) represent the total

number of eDNA publications corresponding each keyword, proportions (%) of which

are shown in parenthesis.

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019

050

100

150

200

Overall 0.0

0.2

0.4

0.6

0.8

1.0

Pro

porti

on o

f pub

licat

ions

Published year

Num

ber o

f pub

licat

ions # Production

# State

# Transport

# Persistence

1

7889

122

166

0 0 3 8 1020

38

(a) (b)

78(14.6)

54(10.1)

31 (5.8)16 (3.0)

356(66.5)

13

Chapter 2. Effect of water temperature and fish biomass on environmental DNA

shedding, degradation, and size distribution.

2.1. Introduction

Environmental DNA (eDNA) analysis is a new method that has been developed to

improve the environmental management and assessment of aquatic ecosystems (Ficetola

et al., 2008; Minamoto et al., 2012; Taberlet et al., 2012; Thomsen & Willerslev, 2015).

Environmental DNA, which is the DNA obtained directly from environmental samples

such as water and sediments (Ficetola et al., 2008; Turner et al., 2015), is thought to

derive from feces, mucus, skin, and gametes (Martellini et al., 2005; Ficetola et al.,

2008; Merkes et al., 2014; Bylemans et al., 2017). The presence of a target species can

be estimated by detecting the eDNA in water samples instead of locating or capturing

individuals (Lodge et al., 2012). These advantages have enabled non-invasive, quick,

and wide-ranging assessments of the presence/absence of species and their biodiversity

and abundance in freshwater (Fukumoto et al., 2015; Deiner et al., 2016; Yamanaka &

Minamoto, 2016; Balasingham et al., 2017; Bista et al., 2017) and marine environments

(Thomsen et al., 2012a; 2012b; Sigsgaard et al., 2016; Yamamoto et al., 2017;

Boussarie et al., 2018; Lacoursière-Roussel et al., 2018).

Although various studies over the past decade have demonstrated successful

eDNA detection, there is a lack of basic information about eDNA, such as its origin

(i.e., the sources of eDNA), state, transport, and fate (Barnes & Turner, 2016; Hansen et

al., 2018). These factors affect the interpretation of eDNA monitoring. For example, the

14

detectability and persistence of eDNA in environmental samples are mainly determined

by eDNA shedding, transport, and degradation (Díaz-Ferguson & Moyer, 2014;

Goldberg et al., 2015; Strickler et al., 2015). Furthermore, various interactions between

eDNA and its environment should also be taken into account (Taberlet et al., 2012;

Thomsen & Willerslev, 2015; Barnes & Turner, 2016). To develop effective sampling

methods and improve the reliability of this method, it is necessary to understand and

accumulate basic information about eDNA. This study investigated the factors

associated with eDNA shedding and degradation and the eDNA size distribution.

The degradation of eDNA mainly depends on (a) abiotic factors, such as water

temperature (Strickler et al., 2015), pH (Tsuji et al., 2016), salinity (Dell'Anno &

Corinaldesi, 2004), and ultraviolet (UV) radiation (Pilliod et al., 2014); (b) biotic

factors, such as microbes and extra-cellular enzymes (Barnes et al., 2014); and (c) DNA

characteristics, such as the differences between intra-/extra-cellular DNA (Turner et al.,

2014) and the length of the DNA fragments (Jo et al., 2017). In particular, water

temperature seems to have a significant effect on eDNA; eDNA degradation was

accelerated by higher temperature (Strickler et al., 2015; Eichmiller, et al., 2016; Lance

et al., 2017; Tsuji et al., 2017). Furthermore, it is thought that water temperature does

not directly affect eDNA degradation, such as the denaturation of double-stranded DNA

(Lindahl, 1993), but indirectly affects it through enzymatic hydrolysis by microbes and

extra-cellular nucleases (Levy-Booth et al., 2007; Barnes & Turner, 2016). It is likely

that other factors also affect eDNA degradation by influencing the activity and

abundance of microbes and extra-cellular nucleases. For example, the eDNA decay rate

15

may vary depending on fish biomass because it is thought that higher fish biomass leads

to increases in the abundance of bacteria in their local environment. However, there

have been no studies on the relationship between the biomass of organisms and eDNA

degradation.

The main factors associated with eDNA shedding are (a) the number and the

biomass of organisms (Takahara et al., 2012; Klymus et al., 2015); (b) the

developmental stage of the organisms (Maruyama et al., 2014); (c) the behavior of

organisms (Dunn et al., 2017); and (d) the stress against organisms (Pilliod et al., 2014;

Bylemans et al., 2018a). In addition, considering that feed intake increased the eDNA

shedding (Klymus et al., 2015), eDNA shedding rate is likely to depend on (e) the

metabolism and physiological activity of the organisms. For example, water

temperature plays an important role in the growth and metabolism of fish (Clarke &

Johnston, 1999; Morita et al., 2010; Sandersfeld et al., 2017). For juvenile European sea

bass (Dicentrarchus labrax), feed intake (FI) and efficiency (FE) and total ammonia

nitrogen (TAN) excretion increased at 25 °C, which was the optimum temperature for

the growth of this species (Person-Le Ruyet et al., 2004). Therefore, it is likely that

eDNA shedding increases at the optimum temperature for fish growth. However, there

have been no studies on the relationship between temperature and eDNA shedding.

Although the physiological origins of the material collected as eDNA remain

uncertain (Barnes & Turner, 2016), previous studies have shown that eDNA size varied

between >180 and <0.2 µm, and the most abundant eDNA size range for macro-

organisms was from 1 to 10 µm (Turner et al., 2014; Wilcox et al., 2015; Sassoubre et

16

al., 2016). The different eDNA sizes reflect the various eDNA states (e.g., intra-/extra-

cellular DNA and within live/dead cells). Therefore, eDNA persistence and degradation

could vary depending on the eDNA states. For example, eDNA size distribution might

vary depending on the temperature of the rearing water and the biomass of organisms

and may also temporally vary after removal of the organisms.

The aim of this study was to determine the effect of water temperature and

fish biomass on eDNA shedding, degradation, and size distribution, and to refine the

eDNA analysis method. Japanese jack mackerel (Trachurus japonicus) was used as a

target species due to its use in previous eDNA studies (Yamamoto et al., 2016; Jo et al.,

2017; Yamamoto et al., 2017) and due to its economic importance as one of the most

consumed fish species in Japan. It is therefore critical to understand and accumulate

such basic information on eDNA for this species.

2.2. Materials and methods

2.2.1. Tank experiment

2.2.1.1. Experimental design

The experiments took place at the Maizuru Fisheries Research Station of Kyoto

University, Japan, which is in front of Maizuru Bay, from June 2016 to July 2017.

Polycarbonate 200-L tanks were assigned four water temperatures (13, 18, 23, and

28 °C) and three fish biomass levels (Small, Medium, and Large; see below for fish size

details), which resulted in twelve treatment levels in this study. Four temperature levels

were selected based on the preference temperature of target fish (i.e., around 20 °C;

17

Nakamura & Hamano, 2009) and within the range of bottom water temperature when

this species is recorded at the sampling site (Masuda, 2008). Four tank replicates were

prepared for each treatment level. Two experimental tanks were placed in each water

bath and heated using a 100 V-500 W heater (Mitsubishi, Japan). The temperature was

regulated using a thermostat (Nitto, Derthermo, Japan). The tanks were kept at a

constant water temperature throughout the experiment and were aerated using a pump.

The water temperature was measured every morning using a digital thermometer (Tetra,

Spectrum Brands Japan). Filtered seawater used in the experiment was pumped from 6

m depth off the Research Station where the water quality is scarcely impacted by

rainfall and other environmental factors. Before use, it was filtered by passing through

five different materials starting with coarse polyvinyl fabric (Saranlock OM-150, Asahi

Kasei, Japan) and ending with fine sand of around 0.6 mm in diameter (5G-ST, Nikkiso

Eiko, Japan). Inlet water was poured at a rate of 600 mL/min into each tank.

After the experimental tanks had been prepared, three Japanese jack

mackerels were added to each tank and they were left in the tank for about 1 week prior

to the experiments for the acclimation (Takahara et al., 2012; Sassoubre et al., 2016).

For the tank experiment using Medium-sized fish, all Japanese jack mackerels were

used only once. On the other hand, for the tank experiment using Large- and Small-

sized fish, some of the fish were used more than once. In these experimental periods, all

the fish which had survived were repeatedly used, and replacements were supplied for

the dead or dying fish. The fish were fed a small amount of krill every morning until the

day before water sampling. The bottom of each tank was cleaned an hour after feeding

18

to eliminate the effect of the feces, and, on the sampling day, the fish were starved. After

1 week, the Japanese jack mackerels were quickly removed from each tank, and their

total length (TL) and wet weight were measured. A water sample from each tank was

also collected (water sampling details are described below). The TLs and wet weights

for each fish biomass level were 6.2 ± 0.4cm and 2.3 ± 0.5 g (Small), 11.7 ± 1.2 cm and

13.4 ± 4.2 g (Medium), and 21.4 ± 3.1 cm and 106.5 ± 48.4 g (Large; both mean ± 1

SD; Table 2-1). There were no significant differences in TLs and wet weights among

fish within each fish size group (ANOVA, P > 0.1).

2.2.1.2. eDNA sampling

The eDNA was sampled using two different methods. The first method used a 47-mm-

diameter glass microfiber filter GF/F (nominal pore size 0.7 µm; GE Healthcare Life

Science, Little Chalfont, U.K.) to estimate eDNA shedding and decay rates, and the

second method used a series of 47-mm-diameter polycarbonate membrane filters (pore

size 10, 3, 0.8, and 0.4 or 0.2 µm; MILLIPORE, U.S.) to estimate eDNA size

distribution. Disposable gloves were worn when collecting water samples, and the

outside of the sampling bottles was washed with tap water after the samples were

collected. This was to prevent contamination during water sampling and filtration. The

filtering devices (i.e., filter funnels [Magnetic Filter Funnel, 500 mL capacity; Pall

Corporation, Westborough, MA, U.S.], plastic holders [ADVANTEC, Japan], nipple

joints [ADVANTEC, Japan], hoses [TOYOX, Japan], 1-L beakers, tweezers, and

sampling bottles used for water sampling) were bleached after every use in 0.1 %

19

sodium hypochlorite solution for at least 5 min.

eDNA sampling (i): estimation of eDNA shedding and decay rates

The aim of this sampling was to estimate Japanese jack mackerel eDNA shedding and

decay rates and to investigate how they were affected by water temperature and fish

biomass. The time just after removing the fish from each tank was defined as time 0,

and more than 1 L of water was collected from each tank at 0, 2, 4, 8, 16, 24, 48, 72,

and 96 hours (these time points are referred as time 0 to time 96). An additional water

sample was collected at 120 and 216 hours after time 0 (i.e., time 120 and 216) in the

tank experiments containing Medium-sized fish to measure eDNA persistence in the

tank, which was the first experimental period in the overall study. Water samples were

also collected the day before removing the fish from the tanks to measure the eDNA

concentrations at a steady state. This was defined as time before fish removal (i.e., time

bfr). The term “steady state” was defined as being when eDNA shedding was in

equilibrium with total eDNA degradation and dilution in each tank after the eDNA

concentration had stabilized (Sassoubre et al., 2016; Sansom & Sassoubre, 2017).

After water collection, the 1 L water samples were immediately filtered with a

GF/F filter. At each sampling time, 1 L of distilled water was also filtered as a filtration

negative control. Furthermore, 1 L of inlet water was sampled from each tank at time 24

to evaluate the background Japanese jack mackerel eDNA concentration in the inlet

water. Note that the experimental tanks were flown-through until removing the fish

from the tank, while inlet water was stopped once fish were removed. All filter samples

20

were kept at -20 °C after filtration until needed for eDNA extraction.

eDNA sampling (ii): estimation of eDNA size distribution

The aim in this sampling was to estimate the eDNA size distribution for Japanese jack

mackerels and to investigate the effect of temperature, fish biomass, and the time

passage on eDNA size distribution. Sequential filtration was performed using a

combination of plastic holders, nipple joints, and hoses. The water samples were 500

mL in volume, and they were filtered using four polycarbonate membrane filters (except

for the Large fish biomass level in the 28 °C treatment where 250 mL water samples

were taken due to filter clogging). At each sampling time, 500 mL of distilled water was

also sequentially filtered as a filtration negative control.

For all fish biomass levels, the water samples were collected at time bfr using

a series of polycarbonate membrane filters with 10, 3, 0.8, and 0.4 µm pore sizes. For

the Small and Large fish biomass tank experiments, the water samples were also

collected at time bfr, 0, 6, 12, and 18 using the same filters with 10, 3, 0.8, and 0.2 µm

pore sizes (Figure 2-1). All filter samples were kept at -20 °C until eDNA extraction.

2.2.1.3. DNA extraction

The total eDNA on each filter was extracted using a DNeasy Blood and Tissue Kit

(Qiagen, Hilden, Germany), and all eDNA extracts were placed in a freezer (-20 °C)

until quantitative PCR analysis. The DNA was extracted from the GF/F filters by a

method used in a previous study (Jo et al., 2017). Briefly, a filter sample was placed in

21

the suspended part of a Salivette tube (Sarstedt, Nümbrecht, Germany). Then, 420 µL of

a solution containing 20 µL proteinase K, 200 µL buffer AL, and 200 µL pure water was

placed on the filter and the tube was incubated at 56 °C for 30 min. After incubation, the

liquid held in the filter was collected by centrifugation at 5,000 g for 3 min. To increase

the eDNA yield, the filter was re-washed with 200 µL TE buffer for 1 min and the liquid

was again collected after centrifugation at 5,000 g for 3 min. Then, 500 µL ethanol was

added to the collected liquid and the mixture transferred to a spin column. Subsequently,

the total eDNA was eluted in 100 µL AE buffer following the manufacturer's

instructions.

The DNA was extracted from the polycarbonate membrane filters using a

DNeasy Blood & Tissue Kit with slight modifications to its protocol (Matsuhashi et al.,

unpublished). Briefly, tweezers were used to place a filter sample in a spin column.

Then, 320 µL of a solution containing 20 µL proteinase K, 150 µL buffer AL, and 150

µL TE buffer was added to the sample and the mixture was incubated it at 56 °C for 30

min. After incubation, 150 µL ethanol was added to the filter sample, and the mixture

centrifuged in a spin column at 6,000 g for 1 min. To increase the eDNA yield, the filter

was re-washed with a 300 µL solution that contained 100 µL TE buffer, 100 µL buffer

AL, and 100 µL ethanol, for 1 min, and then, the mixture was centrifuged 6,000 g for 1

min. The sample filter was removed from the spin column, and the total eDNA was

eluted in 100 µL AE buffer following the manufacturer's instructions.

22

2.2.1.4. Quantification of eDNA using qPCR

The amount of eDNA derived from Japanese jack mackerel at each time point was

evaluated by quantifying the CytB gene copy numbers using real-time TaqMan PCR

and the StepOnePlus Real-Time PCR system (Applied Biosystems, Foster City, CA,

U.S.). The primers/probe set in this study specifically amplified the Japanese jack

mackerel DNA and targeted a 127-bp fragment of the mitochondrial CytB gene

(Yamamoto et al., 2016). The number of Japanese jack mackerel CytB genes in each 2

µL eDNA solution sample was quantified by simultaneously performing qPCR using a

dilution series of standards containing 3 × 101 - 3 × 104 copies of a linearized plasmid

that contained synthesized artificial DNA fragments of the full CytB gene sequence for

Japanese jack mackerel (Jo et al., 2017). In addition, a 2 µL pure water sample was

analyzed as a PCR-negative control. Each 20 µL TaqMan reaction contained 2 µL DNA

extract, a final concentration of 900 nM of forward and reverse primers, and 125 nM of

TaqMan probe in 1 × TaqMan Gene Expression PCR Master Mix (Thermo Fisher

Scientific, Waltham, MA, U.S.). Quantitative PCR was performed with the following

conditions: 2 min at 50 °C, 10 min at 95 °C, 55 cycles of 15 s at 95 °C, and 1 min at

60 °C. All the qPCRs for eDNA extracts, standards, and negative controls were

performed in triplicate. The DNA concentrations in the water samples were calculated

by averaging the triplicate. All positive replicates were treated as having been

successfully quantified (i.e., no “limit of quantification” was set) following the previous

studies not setting the limit of quantification (Thomsen et al., 2012a; 2012b; Pilliod et

al., 2014; Minamoto et al., 2017a). Each replicate showing non-detection (PCR-

23

negative) was regarded as containing 0 copies (Ellison et al., 2006). PCR inhibition in

all PCR runs were not tested because it is unlikely that PCR inhibition occurred using

the water samples derived filtered seawater (Yamamoto et al., 2016; 2017).

2.2.2. Data analysis

R version 3.2.4 (R Core Team, 2016) was used to perform the statistical analyses. One

of the tanks containing Large fish at 28 °C was excluded from the statistical analysis

due to fish mortality. The statistical analyses are in detail described in the sections

below.

2.2.2.1. Environmental DNA shedding and decay rates

The Japanese jack mackerel eDNA decay rates were estimated from the eDNA decay

curves obtained from each experimental tank. Previous studies have estimated eDNA

decay rates by fitting an exponential decay model (Thomsen et al., 2012a; 2012b;

Eichmiller et al., 2016; Sassoubre et al., 2016; Minamoto et al., 2017a; Sansom &

Sassoubre, 2017; Tsuji et al., 2017) as follows:

"# = "%&'(#

where "# is the eDNA concentration at time ) (copies/L), "% is the eDNA

concentration at time 0, and * is the decay rate constant (/hour). After referring to Tsuji

et al. (2017), the model was extended to include the effect of water temperature and/or

fish biomass in the tank. The fitness of each regression model was then compared.

These models were as follows:

24

"# = "%&'(,-./)#

"# = "%&'(12./)#

"# = "%&'(,-.12./)#

where 3 is the water temperature (°C), 4 is the total wet weight of Japanese jack

mackerels in each 200 L tank (g/200 L), and 5, 6, and 8 are constants, which were

estimated by analyzing the nonlinear least-squares regression of the nls function in R.

The eDNA concentrations at each time point were adjusted by the eDNA concentration

at time 0 (i.e., "% in each tank was regarded as 1), and the total wet weight of Japanese

jack mackerels was log-transformed. The effects of water temperature and fish biomass

on the eDNA decay rate were investigated by comparing the four models using Akaike's

Information Criterion (AIC), and the model with the smallest AIC values was accepted

as the most supported model. The estimated parameters of this model were used to

calculate the eDNA decay rates at each treatment level.

Methods used in previous studies (Maruyama et al., 2014; Sassoubre et al.,

2016; Sansom & Sassoubre, 2017), with some modifications, were used to estimate

Japanese jack mackerel eDNA shedding rates per tank. This is expressed using the

following equation:

9 = :* +<

=> × "1@AB. × D

where 9 is the eDNA shedding rate in each tank (copies/hour), * is the estimated

eDNA decay rate in each tank (/hour; see above), "1@AB. is the eDNA concentration at a

steady state (i.e., at time bfr; copies/L), E is the flow rate of the inlet water (L/hour),

25

and D is the volume of the experimental tanks (L). Therefore, <= is the dilution rate in

the experimental tanks (/hour). This equation is derived from an ordinary differential

equation representing the change in the abundance of eDNA with time as follows:

DFG

F#= 9 − I × " × D

Briefly, at steady state (i.e., time bfr), it is assumed that eDNA shedding was

in equilibrium with total eDNA degradation and dilution (i.e., I = * + J

K) in each tank.

Thus, LMLN= 0 and 9 = I × " × D = P* + J

KQ × "1@AB. × D. The eDNA shedding rate

per fish body weight (copies/hour/g) was estimated by dividing the eDNA shedding

rates per tank by the total wet weight of the fish in the tank. These shedding rates were

log-transformed, and a two-way ANOVA and a post-hoc Tukey-Kramer test were

performed to investigate the effects of water temperature, fish size, and their interaction.

2.2.2.2. Environmental DNA size distribution

The eDNA concentrations in each size fraction were converted to a percentage of total

sequential filtration (%). The percentage of eDNA calculated above was arc-sin

transformed to reduce skewness and to meet the normality criteria (Cook & Heyse,

2000). Any eDNA particles smaller than 0.4 or 0.2 µm were not assessed because the

amount of eDNA in this size fraction seemed to be very small (Turner et al., 2014).

First, the samples that had passed through a sequential filter with 10, 3, 0.8,

and 0.4 µm pore sizes at time bfr were used to verify the effect of water temperature and

fish biomass on eDNA size distribution at the steady state. The Spearman's rank

26

correlation coefficients between the percentage of eDNA and water temperature at each

size fraction were calculated, where total fish biomass levels were not considered (i.e.,

these correlations were not analyzed at each fish biomass level). In addition, a one-way

ANOVA and a post-hoc Tukey-Kramer test were performed to verify the difference of

the percentage of eDNA among fish biomass levels at each size fraction, where

temperature levels were not considered (i.e., these tests were not analyzed at each

temperature level).

Second, the samples that had passed through a sequential filter with 10, 3, 0.8,

and 0.2 µm pore sizes at times bfr to 18 were used to compare the eDNA size

distribution at each time point. Wilcoxon's rank sum tests were performed between the

percentage of eDNA before and after removing the fish from the tanks (i.e., time bfr vs.

time 0) at each size fraction. In addition, the Spearman's rank correlation coefficients

were calculated between the percentage of eDNA and time point (time 0 to 18) at each

size fraction. For these analyses, all fish biomass and temperature levels were put

together. It was hypothesized that (a) eDNA size distribution would change before and

after the fish removal because the handling stress might lead the fish to shed more

DNA; and (b) eDNA size distribution would temporally change after the fish removal

because the persistence of eDNA might vary depending on the state and size of eDNA.

2.3. Results

In all the qPCR runs, the R2 values, slope, Y-intercept, and PCR efficiency of the

calibration curves were 0.994 ± 0.004, -3.467 ± 0.101, 42.650 ± 0.852, and 94.410 ±

27

3.821, respectively (mean ± 1 SD; Table 2-2). The amplification of target eDNA was

seen in some of inlet water samples and in the filtration negative controls. This means

that some contamination was mainly derived from the process of water filtering.

However, these copy numbers were much lower than in the samples taken from the

experimental tanks. Therefore, the Japanese jack mackerel eDNA in the inlet water and

low-level cross-contamination among samples is not likely to have affected the results.

2.3.1. Effect of water temperature and fish biomass on eDNA shedding and decay rates

The eDNA concentration at time 0 (when the fish were removed) increased by 10 to 100

times compared to the steady state (i.e., time bfr), which could be due to the handling

stress when removing the fish. After removal, the eDNA concentration decreased

exponentially (Figure 2-2). This tendency was consistently observed in all treatments.

The most supported model for the eDNA decay curves based on AIC values

was model 4, which included both water temperature and fish biomass in the tank as

explanatory variables (Table 2-3; "# = "%&'(,-.12./)#). The eDNA decay rates for

each treatment level were calculated based on these parameters, and the results showed

that Japanese jack mackerel eDNA decay increased as the temperature and fish biomass

in the experimental tanks rose (Table 2-4). The two-way ANOVA and post-hoc Tukey-

Kramer test results showed that both fish biomass and temperature significantly affected

eDNA shedding rates per each treatment (P < 0.05; Figure 2-3), and both partly affected

eDNA shedding rates per fish body weight (P < 0.05). Their interaction was not

significant (P > 0.1).

28

2.3.2. Effect of water temperature and fish biomass on eDNA size distribution

Japanese jack mackerel eDNA size distribution at the steady state varied depending on

water temperature and fish biomass. The 0.8 - 3 µm and 0.4 - 0.8 µm eDNA size

proportions showed significant positive correlations with water temperature (P < 0.01;

Figure 2-4), while there were no significant correlations between the percentage of

eDNA and water temperature at >10 µm and 3 - 10 µm size fraction (P > 0.05). Each

eDNA size fraction, except for the >10 µm size fraction, was significantly affected by

the three different fish biomass levels. The highest eDNA proportion was 3 - 10 µm for

the Medium fish size (P < 0.05), whereas it was 0.8 - 3 µm for the Small fish size (P <

0.05), and 0.4 - 0.8 µm for the Large fish size (P < 0.01; Figure 2-4). The difference of

the percentage of eDNA at >10 µm size fraction was not significant but marginal among

the three fish biomass levels (P = 0.0862), and the mean >10 µm eDNA proportion was

highest for the Large fish size (Figure 2-4).

2.3.3. Temporal dynamics of eDNA size distribution

The Japanese jack mackerel eDNA size distribution temporal change varied

considerably. At the steady state (i.e., time bfr), most of the eDNA was in the 3 - 10 µm

size fraction. Just after removing the fish from the tanks (i.e., time 0), the percentage of

eDNA in the >10 µm size fraction increased considerably, whereas the percentages of

eDNA at other size fractions decreased (Figure 2-5). Between time bfr and 0, there were

significant differences of the percentage of eDNA at all size fraction (P < 0.05; Figure

29

2-5). After time 0, the percentage of eDNA in the >10 µm size fraction was significantly

negatively correlated with sampling time (ρ = -0.4433, P < 0.0001), whereas the

percentages of eDNA in the 0.8 - 3 µm and 0.2 - 0.8 µm size fractions were significantly

positively correlated with sampling time (ρ = 0.2507, P < 0.01; ρ = 0.3000, P < 0.001,

respectively). There was no significant correlation between the percentage of eDNA at

the 3 - 10 µm size fraction and sampling time (P = 0.3297).

2.4. Discussion

2.4.1. Factors affecting the degradation of eDNA

The regression analysis results showed that higher water temperatures and higher fish

biomass accelerated eDNA degradation. These results supported previous studies that

had also shown water temperature-dependent degradation of eDNA (Strickler et al.,

2015; Eichmiller et al., 2016; Lance et al., 2017; Tsuji et al., 2017). However, this is the

first study to show that eDNA degradation is associated with fish biomass. It is also the

first to show the water temperature-dependent degradation of marine fish eDNA. As

moderately higher temperatures (<50 °C) stimulate microbial metabolism and

exonuclease activity (Corinaldesi et al., 2008; Poté et al., 2009), and high fish density

can lead to the increase in microbial activity (Barnes et al., 2014; Bylemans et al.,

2018a), these results are likely to support the hypothesis that the activity and abundance

of microbes and extracellular nucleases significantly affect eDNA degradation (Levy-

Booth et al., 2007; Nielsen et al., 2007; Barnes & Turner, 2016).

Several previous studies on eDNA decay rates addressed the persistence and

30

degradation of marine fish eDNA. Sassoubre et al. (2016) had a similar experimental

design to this study and targeted marine fish (Northern anchovy [Engraulis mordax],

Pacific sardine [Sardinops sagax], and Pacific chub mackerel [Scomber japonicus]).

They reported that eDNA decay rates were 0.055 to 0.101 (/hour), which was within the

range reported by the present study (0.035 to 0.485 [/hour]). The wider decay rate range

in the present study may be due to the effect of fish biomass in experimental tanks. In

Sassoubre et al. (2016), the density of three marine fish ranged from 0.2 to 2.0 g/L,

whereas the range was from 0.03 to 2.3 g/L in the present study. It would be common

that fish biomass affects eDNA degradation in seawater. Further study would be needed

to reveal the relationship between the abundance/ biomass of organisms and eDNA

concentrations.

2.4.2. Factors affecting the shedding of eDNA

The eDNA shedding rates varied according to fish biomass, which supports previous

studies (Doi et al., 2015; Klymus et al., 2015; Doi et al., 2016). It was not expected that

the eDNA shedding rates per fish body weight at some temperature levels were also

positively correlated with fish biomass, as the surface area per fish body weight was

negatively correlated with fish body weight (Bergmann, 1847). One explanation could

be the excessive effect of fish density in tanks, particularly for the Large fish biomass

level. For example, the fish might have touched each other more often or rubbed up

against the net in the tanks.

The present study demonstrated that eDNA shedding rate depended on water

31

temperature. Some studies have shown that eDNA concentration did not depend on

water temperature (Takahara et al., 2012; Klymus et al., 2015). However, they did not

estimate the true eDNA shedding rate (i.e., they estimated the accumulated amount of

eDNA) and thus could not divide the effects of eDNA shedding and degradation. It is

important to investigate how water temperature influences not only the amount of

eDNA detected in the field but also the eDNA shedding rate. As mentioned above, the

metabolism of fish greatly depends on water temperature (Person-Le Ruyet et al., 2004;

Morita et al., 2010), which means that high water temperatures can be stressful for fish

(Barton, 2002; Takahara et al., 2014). Therefore, Japanese jack mackerel eDNA

shedding rate would be expected to increase at around 20 °C, which is the optimal

temperature for this species (Nakamura & Hamano, 2009) or at 28 °C, which was the

highest water temperature in these experiments. The results showed that both eDNA

shedding rates per each treatment and per fish body weight tended to increase at higher

temperatures, which confirmed the above expectations.

2.4.3. Environmental DNA size distribution

The results showed that the percentage of eDNA at the 0.8 - 3 µm and 0.4 - 0.8 µm size

fractions increased with higher water temperatures. As the primers/probe set in this

study targeted mitochondrial DNA, the eDNA detected at these small size fractions was

considered to be mainly mitochondria itself (0.5 to 2 µm diameter; Wrigglesworth et al.,

1970; Ernster & Schatz, 1981) or extra-cellular DNA, rather than cell or tissue DNA.

One possible explanation is that microbial activity increases as water temperature

32

increases, allowing degradation of mitochondrial double cell membranes and the

mitochondrial DNA within. Furthermore, such a reduction of eDNA size with higher

temperature might contribute to the water temperature-dependent degradation of eDNA.

For example, the nominal pore size of the GF/F filter, which were used for estimation of

eDNA decay rates, was 0.7 µm, which means that the filter cannot capture eDNA

smaller than 0.7 µm. A decrease in the amount of eDNA larger than the filter pore size

as temperature increased might result in such water temperature-dependent degradation

of eDNA.

The results showed that the most abundant size fraction was 3 - 10 µm for the

Medium fish size, 0.8 - 3 µm for the Small fish size, and 0.4 - 0.8 µm for the Large fish

size. The percentage of eDNA at the >10 µm size fraction was not significantly, but

statistically marginally different among fish biomass levels. Such differences might

partly reflect the effect of fish density. For example, the percentage of eDNA at 0.4 - 0.8

µm was larger for Large size level than for other size levels, which might be caused by

the increase in microbial activity due to the increase in fish biomass in the tank. In

addition, this result might suggest that the eDNA origin, state, and their component ratio

could vary depending on fish biomass or, possibly, their development stage. Further

study would be needed to clarify the relationships between the developmental stage and

aforementioned eDNA characteristics.

The results showed that eDNA size distribution varied with time passage. At

first, the percentage of eDNA at >10 µm size fraction dramatically increased just after

the fish removal. Considering that such handling stress could cause the fish to shed

33

large-sized DNA, such as their scale and mucus (Merkes et al., 2014; Sassoubre et al.,

2016), this could be reasonable. In addition, the percentages of eDNA at small size

fractions increased with a time passage and that at >10 µm decrease. These temporal

shifts in eDNA size distribution to smaller size fractions might represent the dynamics

of eDNA described above. These results demonstrated that the states of eDNA changed

with time passage after it is released from organisms. Further study would be needed to

reveal the relationship between the persistence of eDNA and its state (i.e., intra-/extra-

cellular and within live/dead cells).

2.5. Conclusions

In conclusion, water temperature and fish biomass facilitated eDNA shedding and

degradation. The higher eDNA decay rates with larger biomass could reflect the activity

and abundance of microbes and extra-organism nucleases in the water, and the higher

eDNA shedding rates with higher temperature might be due to higher metabolism and

physiological activity of organisms. In addition, eDNA size distribution also varied

depending on water temperature, fish biomass, and time passage. The increases of

smaller sized fractions of eDNA with higher temperature and the difference in eDNA

size distribution among fish biomass might reflect the microbial activity in the water.

Furthermore, the temporal changes of eDNA size distribution showed that the state of

eDNA could vary with time passage due to degradation caused by various

environmental factors after release into the environment.

Although this study clarified some of the eDNA dynamics, the research area

34

needs further study. For example, although the findings imply that microbes and extra-

organism nucleases are involved in eDNA degradation, and that metabolism affects the

eDNA shedding rate, these aspects were not demonstrated directly in this study. In

addition, the effect of seasonal change in the seawater (e.g., nutrient load, salinity,

chlorophyll) could not be assessed despite the experimental periods over different

season. There is therefore a possibility that certain chemical and microbial conditions

could influence the behavior of fish individuals as well as that of eDNA, and these

could be subjects of future studies. Moreover, there has been little research on the

physiological source of eDNA production and the physical aspects of eDNA such as its

structure and length (Barnes & Turner, 2016). A greater understanding and

accumulation of basic information on eDNA would improve eDNA analysis and enable

researchers to maximize the potential of future eDNA applications. This study would

lay a groundwork that can be used in further eDNA research.

35

2.6. Tables

Table 2-1. Total lengths (TL) and weights of all Japanese jack mackerel used in the tank

experiments.

Fish size Temperature Tank Total length (cm) Wet weight (g)

Fish1 Fish2 Fish3 Mean Fish1 Fish2 Fish3 Mean

Small 13 °C Tank1 6.4 7.0 5.9 6.4 2.4 3.0 2.1 2.5 Small 13 °C Tank2 6.4 6.3 6.3 6.3 2.3 2.3 2.2 2.2 Small 13 °C Tank3 6.2 6.0 5.5 5.9 2.0 2.4 2.1 2.1 Small 13 °C Tank4 6.5 6.0 6.2 6.2 2.4 1.9 2.2 2.2 Small 18 °C Tank1 7.1 6.5 5.6 6.4 3.1 2.7 1.6 2.5 Small 18 °C Tank2 6.4 6.5 6.6 6.5 2.2 2.4 2.6 2.4 Small 18 °C Tank3 5.5 6.2 6.2 6.0 1.7 1.9 2.2 1.9 Small 18 °C Tank4 6.5 5.5 6.2 6.1 2.3 1.4 2.2 2.0 Small 23 °C Tank1 5.5 6.5 6.1 6.0 1.8 2.1 2.1 2.0 Small 23 °C Tank2 6.7 6.1 5.5 6.1 3.2 2.0 2.0 2.4 Small 23 °C Tank3 6.2 6.5 6.7 6.5 2.8 2.1 3.1 2.7 Small 23 °C Tank4 7.1 5.6 6.4 6.4 3.0 1.9 3.0 2.6 Small 28 °C Tank1 5.5 6.1 6.8 6.1 1.5 2.4 2.5 2.1 Small 28 °C Tank2 6.5 5.8 5.7 6.0 2.5 2.3 1.3 2.0 Small 28 °C Tank3 6.0 6.5 5.9 6.1 2.2 3.1 2.8 2.7 Small 28 °C Tank4 6.7 5.6 5.8 6.0 2.8 1.9 1.8 2.1

Medium 13 °C Tank1 11.0 12.5 11.1 11.5 11.1 14.0 10.3 11.8 Medium 13 °C Tank2 9.5 13.3 13.9 12.2 6.1 18.5 22.8 15.8 Medium 13 °C Tank3 12.1 12.0 13.1 12.4 11.9 15.4 17.9 15.1 Medium 13 °C Tank4 11.3 11.6 9.9 10.9 10.7 13.2 7.3 10.4 Medium 18 °C Tank1 12.2 13.0 12.5 12.5 16.1 18.9 15.5 16.8 Medium 18 °C Tank2 12.4 12.5 10.6 11.8 15.2 16.5 9.2 13.6 Medium 18 °C Tank3 12.9 13.2 11.6 12.6 15.9 19.5 12.7 16.0 Medium 18 °C Tank4 11.7 9.9 12.4 11.3 13.6 8.2 17.4 13.1 Medium 23 °C Tank1 12.3 13.4 10.5 12.0 16.9 19.3 10.8 15.7 Medium 23 °C Tank2 12.5 11.2 11.3 11.6 14.8 11.1 12.1 12.6 Medium 23 °C Tank3 12.9 11.5 10.7 11.7 16.4 10.9 8.5 12.0 Medium 23 °C Tank4 12.2 12.6 10.4 11.7 17.1 18.0 9.1 14.7 Medium 28 °C Tank1 9.5 12.4 10.0 10.6 6.6 17.1 8.8 10.8 Medium 28 °C Tank2 12.6 11.4 11.2 11.7 17.8 10.8 11.9 13.5 Medium 28 °C Tank3 12.2 13.3 11.7 12.4 16.2 17.4 12.0 15.2 Medium 28 °C Tank4 11.5 9.5 10.0 10.3 10.1 6.0 7.0 7.7

36

Large 13 °C Tank1 26.1 25.8 19.5 23.8 174.4 150.0 72.9 132.4 Large 13 °C Tank2 20.2 23.3 24.5 22.7 76.3 114.0 140.1 110.1 Large 13 °C Tank3 23.2 21.1 18.9 21.1 127.3 83.8 60.8 90.6 Large 13 °C Tank4 23.7 23.0 20.4 22.4 145.5 115.5 86.6 115.9 Large 18 °C Tank1 21.6 23.0 21.2 21.9 96.4 121.0 87.1 101.5 Large 18 °C Tank2 19.8 25.5 25.2 23.5 87.6 172.2 171.4 143.7 Large 18 °C Tank3 18.2 26.1 21.3 21.9 65.4 172.6 93.8 110.6 Large 18 °C Tank4 24.5 20.6 17.9 21.0 133.8 83.8 52.7 90.1 Large 23 °C Tank1 19.6 23.3 17.4 20.1 75.6 111.0 59.1 81.9 Large 23 °C Tank2 19.0 23.8 18.3 20.4 70.1 132.4 80.2 94.2 Large 23 °C Tank3 20.6 24.2 20.6 21.8 91.3 146.7 89.3 109.1 Large 23 °C Tank4 19.5 21.5 17.7 19.6 77.9 105.6 68.2 83.9 Large 28 °C Tank1 Large 28 °C Tank2 17.5 22.9 14.7 18.4 60.0 130.5 40.0 76.8 Large 28 °C Tank3 23.5 21.3 16.5 20.4 150.0 110.0 50.0 103.3 Large 28 °C Tank4 29.5 22.0 16.0 22.5 310.0 110.1 40.0 153.4

37

Table 2-2. R2 values, slopes, and Y intercepts of the calibration curves, and the PCR

efficiencies (mean ± 1 SD) for each qPCR experiment performed in this study.

Fish size N R2 Slope Y-intercept PCR efficiency

Small 26 0.994 ± 0.004 -3.473 ± 0.076 42.446 ± 1.055 94.159 ± 2.848 Medium 11 0.994 ± 0.003 -3.479 ± 0.149 42.534 ± 0.873 94.154 ± 5.504 Large 30 0.994 ± 0.004 -3.472 ± 0.079 42.958 ± 0.439 94.179 ± 2.943

Note: N means the number of PCR plates.

38

Table 2-3. The eDNA decay curve model results for the tank experiments estimated by the nls function in R.

Model C0 b c a AIC ⊿AIC

C(t)=C0*exp(b*t) 0.9590 *** -0.1876 *** 70.5144 127.4962

C(t)=C0*exp{(b*T + a)*t} 0.9737 *** -0.0176 *** 0.1415 *** 6.1045 63.0863

C(t)=C0*exp{(c*D + a)*t} 0.9455 *** -0.1004 *** -0.0260 48.9897 105.9715

C(t)=C0*exp{(b*T + c*D + a)*t} 1.0029 *** -0.0173 *** -0.1027 *** 0.2732 *** -56.9818 0.0000

Note. The AIC values (bold) were used to identify the most supported model for the eDNA decay curves. Asterisks *** show the

significant effects (P < 0.001) of each parameter. The best model included both water temperature (T) and fish density (D, log-transformed) as explanatory variables, which indicated that both water temperature and fish density influence eDNA degradation.

39

Table 2-4. Japanese jack mackerel eDNA decay rate results when estimated by the best

model.

eDNA decay rates (/hour)

Temperature Small Medium Large

13 °C 0.0372 ± 0.0028 0.1154 ± 0.0077 0.2110 ± 0.0061

18 °C 0.1219 ± 0.0049 0.2074 ± 0.0047 0.2969 ± 0.0077

23 °C 0.2126 ± 0.0052 0.2903 ± 0.0049 0.3753 ± 0.0051

28 °C 0.2959 ± 0.0052 0.3689 ± 0.0115 0.4686 ± 0.0126

Note: Values for eDNA decay rates are the mean ± 1 SD (average of four tank

replicates, except for the Large size at 28 °C). Note that the treatment of 28 °C -Large

fish biomass level had only three tank replicates due to fish mortality.

40

2.7. Figures

Figure 2-1. Diagram showing eDNA sampling for the estimation of eDNA size

distribution. Targeting all fish biomass levels, water samples were filtered only at time

bfr using a series of polycarbonate membrane filters with 10, 3, 0.8, and 0.4 µm pore

size. Besides, targeting Small and Large fish biomass levels, water samples were

temporally filtered at time bfr, 0, 6, 12, 18 using same filters with 10, 3, 0.8, and 0.2 µm

pore size.

41

Figure 2-2. Decay curves for Japanese jack mackerel eDNA in the experimental tanks. Dots show eDNA concentrations per liter of tank

water at each time point (Small: square, Medium: circle, Large: triangle; average of four tank replicates, except for the Large at 28 °C).

Error bars show the standard deviations (SD).

0 100 200

050000

100000

150000

200000

Small

0 100 200

0500000

100000015000002000000 Medium

0 100 2000e+00

1e+07

2e+07

3e+07

4e+07 Large

13°C

eDN

A c

onc.

[cop

ies/

L ta

nk w

ater

]

0 100 200

050000

150000

250000

Small

0 100 2000.0e+00

5.0e+06

1.0e+07

1.5e+07 Medium

0 100 2000.0e+00

5.0e+07

1.0e+08

1.5e+08

2.0e+08 Large

18°C

eDN

A c

onc.

[cop

ies/

L ta

nk w

ater

]

0 100 2000e+00

2e+05

4e+05

6e+05 Small

0 100 2000e+00

2e+06

4e+06

6e+06 Medium

0 100 2000e+00

1e+07

2e+07

3e+07

4e+07

5e+07 Large

23°C

eDN

A c

onc.

[cop

ies/

L ta

nk w

ater

]

0 100 200

050000

100000

150000

Small

0 100 200

0500000

1000000

1500000 Medium

0 100 2000e+00

2e+07

4e+07

6e+07

8e+07

1e+08 Large

28°C

time point [hour]

eDN

A c

onc.

[cop

ies/

L ta

nk w

ater

]

0 100 200

050000

100000

150000

200000

Small

0 100 200

0500000

100000015000002000000 Medium

0 100 2000e+00

1e+07

2e+07

3e+07

4e+07 Large

13°C

eDN

A c

onc.

[cop

ies/

L ta

nk w

ater

]

0 100 200

050000

150000

250000

Small

0 100 2000.0e+00

5.0e+06

1.0e+07

1.5e+07 Medium

0 100 2000.0e+00

5.0e+07

1.0e+08

1.5e+08

2.0e+08 Large

18°C

eDN

A c

onc.

[cop

ies/

L ta

nk w

ater

]

0 100 2000e+00

2e+05

4e+05

6e+05 Small

0 100 2000e+00

2e+06

4e+06

6e+06 Medium

0 100 2000e+00

1e+07

2e+07

3e+07

4e+07

5e+07 Large

23°C

eDN

A c

onc.

[cop

ies/

L ta

nk w

ater

]

0 100 200

050000

100000

150000

Small

0 100 200

0500000

1000000

1500000 Medium

0 100 2000e+00

2e+07

4e+07

6e+07

8e+07

1e+08 Large

28°C

time point [hour]

eDN

A c

onc.

[cop

ies/

L ta

nk w

ater

]

0 100 200

050000

100000

150000

200000

Small

0 100 200

0500000

100000015000002000000 Medium

0 100 2000e+00

1e+07

2e+07

3e+07

4e+07 Large

13°C

eDN

A c

onc.

[cop

ies/

L ta

nk w

ater

]

0 100 200

050000

150000

250000

Small

0 100 2000.0e+00

5.0e+06

1.0e+07

1.5e+07 Medium

0 100 2000.0e+00

5.0e+07

1.0e+08

1.5e+08

2.0e+08 Large

18°C

eDN

A c

onc.

[cop

ies/

L ta

nk w

ater

]

0 100 2000e+00

2e+05

4e+05

6e+05 Small

0 100 2000e+00

2e+06

4e+06

6e+06 Medium

0 100 2000e+00

1e+07

2e+07

3e+07

4e+07

5e+07 Large

23°C

eDN

A c

onc.

[cop

ies/

L ta

nk w

ater

]

0 100 200

050000

100000

150000

Small

0 100 2000

500000

1000000

1500000 Medium

0 100 2000e+00

2e+07

4e+07

6e+07

8e+07

1e+08 Large28°C

time point [hour]

eDN

A c

onc.

[cop

ies/

L ta

nk w

ater

]

42

Figure 2-3. Results for eDNA shedding rate per treatment (upper) and per fish body

weight (lower). Both boxplots show the comparison of eDNA shedding rates among

four temperature and three biomass levels (average of four tank replicates, except for

the Large size at 28 °C). Factor levels with different letters are statistically significantly

different (P < 0.05) based on post-hoc Tukey-Kramer tests.

56

78

910

a b c a bc c ab bc c a bc c

S M L S M L S M L S M L

13°C 18°C 23°C 28°Cwater temperature

log1

0(sh

eddi

ng ra

te p

er tr

eatm

ent)

[cop

ies/

hour

]

45

67

8

a abcd abcd abc abcd abcd abc cd abcd ab bcd d


13°C 18°C 23°C 28°Cwater temperature

log1

0(sh

eddi

ng ra

te p

er fi

sh b

ody

wei

ght)

[cop

ies/

hour

/g]

43

Figure 2-4. Results for eDNA size distributions at the steady state. Upper boxplots show

a comparison between the four water temperature levels (13, 18, 23, and 28 °C) when

all fish biomass levels (Small, Medium, and Large) are combined. The lower boxplots

show comparisons between the three fish biomass levels when all water temperature

levels are combined. Stars show significant differences in the percentage of eDNA for

each fish biomass level based on post-hoc Tukey-Kramer tests. Only significant

correlations (P < 0.05) are shown in the boxplots.

13 18 23 28

020

4060

80100

>10 µm

13 18 23 28

020

4060

80100

3-10 µm

13 18 23 28

020

4060

80100

0.8-3 µm

rho = 0.5263

13 18 23 28

020

4060

80100

0.4-0.8 µm

rho = 0.4089

pore size

water temperature [°C]

eDN

A c

onc.

[%]

S M L

020

4060

80100

>10 µm

n.s.

S M L

020

4060

80100

3-10 µm

**

S M L

020

4060

80100

0.8-3 µm

**

S M L

020

4060

80100

0.4-0.8 µm

**

pore size

fish size

eDN

A c

onc.

[%]

13 18 23 28

020

4060

80100

>10 µm

13 18 23 28

020

4060

80100

3-10 µm

13 18 23 28

020

4060

80100

0.8-3 µm

rho = 0.5263

13 18 23 280

2040

6080

100

0.4-0.8 µm

rho = 0.4089

pore size

water temperature [°C]

eDN

A c

onc.

[%]

S M L

020

4060

80100

>10 µm

n.s.

S M L

020

4060

80100

3-10 µm

**

S M L

020

4060

80100

0.8-3 µm

**

S M L

020

4060

80100

0.4-0.8 µm

**

pore size

fish size

eDN

A c

onc.

[%]

44

Figure 2-5. Result for temporal dynamics of eDNA size distribution from time bfr

(bright pink) to 18 (dark pink). Sum of the same colors at each pore size gives 100 %.

Boxplots show the temporal dynamics for the different Japanese jack mackerel eDNA

percentages at each pore size when all water temperature levels (13, 18, 23, and 28 °C)

and fish biomass levels (Small, Medium, and Large) are combined. The figure “-24”

below means the time bfr. Stars show significant differences (P < 0.05) in the eDNA

concentration proportions between time bfr and 0. Only significant correlations

(positive in red and negative in blue) from time 0 to 18 are shown in the boxplots.

-24 6 18

020

4060

80100

>10 µm

* rho = -0.4433

-24 6 18

020

4060

80100

3-10 µm

*

-24 6 180

2040

6080

100

0.8-3 µm

* rho = 0.2507

-24 6 18

020

4060

80100

0.2-0.8 µm

* rho = 0.3000

pore size

time point [hour]

eDN

A c

onc.

[%]

45

Chapter 3. Estimating shedding and decay rates of environmental nuclear DNA with

relation to water temperature and biomass.

3.1. Introduction

During the last decade, environmental DNA (eDNA) analysis has been developed as a

novel tool for the assessment and management of aquatic ecosystems (Ficetola et al.,

2008; Minamoto et al., 2012; Taberlet et al., 2012; Bohmann et al., 2014; Thomsen &

Willerslev, 2015). Organisms release DNA into the environment in the form of mucus,

feces, scales, and gametes (Martellini et al., 2005; Merkes et al., 2014; Sassoubre et al.,

2016; Bylemans et al., 2017), and this genetic material is called eDNA. The analysis of

eDNA has enabled us to obtain information on species distribution and composition

quickly, extensively, and non-invasively (Biggs et al., 2015; Fukumoto et al., 2015;

Balasingham et al., 2017; Yamamoto et al., 2017).

To date, most eDNA analyses relating to macro-organisms have targeted

mitochondrial DNA (mtDNA) as a genetic marker (Ficetola et al., 2008; Takahara et al.,

2012; Goldberg et al., 2013; Dougherty et al., 2016; Ushio et al., 2018). This is mainly

because a single cell has multiple mitochondrial genomes (tens to thousands of mtDNA

copies), contrary to the nuclear genome (Robin & Wong, 1988; Foran, 2006). However,

some studies have suggested the use of nuclear DNA (nuDNA) markers, particularly the

markers targeting multiple copies of ribosomal RNA genes such as internal transcribed

spacer (ITS) regions, and reported that the regions could be sensitive genetic markers

for eDNA analyses (Minamoto et al., 2017b; Dysthe et al., 2018; Gantz et al., 2018).

46

The mtDNA copy numbers per cell may vary depending on individual body condition

and cell type, whereas those of nuDNA do not depend on such factors (Long & Dawid,

1980). In addition, some regions of nuDNA have high interspecific variation (Booton et

al., 1999), which could be useful for distinguishing closely related species via eDNA

analysis. Therefore, testing the availability of nuDNA marker is important for the

expansion of eDNA applicability in the field.

Given the possibility and prospect of using nuDNA marker in eDNA analyses,

it is important to understand the characteristics and dynamics of nuclear and

mitochondrial eDNA (nu-eDNA and mt-eDNA, respectively). For example, some

studies have examined how various environmental factors may influence the shedding

and degradation of eDNA (Strickler et al., 2015; Barnes & Turner, 2016; Hansen et al.,

2018). For mt-eDNA, previous studies reported that its shedding is mainly affected by

the biomass/abundance of organisms and temperature (Takahara et al., 2012; Maruyama

et al., 2014; Klymus et al., 2015; Jo et al., 2019a), whereas eDNA degradation is

affected by different water chemistries, temperature, and microbial activity (Barnes et

al., 2014; Strickler et al., 2015; Eichmiller et al., 2016; Seymour et al., 2018; Jo et al.,

2019a). Although some studies have examined the detectability, amount, and persistence

of eDNA among different DNA markers (Bylemans et al., 2017; Minamoto et al.,

2017b; Bylemans et al., 2018a; Gantz et al., 2018), the influence of environmental

factors on the shedding and degradation of nu-eDNA has not been formally evaluated,

and such information is needed to evaluate the feasibility of using nu-eDNA in future

studies.

47

Using Japanese jack mackerel (Trachurus japonicus) as a model species, the

present study estimated the shedding and decay rates of nu-eDNA and investigated the

effects of water temperature, biomass of organisms, and type of DNA marker (nuclear

or mitochondrial) on eDNA shedding and degradation. For this, a novel primers/probe

set that specifically amplified a nuDNA fragment of Japanese jack mackerel, an

economically important marine fish in East Asia (Zhang & Lee, 2001; Sassa & Konishi,

2006), was first developed. Considering that a large proportion of eDNA exists as intra-

cellular DNA, such as cell and tissue fragments in water (Turner et al., 2014; Jo et al.,

2019a), it was expected that the tendencies of eDNA shedding and degradation would

be similar between nu- and mt-eDNA.


3.2.1. Experimental design

All extracted eDNA samples used were from Jo et al. (2019a). Briefly, 200-L acrylic

tanks were assigned to four water temperatures (13, 18, 23, and 28 °C) and three fish

biomass levels (Small, Medium, and Large), resulting in twelve treatment levels (Figure

3-1). Four tank replicates were prepared per treatment. Temperature levels were set

based on previous studies that reported the range of water temperature when the species

was recorded at the sampling site (Masuda, 2008) and the preferred temperature of

model species (i.e., around 20 °C; Tsuchida, 2002; Nakamura & Hamano, 2009). Fish

biomass levels were determined by the difference in fish body size. All tanks housed the

same number of fish individuals. Water temperature was kept constant for each tank

48

throughout the experiment. All tanks were individually aerated using a pump and flown-

through until the fish were removed from the tank. Filtered seawater was pumped from

a 6 m depth at the Research Station and used as the inlet water for each tank (flow

velocity: 600 mL/min).

Three Japanese jack mackerels were added to each tank and kept there for 1

week. A small amount of krill was used to feed the fish every morning until the day

before water sampling. The bottom of each tank was cleaned an hour after feeding to

remove the effect of the feces from the analyses. The fish were starved on the sampling

day. Total lengths (TLs) and wet weights of the fish were 6.2 ± 0.4 cm and 2.3 ± 0.5 g

(Small), 11.7 ± 1.2 cm and 13.4 ± 4.2 g (Medium), and 21.4 ± 3.1 cm and 106.5 ± 48.4

g (Large) (mean ± 1 SD). In addition, the age of each fish was estimated by the growth

model for Japanese jack mackerel (Mitani & Ida, 1964). Ages were 0.16 ± 0.02 year

(Small), 0.37 ± 0.06 year (Medium), and 1.04 ± 0.26 year (Large) (mean ± 1 SD).

3.2.2. eDNA sampling and extraction

After a 1-week acclimation period, the fish were quickly and carefully removed from

each tank using a net. Flow-through was switched off after removing the fish. The time

point immediately after removing the fish from each tank was defined as time 0. Water

samples were collected with plastic bottles from the tanks 0, 2, 4, 8, 16, 24, 48, 72, and

96 hours after time 0; these time points are referred as times 0 to 96. At each time point,

1 L of rearing water was collected from each tank and filtered using a 47-mm diameter

glass microfiber filter GF/F (nominal pore size 0.7 µm; GE Healthcare Life Science,

49

Little Chalfont, U.K.). Water samples were also collected the day before time 0, which

was defined as the time before fish removal (time bfr), to measure eDNA concentrations

at a steady state (i.e., the time at which eDNA shedding was in equilibrium with total

eDNA degradation; Sassoubre et al., 2016; Jo et al., 2019a). Besides, 1 L of distilled

water was filtered at each time point as a filtration negative control, and 1 L of inlet

water put into each tank was filtered at time 24, when flow-through had already been

switched off, to evaluate the background Japanese jack mackerel eDNA concentration

in it.

Disposable gloves were used during water samplings, and the outer part of

sampling bottles was washed with tap water after water samplings. Filtering devices

(i.e., filter funnels [Magnetic Filter Funnel, 500 ml capacity; Pall Corporation,

Westborough, MA, U.S.], 1 L beakers, tweezers, and plastic bottles) used for water

sampling were bleached after every use in 0.1 % sodium hypochlorite solution for at

least 5 min (Yamanaka et al., 2017). All filter samples were kept at -20 °C until DNA

extraction. Total DNA from each filter was extracted using a DNeasy Blood and Tissue

Kit (Qiagen, Hilden, Germany) (Jo et al., 2019a). All eDNA samples were kept at -

20 °C until quantitative PCR analysis.

3.2.3. Primers and probe development

A novel primers/probe set that specifically amplified the DNA fragment of the nuclear

internal transcribed spacer-1 (ITS1) region of Japanese jack mackerel were designed.

This region was targeted because from tens to tens of thousands of copies of ribosomal

50

RNA genes, including ITS1 regions, are present in the nuclear genome (Prokopowich et

al., 2003), which is fixed regardless of the individual's body condition or cell type

(Long & Dawid, 1980; Hillis & Dixon, 1991). Because of the paucity of publicly

available sequence data for this region, the ITS1 region of the model species and related

species from Maizuru Bay (Amberfish [Decapterus maruadsi], Amberjack [Seriola

quinqueradiata], and Greater amberjack [Seriola dumerili]) were sequenced and used as

reference sequences for primers and probe development. In addition, eight Japanese

jack mackerels were newly captured in the west Maizuru Bay (Nagahama, Maizuru,

Kyoto, Japan; 35°29′N and 135°22′E) in June 2018, and tissue samples were collected.

Tissue samples of the related fishes were obtained from the fish collection of Kyoto

University (FAKU). Total DNA was extracted from the tissues using the DNeasy Blood

and Tissue Kit following manufacturers’ guidelines. These DNA extracts were amplified

in a Veriti Thermal Cycler (Applied Biosystems) using the universal ITS1 primer pair

(forward primer: 5′-TCCGTAGGTGAACCTGCGG-3′; reverse primer: 5′-

CGCTGCGTTCTTCATCG-3′), which was designed to amplify the ITS1 region of a

wide variety of marine animals (Chow et al., 2009). Each 25 µL PCR reaction contained

2 µL of DNA extract, 0.4 µM of each primer, 0.1 mM of dNTPs, and 1 U of ExTaqTM

DNA polymerase (Takara Bio, Tokyo, Japan) in 1 × ExTaq Buffer (Takara Bio, Tokyo,

Japan). PCR was performed with the following conditions: 2 min at 94 °C, 55 cycles of

30 s at 96 °C, 30 s at 50 °C, and 1.5 min at 72 °C, and 7 min at 72 °C. PCR products

were visualized using electrophoresis on 1.5 % agarose gels stained with Midori Green

(NIPPON Genetics Co, Ltd., Japan). The agarose gels with the band of target length

51

were then purified using Wizard® SV Gel and PCR Clean-Up System (Promega,

Madison, U.S.). These products were commercially Sanger-sequenced using the 3130xl

Genetic Analyzer (Applied Biosystems) and BigDye Terminator v3.1 (Thermo Fisher

Scientific, Waltham, MA, U.S.) to obtain the reference sequences for the primers and

probe development.

Using the produced sequences and those available in the National Center for

Biotechnology Information (NCBI) database (Table 3-1), a species-specific

primers/probe set was designed using Primer Express 3.0 (Thermo Fisher Scientific)

with default settings. In vitro specificity of the assay was then checked using the

StepOnePlus Real-Time PCR system (Applied Biosystems). Each 20 µL TaqMan

reaction contained 100 and 10 pg of template DNA (from one individual of Japanese

jack mackerel or of a related species described above), a final concentration of 900 nM

of forward and reverse primers, and 125 nM TaqMan probe in a 1 × TaqMan Gene

Expression PCR Master Mix (Thermo Fisher Scientific). PCR was performed with the

following conditions: 2 min at 50 °C, 10 min at 95 °C, and 55 cycles of 15 s at 95 °C

and 1 min at 60 °C. A 2 µL pure water sample was simultaneously analyzed as a PCR

negative control.

3.2.4. Quantification of eDNA samples

The amount of Japanese jack mackerel nu-eDNA in water samples was evaluated by

quantifying the ITS1 region copy number using the StepOnePlus Real-Time PCR

system. Each 20 µL TaqMan reaction contained 2 µL of template DNA, a final

52

concentration of 900 nM of forward and reverse primers, and 125 nM of TaqMan probe

in a 1 × TaqMan Gene Expression PCR Master Mix. Thermal conditions of the

quantitative real-time PCR were the same as described above. The ITS1 region copy

number in each 2 µL template DNA was quantified by simultaneously performing a

qPCR with a dilution series of standards containing 3 × 101 - 3 × 104 copies of a

linearized plasmid that contained synthesized artificial DNA fragments of the partial

sequence of the ITS1 region (237 bp) of Japanese jack mackerel. A negative PCR

control was included by simultaneously analyzing 2 µL of pure water. All quantitative

mt-eDNA data used for comparisons were obtained from Jo et al. (2019a). All qPCRs

for eDNA extracts, standards, and negative controls were performed in triplicate, and

the eDNA concentrations were calculated by averaging the triplicate. Each PCR

negative replicate (indicating non-detection) was regarded as containing zero copies

(Ellison et al., 2006).

3.2.5. Statistical analyses

R version 3.2.4 (R Core Team, 2016) was used for all statistical analyses. One of the

tanks (treatment 28 °C/Large fish biomass level) was excluded from all analyses

because of fish mortality. Japanese jack mackerel eDNA decay rates were first estimated

using the time-series change of their eDNA concentrations after fish removal from each

tank. Previous studies estimated eDNA decay rates by fitting a first-order exponential

decay model (Thomsen et al., 2012; Eichmiller et al., 2016; Sassoubre et al., 2016;

Minamoto et al., 2017a; Collins et al., 2018) as follows:

53

#$ = #&'()$

where #$ is eDNA concentration at time * (copies/L), #& is eDNA concentration at

time 0, and + is the decay rate constant (/hour). This model was expanded to include

the effects of water temperature and total fish biomass in the tank (Jo et al., 2019a) as

follows:

#$ = #&'((-./01/2)$

where 4 is water temperature (°C), 5 is total wet weight of Japanese jack mackerels

in each 200-L tank (log-transformed, g/200 L), and 6, 7, and 8 are constants

estimated using the nonlinear least-squares regression of the function nls in the R

software. The eDNA concentrations at each time point were adjusted with those at time

0 (i.e., #& was regarded as 1). All eDNA samples whose concentrations were below

one copy per reaction (Takahara et al., 2012; Doi et al., 2017; Katano et al., 2017) were

excluded from model fitting. In addition, the eDNA samples with concentrations below

the background eDNA signal, as measured from the inlet water, were excluded from

model fitting. Using these parameters and constants, nu-eDNA and mt-eDNA decay

rates were, respectively, calculated for each tank.

Japanese jack mackerel eDNA shedding rates per treatment were then

estimated following Jo et al. (2019a). The ordinary differential equation was assumed to

represent the change with time of eDNA abundance in the tank (Thomsen et al., 2012;

Maruyama et al., 2014; Sassoubre et al., 2016) as follows:

9:;

:$= < − > × # × 9

where 9 is the volume of the tank (L), # is eDNA concentration from Japanese jack

54

mackerel (copies/L), < is eDNA shedding rate (copies/ hour), and > is total eDNA

degradation rate (/hour). > included eDNA decay rates estimated above (+) and eDNA

dilution rates resulting from a flow-through system (i.e., @A; B is the flow rate of the

inlet water [L/hour]). At steady state (i.e., time bfr), eDNA shedding (<) was assumed to

be in equilibrium with total eDNA degradation (> = + +D

E), which resulted in FG

FH= 0.

The equation above can therefore be expressed as follows:

< = J+ +@

AK × #0LMN. × 9

where #0LMN. is eDNA concentrations at time bfr (copies/L). Using this equation,

eDNA shedding rates were calculated for each tank. A three-way ANOVA was

performed to investigate the effects of temperature (°C), fish biomass (Small, Medium,

and Large), type of DNA markers, and their interaction on eDNA shedding rates, where

eDNA shedding rates were log-transformed to reduce skewness.

Furthermore, B6*PQR and B6*PQ; were calculated for each tank as follows:

B6*PQR =SQT10(<QVW* − '5XY)

SQT10(<QVZ[ − '5XY)

B6*PQ; =SQT10(#0LMN.QVW* − '5XY)

SQT10(#0LMN.QVZ[ − '5XY)

where B6*PQR was the ratio between the nu- and mt-eDNA shedding rates, and

B6*PQ;was the ratio between nu- and mt-eDNA concentrations at time bfr. Although a

single-copy nuclear gene would be more suitable, it would be difficult to detect and

quantify the single-copy nuDNA in water samples. Thus, the indices using the copy

number of the ITS1 region instead, whose copy number is fixed among cells, were

55

assumed to be a measurement of the amount of mtDNA per cell. The Kruskal–Wallis

rank sum test and post-hoc Wilcoxon rank sum test with Bonferroni adjustment were

performed to investigate the effects of fish biomass on the ratios of eDNA shedding and

concentration (B6*PQR and B6*PQ;), where eDNA shedding rates and concentrations

were log-transformed in the same manner as above. Four temperature levels were

pooled to increase sample size per biomass level.

3.2.6. Additional experiment for the relationship between eDNA decay rates and its

fragment size

Additional experiment was conducted to confirm whether the difference of nuclear and

mitochondrial eDNA (nu-eDNA, mt-eDNA, respectively) decay rates resulted from

amplicon length. The primers/probe set that specifically amplified 164-bp Japanese jack

mackerel’s DNA fragment of cytochrome b (CytB) gene was designed, where the same

forward primer and TaqMan probe as those in Yamamoto et al. (2016) was used (Table

3-2), and only the reverse primer was exchanged to vary the length of PCR amplicon.

Thus, the reverse primer was newly developed to produce 164-bp DNA fragment of

CytB gene using Primer Express 3.0 with default settings. The sequences of target and

closely related species (Amberfish, Amberjack, and Greater amberjack) in the National

Center for Biotechnology Information (NCBI) was used as references (Table 3-1). In

vitro specificity of the assay was then checked using the StepOnePlus Real-Time PCR

system in the same manner as described above, except for the change that we used only

10 pg of template DNA from target and related species.

56

The amount of 164-bp Japanese jack mackerel mt-eDNA in water samples

were then evaluated by quantifying the CytB gene copy number using the StepOnePlus

Real-Time PCR system. Here, two eDNA samples per treatment level at time bfr were

used, which had the highest and lowest nu-eDNA concentrations in each treatment

level. Other methodology on quantitative PCR was the same as described above. One-

way ANOVA and post-hoc Tukey-Kramer test were performed to investigate the effects

of DNA markers (category: CytB_127 bp, CytB_164 bp, and ITS1_164 bp), where all

the eDNA concentrations were log-transformed.

3.3. Results

A novel primers/probe set to specifically amplify the ITS1 region of nuDNA from the

Japanese jack mackerel was successfully developed (Table 3-2). The in vitro specificity

check showed no PCR amplification of any related species DNA and PCRnegative

controls. In addition, in all qPCR runs of tank experiments for nu-eDNA, the R2 values,

slope, Y-intercept, and PCR efficiency of the standard curves were 0.992 ± 0.008, -

3.779 ± 0.177, 45.034 ± 2.442, and 83.613 ± 4.998, respectively (mean ± 1 SD). PCR

amplifications were confirmed in some of the inlet water samples and filtration negative

controls. Concentrations of nu-eDNA in the inlet water samples ranged from 0.0 to

570.9 copies/reaction, which corresponded to 0.0 to 20.0 % of the water samples at time

bfr. In addition, nu-eDNA concentrations in filtration negative controls ranged from 0.0

to 418.3 copies/reaction, which corresponded to 0.0 to 13.1 % of nuDNA concentration

compared with water samples at the same sampling time point. Thus, the Japanese jack

57

mackerel eDNA in inlet water and low-level cross-contamination among samples were

not likely to have affected the conclusions. No PCR amplification was observed in any

PCR-negative controls.

Concentrations of nu-eDNA at time 0 drastically increased compared with

those at time bfr, which resulted from the stress caused by the removal of fish from the

tanks (Figure 3-2). After fish removal, nu-eDNA concentrations decreased exponentially

in all treatment levels. Coefficients from model fitting showed that higher temperature

and fish biomass significantly increased nu-eDNA decay rates (Table 3-3). In addition,

nu-eDNA decay rates were higher than those of mt-eDNA in all treatment levels (Table

3-3). Moreover, three-way ANOVA showed that eDNA shedding rates were

significantly different among water temperatures, fish biomass, and type of DNA

markers (all P < 0.0001). The interactions between fish biomass and type of DNA

markers (P < 0.001) and temperature and fish biomass (P < 0.05) also significantly

influenced eDNA shedding rates. Other interactions among factors were not significant

(P > 0.1) (Figure 3-3).

The ratios of mt-eDNA to nu-eDNA shedding rates and concentrations at time

bfr (B6*PQR and B6*PQ;) changed depending on fish biomass levels (Figure 3-4). Both

B6*PQR and B6*PQ; were significantly different among fish biomass levels (both P <

0.0001), and they were significantly lower for Large fish biomass level than for Small

(both P < 0.0001) and Medium ones (both P < 0.01). There were no significant

differences between Small and Medium fish biomass levels for B6*PQR (P = 0.6420)

and B6*PQ; (P = 0.3226).

58

Moreover, a primer/probe set to specifically amplify the CytB gene of mtDNA

from the Japanese jack mackerel with 164-bp fragment (reverse primer: 5’-

TTCTTTGTAGAGGTACGAGCCG-3’) were additionally developed. The in vitro

specificity check confirmed the successful PCR amplification of target species, and

showed no PCR amplification of any related species DNA and PCR negative control.

One-way ANOVA showed that there was no statistical difference among DNA markers

for Small (P = 0.7970) and Medium (P = 0.8240) biomass levels (Figure 3-5). Although

the statistical difference was confirmed in Large biomass level, eDNA concentrations of

CytB_164 bp was not statistically different from that of CytB_127 bp (P = 0.9730).

These results mean that the difference of nu- and mt-eDNA concentrations in the

present study did not depend on the PCR amplification length, for which it is concluded

that the effect of fragment size on the difference between nu- and mt-eDNA degradation

was negligible in the study. In the qPCR run for the additional experiment, the R2

values, slope, Y-intercept, and PCR efficiency of the standard curves were 0.999, -

3.408, 42.013, and 96.541, respectively.

3.4. Discussion

Although some studies have focused on the characteristics and dynamics of eDNA, our

understanding of them may still be limited. Moreover, almost all the studies targeted

only mt-eDNA. In the present study, this knowledge gap was addressed by estimating

the shedding and decay rates of nu-eDNA from Japanese jack mackerel and compared

them with those of mt-eDNA. It is found that higher water temperature and larger fish

59

biomass accelerated both shedding and degradation of nu-eDNA, and the observed

patterns were generally similar to those of mt-eDNA (Jo et al., 2019). In addition, the

ratios of mt-eDNA to nu-eDNA shedding and concentration (B6*PQR and B6*PQ;)

changed depending on total fish biomass.

The tendencies of eDNA shedding and degradation for nu-eDNA and mt-

eDNA in different water temperature and biomass levels generally showed similar

patterns; these were accelerated with higher water temperature and larger fish biomass.

The effects of temperature and biomass on mt-eDNA shedding and degradation rates

have been previously investigated. Moderately high temperatures (less than 50 °C) and

high densities of organisms could facilitate the activity of microbes and extracellular

enzymes, therefore increasing mt-eDNA degradation (Levy-Booth et al., 2007; Nielsen

et al., 2007; Strickler et al., 2015; Bylemans et al., 2018a; Jo et al., 2019a). In addition,

mt-eDNA shedding increased with an increase in biomass/abundance/size of organisms,

stress introduction, and, possibly, metabolic activation at higher temperature

(Maruyama et al., 2014; Klymus et al., 2015; Sassoubre et al., 2016; Mizumoto et al.,

2018; Jo et al., 2019a). The findings that the shedding and degradation of nu-eDNA

showed generally similar patterns to those of mt-eDNA may help the understanding of

nu-eDNA properties and the interpretation of nu-eDNA detection in natural

environments. In addition, the findings could support the hypothesis that the majority of

eDNA exists in the form of intra-cellular DNA, such as cell and tissue fragments in

water (Turner et al., 2014; Jo et al., 2019a).

On the other hand, some differences between nu-eDNA and mt- eDNA were

60

found. Nonlinear least-squares regression revealed that nu-eDNA degraded faster than

mt-eDNA in all treatments. Although the fragment size of nu-eDNA (164 bp) was

slightly larger than that of mt-eDNA (127 bp), the additional experiment comparing mt-

eDNA concentrations in water samples between different fragments revealed no

statistical differences in concentrations of 127- and 164-bp mt-eDNA. This suggests that

fragment size did not affect the difference in nu- and mt-eDNA decay rates and that the

difference in decay rates between nu- and mt-eDNA in the present study likely

depended primarily on DNA characteristics, including structure and packaging. The

nuclei in eukaryotic cells have a chromatin structure (Kornberg, 1974; Grunstein, 1997),

which may prevent nuDNA from attack by nucleases, whereas mtDNA has a simple

cyclic structure (Lindahl, 1993; Shadel & Clayton, 1997). In contrast, the linearity of

nuDNA might make it susceptible to exonucleases that do not digest circular DNA

molecules such as mtDNA (Hosfield et al., 1998; Alaeddini et al., 2010). Foran (2006)

reported that degradation of mtDNA was slower than that of nuDNA in tissue samples,

and some eDNA studies also implied that the nu-eDNA of macro-organisms degrades

faster than mt-eDNA (Jo et al., 2019b; Moushomi et al., 2019). On the basis of these

facts, it is likely that the persistence of mtDNA is longer than that of nuDNA in tissues

as well as aquatic environments. However, it remains unknown which environmental

factors the persistence of nuDNA and mtDNA depend on. Further studies are required to

evaluate the structural and cellular differences between nuclei and mitochondria that

may affect eDNA persistence and degradation.

The shedding rates of Japanese jack mackerel nu-eDNA showed similar

61

patterns to those of mt-eDNA. On the other hand, the interaction between fish biomass

and type of DNA markers was detected, where nu-eDNA shedding rates appeared to be

higher than those of mt-eDNA especially for Large fish biomass levels. Thus, I expected

that the ratios of mt-eDNA to nu-eDNA shedding and concentrations (B6*PQR and

B6*PQ;) might change depending on the fish biomass and body size and consequently

confirmed that both B6*PQR and B6*PQ; decreased with larger fish biomass levels. It

implies that the increment of mt-eDNA shedding as increasing fish biomass and body

size is smaller than that of nu-eDNA. Interestingly, the mtDNA copy number per cell or

per gram of tissue is known to decrease with larger body size and/or aging for various

taxa, which is caused by the accumulation of mtDNA point mutations and deletions

(Hayakawa et al., 1991; Montier et al., 2009; Hartmann et al., 2011). In this experiment,

the fish in Small and Medium biomass levels were estimated to be 0+ years and those in

Large biomass level to be 1+ years old (Mitani & Ida, 1964). Thus, the results may

partly reflect the decrease of mtDNA in a cell, and likely free-floating DNA released

from the cell, with maturity and aging. If the physiological phenomenon within

organisms was reflected to environmental samples, the error of eDNA-based estimation

of species biomass/abundance associated with age and developmental stage might be

smaller for nu-eDNA than for mt-eDNA. Considering the higher concentrations and

detectability of nu-eDNA compared with mt-eDNA (Minamoto et al., 2017b; Dysthe et

al., 2018), the result suggests that nuDNA marker may rather be superior to mtDNA

marker for the quantification of eDNA in the field. As Japanese jack mackerels of older

ages (to 5+ year; Mitani & Ida, 1964) were not targeted in this study, further studies are

62

needed to reveal the relationships between the eDNA ratios and body size with wider

age structures.

In conclusion, the present study compared the shedding and degradation of

nu-eDNA and mt-eDNA and showed that both processes were facilitated by high

temperatures and large biomasses, which was generally similar between both markers.

In addition, the ratio of mt-eDNA to nu-eDNA was dependent on the fish biomass.

These findings can contribute to an understanding of the characteristics and dynamics of

eDNA, especially the similarity and difference between DNA markers, which would

lead to the improvement of eDNA analysis (Goldberg et al., 2015; Barnes & Turner,

2016). On the other hand, some issues still need to be verified. Using mtDNA markers,

several studies have reported the relationships between various environmental factors

and eDNA production, persistence, and transport (Barnes et al., 2014; Pilliod et al.,

2014; Eichmiller et al., 2016; Buxton et al., 2017; Seymour et al., 2018). It is necessary

to study more how environmental factors influence eDNA dynamics, as well as evaluate

whether the influence of environmental factors is different for different DNA markers.

These findings may help to interpret the eDNA detection and quantification for different

DNA markers. Moreover, it would be subjects of future studies whether the ratio of mt-

eDNA to nu-eDNA could depend on body size and ages for other species and in the

field. If such relationships are found to be consistent in various situations, the ratio may

offer a suitable index to estimate the age structure of a population from environmental

samples. Alternatively, in combination with other indices (e.g., messenger RNA and

protein), the ratio may help to estimate biological and physiological information

63

(Barnes & Turner, 2016). This information would enable a more cost-effective and non-

invasive conservation tactics for the management of aquatic ecosystems than those

provided by traditional tools. The present study has addressed the groundwork required

for the understanding of different eDNA characteristics and dynamics, and has provided

valuable information to further utilize nuDNA markers for eDNA-based species

conservation and management.

64

3.5. Tables

Table 3-1. Details of sequence information from National Center for Biotechnology

Information (NCBI) for the primer/probe development.

Target region Species Accession number

nuclear ITS1

Japanese jack mackerel (Trachurus japonicus) AB375612

Amberfish (Decapterus maruadsi) AB375618 Amberjack (Seriola quinqueradiata) AB375568

Greater amberjack (Seriola dumerili) AB375605

Target region Species Accession number

mitochondrial CytB (164-bp)

Japanese jack mackerel (Trachurus japonicus) AB018994.1 Amberfish (Decapterus maruadsi) EF512291.1

Amberjack (Seriola quinqueradiata) AB517556.1 Greater amberjack (Seriola dumerili) KF760454.1

65

Table 3-2. The primers/probe set used in this study.

Primer or Probe ID Target region Sequences (5ʹ→3ʹ) Length

(bp) Tm (°C)

Reference

TjaITS1_F nuclear internal transcribed spacer-1

(ITS1)

GCGGGTACCCAACTCTCTTC 164

60.1 This study TjaITS1_R CCTGAGCGGCACATGAGAG 63.2

TjaITS1_P [FAM]-CTCTCGCTTCTCCGACCCCGGTCG-[TAMRA] 70.8

Tja_CytB_F2 mitochondrial cytochrome b

(CytB)

CAGATATCGCAACCGCCTTT 127

58.7 Yamamoto et al. (2016) Tja_CytB_R2 CCGATGTGAAGGTAAATGCAAA 57.6

Tja_CytB_P2 [FAM]-TATGCACGCCAACGGCGCCT-[TAMRA] 67.9

66

Table 3-3. Parameters from the nonlinear least-squares regression for the eDNA decay

curves (upper) and the eDNA decay rates calculated by these parameters (lower, mean ±

1 SD).

Target region Parameter Coefficient SE P value

nuclear ITS1

C0 0.9733 0.0339 *** b 0.0247 0.0050 *** c 0.1092 0.0349 ** a -0.2690 0.0857 **

mitochondrial CytB

C0 1.0025 0.0316 *** b 0.0173 0.0024 *** c 0.1030 0.0173 *** a -0.2744 0.0331 ***

Target region Temperature level

Fish biomass level

Small Medium Large

nuclear ITS1

13 °C 0.1432 ± 0.0034 0.2264 ± 0.0094 0.3280 ± 0.0075 18 °C 0.2650 ± 0.0060 0.3559 ± 0.0058 0.4510 ± 0.0094 23 °C 0.3932 ± 0.0064 0.4757 ± 0.0060 0.5660 ± 0.0062 28 °C 0.5133 ± 0.0064 0.5910 ± 0.0142 0.6969 ± 0.0164

mitochondrial CytB

13 °C 0.0366 ± 0.0032 0.1151 ± 0.0089 0.2109 ± 0.0070 18 °C 0.1215 ± 0.0056 0.2073 ± 0.0055 0.2969 ± 0.0089 23 °C 0.2124 ± 0.0060 0.2903 ± 0.0056 0.3755 ± 0.0059 28 °C 0.2957 ± 0.0060 0.3690 ± 0.0134 0.4690 ± 0.0155

Note: Each parameter is derived from the equation for model fitting to eDNA decay

curves. "# is the adjusted eDNA concentration at time 0, and $, %, and & are

constants estimated by the nonlinear least-squares regression (see Materials and

Methods). Asterisks show that the corresponding coefficients were statistically

significant (** P < 0.01; *** P < 0.001) in the model fitting. All raw mt-eDNA data are

from Jo et al. (2019a).

67

3.6. Figures

Figure 3-1. The overall flowchart of tank experiment. Three Japanese jack mackerels

were introduced into each 200-L tank with twelve different combinations of four

temperature and three fish biomass levels. After the 1-week acclimation, the fish were

removed from each tank. Time-series water sampling was performed on the day before

and after the fish removal. By model fitting to eDNA decay curves from times 0 to 96,

the eDNA decay rates were estimated for each tank. Using the decay rate constants and

eDNA concentrations at time bfr, eDNA shedding rates were calculated for each

treatment.

& RK T SENT VF

( LV

9S VTJ I TS TL L WN

) . () (.

ARGQQ+ . . ( M GS

=KJ R() +& + M GS

GVMK()& + & M GS

)

)

)

)

)

)

)

)

)

)

)

)

KR KVG VK QK KQ7 WN TRGWW QK KQ

.

KRT GQ TL L WN

( -( /.

;K LTV [KK

DG KV WGR Q SM L Q VG TS 1 87 7

T GQ 5 3 K VGI TS

GS L IG TS TL K5 3 IT ] S R KVS IQKGV0 S N W W J]

R TINTSJV GQ0 JG G LVTR :T K GQ (& /

6W RG SM K5 3 JKIG] WNKJJ SM VG KWA G W IGQ GSGQ]W W

4 2 4&K � A 2 C 4C

.

68

Figure 3-2. Overall nu-eDNA (upper) and mt-eDNA (lower) decay curves from the tank experiments. Dots show eDNA concentrations

per PCR reaction at each time point (average of four tank replicates). Error bars show the standard deviations (SD) of tank replicates.

The exponential curve fitted to each eDNA decay curve is based on the mean eDNA concentration at time 0 and the decay rate constant

estimated by the nonlinear least-squares regression for each treatment level. All raw mt-eDNA data are from Jo et al. (2019a).

-20 0 20 40 60 80 100

01000

3000

5000

Small 13°C

nu-e

DN

A c

onc.

[cop

ies/

2µL

tem

plat

e D

NA

]

time point [hour]-20 0 20 40 60 80 100

01000

3000

5000

Small 13°C

nu-e

DN

A c

onc.

[cop

ies/

2µL

tem

plat

e D

NA

]

time point [hour]

-20 0 20 40 60 80 100

01000

3000

5000

Small 28°C

-20 0 20 40 60 80 100

020000400006000080000

Small 23°C

-20 0 20 40 60 80 100

01000

3000

5000

Small 18°C

-20 0 20 40 60 80 100

01000

3000

5000

Small 13°C

nu-e

DN

A c

onc.

[cop

ies/

2µL

tem

plat

e D

NA

]

time point [hour]

-20 0 20 40 60 80 100

050000

150000

250000

Medium 28°C

-20 0 20 40 60 80 1000e+00

2e+05

4e+05 Medium 23°C

-20 0 20 40 60 80 1000e+00

4e+05

8e+05 Medium 18°C

-20 0 20 40 60 80 100

040000

80000

120000

Medium 13°C

-20 0 20 40 60 80 1000e+00

2e+06

4e+06

Large 28°C

-20 0 20 40 60 80 100

0500000

1500000

2500000

Large 23°C

-20 0 20 40 60 80 1000e+001e+072e+073e+074e+07

Large 18°C

-20 0 20 40 60 80 1000e+001e+072e+073e+074e+07

Large 13°C

69

(Figure 3-2)

-20 0 20 40 60 80 100

01000

3000

5000

Small 13°Cm

t-eD

NA

con

c. [c

opie

s/2µ

L te

mpl

ate

DN

A]

time point [hour]

-20 0 20 40 60 80 1000e+002e+054e+056e+058e+05

Large 13°C

-20 0 20 40 60 80 100

01000000

2500000

Large 18°C

-20 0 20 40 60 80 1000e+00

4e+05

8e+05 Large 23°C

-20 0 20 40 60 80 100

0500000

1500000 Large 28°C

-20 0 20 40 60 80 100

010000

30000

Medium 13°C

-20 0 20 40 60 80 100

01000

3000

5000

Small 18°C

-20 0 20 40 60 80 100

05000

10000

15000

Small 23°C

-20 0 20 40 60 80 100

0500

1000

2000

Small 28°C

-20 0 20 40 60 80 100

01000

3000

5000

Small 13°C

mt-e

DN

A c

onc.

[cop

ies/

2µL

tem

plat

e D

NA

]

time point [hour]

-20 0 20 40 60 80 100

01000

3000

5000

Small 13°C

nu-e

DN

A c

onc.

[cop

ies/

2µL

tem

plat

e D

NA

]

time point [hour]

-20 0 20 40 60 80 100

050000

150000

250000

Medium 18°C

-20 0 20 40 60 80 100

050000

100000

150000

Medium 23°C

-20 0 20 40 60 80 100

010000200003000040000

Medium 28°C

70

Figure 3-3. Results of nu-eDNA (upper) and mt-eDNA (lower) shedding rates per

treatment. Fish biomass level: S, Small; M, Medium; L, Large. All raw mt-eDNA data

are from Jo et al. (2019a).

56

78

910

11

S M L S M L S M L S M L13°C 18°C 23°C 28°C

Treatment level

log1

0(eD

NA

she

ddin

g ra

tes

per t

reat

men

t) [c

opie

s/ho

ur]

56

78

910

11


13°C 18°C 23°C 28°C

Treatment level

log1

0(eD

NA

she

ddin

g ra

tes

per t

reat

men

t) [c

opie

s/ho

ur]

71

Figure 3-4. Results of the ratios of mt-eDNA to nu-eDNA shedding rates (left) and

concentrations at time bfr (right); the four temperature levels were pooled to increase

sample size. Factor levels with different letters are significantly different based on a

post hoc Wilcoxon rank sum test with Bonferroni adjustment (P < 0.05).

Small Medium Large

0.8

0.9

1.0

1.1

1.2

a a b

Shedding rates (log10)

Small Medium Large

0.8

0.9

1.0

1.1

1.2

a a b

Concentrations at time bfr (log10)

ratio

of m

t-eD

NA

to n

u-eD

NA

72

Figure 3-5. Comparison of eDNA concentrations with different DNA markers. The

eDNA samples with highest and lowest nu-eDNA concentrations in each treatment level

at time bfr were used in the analysis (each boxplot included the eDNA concentrations of

eight water samples). Factor levels with different letters are significantly different based

on a post-hoc Tukey-Kramer test (P < 0.05).

CytB_127 bp CytB_164 bp ITS1_164 bp

01

23

4

Small

n.s.


23

45

6

Medium

n.s.


34

56

7

Large

a a b

log1

0(eD

NA

con

c.) [

copi

es/2

µL

tem

plat

e D

NA

]

73

Chapter 4. Particle size distribution of environmental DNA from the nuclei of

marine fish.

4.1. Introduction

Environmental DNA (eDNA) analyses have been developed for improving the

conservation and management of aquatic ecosystems in this decade (Ficetola et al.,

2008; Minamoto et al., 2012; Bohmann et al., 2014). Macro-organisms shed their DNA

into the environment as feces, mucus, scales, and gametes (Martellini et al., 2005;

Merkes et al., 2014; Sassoubre et al., 2016; Bylemans et al., 2017), which is termed

eDNA. The presence of target species can be estimated by detecting their eDNA from

environmental media such as water and sediment, allowing more efficient and

noninvasive surveillance of species distribution and composition than traditional

methods (Biggs et al., 2015; Fukumoto et al., 2015; Yamamoto et al., 2017; Li et al.,

2018; Boussarie et al., 2018).

Most eDNA analyses for macro-organisms have targeted mitochondrial DNA

(mtDNA) as a genetic marker due to its abundance in a cell (Ficetola et al., 2008;

Takahara et al., 2013; Deiner et al., 2016; Carraro et al., 2018). However, recent studies

have suggested the applicability of nuclear DNA (nuDNA) marker for eDNA analysis,

which targets multiple copies of the rRNA gene such as internal transcribed spacer

(ITS) regions (Minamoto et al., 2017b; Dysthe et al., 2018; Gantz et al., 2018). The

genetic regions have high interspecific variations and, unlike mtDNA, can provide high

resolutions to discriminate closely related species (Hillis & Dixon, 1991; Booton et al.,

74

1999; Bylemans et al., 2018b). It is likely that nuDNA markers will become an

alternative eDNA tool, whereas the knowledge on the characteristics and dynamics of

eDNA derived from nuclei (nu-eDNA) is scarce.

Researchers have been interested in how eDNA can be produced and exist in

the environment, and therefore have emphasized the necessity to collect such

fundamental information on eDNA (Díaz-Ferguson & Moyer, 2014; Goldberg et al.,

2015; Barnes & Turner, 2016; Hansen et al., 2018; Stewart, 2019). For example,

although there is still much to be verified, several studies have reported the effects of

various biotic/abiotic factors on eDNA detectability and persistence (Barnes et al., 2014;

Strickler et al., 2015; Turner et al., 2015; Eichmiller et al., 2016; Collins et al., 2018;

Seymour et al., 2018; Jo et al., 2019) and the horizontal/vertical transport of eDNA in

various aquatic environments (Deiner & Altermatt, 2014; Jane et al., 2015; Shogren et

al., 2017; Nukazawa et al., 2018; Murakami et al., 2019). However, the information on

the physiological origin and state of eDNA (e.g., living/dead cell, intra-/extra-cellular,

dissolved/free) is relatively limited, which is rather fundamental for understanding the

characteristics and dynamics of eDNA (Barnes & Turner, 2016; Hansen et al., 2018).

These eDNA aspects can influence the transport and fate of eDNA, since larger and

heavier eDNA particles in water can be expected to disperse less and settle more rapidly

(Wotton & Malmqvist, 2001). In addition, DNA molecules within a cell membrane (i.e.,

intra-cellular DNA) should be attacked less efficiently by microbes and extra-cellular

enzymes in the environment than extra-cellular-free DNA (Ahrenholtz et al., 1994;

Matsui et al., 2001; Levy-Booth et al., 2007). Studying how eDNA can be produced and

75

exist in an aquatic environment would substantially contribute to the understanding of

eDNA characteristics and dynamics. However, almost all eDNA studies have targeted

only mitochondrial eDNA (eDNA derived from mitochondria, hereafter, mt-eDNA).

Therefore, the present study focused on the characteristics and dynamics of

nu-eDNA, especially its particle size distribution (PSD) and temporal changes. Previous

studies estimated the PSD of eDNA in natural environments using a mtDNA marker and

found that the largest proportion of fish mt-eDNA was found in the 1 - 10 µm size

fraction (Turner et al., 2014; Wilcox et al., 2015). In addition, Jo et al. (2019) reported

that mt-eDNA PSD from Japanese jack mackerel (Trachurus japonicus) could vary

depending on water temperature and time passages after fish removal. These results

included various eDNA production and degradation processes, and it remains unknown

how each process could contribute to the PSD of eDNA. The state of eDNA (e.g., intra-

to extra- membrane) may vary over time until the material is no longer detectable, and

such a process would influence the persistence of eDNA. In addition, these processes

may differ between nu- and mt-eDNA. In eukaryotic cells, the nuclei have chromatin

structures that are 5 - 10 µm in diameter (Kornberg, 1974; Lloyd et al., 1979), while

mitochondria have simple cyclic structures that are generally smaller (Ernster & Schatz,

1981; Shadel & Clayton, 1997). If the PSDs differ between nu-eDNA and mt- eDNA,

the selective capture of target eDNA might be possible based on their size. The PSDs of

eDNA based on multiple DNA regions or loci would help our understanding of the state

and fate of eDNA.

This study investigated the PSD of nu-eDNA and its temporal changes

76

through a tank experiment. Japanese jack mackerel was used as a model species due to

its frequent use in previous eDNA studies (Yamamoto et al., 2016; Jo et al., 2017; 2019)

and its economic importance in East Asia including Japan (Zhang & Lee, 2001). In

addition, focusing on water temperature and fish biomass density, the effects of these

biotic/abiotic factors on nu-eDNA PSD were examined. Furthermore, these results were

compared with those of mt-eDNA PSD from previous studies (Jo et al., 2019).


4.2.1. Experimental protocol

Tank experiments were conducted at the Maizuru Fisheries Research Station, Kyoto

University, Japan, from June 2016 to July 2017 (Figure 4-1). All the eDNA samples

were from Jo et al. (2019). Briefly, the rearing water were collected from experimental

tanks with different temperatures (13, 18, 23, and 28 °C) and fish biomass levels (Small,

Medium, and Large) with four tank replicates per treatment. Fish biomass levels were

based on the difference of total fish biomass in the tank (g/200 L). Sequential filtrations

were performed using a series of filters with different pore sizes (10, 3, 0.8, and 0.4 or

0.2 µm), extracted total DNA on the filter with DNeasy Blood and Tissue Kit (Qiagen,

Hilden, Germany), and quantified Japanese jack mackerel’s eDNA concentrations at

each size fraction. The concentration of Japanese jack mackerel’s nu-eDNA in water

samples were estimated by quantifying the copy number of nuclear internal transcribed

spacer-1 (ITS1) regions using the StepOnePlus Real-Time PCR system (Thermo Fisher

Scientific, Foster City, CA, U.S.). The primers/probe set that specifically amplified the

77

Japanese jack mackerel’s DNA fragment from the ITS1 region (Jo et al., 2020; Table 4-

1) was used. ITS1 is a part of rRNA genes (rDNA), and multiple copies of ITS1 are

present in the nuclear genome. It is confirmed that the ITS1 primer set amplified only

target species and locus using an in silico specificity check (Jo et al., 2020). Each 20 µL

of TaqMan reaction contained a 2 µL template DNA, a final 900 nM concentration of

forward and reverse primers, and 125 nM of TaqMan probe in 1 × TaqMan Gene

Expression PCR Master Mix (Thermo Fisher Scientific). 2 µL of pure water was

simultaneously analyzed as a PCR negative control. Real-time PCR was performed

using a dilution series of standards containing 3 × 101 to 3 × 104 copies of a linearized

plasmid containing synthesized artificial DNA fragments from a partial ITS1 region

sequence (237 bp) of a target species. All qPCRs for eDNA extracts, standards, and

negative controls were performed in triplicate. Thermal conditions of quantitative real-

time PCR were as follows: 2 min at 50 °C, 10 min at 95 °C, 55 cycles of 15s at 95 °C,

and 1 min at 60 °C. Quantification of the eDNA copy number for the mitochondrial

CytB gene was performed as per the method in Jo et al. (2019). Concentrations of target

eDNA was calculated by averaging the triplicate, and each replicate showing non-

detection (PCR negative) was classified as containing zero copies (Ellison et al., 2006).

The limit of quantification (LOQ) of the qPCR was one copy per reaction with

triplicates following previous studies (Doi et al., 2017; Katano et al., 2017), and any

eDNA concentration below LOQ was classified as a zero copy.

78


R version 3.2.4 was used for all the statistical analyses (R core team, 2006). Before the

analyses, all the eDNA concentrations were log-transformed after adding one to meet

the assumption of normality. Using all samples that had passed through sequential filters

with 10, 3, 0.8, and 0.4 µm pore sizes at time before fish removal (bfr), multivariate

analysis of variance (MANOVA) and post-hoc ANOVAs were performed to investigate

how the PSD of eDNA related to water temperature, fish biomass, and DNA markers. In

the analyses, the eDNA concentrations at each size fraction were included as dependent

variables, and water temperature level, fish biomass, and type of DNA markers (ITS1 or

CytB) were included as factors. There were four tank replicates per treatment level

(except for 28 °C/Large biomasslevels, where three tank replicates were prepared due to

fish mortality). MANOVA can simultaneously evaluate the effects of each factor on

multiple response variables, which can reduce the likelihood of Type I errors and

increase the statistical powers (Fish, 1988; Warne, 2014).

In addition, an ANOVA was performed to investigate how the PSD of eDNA

changed with fish removal using the samples that passed through the sequential filters

with 10, 3, 0.8, and 0.2 µm pore sizes at time bfr and 0 (hour). Concentrations of eDNA

were included as dependent variables, and filter pore size, sampling time point (time bfr

or 0), type of DNA markers, and all the interactions between them were included as

factors. Furthermore, temporal changes of nu- and mt-eDNA PSDs after fish removal

were investigated using the same samples at time 0, 6, 12, and 18 (hour). LMM (linear

mixed model) with the function lmer of the R package lmerTest (Kuznetsova et al.,

79

2017) was performed, and filter pore size, sampling time point, temperature level, fish

biomass level, and type of DNA marker were included as explanatory variables. Water

temperature was considered to be as quantitative values and set each temperature level

as the increment from the lowest temperature level (13 °C). The interactions between

sampling time points and each of the other factors were also included, assuming that the

temporal degradation of eDNA may vary among size fractions, treatment levels, and

DNA markers, and tank replicates as random effects.

4.3. Results and Discussion

In all qPCR analyses for nu-eDNA including filtration negative controls, the R2 values,

slope, Y-intercept, and PCR efficiency (%) of the calibration curves were 0.984 ± 0.017,

-3.586 ± 0.208, 44.940 ± 1.567, and 90.615 ± 7.690, respectively (mean ± 1 SD). PCR

amplifications were confirmed in some inlet water samples which were pumped from a

depth of 6 m at the station, where Japanese jack mackerels are abundant, and filtration

negative controls: nu-eDNA concentrations in inlet water samples were 22.3 ± 84.2

copies/reaction. This corresponded to 5.2 ± 15.9 % of eDNA concentrations relative to

those with the sum of sequential filters in water samples at time bfr (mean ± 1 SD,

respectively). Besides, nu-eDNA concentrations in filtration negative controls were 11.3

± 54.5 copies/reaction, which corresponded to 1.2 ± 9.2 % of eDNA concentrations

relative to those in overall water samples (mean ± 1 SD, respectively). Thus, the

Japanese jack mackerel’s eDNA in inlet water and cross-contamination among samples

is not likely to have affected the results. No PCR amplification was confirmed from any

80

PCR negative controls.

4.3.1. The relationships of eDNA PSD with temperature, fish biomass, and DNA markers

Water temperature, fish biomass, and DNA markers significantly affected the eDNA

concentrations at each size fraction (MANOVA, all P < 0.05; Figures 4-2 and 4-3; Table

4-2). Post-hoc ANOVAs showed that the type of DNA marker was a significant factor

for the 3 - 10 µm size fraction (P < 0.05), water temperature was significant for the 0.8 -

3 and 0.4 - 0.8 µm size fractions (both P < 0.01), and fish biomass was significant for

all size fractions (all P < 0.0001). First, the concentration of nu-eDNA was larger than

that of mt-eDNA for the 3 - 10 µm size fraction. Although both nu- and mt-eDNA could

also be detected at the size of cell or tissue fragments (mainly the >10 µm size fraction

in the study), the results may partly reflect the size differences between nuclei and

mitochondria; nuclei (around 5 - 10 µm in diameter) is generally larger than

mitochondria (around 0.5 - 2 µm) in eukaryotic cells (Wrigglesworth et al., 1970; Lloyd

et al., 1981). This study is the first report to estimate the PSD of nu-eDNA, as well as to

find the differences of PSD between nu- and mt-eDNA. Meanwhile, the fact that much

of nu- and mt-eDNA was detected at >3 µm size fractions would also be meaningful,

because nu-eDNA could be captured as much as, or more than, mt-eDNA using the

filter with the same pore sizes. Further study is needed to examine whether these

similarities and differences are common among taxa.

Second, the concentration of nu- and mt-eDNA generally increased at higher

water temperatures for the 0.8 - 3 µm and 0.4 - 0.8 µm size fractions, whereas

81

temperature was not a significant factor for the >3 µm size fraction. The degradation of

eDNA would be likely promoted with higher water temperatures in all size fractions

(Strickler et al., 2015; Eichmiller et al., 2016) and the warmer temperature could

influence fish behavior and increase eDNA shedding (Jo et al., 2019). On the other

hand, it is also likely that the outer cell membrane of large-sized eDNA (such as cells

and tissues) broke down, and part of them was turned into small-sized eDNA (such as

nuclei, mitochondria, and their extra-cellular DNA). The DNA release from prokaryotic

cells occurs following viral attacks or enzymatic activity (Levy-Booth et al., 2007;

Arnosti, 2014; Torti et al., 2015). Besides, the activity of microbes and extra-cellular

enzymes can be stimulated by moderately higher temperatures (< 50 °C) (Ahrenholtz et

al., 1994; Price & Sowers, 2004). Thus, it is possible that, through the enzymatic

activity, higher temperature facilitates the release of such small-sized eDNA out of the

cell membrane. The decrease of eDNA due to degradation at smaller size fractions

might be buffered by an increase of eDNA production from larger to smaller size

fractions.

Third, the concentration of eDNA was much larger in the Large biomass level

than the other biomass levels for all size fractions. Interestingly, there was almost no

difference of eDNA concentrations between Small and Medium biomass levels. The

growth model for Japanese jack mackerel (Mitani & Ida, 1964) estimated the ages of

both small- and medium-sized fishes to be 0+ year, while those of large-sized fish to be

almost 1+ year. The release of eDNA might be similar within the same age group.

Further investigation would be needed to understand the relationship between eDNA

82

release and the age/developmental stage of organisms (Maruyama et al., 2014). Besides,

it might be accounted by the effect of fish biomass density in experimental tanks. For

example, Sassoubre et al. (2016) reported that eDNA shedding rates per individual of

Pacific sardine (Sardinops sagax) and Pacific chub mackerel (Scomber japonicus)

increased with larger fish biomass density in the tanks. In this experiment, large-sized

fish might have touched each other more often.

4.3.2. Temporal changes of eDNA PSD

Temporal changes of eDNA PSD were also studied (Figures 4-4 and 4-5). At first,

immediately after fish removal, the concentration of eDNA increased for all size

fractions, which could be due to the handling stress at fish removal. The eDNA

concentrations significantly depended on sampling time and filter pore size (ANOVA,

all P < 0.001; Table 4-3). The interaction between sampling time and filter pore size

was also significant (P < 0.01); eDNA increases were not similar among size fractions

but were emphasized in >10 µm size fraction. Previous studies have suggested that

physical and environmental stresses on organisms could stimulate eDNA release, which

could originate from scales and mucus (Pilliod et al., 2014; Sassoubre et al., 2016;

Bylemans et al., 2018a). The type of DNA marker and other interactions were not

significant (all P > 0.1), suggesting that, due to fish removal, there was no difference of

eDNA release between nu- and mt-eDNA. Most of the eDNA just after release from

aquatic organisms might be intracellular DNA such as cells and tissues rather than

extracellular DNA. Further study is needed to verify the physical forms of eDNA

83

released into natural environments.

Following fish removal, the concentration of eDNA decreased over time for

all size fractions, while eDNA degradation was suppressed in the smaller size fractions

(Figures 4-4 and 4-5). The eDNA concentrations were significantly affected by filter

pore size and temperature positively, and time point and fish biomass negatively (LMM,

all P < 0.0001; Table 4-3) but did not significantly change with DNA marker (P =

0.8175). Besides, all interactions in the analyses were significant (P < 0.05). Thus, the

significance of main effects of each variable might be restrained. The significant

interactions between filter pore size and time point could reflect the reduction of large-

sized eDNA toward smaller size fractions as above; some of the eDNA at larger size

fractions broke down, changed their physical forms, and turned into small-sized eDNA.

Especially at the 0.2 - 0.8 µm size fraction, there were some treatment levels at which

the concentration of eDNA seemed to rather increase over time. These results imply

that, depending on the size fraction, the production of eDNA from larger to smaller size

fractions might sometimes surpass the reduction of eDNA (i.e., non-detection by PCR).

If the experiment had been continued another a few days, the shift of eDNA PSD

toward smaller size fractions might have been more obvious. Because of the differences

of physical forms, small-sized eDNA such as organelles and extra-membrane DNA

would likely be more sensitive to enzymatic activity in environment than large-sized

eDNA such as cells and tissues. The present study, however, suggested that the

production of eDNA from larger to smaller size fractions could occur, which could

buffer the degradation of small-sized eDNA and prolong its “apparent” persistence in

84

water. The findings imply that the size, and the state, of eDNA could vary over time,

which would contribute to the elucidation on the state and fate of eDNA in aquatic

environments. On the other hand, there might be some difference of eDNA PSDs

between experimental tanks and natural environment. I suggest future study of eDNA

PSDs with various environmental conditions (e.g., pH, trophic state, and fish density)

such as natural conditions and temporal changes of eDNA PSDs in environmental water

samples. This might help link the results of the present study with actual eDNA

dynamics in an aquatic environment.

Other significant interactions in the LMM analysis offer interesting

interpretations. The negative interaction between time point and temperature indicates

that eDNA degradation was accelerated with higher temperatures, which has been found

in previous studies (Strickler et al., 2015; Eichmiller et al., 2016). The positive

interaction between time point and fish biomass shows that eDNA degradation was

suppressed for small contrary to large biomass level. This might be due to an increase of

microbial density with fish biomass density in the experimental tanks (Barnes et al.,

2014; Jo et al., 2019). Further studies are needed to show how the relationships between

eDNA persistence and various biotic/abiotic factors depend on the size and state of

eDNA.

4.3.3. Implications and Perspectives

Through the study, by estimating the PSD of nu-eDNA, important implications on the

state and fate of eDNA derived from macro-organisms in aquatic environment were

85

obtained. On the basis of the present and previous studies, I summarized on the state

and fate of eDNA from fish in water (Figure 4-6). First, much of eDNA would be

released from organisms as relatively large-sized particles (>10 µm in diameter),

originating as intra-cellular DNA such as cell and tissue fragments (Figure 3). These

eDNA could be released into the environment with mucus and scales, which may

increase the average eDNA size. It is less likely that organisms would directly shed their

nuclei, mitochondria, and their intra-membrane DNA; rather, the part of eDNA

especially at larger size fractions could break down (e.g., the lysis and fragmentation of

cell membrane through the activity of microbes and exonucleases), which might change

their physical state and structure, and thus turn them into smaller-sized eDNA. In this

study, the degradation of both nu- and mt-eDNA was suppressed in the smaller size

fractions, which is likely due to the breakdown of large-sized eDNA. This tendency

might be facilitated by an increase of water temperature and species biomass density

since these factors can promote microbial activity. Moreover, because of the size

differences between nuclei and mitochondria, nu-eDNA was more detected than mt-

eDNA, especially at >3 µm size fractions, which might contribute to the difference of

eDNA PSDs between DNA markers.

The present study clarified some aspects of particle size characteristics of fish

eDNA, though there are still knowledge gaps that must be verified before this tool can

be used in environmental applications. Regardless of the increase of eDNA applications

with various taxa (Katano et al., 2017; Carraro et al., 2018; Seymour et al., 2018), the

PSD of eDNA has not been reported for taxa other than fish. It could be possible that

86

eDNA PSDs are different among taxa. In addition, PCR efficiencies tended to be

slightly lower for nu-eDNA (90.615 ± 7.690 %) than mt-eDNA (93.789 ± 3.794 %;

mean ± 1 SD). This might partly be due to the difference of amplification length

between primers/probe sets (ITS1, 164 bp; CytB, 127 bp). When comparing the results

of eDNA detection between different DNA regions or fragment sizes, equalizing PCR

efficiencies would be ideal. Furthermore, it will be necessary to understand the

physiological and cytological characteristics of eDNA other than its PSD. For example,

chromatin structure in nuclei (Kornberg, 1974) and the fission and fusion of

mitochondria for the maintenance of its integrity (Suen et al., 2008) might influence the

detectability and persistence of eDNA. A greater understanding of such fundamental

information on eDNA would improve the efficiency of eDNA analyses, and contribute

to the validation of its use in natural environments. The present study can be the basis

for future eDNA studies, and may help facilitate the use of eDNA analyses as an

efficient tool for improving the conservation and management of aquatic ecosystems.

87

4.4. Tables

Table 4-1. Primers/probe set used in this study.

Primer or Probe ID Target region Sequences (5’→3’) Length

(bp)

Tm

(°C) Reference

TjaITS1_F nuclear

internal transcribed spacer-1

(ITS1)

GCGGGTACCCAACTCTCTTC

164

60.1

Jo et al. (2020) TjaITS1_R CCTGAGCGGCACATGAGAG 63.2

TjaITS1_P [FAM]-CTCTCGCTTCTCCGACCCCGGTCG-[TAMRA] 70.8

Tja_CytB_F2 mitochondrial

cytochrome b

(CytB)

CAGATATCGCAACCGCCTTT

127

58.7

Yamamoto et al. (2016) Tja_CytB_R2 CCGATGTGAAGGTAAATGCAAA 57.6

Tja_CytB_P2 [FAM]-TATGCACGCCAACGGCGCCT-[TAMRA] 67.9

88

Table 4-2. Results of MANOVA (upper) and post-hoc ANOVAs (lower) for the

relationships between eDNA concentrations at each size fraction and each factor.

Response Factor P value

eDNA conc.

(for all size fractions)

Temperature 0.0000 ***

Fish biomass 0.0000 ***

DNA marker 0.0353 *


eDNA conc.

(> 10 µm size fraction)

Temperature 0.8906


DNA marker 0.2059

eDNA conc.

(3 - 10 µm size fraction)

Temperature 0.7147


DNA marker 0.0254 *

eDNA conc.

(0.8 - 3 µm size fraction)

Temperature 0.0012 **


DNA marker 0.9596

eDNA conc.

(0.4 - 0.8 µm size fraction)

Temperature 0.0040 **


DNA marker 0.7663

Note: Asterisks show the corresponding factors that are statistically significant (* P <

0.05; ** P < 0.01; *** P < 0.001). All eDNA concentrations were log-transformed.

89

Table 4-3. Results of the statistical analyses for temporal change of eDNA PSDs.


eDNA conc.

Time point 0.0000

Pore size 0.0000

DNA marker 0.4063

Time point: Pore size 0.0028

Time point: DNA marker 0.1003

Pore size: DNA marker 0.1903

Time point: Pore size: DNA marker 0.5425

Response Explanatory Estimate SE P value

eDNA conc.

Intercept 2.3436 0.1302 ***

Time point -0.0419 0.0095 ***

Pore size 0.1594 0.0116 ***

Temperature 0.0342 0.0082 ***

Fish biomass (S) -0.8641 0.0908 ***

DNA marker (ITS1) 0.0209 0.0906

Time point: Pore size -0.0050 0.0010 ***

Time point: Temperature -0.0026 0.0007 ***

Time point: Fish biomass (S) 0.0162 0.0081 *

Time point: DNA marker (ITS1) -0.0258 0.0081 **

Note: The upper table shows the results of ANOVA for the differences in eDNA

concentrations between time bfr and 0, where bold values represent the statistical

significances of these factors (P < 0.05). The lower table shows the results of LMM for

the relationships between eDNA concentrations and each factor, where asterisks

represent the statistical significances of the parameters (* P < 0.05; ** P < 0.01; *** P

< 0.001). In the LMM, variables ‘Fish biomass (S)’ and ‘DNA marker (ITS1)’ represent

the fixed effects of Small against Large biomass levels, and the markers for nuclear

DNA against mitochondrial DNA, respectively.

90

4.5. Figures

Figure 4-1. Overall flowchart for the tank experiments. Three Japanese jack mackerels

were kept in 200 L tanks with four temperature and three biomass levels. After 1 week,

the fish were removed from each tank. Water sampling and sequential filtration were

conducted the day before and after fish removal. For all fish biomass levels, water

samples were filtered only at time bfr using polycarbonate (PC) filters with 10, 3, 0.8,

and 0.4 µm pore sizes. For small and large fish biomass levels, water samples were

filtered at times bfr, 0, 6, 12, and 18 using PC filters with 10, 3, 0.8, and 0.2 µm pore

sizes.

KN QPKAJP S

( D S

5 SPF KP P KTJ

) - () (-

NCMM- - ( I C L

9 FK N() & I C L

8CSI()& ,& I C L

])

])

])

])

])

])

])

])

])

])

])

])

NQ SC S M W MT

4KTJ DKPNCTT M W MT

,-

NPWCM P KTJ

& ( --

Q PS L

,

C S TCNQMK I T R KCM KM SC KP && N80 ;2

P CM 3:1 SC KP

C K K C KP P 3:1 PQ ND SM CS/ K JKT T F

NK P JP FSKCM/ 6P CM (& .

C KT K CM C CM TKT

&

A N

4KM SQPS TK[ ) & - &

-20 1- .

C S

1 KN D S P M

4PS CMM KTJ DKPNCTT M W MT

&

A N

4KM SQPS TK[ ) & - & (

-20 1- .

C S

1 KN D S & , ( -

4PS NCMM 8CSI KTJ DKPNCTT M W MT

91

Figure 4-2. Results of the PSDs of Japanese jack mackerel nu-eDNA at time bfr. Upper

boxplots show the eDNA PSD at each temperature level (lightblue, 13 °C; blue, 18 °C;

purple, 23 °C; and red, 28 °C), where fish biomass levels are pooled. The lower

boxplots show the eDNA PSD at each fish biomass level (cyan, S; skyblue, M; and

pink, L), where water temperature levels are pooled.

13 18 23 28

01

23

45

6

>10 µm

13 18 23 28

01

23

45

6

3-10 µm

13 18 23 28

01

23

45

6

0.8-3 µm

13 18 23 28

01

23

45

6

0.4-0.8 µm

Water temperature [°C]

log1

0(eD

NA

con

cent

ratio

n)[c

opie

s/2 µL

tem

plat

e D

NA

]

S M L

01

23

45

6

>10 µm

S M L

01

23

45

6

3-10 µm

S M L

01

23

45

6

0.8-3 µm

S M L

01

23

45

6

0.4-0.8 µm

Fish biomass level

log1

0(eD

NA

con

cent

ratio

n)[c

opie

s/2 µL

tem

plat

e D

NA

]

92

Figure 4-3. Results of the PSDs of Japanese jack mackerel mt-eDNA at time bfr (data

from Jo et al., 2019). Upper boxplots show the eDNA PSD at each temperature level

(lightgreen, 13 °C; green, 18 °C; yellow, 23 °C; and orange, 28 °C), where fish biomass

levels are pooled. The lower boxplots show the eDNA PSD at each fish biomass level

(gray, S; darkgreen, M; and darkred, L), where water temperature levels are pooled.

13 18 23 28

01

23

45

6

>10 µm

13 18 23 28

01

23

45

6

3-10 µm

13 18 23 28

01

23

45

6

0.8-3 µm

13 18 23 28

01

23

45

6

0.4-0.8 µm

Water temperature [°C]

log1

0(eD

NA

con

cent

ratio

n)[c

opie

s/2 µL

tem

plat

e D

NA

]

S M L

01

23

45

6

>10 µm

S M L

01

23

45

6

3-10 µm

S M L

01

23

45

6

0.8-3 µm

S M L

01

23

45

6

0.4-0.8 µm

Fish biomass level

log1

0(eD

NA

con

cent

ratio

n)[c

opie

s/2 µL

tem

plat

e D

NA

]

93

Figure 4-4. Results of the temporal changes of Japanese jack mackerel nu-eDNA

(upper) and mt-eDNA (lower) PSDs. All temperature and fish biomass levels are pooled

for both boxplots. Note that the smallest size fraction here is 0.2 - 0.8 µm.

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

nu-eDNA (ITS1)

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

mt-eDNA (CytB)

94

Figure 4-5. Results of the time-series changes of Japanese jack mackerel eDNA PSDs

for each treatment level. Blue boxplots represent the PSDs of nu-eDNA, and green ones

do the PSDs of mt-eDNA. Dataset of mt-eDNA is from Jo et al. (2019).

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

13 °C - Small

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

18 °C - Small

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

60.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

23 °C - Small

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

28 °C - Small

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

13 °C - Large

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

18 °C - Large

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

23 °C - Large

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

28 °C - Large

95

(Figure 4-5)

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 180

12

34

56

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

13 °C - Small

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

18 °C - Small

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

23 °C - Small

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

28 °C - Small

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

13 °C - Large

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

18 °C - Large

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

23 °C - Large

-24 0 6 12 18

01

23

45

6

>10 µm

-24 0 6 12 18

01

23

45

6

3-10 µm

-24 0 6 12 18

01

23

45

6

0.8-3 µm

-24 0 6 12 18

01

23

45

6

0.2-0.8 µm

Time point [hour]

log1

0(eD

NA

con

cent

ratio

n)

[cop

ies/

2 µL

tem

plat

e D

NA

]

28 °C - Large

96

Figure 4-6. Schematic depiction of the state and fate of eDNA in water. Macro-organism

eDNA can exist in aquatic environments in various sizes and states, most being 1-10 µm

in diameter (i). At 3-10 µm size fractions, nu-eDNA can be more detected than mt-

eDNA. Just after being released into the water, most eDNA could be intra-cellular DNA

within cells and tissues (ii). After the eDNA is released into the water, it could break

down by various degradation processes, such as hydrolysis and take-up by extra-cellular

enzymes and viral attack (iii), which would result in the shedding of nuclei and other

organelles out of degraded cell membrane (iv). Likewise, the outer nuclei membranes of

nuclei and mitochondria could also break down by environmental factors (v), and their

DNA molecules would be released (vi). These extra-cellular DNA could also be

degraded and eventually become undetectable.

A .

-

-- .

- -

-

-

- .

-

- .

--

& D - )

A()

-

. . .-

. . .-

. . .- .

97

Chapter 5. Rapid degradation of longer DNA fragments enables the improved

estimation of distribution and biomass using environmental DNA.

5.1. Introduction

Global biodiversity loss is currently one of the most critical ecological challenges,

particularly in the ocean (Dulvy et al., 2003; Worm et al., 2006), but it is generally

difficult to obtain accurate information about species distribution and population size.

For example, traditional survey methods such as visual surveys, capturing and tracking

with biotelemetry require substantial efforts and costs (Henderson et al., 1966; Brill et

al., 1993). Moreover, the accuracy of species identification depends on the observer’s

ability.

Environmental DNA (eDNA) analysis is a new monitoring method that can

overcome such problems (Ficetola et al., 2008; Minamoto et al., 2012; Takahara et al.,

2012). Environmental DNA, which is the DNA shed by organisms into the environment

(Ficetola et al., 2008; Lodge et al., 2012; Thomsen et al., 2012a), is thought to derive

from skin, urine, feces, and mucus (Martellini et al., 2005; Ficetola et al., 2008; Merkes

et al., 2014; Barnes & Turner, 2016). The presence of a target species can be estimated

by detecting eDNA from water samples without locating or capturing individuals

(Lodge et al., 2012). These advantages of eDNA analysis have enabled quick and wide-

range assessments of species presence/absence, biodiversity, and abundance in

freshwater (Thomsen et al., 2012a; Fukumoto et al., 2015; Dougherty et al., 2016;

Yamanaka & Minamoto, 2016) and marine environments (Foote et al., 2012; Thomsen

98

et al., 2012b; Port et al., 2016; Yamamoto et al., 2016).

However, some technical challenges still remain unexplored in eDNA

methodologies. For example, it is difficult to know when the detected eDNA was

released from an individual: how many hours have passed since the eDNA was shed?

Environmental DNA has been shown to persist in aquatic environments or terrestrial

soils for hours to months (Dejean et al., 2011; Goldberg et al., 2013; Barnes et al., 2014;

Merkes et al., 2014). Thus, the species that released the detected eDNA might already

be absent at the time of eDNA detection. In addition, applications of eDNA analysis to

migratory fish species require knowledge of timescale information because precise

timing and location information is required to monitor these species.

Previous studies might suggest the answer to this problem. It has been shown

that the detected copy number decreases exponentially or biphasically after removal of

the target species (Dejean et al., 2011; Barnes et al., 2014; Maruyama et al., 2014;

Eichmiller et al., 2016; Minamoto et al., 2017a), that there is a negative correlation

between the length of DNA fragments and the detected copy number (Deagle et al.,

2006) and that the difference in detection using eDNA metabarcoding might be a result

of longer persistence of the shorter 12S rRNA fragment (~100 bp) than the longer

cytochrome b (CytB) fragment (460 bp) in lake water (Hanfling et al., 2016). According

to these findings, it can be hypothesized that the decay rate of eDNA varies depending

on the length of DNA fragments: a longer DNA fragment decays more rapidly than a

shorter one. To test this hypothesis, this study compared temporal changes in the copy

number of a long eDNA fragment (719 bp) with that of a short eDNA fragment (127

99

bp), using Japanese jack mackerel (Trachurus japonicus) as a model species. The

primers and probe that targeted a longer DNA fragment than previous studies of eDNA

did were first developed. Then, rearing water from the target fish were isolated and the

copy numbers of the long and short eDNA fragments in water samples were monitored

for 48 hr. In addition to the tank experiment, longer eDNA fragments in field samples

obtained in a previous survey (Yamamoto et al., 2016) were quantified, which were

compared with the distribution of biomass estimated from echo sounder data.

5.2. Materials and Methods


In this study, two primer/probe sets that specifically amplified the Japanese jack

mackerel DNA were used, targeting two different DNA fragments of the same gene

CytB. One set of primers and probe, which targeted a short DNA fragment (hereafter

“Primer S”), was taken from Yamamoto et al. (2016). Primer S was designed to

specifically amplify a 127-bp fragment of the mitochondrial CytB gene: forward primer,

5’-CAGATATCGCAACCGCCTTT-3’ ; reverse primer, 5’-

CCGATGTGAAGGTAAATGCAAA-3’ ; probe, 5’-[FAM]-

TATGCACGCCAACGGCGCCT-[TAMRA]-3’ (Yamamoto et al., 2016). Another set of

primers and probe, which targeted a long DNA fragment (hereafter “Primer L”), was

designed to specifically amplify a 719-bp fragment of the mitochondrial CytB gene with

Primer Express 3.0 (Thermo Fisher Scientific, Waltham, MA, U.S.) with default

settings, using sequences of the Japanese jack mackerel CytB gene, which was used in

100

the previous study (Yamamoto et al., 2016), from the National Center for Biotechnology

Information.

Then the specificity of both primers was checked as follows. Each 20 µL

TaqMan reaction contained 2 µL DNA extract (one individual of Japanese jack

mackerel or Amberfish [Decapterus maruadsi], the species most closely related to the

target species in the surveyed area, was used as a template), a final concentration of 900

nM forward and reverse primers and 125 nM TaqMan probe in 1×TaqMan Gene

Expression PCR Master Mix (Thermo Fisher Scientific). Using both primer sets,

quantitative PCR (qPCR) was performed with the following conditions: 2 min at 50 °C,

10 min at 95 °C, 40 cycles of 15 s at 95 °C and 1 min at 60 °C. For each DNA sample,

qPCR was performed in duplicate. In addition, a 2 µL pure water sample was analyzed

simultaneously, in duplicate, as a negative control (PCR negative control). Quantitative

PCR was performed using a StepOnePlus Real-Time PCR system (Thermo Fisher

Scientific). Additionally, qPCR products were verified on 2 % agarose gels stained with

Midori Green (NIPPON Genetics Co, Ltd., Japan).

5.2.2. Tank experiment

5.2.2.1. Experimental set-up and water sampling

Tank experiments were conducted to verify that the decay rate of eDNA varies

depending on the length of the DNA fragments. The experiment was conducted at the

Maizuru Fisheries Research Station of Kyoto University on 9 to 11 August 2015. Three

black polycarbonate 200-L tanks were prepared and three Japanese jack mackerels were

101

kept in each tank for 1 week prior to the experiments. Total length (TL) and weight of

each Japanese jack mackerel used for this experiment were measured after the

experiment (Table 5-1). Filtered seawater, which was pumped up from 6 m depth at the

station, was used as inlet water into each tank (900 mL/min). In each tank, the

temperature was kept constant using a chiller, and aeration was performed using a

pump. Fish were fed a small amount of krill every morning until the day before water

sampling. The bottom of each tank was cleaned an hour after feeding to eliminate the

effect of the feces, and on the sampling day, the fish were starved. For sampling, 100 L

of each rearing water was transferred to other tanks from which we sampled. Soon after

isolating rearing water, 1 L of sampling tank water was collected. The time when the

first water sampling was started was defined as time 0, and the water was sampled at

0.5, 1, 1.5, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 28, 32, 36, 40, 44, and 48 hr after

time 0 (hereafter, those time points are referred as time 0.5 to 48). There were 22 total

sampling time points. At each sampling time, we also filtered 1 L of artificial seawater

as a filtration negative control. Moreover, 1 L of inlet water was sampled from each

tank at time 24 to evaluate the background Japanese jack mackerel eDNA concentration

in the inlet water, because the seawater was collected from the sea, where Japanese jack

mackerel potentially occur.

At each sampling time, the 1 L sample was immediately filtered through a 47-

mm-diameter glass microfiber filter GF/F (nominal pore size 0.7 µm; GE Healthcare

Life Science, Little Chalfont, UK). Filtering devices (i.e., filter funnels (Magnetic Filter

Funnel, 500 mL capacity; Pall Corporation, Westborough, MA, U.S.), 1-L beakers,

102

tweezers and sampling bottles used for water sampling) were bleached after every use,

using 0.1 % sodium hypochlorite solution for at least 5 min. The filters were placed in a

freezer immediately after filtration until eDNA extraction.

5.2.2.2. DNA extraction

Total eDNA was extracted from each filter using a DNeasy Blood and Tissue Kit

(Qiagen, Hilden, Germany). Briefly, a sample filter was placed in the suspended part of

a Salivette tube (Sarstedt, Numbrecht, Germany). Then, 420 µL solution, composed of

20 µL Proteinase K, 200 µL Buffer AL, and 200 µL pure water, was put on the filter and

the tube was incubated at 56 °C for 30 min. After incubation, the liquid held in the filter

was collected by centrifugation. To increase the yield of eDNA, the filter was rewashed

with 200 µL TE buffer for 1 min and the liquid was again gathered by centrifugation.

500 µL ethanol was added to the collected liquid and transferred the mixture to a spin

column. Subsequently, following the manufacturer’s instructions, total eDNA was

eluted in 100 µL AE buffer. The eDNA samples were placed in a freezer until

quantitative PCR.

5.2.2.3. Quantification of eDNA using qPCR

To evaluate the amount of eDNA derived from Japanese jack mackerel at each time

point, quantification of the copy number of CytB genes was performed using real-time

TaqMan PCR with the StepOnePlus Real-Time PCR system. To quantify the number of

Japanese jack mackerel CytB genes in each 2 µL eDNA solution sample, qPCR was

103

simultaneously performed by using a dilution series of standards containing 3 × 101 to

104 copies of a linearized plasmid that contained synthesized artificial DNA fragments

of the full CytB gene sequence of Japanese jack mackerel. In addition, a 2 µL pure

water sample was analyzed simultaneously as a negative control in the PCR (PCR

negative control). Each 13.3 µL TaqMan reaction contained 2 µL DNA extract, a final

concentration of 900 nM forward and reverse primers, and 125 nM TaqMan probe in 1

× TaqMan Gene Expression PCR Master Mix. Quantitative PCR with Primer S was

performed with the following conditions: 2 min at 50 °C, 10 min at 95 °C, 40 cycles of

15 s at 95 °C and 1 min at 60 °C. The qPCR with Primer L was performed with the

following conditions: 2 min at 50 °C, 10 min at 95 °C, 55 cycles of 15 s at 95 °C, 30 s at

60 °C and 1 min at 72 °C. All qPCRs for eDNA extract, standards and PCR negative

control were performed in triplicate. The DNA concentration of each water sample was

calculated by averaging the triplicate. All positive replicates were treated as successfully

quantified (no “limit of quantification” was set). Each replicate with non-detection

(PCR negative) was regarded as containing 0 copies (Ellison et al., 2006). The

performance of the qPCR assays is shown in Table 5-2.

A linear mixed model (LMM) was used to evaluate the differences in the

decay rate of eDNA depending on the amplification target length of each primer set

with R version 3.2.4 (R Core Team, 2016) using the function LMER of the R package

lme4 (Bates et al., 2015). In this model, log-transformed eDNA concentrations in each

tank were included as the dependent variable, and each time point (hr) and primer set

(Primer S or L) were included as explanatory variables. Tank replicates were included

104

as random effects. The slopes of the two regression lines, one based on each primer set,

should be different if a significant interaction effect of the explanatory variables is

observed. Note that, as the temperature of each tank before time 2 was higher than it

was after time 4 (Figure 5-1), the model using only the data after time 4 was also run,

because it has been shown that eDNA degrades rapidly in warmer environments

(Strickler et al., 2015; Lacoursiere-Roussel et al., 2016). The significance threshold was

set at 0.05.

5.2.3. Application to field samples

Quantification of Japanese jack mackerel’s eDNA was performed using qPCR with

Primer L. The eDNA samples used here were those used in Yamamoto et al. (2016), and

thus, eDNA concentrations with Primer S were cited from Yamamoto et al. (2016).

Seawater sampling was conducted on 18 June 2014 in west Maizuru Bay, Japan.

Seawater samples (1 L) for eDNA analyses were collected both from the sea surface

using buckets and from ~1.5 m above the bottom of the sea using Van Dorn water

samplers at 47 sites. Quantitative PCR conditions for Primer L were the same as above.

Quantitative PCR for seawater samples, standards and PCR negative control were

performed in duplicate. The DNA concentration in each water sample was calculated by

averaging the duplicates. All positive replicates were treated as successfully quantified.

Each replicate with non-detection was regarded as containing 0 copies (Ellison et al.,

2006). The performances of the qPCR assays are shown in Table 5-2. Three of the

detected DNA samples were commercially sequenced, and all were confirmed as target

105

sequences.

The correlation coefficients between echo intensities and DNA concentrations

of each primer set were calculated with R version 3.2.4. Here, echo intensity data were

also cited from Yamamoto et al. (2016), who obtained echo intensity, using a calibrated

quantitative echo sounder, as a biomass index of Japanese jack mackerel. An acoustic

survey was also conducted on 18 June 2014 in west Maizuru Bay, Japan. The echo

sounder surveys started from the mouth of the bay and moved southwest to the end of

the bay (the location of Maizuru Bay is shown in Figure 5-2). It can be assumed that

signals detected via echo sounder in June in Maizuru Bay predominantly indicated

Japanese jack mackerels (see Yamamoto et al. 2016 for detail). Five levels of horizontal

range (buffer area) and four levels of vertical range were set to define the water columns

reflecting the spatial pattern of eDNA concentration inside the bay. Horizontal ranges

were within a 10, 30, 50, 150, and 250 m radius from each sampling station, and vertical

ranges were within 2, 5 and 10 m from both the surface and bottom at each sampling

station, as well as the entire vertical range of the sea. Because neither surface nor

bottom distribution of eDNA satisfied the normality and homoscedasticity assumptions,

which was verified by performing Shapiro-Wilk and Bartlett tests (P < 0.05), Spearman

rank correlation coefficients were used for the comparison of eDNA data and Japanese

jack mackerel’s distribution. The significance threshold was the same as above. In this

analysis, the sites where no eDNA was detected with either primer set were eliminated.

106

5.3. Results


Primer L was designed as below:

forward primer, 5’-AATCCTCACAGGTCTTTTCCTAGCTA-3’;

reverse primer, 5’-ATTGATCGGAGAATGGCGTATG-3’;

probe, 5’-[FAM]-TACCATTCGTCATTGCAGCCTTCTTTGTTC-[TAMRA]-3’,

producing a 719-bp amplicon. As a result of qPCR and agarose gel electrophoresis,

Japanese jack mackerel DNA was amplified by both S and L primer sets, while

amplification of Amberfish DNA was not observed. The primer specificity was checked

using NCBI Primer Blast, and only CytB gene sequences of Japanese jack mackerel

were hit as complete match sequences to the designed primers.

5.3.2. Degradation curves for long and short amplicons

Depending on the length of the DNA fragments, slopes of the two regression lines based

on all eDNA concentrations at each time point differed significantly (P < 0.05).

Although one of filtration negative controls (at time 8) and one of the inlet water

samples showed eDNA amplification, these copy numbers were much fewer than those

of experimental tanks. In addition, all of the PCR negative controls showed no eDNA

amplification. Thus, the effects of Japanese jack mackerel eDNA included in the inlet

water and cross-contamination among samples during filtration and qPCR could be

neglected. In another model, which used only the data after time 4, the slopes of the two

regression lines also differed significantly (P < 0.001). The decay curves of primers S

107

and L were estimated as CS(t) = 507.3e-0.044t and CL(t) = 158.74e-0.09t, respectively,

where Ci(t) is eDNA concentration at time t as measured by the Primer i (S or L)

(Figure 5-3).

5.3.3. Comparison of eDNA and echo intensity in the field survey

The qPCR data from seawater samples with each primer set and echo intensity data

were compared. The distribution of Japanese jack mackerel eDNA concentrations in

west Maizuru Bay is shown in Figure 2. The copy number of eDNA differed

significantly between the surface and the bottom with both primer sets (Wilcoxon

signed rank test; P < 0.05). With Primer L, Japanese jack mackerel eDNA was detected

at 15 of 47 sites (surface) and 8 of 47 sites (bottom), while it was detected with Primer

S at 46 of 47 sites (surface) and 40 of 47 sites (bottom). For Primer S, eDNA

concentrations of surface samples were significantly higher than those of bottom

samples (P < 0.05), while for Primer L, eDNA concentrations between the surface and

the bottom showed a marginally significant difference (P = 0.051). The average

concentrations with Primer L were 25.4 copies/L (surface) and 4.7 copies/L (bottom),

while those with Primer S were 479.1 copies/L (surface) and 317.9 copies/L (bottom).

Spearman rank correlation coefficients between eDNA concentration and

echo intensity are shown in Table 5-3. On the surface, eDNA concentrations with

Primer L showed a significantly positive correlation with echo intensity of 150 or 250 m

in radius horizontally and the entire water column vertically (i.e., from surface to

bottom). These correlation coefficients were 0.61 (P = 0.02) and 0.59 (P = 0.02) for the

108

radius of 150 and 250 m, respectively. On the other hand, eDNA concentrations with

Primer S showed no significant correlation with any echo intensity data sets. For bottom

collected samples, those eDNA concentrations found with Primer L had no significant

correlations with any echo intensity data sets, while those with Primer S had a

significant negative correlation with echo intensity of 50 m in radius horizontally and 2

m vertically, and the correlation coefficient was -0.35 (P = 0.03). However, there was no

correlation between them when excluding the two outlier sites (see Discussion).

5.4. Discussion

The present study successfully showed that decay rate of eDNA varied depending on the

length of the DNA fragment. Previously, some studies have indicated that although

longer DNA fragments are present at lower concentrations in the field, they may

represent more recent biological information (Hanfling et al., 2016; Bista et al., 2017).

However, this study is the first to directly measure the degradation rates of shorter and

longer eDNA fragments. The results might expand the application of eDNA techniques

such as monitoring in time series and estimating population abundance and biomass.

In the tank experiment, a linear mixed model was used to evaluate the

differences in the decay rate of eDNA depending on the length of DNA fragments,

except the data sets before time 2 because the temperature of each tank before time 2

was higher than at later times, so I considered that eDNA data before time 2 should be

divided from those of later. Actually, eDNA decay in this experiment showed a period

of rapid decay (i.e., the initial 2 hr) followed by a period of slower decay, which is

109

considered to correspond with the change in temperature. The effect of temperature on

eDNA degradation has been shown previously (Strickler et al., 2015; Lacoursiere-

Roussel et al., 2016), and eDNA decay rate is correlated with water temperature. On the

other hand, Eichmiller et al. (2016) showed that common carp eDNA exhibited biphasic

exponential decay, characterized by rapid decay for 3 to 8 days followed by slow decay,

in spite of a constant temperature during the experiment. Further study would be needed

to clarify the underlying mechanisms.

Under the assumption that eDNA decay starts after it is shed from individuals,

eDNA concentration at time 0 should theoretically be the same regardless of the length

of the DNA fragment, but eDNA concentration at time 0 estimated with Primer S was

about 10 times as much as that estimated with Primer L. This difference in eDNA

concentration at time 0 suggests that eDNA had already degraded before it was released

into the environment. For instance, if feces are the origin of eDNA, the DNA must have

already degraded when the feces were released from the body. The two exponential

decay curves based on Primer S and L intersect with each other at t = -25.3 hr,

indicating that eDNA started to degrade the day before sampling. For example, gut cell

DNA included in feces should already be decayed before release from the body.

Similarly, other hypothetical sources of eDNA, such as mucus and epithelia (Martellini

et al., 2005; Merkes et al., 2014; Barnes & Turner, 2016), might be decayed before

shedding. The findings suggest that the time point at which DNA molecules start to

degrade is not always equal to the point when eDNA is released into the environment

from the individuals.

110

Based on previous studies, it was hypothesized that longer DNA fragments

show lower detection probabilities because longer DNA fragments could be more

damaged by environmental factors. The fragment sizes in this study were 127 and 719

bp, and other fragment sizes were not tested. The length-dependent change of DNA

decay rate could be clarified using other fragment sizes, such as ~300 and ~500 bp;

further studies are needed to clarify this.

In the field survey, targeting short DNA fragments, the copy number of

Japanese jack mackerel eDNA at the surface was significantly higher than that at the

bottom, while there was a marginally significant difference between the copy numbers

at the surface and at the bottom when targeting long DNA fragments. Thus, eDNA of

Japanese jack mackerel is distributed more at the sea surface than at the bottom. It has

been reported that when Japanese jack mackerel larvae were collected in the East China

Sea, over 95% were collected in the upper 30-m layer (Sassa & Konishi, 2006). The

differences of the eDNA distribution between the surface and the bottom in this study

may be correlated with this distribution.

The echo intensity and eDNA concentrations measured with two primer sets

(S and L) was compared to clarify whether the eDNA decay rate varies depending on

the length of DNA fragments in the field, as it was thought that these decay rates should

be the same in the field and in the tank experiment. On the sea surface, eDNA

concentrations with Primer L showed a significantly positive correlation with echo

intensity of 150 or 250 m in radius horizontally and the entire water column vertically,

while those with Primer S showed no significant correlation with any echo intensity

111

data sets. This result suggests that detection of longer eDNA can improve the accuracy

of estimations of fish distribution or biomass. Yamamoto et al. (2016) considered a

wholesale fish market in Maizuru Bay as an additional source of Japanese jack mackerel

eDNA, and they were able to evaluate a partial correlation between eDNA

concentrations and echo intensity by including the inverse of the distance of each

sampling station from the fish market as an explanatory variable in their statistical

models. On the other hand, the present study was able to evaluate a correlation without

considering any effects of the fish market. Primer S targets shorter DNA fragments that

would include “old” or “nonfresh” eDNA. Therefore, it should be more affected by

eDNA contamination from the fish market. Whereas Primer L, which targets longer

DNA fragments, can detect relatively “fresh” eDNA compared to that detected by

Primer S. Environmental DNA from the fish market should be more degraded, and

therefore, the relationships between eDNA concentration with Primer L and echo

intensity could be observed, excluding the effect of the fish market. On the sea bottom,

eDNA concentrations with Primer L showed no significant correlation with any echo

intensity data sets, while those with Primer S showed a negative correlation with echo

intensity of 50 m in radius horizontally and 5 m vertically. However, a significant

correlation was not observed for Primer S when excluding two outlier sites (St. 2 and

27). At these sites, there were much higher eDNA concentrations than at other sites,

which was referred to as “exogenous DNA” in Yamamoto et al. (2016). In particular, St.

2 is close to the fish market, which was considered a major source of Japanese jack

mackerel eDNA (Yamamoto et al., 2016). Also at St. 27, for instance, eDNA might be

112

released from dead individuals that may accumulate there due to the specific features of

the site such as seafloor dips or rocks. It has already been reported that high

concentrations of eDNA from silver carp carcasses can be detected for at least 28 days

(Merkes et al., 2014), so the release of eDNA from carcasses might be possible.

Contrastingly, eDNA concentrations with Primer L at these sites were very low or zero,

suggesting that this is “nonfresh” eDNA; eDNA from carcasses or from the fish market

has already been degraded when released.

Previous studies have focused on the influences of environmental factors on

eDNA persistence (Dejean et al., 2011; Thomsen et al., 2012a; Barnes et al., 2014;

Strickler et al., 2015). In this study, for instance, eDNA degradation might have been

slowed at lower temperatures (Strickler et al., 2015), UV radiation might damage DNA

nucleic acids (Pilliod et al., 2014), and water chemistry might also influence eDNA

persistence (Barnes et al., 2014; Eichmiller et al., 2016). However, it remains unknown

how these environmental factors can influence eDNA persistence in the field, especially

in marine environments. Answering these questions would be important when applying

eDNA analysis to field surveys.

The present study successfully showed that the decay rate of eDNA varied

depending on the length of the DNA fragment, and the findings showed the possibility

of obtaining timescale information from eDNA. With primer sets that target longer

DNA fragments than in previous eDNA studies, newly released eDNA can selectively

be detected. Such longer eDNA fragments indicate fresher biological information in the

field. Thus, by selecting the detected fragment length, we can extract timescale

113

information from eDNA. For instance, detection of longer eDNA fragments enables us

to obtain more accurate distribution information (Hanfling et al., 2016; Bista et al.,

2017), which would contribute to revealing the route of migratory organisms. Various

fish species are known to migrate on different scales (Heard, 1991; Arai et al., 1999;

Yamanaka & Minamoto, 2016). The timescale information obtained using the results of

the study may enable us to understand the details of fish migration. On the other hand,

the primer/probe sets in this study targeted a CytB gene of Japanese jack mackerel that

might be too long to be sufficiently informative. The primer/probe sets that target a

shorter fragment size than Primer L and longer than Primer S (e.g., 300 to 500 bp)

would be more informative and also detectable for a reasonable period of time.

Detection of longer eDNA fragments might be able to dramatically improve the study of

ecological monitoring.

114

5.5. Tables

Table 5-1. Total length (TL) and weight of Japanese jack mackerel in the tank

experiment.

Tank ID TL [cm] weight [g]

1 14.9 ± 0.75 38.82 ± 6.32

2 15.0 ± 1.11 45.01 ± 5.18

3 14.7 ± 1.59 39.28 ± 3.65

Values of TL and weight are represented as mean ± 1 SD.

115

Table 5-2. R2 values, slopes, and Y intercepts of the calibration curves and the polymerase chain reaction (PCR) efficiencies for each

qPCR experiment performed in this study.

N R2 Slope Y intercept PCR efficiency [%]

Tank experiment 4 0.996 ± 0.002 -3.414 ± 0.037 42.013 ± 0.120 96.319 ± 1.425

(Primer S) Tank experiment

4 0.963 ± 0,022 -3.705 ± 0.167 45.379 ± 0.803 86.451 ± 4.945 (Primer L) Field survey

3 0.957 ± 0.004 -4.200 ± 0.254 47.310 ± 1.524 73.487 ± 5.919 (Primer L)

These values are represented as mean ± 1 SD.

116

Table 5-3. Spearman’s rank correlation coefficients between eDNA concentration and echo intensity of various water column size

(horizontal range/vertical range) for surface (left) and bottom (right).

Surface 10 m/2 m 10 m/5 m 10 m/10 m 10 m/Ec Bottom 10 m/2 m 10 m/5 m 10 m/10 m 10 m/Ec

Primer L r -0.04 0.06 -0.15 -0.43

Primer L r 0.21 0.14 0.14 0.14

P 0.88 0.82 0.60 0.11 P 0.62 0.75 0.75 0.75

Primer S r -0.02 -0.09 -0.15 -0.20

Primer S r -0.23 -0.24 -0.29 -0.28

P 0.88 0.55 0.32 0.17 P 0.15 0.13 0.07 0.08

30 m/2 m 30 m/5 m 30 m/10 m 30 m/Ec 30 m/2 m 30 m/5 m 30 m/10 m 30 m/Ec

Primer L r 0.05 0.15 0.31 0.23

Primer L r 0.36 0.24 0.19 -0.26

P 0.85 0.58 0.26 0.41 P 0.39 0.58 0.66 0.54

Primer S r -0.05 -0.05 -0.02 0.00

Primer S r -0.26 -0.15 -0.10 0.08

P 0.75 0.74 0.88 1.00 P 0.11 0.36 0.53 0.62


Primer L r -0.09 0.26 0.20 0.34

Primer L r 0.33 0.24 0.21 -0.24

P 0.74 0.35 0.47 0.22 P 0.43 0.58 0.62 0.58

Primer S r -0.01 0.03 0.07 -0.07

Primer S r -0.35 -0.26 -0.04 -0.04

P 0.93 0.84 0.66 0.63 P 0.03 0.11 0.81 0.80


117

Primer L r 0.09 0.18 0.24 0.61

Primer L r 0.12 0.10 0.19 -0.26

P 0.75 0.52 0.38 0.02 P 0.79 0.84 0.66 0.54

Primer S r 0.10 0.20 0.20 0.09

Primer S r -0.16 -0.20 0.03 -0.03

P 0.51 0.19 0.19 0.53 P 0.34 0.23 0.84 0.84


Primer L r 0.33 0.15 0.20 0.59

Primer L r 0.19 0.19 -0.02 -0.24

P 0.24 0.60 0.47 0.02 P 0.66 0.66 0.98 0.58

Primer S r 0.01 0.07 0.26 0.24

Primer S r 0.15 0.12 0.25 0.13

P 0.97 0.63 0.07 0.11 P 0.36 0.47 0.12 0.43

Note: ‘r’ means Spearman’s rank correlation coefficients between target eDNA concentration and echo intensity, and ‘P’ means the P

values of corresponding Spearman’s rank correlation coefficients. Statistically significant correlations are shown in bold.

118

5.6. Figures

Figure 5-1. The shift of water temperature in the tank experiment. Each line (solid red, dotted blue, and dashed green) shows the shift in water temperature in each tank.

Because the inlet water was drawn from the sea, the water temperature in the tanks

fluctuated with changes in the temperature of the sea water.

0 10 20 30 40 50

2526

2728

2930

0 10 20 30 40 50

2526

2728

2930

0 10 20 30 40 50

2526

2728

2930

time point [h]

wat

er te

mp.

[°C

]

119

Figure 5-2. The distribution of Japanese jack mackerel eDNA concentrations and echo

intensity in west Maizuru Bay (surface and bottom). The level of the estimated eDNA

concentrations is indicated by colours between red (relatively high concentration) and

blue (low concentration or zero), as well as the echo intensity by echo sounder as

indicated by colours between dark yellow (relatively high intensity) and white (low

intensity or zero). Grey areas indicate land masses. Spatial approximation was

performed using a regularized spline with a tension parameter of 40.

120

Figure 5-3. Decay curves for Japanese jack mackerel eDNA in the tank experiments.

Dots show eDNA concentrations (average of triplicate) at each time point (blue: Primer

S, red: Primer L), and solid lines show regression curves excluding the initial 2 hr of

data. Error bars show standard deviation (SD).

0 10 20 30 40 50

-10

12

34

0 10 20 30 40 50

-10

12

34

0 10 20 30 40 50

-10

12

34

0 10 20 30 40 50

-10

12

34

time point [h]

log1

0(eD

NA

con

c.) [

copi

es/2

µL]

121

Chapter 6. Selective collection of environmental DNA with long fragment using

larger filter pore size.

6.1. Introduction

For the rapid and extensive detection of threatened rare and invasive species in the

aquatic environment, environmental DNA (eDNA) analysis has recently been developed

(Takahara et al., 2012; Bohmann et al., 2014; Deiner et al., 2017a). Environmental DNA

is defined as the genetic materials in environment derived from mucus, feces, skin,

scale, and gametes of organisms (Barnes & Turner, 2016). The detection of eDNA infers

the presence of target species without capturing or observing them, and thus analyzing

eDNA is less-invasive and more cost-effective than traditional methods (Darling &

Mahon, 2011; Thomsen & Willerslev, 2015). Ever since the first inception of eDNA

analysis (Ficetola et al., 2008), its applicability has been demonstrated for various taxa

and environments (Minamoto et al., 2012; Thomsen et al., 2016; Bista et al., 2017;

Ushio et al., 2017; Carraro et al., 2018; Nichols & Marko, 2019).

Some studies have previously reported that aqueous eDNA can be detected at

various size fractions (<0.2 to >180 µm) (Turner et al., 2014; Wilcox et al., 2015; Jo et

al., 2019a). This implies that eDNA can exist with various physical and physiological

states in water (Barnes & Turner, 2016). With regards to macro-organisms' eDNA, not

only cell and tissue fragments but also nuclei and mitochondria, and even extra-

membrane nuclear and mitochondrial DNA can potentially be detected. Among them,

cell and tissue fragments are likely to be detected as intra-cellular DNA and at larger

122

size fractions than extra-cellular DNA, which in contrast would be detected at smaller

size fractions (Jo et al., 2019a). In addition, owing to the presence of a cellular

membrane, such large-sized and intra-cellular eDNA may be protected from various

DNA degradation processes (e.g., enzymatic and mechanical fragmentation by

microbes) compared with small-sized and extra-cellular eDNA (Nielsen et al., 2007;

Torti et al., 2015). Deiner et al. (2017b) reported the successful sequencing of fish

mitochondrial genomes (>16 kbp) from water samples, which might have attributed to

the presence of large-sized eDNA that was covered with the cellular membrane. Thus, it

can be hypothesized that selective collection of large-sized eDNA results in the effective

collection of the less-degraded eDNA.

Here, the present study verified the aforementioned hypothesis by the

filtration using filters with different pore sizes (described below). That is, water

filtration with a larger pore size filter leads to the selective collection of the eDNA at

larger size fractions. Water samples were collected from a tank in which Japanese jack

mackerels (Trachurus japonicus) were kept, and the copy numbers of short and long

mitochondrial DNA fragments were quantified in water samples. It is expected that the

filtration with a larger pore size filter would increase the relative yield of long DNA

fragments (i.e., the ratio of long to short DNA fragments) from water samples.

Moreover, the copy number of short nuclear DNA fragment was also quantified in water

samples. The findings of some previous studies have indicated that the persistence of

nuclear eDNA in water could be lower than that of mitochondrial eDNA (Jo et al.,

2019a, 2019b), and that the persistence of nuclear and mitochondrial DNA could differ

123

between intra- and extra-cellular environments using tissue samples (Foran, 2006).

Accordingly, it is conceivable that the yields of nuclear and mitochondrial eDNA and

their ratios may differ depending on size fractions and filter pore sizes.


6.2.1. Water sampling

Seawater was sampled from a 500-L tank, in which around 30 individuals of Japanese

jack mackerels were kept, at the Maizuru Fisheries Research Station (MFRS) of Kyoto

University, Japan, in September 2019 (Figure 6-1). This species was used because the

primers/probe sets targeting its mitochondrial and nuclear DNA with different fragment

sizes were available (Jo et al., 2017, 2019b; Yamamoto et al., 2016). The tank was

aerated by a pump, and filtered seawater, which was pumped from 6 m depth off the

coast of the station, was used as the inlet water into the tank. 10 replicates of 100, 250,

500, and 1000 mL of tank water samples were collected using 1.3 L plastic bottles. The

five replicates of each volume of water samples were randomly filtered with a 47 mm-

diameter glass microfiber filter GF/F (nominal pore size 0.7 µm; GE Healthcare Life

Science, U.K.), and the other five replicates with a 47 mm-diameter glass microfiber

filter GF/D (nominal pore size 2.7 µm; GE Healthcare Life Science). The water

temperature was 25.5 °C when collecting water samples. 500 mL of distilled water was

simultaneously filtered as filtration negative controls and 500 mL of inlet water into the

tank was filtered to evaluate the background concentrations of target eDNA using both

filters. Throughout the sampling, disposable gloves were worn, and the filtering devices

124

(i.e., filter funnels [Magnetic Filter Funnel, 500 mL capacity; Pall Corporation,

Westborough, MA, U.S.], 1 L beakers, tweezers, and sampling bottles) were bleached

before every use in 0.1% sodium hypochlorite solution for at least 5 min (Yamanaka et

al., 2017). We kept all the filter samples at -20 °C until DNA extraction.

6.2.2. DNA extraction and quantitative real-time PCR

Total eDNA on the filter was extracted by DNeasy Blood and Tissue Kit (Qiagen,

Germany) following Jo et al. (2017). Japanese jack mackerel eDNA concentration in

water samples was estimated by quantifying the copy number of mitochondrial

cytochrome b (CytB) genes and nuclear internal transcribed spacer-1 (ITS1) regions of

ribosomal RNA genes using the StepOnePlus Real-Time PCR system (Thermo Fisher

Scientific, U.S.). We used three primers/probe sets that specifically amplified the 164-

bp fragment of CytB gene (mtS), the 682-bp fragment of CytB gene (mtL), and the 164-

bp fragment of ITS1 region (nuS) from the target species (Table 6-1). The species-

specificity of each primers/probe set had already been in vitro checked (Yamamoto et

al., 2016; Jo et al., 2017, 2019b). Each 20 µL of TaqMan reaction contained a 2 µL

template DNA, a final 900 nM concentration of each forward and reverse primer, and

125 nM of TaqMan probe in 1 × TaqPathTM qPCR Master Mix, CG (Thermo Fisher

Scientific). 2 µL of pure water was simultaneously analyzed as PCR negative controls.

qPCRs were performed by using a dilution series of standards containing 3 × 101 to 3 ×

104 copies of a linearized plasmid containing synthesized artificial DNA fragments from

a full CytB gene (1141 bp) or partial ITS1 region (237 bp) of the target species (Jo et

125

al., 2019b). All qPCRs for eDNA samples, standards, and negative controls were

performed in triplicate. Thermal conditions of quantitative real-time PCR were as

follows: 2 min at 50 °C, 10 min at 95 °C, 55 cycles of 15 s at 95 °C , and 1.5 min at

60 °C for mtS and nuS (2-step PCR), and 2 min at 50 °C, 10 min at 95 °C, 55 cycles of

15 s at 95 °C, 30 s at 60 °C, and 1 min at 72 °C for mtL (3-step PCR). Concentrations of

target eDNA were calculated by averaging the triplicate, and each PCR-negative

replicate (indicating non-detection) was regarded as containing zero copies (Ellison et

al., 2006).


For each type of eDNA (mtS, mtL, and nuS), linear relationships were characterized

between eDNA concentrations (log-transformed) and the volume of water filtration

(log-transformed). The interactions between covariates (filtration water volume) and

factors (filter types) were significant or marginal (see Results and Discussion), which

meant that the effect of covariates on eDNA concentrations was different between

factors. Thus, analyses of covariance (ANCOVA) could not be applied to the dataset

and, instead, log-transformed eDNA concentrations between filter pore sizes were

compared for each filtration water volume using Student's t-test. In addition, the ratios

of long to short mitochondrial eDNA (i.e., mtL: mtS) and short nuclear to short

mitochondrial eDNA (i.e., nuS: mtS) concentrations were calculated, and the ratios were

compared between the filters using Mann-Whitney's U test. Moreover, the coefficients

of variations (CVs; standard deviations divided by mean values) were calculated for

126

each type of filter and eDNA. For the calculation of the ratios and CVs, the raw eDNA

concentrations were used and the eDNA data from different filtration volumes was

pooled to increase the sample size. All the statistical analyses were performed by R

version 3.6.1 (R Core Team, 2019).

6.3. Results and Discussion

The eDNA concentrations in inlet water samples and filtration negative controls were at

most 28.8 and 1.0 copies per PCR reaction respectively, which was negligible relative to

those in tank water samples (Table 6-2). No amplification was observed in any of the

PCR negative controls. DNA concentrations in all tank water samples were larger than

30 copies/reactions, which is the lowest value of a dilution series of standards. PCR

information for each type of eDNA is shown in Table 6-3.

6.3.1. The ratio of long to short mitochondrial eDNA

The ratio of long to short mitochondrial eDNA concentrations (mtL: mtS) was

significantly higher for GF/D (2.7 µm pore size) than GF/F filters (0.7 µm pore size) (P

= 0.0020; Figure 6-2). As expected, the use of a larger pore size filter increased the

relative yield of long DNA fragments, which is the most important finding in the study.

After released by the organisms, aqueous eDNA is degraded by mainly microbes and

extra-cellular enzymes (Barnes et al., 2014; Collins et al., 2018). Its persistence in water

can be lower for longer DNA fragments (Bista et al., 2017; Jo et al., 2017), while it is

considered that, due to its cellular membrane, DNA fragmentation can be suppressed for

127

intra-cellular DNA relative to extra-cellular DNA in natural environment (Nielsen et al.,

2007; Torti et al., 2015). Thus, the result would indicate the increase of the relative yield

of less-degraded eDNA by the selective collection of large-sized and intra-cellular

DNA. Such long DNA fragments in water may have the potential to improve the

identification of closely related species and the evaluation of intra-specific genetic

diversity based on eDNA analysis (Uchii et al., 2016; Sigsgaard et al., 2017; Williams et

al., 2019). Also, the detection of long DNA fragments may provide a more precise

temporal inference of an organism's biomass/abundance contrary to short DNA

fragments (Jo et al., 2017). As only two kinds of filter pore sizes were tested in this

study, future studies using various pore sizes of filters will strengthen the findings.

6.3.2. The ratio of nuclear to mitochondrial eDNA

The ratio of nuclear to mitochondrial eDNA concentrations (nuS: mtS) was significantly

lower for GF/D than GF/F filters (P < 0.0001; Figure 6-3). There was little difference in

yields between nuclear and mitochondrial eDNA using GF/F, whereas the yield of

nuclear eDNA tended to be lower than that of mitochondrial eDNA using GF/D (Table

2). Considering that a GF/D filter (2.7 µm pore size) could mainly capture the intra-

cellular eDNA while a GF/F filter (0.7 µm pore size) could capture both intra- and

extra-cellular eDNA, it is likely that extra-cellular nuclear eDNA from target species is

more abundant in rearing water than extra-cellular mitochondrial eDNA. So far as being

covered with its non-porous membranes, mitochondrial DNA could be more persistent

to enzymatic activities than nuclear one, which is covered with porous membrane

128

(Ernster & Schatz, 1981; Ellenberg et al., 1997). However, once having lost their

membranes, extra-cellular mitochondrial DNA might be degraded faster than extra-

cellular nuclear DNA in water. Foran (2006) reported that the degradation of tissue-

derived DNA was faster for nuclear than mitochondrial DNA without homogenization

(assuming intra-cellular DNA), whereas the result was reversed in homogenized tissues

(assuming extra-cellular DNA) and the degradation of mitochondrial DNA was faster

than that of nuclear DNA. Therefore, the result in the study might be partly attributed to

the reversal in the degradative vulnerability of mitochondrial and nuclear DNA between

intra- and extra-cellular environments (Foran, 2006).

Alternatively, the fusion of mitochondria in cells to form dynamic inter-

connecting networks for the maintenance of their integrities (Koshiba et al., 2004; Suen

et al., 2008) would possibly have brought the result; there could be some mitochondria

larger than the nuclei due to the fusion, which might decrease the omission of

mitochondrial eDNA with an increase in filter pore size relative to nuclear one. Further

studies would be needed to investigate how the cellular environment and DNA

structure, could physically, chemically, and biologically influence the persistence of

mitochondrial and nuclear DNA. It would help understand the characteristics and

dynamics of nuclear and mitochondrial eDNA in water.

6.3.3. The difference of eDNA capture efficiencies between filters

According to results of Student's t-tests, all types of eDNA concentrations were

significantly higher for GF/F (pore size: 0.7 µm) than GF/D filters (pore size: 2.7 µm) in

129

1000 mL of filtration water volume (all P < 0.01; Table 6-4). In addition, short nuclear

eDNA concentrations were significantly higher for GF/F than GF/D filters in any of the

filtration volumes (all P < 0.05), while short and long mitochondrial eDNA

concentrations did not significantly differ between filter pore sizes in <500 mL of

filtration volume (P > 0.1). All types of eDNA concentrations were generally higher for

the smaller pore size filter, whereas the differences of eDNA concentrations tended to

be unclear when smaller volume of water samples was filtered (Figure 6-4). In a natural

environment especially with high turbidity, larger filter pore size enables to prevent a

filter clogging and to increase the filtration efficiencies (Robson et al., 2016; Wilson et

al., 2014), and thus the lower capture efficiency of GF/D may not be the major problem.

For example, the yield of eDNA by GF/F filtrations of 100 mL water samples can be

recovered by GF/D filtrations of at most 250 mL water samples (Figure 6-4). It would

also result in the increase of the relative yield of long DNA fragments from water

samples.

In contrast, higher CVs for GF/D relative to GF/F filters were observed in all

types of eDNA (Table 6-5). A similar result was reported previously (Minamoto et al.,

2016), and it may decrease the precision of biomass estimation based on eDNA analysis

(Mauvisseau et al., 2019). The heterogeneous distribution of aqueous eDNA is likely

attributed to the large-sized eDNA such as aggregation of cells and tissues (Turner et al.,

2014; Furlan et al., 2016; Song et al., 2017). Therefore, selective collection of such

large-sized eDNA might possibly disperse the eDNA concentrations; however, the

present study was unable to determine the statistical significance of differences, as the

130

findings were based on samples collected from a single experimental tank. For the

improvement of eDNA-based biomass estimation, the effect of filter pore size and

materials on the precision of eDNA quantification would be needed, which will be a

focus of future studies.

6.4. Conclusions

So far, the filter with larger pore size has been used in eDNA studies to prevent the filter

clogging and to increase the filtration volume (Wilson et al., 2014; Robson et al., 2016)

except for Fremier et al. (2019), which focused on the transport of only intra-cellular

DNA using a larger pore size filter. However, the present study showed that the use of

larger pore size filter could improve the relative capture efficiency of long DNA

fragments from water samples. The study suggests that the selective collection of the

specific size, and possibly the state, of aqueous eDNA may allow to improve the eDNA-

based taxonomy and biomass/abundance estimation. In addition, the ratio of nuclear to

mitochondrial eDNA concentrations varied depending on filter pore sizes. Although

further study would be needed from physiological and cytological aspects, the findings

may reflect the potential difference of nuclear and mitochondrial DNA persistence

between cellular and the aquatic environment.

There remain some issues to be verified in this study. First, all the eDNA data

was based on a single experimental tank, and thus the results might be less statistically

robust, even though a large number of water filtration and PCR replicates were

assessed. Second, it would be necessary to verify whether the approach used in the

131

present study is practically applicable to natural environments. As mentioned above, in

the environment difficult for water filtration (e.g., high turbidity), the use of larger pore

size filter to improve the relative capture efficiency of long DNA fragments might be

more efficient. Third, as mentioned above, the findings of the present study were based

on the analysis using filters of only two pore sizes, and thus future analyses using filters

of a larger range of pore sizes would conceivably contribute to the robustness of the

results.

Nevertheless, as far as I know, this is the first report to show the possibility to

control the ecological, and possibly physiological, information from eDNA by utilizing

the knowledge of its size and state in water. Some studies suggested that the

combination of nuclear and mitochondrial eDNA could imply the spawning activity and

the age structure of fish (Bylemans et al., 2017; Jo et al., 2019b), and that the

combination of long and short DNA fragments in water allowed to remove the effect of

eDNA from carcasses and to obtain fresher ecological information (Jo et al., 2017). By

targeting the large-sized and less-degraded eDNA, the results shown in these studies

might be more obvious not only in mesocosm but also in a natural environment. The

selective collection of eDNA based on its size and state would be able to extend the

eDNA applications for ecological monitoring in the future.

132

6.5. Tables

Table 6-1. Primers/probe sets used in this study.

ID Target region Sequences (5ʹ → 3ʹ)

Amplicon size

with the forward

primer (bp)

Tm

(°C) Reference

Tja_CytB_F

mitochondrial

cytochrome b (CytB)

CAG-ATA-TCG-CAA-CCG-CCT-TT 58.7 Yamamoto et al. (2016)

Tja_CytB_R164 TTC-TTT-GTA-GAG-GTA-CGA-GCC-G 164 59.8 Jo et al. (2019a)

Tja_CytB_R682 ATT-GAT-CGG-AGA-ATG-GCG-TAT 682 57.3 Jo et al. (2017)

Tja_CytB_P [FAM]-TAT-GCA-CGC-CAA-CGG-CGC-CT-[TAMRA] 67.9 Yamamoto et al. (2016)

Tja_ITS1_F nuclear

internal transcribed spacer-1

(ITS1)

GCG-GGT-ACC-CAA-CTC-TCT-TC 60.1

Jo et al. (2019a) Tja_ITS1_R CCT-GAG-CGG-CAC-ATG-AGA-G 164 63.2

Tja_ITS1_P [FAM]-CTC-TCG-CTT-CTC-CGA-CCC-CGG-TCG-[TAMRA] 70.8

Note that the forward primer and TaqMan probe were shared for the primers/probe sets targeting mtS and mtL, and only reverse primers

were exchanged to adjust the length of PCR amplicon.

133

Table 6-2. PCR information for each type of eDNA (mtS, mtL, and nuS) (mean ± 1 SD).

Type of eDNA Slope Y-intercept R2 value Efficiency [%]

Short mitochondrial -3.351 ± 0.009 38.456 ± 0.014 0.999 ± 0.001 98.808 ± 0.361 Long mitochondrial -3.415 ± 0.147 40.894 ± 0.481 0.982 ± 0.006 96.578 ± 5.724 Short nuclear -3.512 ± 0.048 40.202 ± 0.574 0.997 ± 0.000 92.655 ± 1.721

134

Table 6-3. Results of Student's t-tests for comparisons of eDNA yields between filter

pore sizes (GF/F and GF/D).

eDNA type Filtration volume P value

Short mt-eDNA

100 mL 0.1359 250 mL 0.8455 500 mL 0.1130 1000 mL 0.0016 **

Long mt-eDNA

100 mL 0.2467 250 mL 0.6721 500 mL 0.1474 1000 mL 0.0051 **

Short nu-eDNA

100 mL 0.0108 * 250 mL 0.0312 * 500 mL 0.0063 ** 1000 mL 0.0007 ***

Note: Asterisks represent the statistical significances of the parameters (∗∗∗, P < 0.001;

**, P < 0.01; ∗, P < 0.05). All eDNA concentrations were log-transformed.

135

Table 6-4. The results of the coefficients of variations (CVs) for each type of eDNA and

filter.

Type of eDNA Type of filters CVs [%]

Short mitochondrial GF/F 40.0 GF/D 46.5

Long mitochondrial GF/F 38.2 GF/D 52.0

Short nuclear GF/F 39.8 GF/D 44.5

136

6.6. Figures

Figure 6-1. Overall flowchart of the tank experiment. 1000, 500, 250, and 100 mL of

rearing water samples were collected from a 500-L tank, in which Japanese jack

mackerels were kept, and then the samples were randomly assigned into two groups:

one filtered by GF/F (nominal pore size of 0.7 µm) and the other filtered by GF/D

(nominal pore size of 2.7 µm). After the filtration, the copy number of short

mitochondrial, long mitochondrial, and short nuclear eDNA were quantified in filter

samples.

Water volume: 500 L

1000 mL×10 rep.

500 mL×10 rep.

250 mL×10 rep.

100 mL×10 rep.

Water sampling

Water filtration

×5 rep.

GF/F 0.7 µm pore size

×5 rep.

GF/D 2.7 µm pore size

Model speciesJapanese jack mackerel(Trachurus japonicus)

137

Figure 6-2. The comparison of the ratio of long (682-bp) to short (164-bp)

mitochondrial eDNA concentrations between GF/F and GF/D. The boxplots were drawn

based on the raw eDNA concentrations. The double star represents the significant

difference (P < 0.01) between filter types by Mann-Whitney's U test.

GFD GFF

0.4

0.5

0.6

0.7

0.8

Rat

io o

f mt-e

DN

A (6

82 b

p / 1

64 b

p)

Filter type

**

GFD GFF

0.4

0.5

0.6

0.7

0.8

Rat

io o

f mt-e

DN

A (6

82 b

p / 1

64 b

p)

Filter type

**

GF/D (pore size 2.7 µm)

GF/F(pore size 0.7 µm)

138

Figure 6-3. The comparison of the ratio of nuclear to mitochondrial (both 164-bp)

eDNA concentrations between GF/F and GF/D. The boxplots were drawn based on the

raw eDNA concentrations. The triple star represents the significant difference (P <

0.001) between filter types by Mann-Whitney's U test.

GFD GFF

0.0

0.5

1.0

1.5

2.0

2.5

Rat

io o

f nu-

eDN

A to

mt-e

DN

A (1

64 b

p)

Filter type

***

GFD GFF

0.4

0.5

0.6

0.7

0.8

Rat

io o

f mt-e

DN

A (6

82 b

p / 1

64 b

p)

Filter type

**

GF/D (pore size 2.7 µm)

GF/F(pore size 0.7 µm)

139

Figure 6-4. Relationships between eDNA concentrations for each type (mtS, mtL, and

nuS) and filtration water volume (both log-transformed) using different filters, GF/F

and GF/D. Blue and red plots are derived from GF/F and GF/D, respectively.

Regression lines and their 95 % confidence intervals (CIs) for each plot are shown by

solid and dotted lines, respectively. R2 values represent the fitness of each regression

line.

2.0 2.2 2.4 2.6 2.8 3.0

2.5

3.0

3.5

4.0

4.5

5.0

2.0 2.2 2.4 2.6 2.8 3.0

2.5

3.0

3.5

4.0

4.5

5.0 Short mt-eDNA

2.0 2.2 2.4 2.6 2.8 3.0

2.5

3.0

3.5

4.0

4.5

5.0

2.0 2.2 2.4 2.6 2.8 3.0

2.5

3.0

3.5

4.0

4.5

5.0 Long mt-eDNA

2.0 2.2 2.4 2.6 2.8 3.0

2.5

3.0

3.5

4.0

4.5

5.0

2.0 2.2 2.4 2.6 2.8 3.0

2.5

3.0

3.5

4.0

4.5

5.0 Short nu-eDNA

log1

0(eD

NA

con

c.) [

copi

es/2

µL

tem

plat

e D

NA

]

log10(filtration volume) [mL]

GF/F (0.7 µm)R2 = 0.8949

GF/D (2.7 µm) R2 = 0.7285

GF/F (0.7 µm)R2 = 0.8934

GF/D (2.7 µm)R2 = 0.7001

GF/F (0.7 µm)R2 = 0.8875

GF/D (2.7 µm)R2 = 0.7427

140

Chapter 7. Complex interactions between environmental DNA (eDNA) state and

water chemistries on eDNA persistence suggested by meta-analyses.

7.1. Introduction

Organisms release their DNA molecules into their surroundings, which are termed as

environmental DNA (eDNA) (Levy-Booth et al., 2007; Nielsen et al., 2007; Taberlet et

al., 2012). The analysis of eDNA has recently been applied to monitor the abundance

and composition of macro-organisms, such as fish and amphibians (Ficetola et al., 2008;

Minamoto et al., 2012; Bohmann et al., 2014; Deiner et al., 2017; Jo et al., 2020a).

Detection of eDNA in water samples does not involve any damage to the target species

and their habitats, thus enabling non-invasive and cost-effective monitoring of species

in aquatic environments, contrary to traditional monitoring methods such as capturing

and observing (Darling & Mahon, 2011). However, the characteristics and dynamics of

eDNA are not yet completely understood, and thus, the spatiotemporal scale of eDNA

signals at a given sampling time and location is not certain, which can result in false-

positive or false-negative detection of eDNA in natural environments (Darling &

Mahon, 2011; Hansen et al., 2018; Beng & Corlett, 2020).

To determine the spatiotemporal scale of eDNA signals and accurately

estimate species presence/absence and abundance in the environment, understanding the

processes of eDNA persistence and degradation is important. Aqueous eDNA is

detectable from days to weeks (Barnes & Turner, 2016; Collins et al., 2018), depending

on various environmental factors. For example, moderately high temperature (Strickler

141

et al., 2015; Eichmiler et al., 2016; Lance et al., 2017; Jo et al., 2020b) and low pH

(Strickler et al., 2015; Lance et al, 2017; Seymour et al., 2018) accelerate eDNA

degradation. In addition, eDNA decay rates are higher in environments with higher

species biomass density (Bylemans et al., 2018a; Jo et al., 2019a). These abiotic and

biotic factors contribute to the increase in microbial activities and abundance in water,

thus indirectly affecting eDNA degradation (Strickler et al., 2015). Moreover, eDNA

decay rates were found to be different between the trophic states of studied lakes, and

were negatively correlated with the dissolved organic carbon (DOC) concentrations

(Eichmiller et al., 2016). This may be attributed to the binding of DNA molecules to

humic substances, protecting eDNA from enzymatic degradation.

However, apart from the effects of such environmental conditions, little is

known about the influence of the physiochemical and molecular states of eDNA on its

persistence and degradation. Fish eDNA has been detected at various size fractions

(<0.2 µm to >180 µm in diameter; Turner et al., 2014; Jo et al., 2019b) in water,

suggesting that eDNA is present as various states and cellular structures, from larger-

sized and intra-cellular DNA (e.g., cell and tissue fragments) to smaller-sized and extra-

cellular DNA (e.g., organelles and dissolved DNA). Enzymatic and chemical

degradation of DNA molecules in the environment depends on the presence of cellular

membranes around the DNA molecules, and thus, the persistence of eDNA is likely to

be linked to its state. In addition, eDNA persistence may be different depending on the

target genetic regions. Recent studies have suggested that eDNA decay rates may vary

between mitochondrial and nuclear DNA (Bylemans et al., 2018a; Moushomi et al.,

142

2019; Jo et al., 2020b). Moreover, studies comparing eDNA degradation between

different target DNA fragment lengths (i.e. PCR amplification length) have yielded

inconsistent conclusions; Jo et al. (2017) and Wei et al. (2018) reported higher eDNA

decay rates for longer DNA fragments, whereas Bylemans et al. (2018a) did not observe

any difference in the eDNA decay rates of different DNA fragment sizes. Notably, Jo et

al. (2020c) reported that selective collection of larger-sized eDNA using a larger pore

size filter increased the ratio of long to short eDNA concentrations and altered the ratio

of nuclear to mitochondrial eDNA concentrations; however, such reports linking eDNA

state to its persistence are scarce.

Although our understanding of the relationship between eDNA state and

persistence is currently limited, this relationship can be inferred by integrating previous

findings of eDNA persistence and degradation. Here, meta-analyses were used to

examine the relationship between eDNA states and persistence. The present study

extracted data on filter pore size, DNA fragment size, target gene, and environmental

parameters from previous studies estimating first-order eDNA decay rate constants, and

investigated the influence of these factors on eDNA degradation. By assembling and

integrating the results of previous eDNA studies, the meta-analyses revealed the hitherto

unknown relationships between eDNA state and persistence, which could not have been

observed in the individual studies. Furthermore, the validity of the findings of the meta-

analyses was assessed by re-analyzing the dataset from a previous tank experiment (Jo

et al., 2019b).

143


7.2.1. Literature search and data extraction

I searched for literature relating to eDNA persistence and degradation, published during

2008 to 2020 (final date for the literature search was 20 Jun 2020), using Google

Scholar (https://scholar.google.co.jp/). The terms “eDNA” or “environmental DNA”,

included in the title and/or text, were used for the literature search. I then filtered and

selected papers that (i) targeted eDNA from macro-organisms (i.e. not from microbes,

fungi, plankton, virus, and bacteria), (ii) were written in English, (iii) were peer-

reviewed (i.e. not preprints), and (iv) described aqueous eDNA decay rate constants

using a first-order exponential decay model (#$ = #&'()$, where #$ is the eDNA

concentration at time *, #& is the initial eDNA concentration, and + is the first-order

decay rate constant). The eDNA decay rate constants estimated using multi-phasic

exponential decay models (e.g. biphasic or Weibull models) (Eichmiller et al., 2016;

Bylemans et al., 2018a; Wei et al., 2018) were not included in the meta-analyses,

because of the limited number of such studies and difficulty in directly comparing the

constants between first-order and multi-phasic models.

From the filtered eDNA studies, I extracted data on the eDNA decay rate

constant (per hour), filter pore size used for water filtration (µm), target DNA fragment

size (base pair; bp), and target gene (mitochondrial or nuclear). The decay rate constant

was converted to “per hour” if it was originally described as “per day”. Different eDNA

decay rate constants based on different experimental conditions within the same study

(e.g. species, temperature, pH, and biomass density) were treated separately. The filter

144

pore size in studies involving aqueous eDNA collection via ethanol precipitation or

centrifugation was regarded as 0 µm. In addition, I extracted information on the water

temperature (°C), water source used for experiments, and target species and taxa.

Although other biotic and abiotic factors are known to affect eDNA degradation, only

temperature and water source data were extracted, because of their consistent and

informative descriptions in all selected papers (i.e. other water physicochemical

parameters such as pH, conductivity, and dissolved oxygen were sometimes not

specified in the paper). If necessary, the mean temperature was used by averaging the

maximum and minimum temperatures during the experimental period. Water source

was classified as ‘artificial’, including tap water and distilled water (DW); ‘freshwater’,

including wells, ponds, lakes, and river water; and ‘seawater’, including harbour,

inshore, and offshore seawaters.

Because Moushomi et al. (2019) had estimated decay rates of Daphnia magna

eDNA at each size fraction, I re-estimated eDNA decay rates based on qPCR raw data

in Moushomi et al. (2019) (https://doi.org/10.6084/m9.figshare.9699143). By

performing sequential filtrations of 10, 1, and 0.2 µm pore size filters followed by

ethanol precipitation of the final filtrations, the study quantified mitochondrial and

nuclear eDNA concentrations from Daphnia magna at four size fractions in

experimental tanks, and estimated eDNA decay rates at 0 - 0.2 and 0.2 - 1 µm size

fractions. I calculated total eDNA concentrations collected by a 0.2 µm pore size filter

and ethanol precipitation as follow:

Ctotal.0.2 = C0.2-1 + C1-10 + C>10

145

Ctotal.0 = C<0.1 + C0.2-1 + C1-10 + C>10

where Ctotal.n means total eDNA concentrations collected by a 0.2 µm pore size filter and

ethanol precipitation (i.e., 0 µm), CX means eDNA concentrations at X size fractions.

Linear regressions were performed between eDNA concentrations (log-transformed)

and sampling time points (hour) for each target gene and size fraction to estimate the

slope (i.e., eDNA decay rate constant) using lm functions in R. Referring to Moushomi

et al. (2019), I did not include the data on days 17 and 31 due to non-detection of target

eDNA. All linear regressions were statistically significant (P < 0.05).


All statistical analyses were performed in R version 3.6.1 (R Core Team, 2019). A

generalized linear model (GLM) with Gaussian distribution was performed to assess the

relationship between eDNA persistence, eDNA state, and environmental conditions. The

eDNA decay rate constants (per hour) were treated as the dependent variable, and the

filter pore size (µm), DNA fragment size (bp), target gene (mitochondrial or nuclear),

water temperature (°C), water source (artificial, freshwater, or seawater), and their

primary interactions were included as the explanatory variables. I first confirmed that

the multi-collinearity among the variables was negligible (1.028 to 1.096), by

calculating the generalized variance inflation factors (GVIF). I then selected models

based on Akaike’s Information Criterion (AIC), using the dredge function in the

‘MuMIn’ package in R (Bartoń, 2019). I adopted the model with the smallest AIC value,

146

and all models with ⊿AIC (i.e. difference in the AIC value) less than two were selected

as the supported models (Burnham & Anderson, 2002).

An additional meta-analysis was performed to examine the relationship

between the DNA fragment size and eDNA decay rate constant. Most eDNA studies

conducted to date have targeted short DNA fragments (<200 bp), and only three papers

have reported eDNA decay rates targeting longer DNA fragments (>200 bp); however,

they yielded inconsistent conclusions. Taking this into consideration and targeting

eDNA decay rate constants derived from <200 bp DNA fragments, I performed a linear

regression to assess the effect of DNA fragment size on eDNA degradation.

7.2.3. Re-analysis of the time-series changes in eDNA particle size distribution

To assess the validity of the findings of the meta-analyses, I re-analyzed the dataset

from a previous study investigating the particle size distribution of eDNA derived from

the mitochondria and nuclei of Japanese jack mackerel (Trachurus japonicus) and the

time-series changes therein, after fish removal from tanks (Jo et al., 2019b). In the

aforementioned study, mitochondrial and nuclear eDNA degradation was examined

under multiple size fractions, and both degradations tended to be suppressed at smaller

size fractions. I estimated the eDNA decay rate constants at different size fractions

using the dataset from the above study, and assessed the variation in eDNA decay rates

depending on the eDNA particle size, target gene, and water temperature. Detailed

information on the experimental design, water sampling, and molecular analyses can be

found in Jo et al. (2019b).

147

I included all eDNA samples that could pass through sequential filters with

10, 3, 0.8, and 0.2 µm pore sizes at 0, 6, 12, and 18 hours, which yielded four eDNA

size fractions, i.e. >10, 3-10, 0.8-3, and 0.2-0.8 µm. Linear regressions were performed

between eDNA concentrations (original concentration + 1 followed by log-

transformation) and sampling time points for each size fraction, target gene

(mitochondrial or nuclear), and temperature level (13, 18, 23, or 28 °C), to estimate the

slope (i.e. eDNA decay rate constant) and the corresponding 95 % CI, using lm and

confint functions in R, respectively. Here, the two fish biomass levels (Small and Large;

see Jo et al. (2019b)) were pooled to increase the sample size. I then performed ANOVA

to assess the relationship between eDNA degradation, particle size, target gene, and

temperature. We included the median of the slope (eDNA decay rate) as the dependent

variable, and the filter pore size, target gene, water temperature, and their primary

interactions as the explanatory factors.

7.3. Results

7.3.1. Literature review

26 published papers were selected in total (Table 7-1), including 106 eDNA decay rate

constants, ranging from 0.0005 to 0.6969 (per hour). The filter pore size, DNA fragment

size, and water temperature ranged from 0 to 3 µm, 70 to 719 bp, and -1.0 to 36.0°C,

respectively. The number of eDNA decay rate constants derived from mitochondrial and

nuclear genes were 89 and 17, respectively, and those derived from artificial water,

freshwater, and seawater sources were 31, 15, and 60, respectively. Most studies

148

reported eDNA decay rates targeting freshwater and marine fishes, whereas only few

papers reported decay rates targeting amphibians and invertebrates.

7.3.2. Model selection

In the full model, interactions between filter pore size and water temperature and

between target gene and water temperature were statistically significant (both P < 0.05),

and effects of the filter pore size and interaction between fragment size and water

source were marginally significant (both P < 0.1) (Table 7-2). All supported models

resulting from model selection included the effects of filter pore size, target gene, and

water source, whereas the effects of fragment size and temperature were uncertain,

owing to their small coefficient and large SE. However, the effects of the interactions

among variables should be focused on; all supported models included interactions

between filter pore size and temperature (Figure 7-1) and between target gene and

temperature (Figure 7-2). In addition, 11 of the 13 models included the interaction

between target gene and water source (Figure 7-3), and four models included the

interaction between filter pore size and water source (Figure 7-4). Other interactions

were included in less than three supported models, and the uncertainties of the

corresponding coefficients were relatively large.

Although DNA fragment size was included in most supported models, its

effect was relatively small due to its high variability (Table 7-2). Considering the

smaller number of eDNA decay rate constants targeting longer DNA fragments as

mentioned previously, I instead assessed the relationship between the eDNA decay rate

149

and shorter DNA fragment size (<200 bp). Consequently, the fragment size was found

to have a significantly positive effect on the decay rate (P < 0.01; Figure 7-5).


The ANOVA test showed that all factors significantly affected the eDNA decay rate

constants (all P < 0.001, Table 7-3). Decay rate constants tended to be lower in smaller

size fractions and at lower temperature levels, and were higher for nuclear than for

mitochondrial genes (Figure 7-6). In addition, the interaction between filter pore size

and temperature was a significant factor affecting the decay rate constant (P < 0.01),

and interaction between target gene and temperature was marginally significant (P =

0.0902). Decay rates of eDNA were smaller for smaller size fractions, and there was a

greater tendency to decay at higher temperature levels than at lower levels.

7.4. Discussion

Most studies conducted in the past decade have focused on the relation of eDNA

persistence with environmental conditions, and little attention has been paid to the

relationship between the persistence of eDNA and its cellular states and molecular

structures. The present study integrated the findings of previous reports on eDNA and

provided new insights into the relationship between the persistence and state of eDNA.

The findings indicated significant influences of the complex interactions between eDNA

states and environmental factors on eDNA persistence.

150

7.4.1. Meta-analyses of eDNA literature

The present meta-analyses showed that filter pore size, water temperature, target gene,

and water source could influence eDNA degradation, not as individual parameters but in

conjunction. I focused on three substantial interactions that were included in almost all

supported models. Firstly, the interaction between filter pore size and water temperature

influenced eDNA decay rates. Considering that a larger pore size filter can selectively

collect eDNA particles in larger size fractions, our result implied that higher water

temperature could accelerate the degradation of eDNA in larger size fractions by a

greater degree than that in smaller size fractions. However, it is unlikely that smaller-

sized eDNA itself is less affected by higher temperature-mediated degradation, and its

apparent persistence can be increased by the inflow of eDNA from larger to smaller size

fractions, as described in Jo et al. (2019b). Organic matter in water, including eDNA, is

degraded by microbes and extra-cellular enzymes in the environment for uptake, and

their activities are promoted by moderately high temperatures (less than 50 °C) (Price &

Sowers, 2004; Nielsen et al., 2007; Arnosti, 2014; Strickler et al., 2015). During the

degradation processes, aqueous eDNA in larger size fractions, such as intra-cellular

DNA, is believed to flow into smaller size fractions, such as extra-cellular DNA. This

suggests that water temperature does not uniformly influence the apparent degradation

of eDNA among the different size fractions, and the effect of temperature on eDNA

degradation might be buffered in smaller-sized eDNA particles. Thus, the effect of

temperature on eDNA degradation would be smaller when using a smaller pore size

filter and collecting eDNA particles at various size fractions.

151

Secondly, the interaction between the target gene (nuclear or mitochondrial)

and water temperature influenced the eDNA decay rates; higher water temperature

could accelerate the degradation of nuclear eDNA by a greater extent when compared

with mitochondrial DNA. This may be attributed to the difference in the protection

conferred to the DNA molecules against the attack of extra-cellular enzymes in the

environment by the outer nuclear and mitochondrial membranes. In contrast to

mitochondrial DNA, which is surrounded by a non-porous outer membrane (Ernster &

Schatz, 1981), nuclear DNA is enclosed in a porous membrane (45-50 nm in diameter;

Fahrenkrog & Aebi, 2003), rendering it more susceptible to environmental extra-cellular

enzymes, and thus, more likely be degraded by a greater degree at higher temperatures

(Price & Sowers, 2004; Strickler et al., 2015). However, these results should be

interpreted with caution, because the number of nuclear eDNA decay rate constants (n =

17) included was considerably lower than that of mitochondrial eDNA decay rate

constants (n = 89). It is necessary to estimate nuclear eDNA decay rates in various

environmental and experimental conditions in the future, which would enable a more

robust comparison of eDNA degradation between nuclear and mitochondria DNA.

Thirdly, the interaction between the target gene and water source influenced

the eDNA decay rates. Although the effects of water source on eDNA degradation

differed between nuclear and mitochondrial DNA, it was evident that eDNA

degradation was suppressed in artificial waters, such as tap water and DW, when

compared to that in natural waters. Eichmiller et al. (2016) compared the degradation of

common carp (Cyprinus carpio) eDNA in natural waters with different trophic states,

152

and found that eDNA decay rates in well water were lower than those in eutrophic and

oligotrophic waters, which could be attributed to the lower microbial activity in the

former. The results were generally consistent with those of Eichmiller et al. (2016).

Using tap water and DW as water sources can lead to underestimation of eDNA

persistence in the natural environment. Moreover, no significant difference could be

observed in the eDNA decay rates between freshwater and seawater. The difference in

eDNA persistence between freshwater and seawater has previously been reported; some

studies indicated faster eDNA degradation in seawater than in freshwater (Thomsen et

al., 2012; Sassoubre et al., 2016), whereas Collins et al. (2018) showed that eDNA

degradation was higher in terrestrially-influenced inshore waters than in ocean-

influenced offshore environments. Marine systems are generally characterized by higher

salinity and ionic content, higher pH, and more stable temperatures when compared

with freshwater systems, which can promote DNA preservation in water (Okabe &

Shimazu, 2007; Schulz & Childers, 2011; Collins et al., 2018). However, the direct

effects of microbial abundance and composition and other physicochemical parameters

of water were not included in the meta-analyses. Thus, greater variations in eDNA

decay rates in seawater when compared with artificial water and freshwater observed in

our meta-analyses might partly be explained by such microbial and physicochemical

conditions. The effects of various nutrient salts and microbial activities on eDNA

persistence and differences in the eDNA degradation processes between freshwater and

seawater systems require further investigation.

153

The interaction between filter pore size and water source influenced the

eDNA decay rates in some supported models; however, its effect was relatively smaller

when compared with those of the interactions discussed above. The water source might

affect the apparently longer persistence of smaller-sized eDNA described previously.

Although no linear regressions were statistically significant, the increase in eDNA

decay rates with larger filter pore sizes appeared to be greater in seawater than in

artificial water, which might also be attributed to the differences in microbial activities

among the different water sources.

Contrary to these four factors, model selection in the present study did not

strongly support the effects of DNA fragment size and its interactions with other

variables on the eDNA decay rate, which may be due to the potential bias of DNA

fragment sizes in the eDNA studies included in the meta-analysis. Only three studies

have previously estimated eDNA decay rates in water targeting longer DNA fragments

(>200 bp) (aqueous eDNA; Jo et al. 2017; Weltz et al., 2017; Bylemans et al., 2018a),

and there was no consensus on the relationship between eDNA degradation and DNA

fragment size among these studies. Although the additional meta-analysis, which

targeted only shorter DNA fragments (70 to 190 bp), supported rapid eDNA degradation

in longer DNA fragments, as suggested by Jo et al. (2017) and Wei et al. (2018), the

analysis might be considered slightly arbitrary, and thus, the validity of the result would

need to be tested in the future. Interactions between DNA fragment size and other

factors may become evident when more information is available on eDNA persistence

and degradation at different fragment sizes.

154


The present meta-analyses provided new insights into the relationship between eDNA

persistence and its state. I then re-analyzed the dataset from a previous tank experiment

(Jo et al., 2019b) to estimate mitochondrial and nuclear eDNA decay rates at multiple

size fractions and water temperature levels. The results of the re-analysis appeared to be

generally consistent with those of the meta-analyses; as indicated by the meta-analyses,

eDNA persistence depended on the interactions between its size fraction, type of the

target gene, and water temperature. In particular, a significant interaction between filter

pore size and temperature indicated that inflow of the degraded, larger-sized eDNA into

smaller size fractions could buffer the effect of temperature on eDNA degradation in

these smaller size fractions, as described in previous sections. The dependence of eDNA

degradation on water temperature would likely be smaller when targeting smaller-sized

eDNA or using a smaller pore size filter.

Some recent studies attempted to estimate species biomass and abundance by

integrating quantitative eDNA analysis and hydrodynamic modelling, allowing the

consideration of eDNA dynamics, such as its production, transport, and degradation

(Carraro et al., 2018; Tillotson et al., 2018; Fukaya et al., 2020). For a more accurate

estimation, environmental parameters affecting these eDNA dynamics may be included

in the statistical modelling framework. The effect of temperature on eDNA degradation

can be minimized during statistical modelling by considering eDNA particles at smaller

fraction sizes, which will allow simplification of the modelling procedure while

155

retaining its accuracy and reliability. However, considering the apparent suppression of

eDNA degradation in smaller size fractions, owing to the inflow of the degraded larger-

sized eDNA, it is possible that such smaller-sized eDNA yield ‘older and less fresh’

biological signals than the larger-sized eDNA. Such non-fresh eDNA signals can result

in false-positives during eDNA detection (Yamamoto et al, 2016; Jo et al., 2017), in

which case the use of eDNA particles in the smaller size fractions would be

disadvantageous for eDNA-based biomass or abundance estimation. The applicability

of smaller-sized eDNA for such estimations can be verified by comparing the

correlation between eDNA quantification and species biomass and abundance, and the

availability of longer eDNA fragments among the filter pore sizes or eDNA particle

sizes, for which meta-analyses such as the present study may be suitable.

7.4.3. Limitations and perspectives

I noted some potential biases and limitations of the dataset used in the meta-analyses.

Firstly, studies estimating the decay rates of nuclear eDNA were substantially fewer

when compared with those on mitochondrial eDNA, particularly in freshwater systems

(Figure 3), which might limit the ability to infer the effect of water source on eDNA

degradation between the target genes. In addition, eDNA decay rates targeting longer

DNA fragments (>200 bp) and taxa other than fish were relatively scarce. Moreover,

estimation of eDNA decay rates using a 0.7 µm pore size filter appeared to be relatively

more common, which suggests greater knowledge of eDNA persistence in this filter

pore size, and a potential bias in the meta-analyses. It is expected that eDNA analysis

156

will be applied to ecological monitoring of more varied taxa and environments in the

future, and will have to be developed accordingly to determine the spatiotemporal scale

of eDNA signals and to maximize the biological information obtained from eDNA

samples. More information on eDNA persistence and degradation should therefore be

collected, by targeting different taxa and environments and using various collection and

analysis methods.

Although the findings and implications require further verification, this study

is the first to propose that the persistence of eDNA from macro-organisms can be

determined by the state of the eDNA and its complex interactions with environmental

conditions, i.e. the mechanism of eDNA persistence and degradation cannot be fully

understood without knowing not only the environmental biotic and abiotic factors

involved in eDNA degradation but also the cellular and molecular states of eDNA

occurring in water. If the findings are correct, the spatiotemporal scale and intensity of

eDNA signals would be different depending on the eDNA particle size and state. The

fact that Weibull or biphasic exponential decay models fit better to eDNA degradation

implies the differences in eDNA persistence depending on its state (e.g., intra- or extra-

cellular, living or dead cells, particulate or dissolved) (Eichmiller et al., 2016; Bylemans

et al., 2018a), which support our results linking eDNA persistence to its state. In

addition, the study by Jo et al. (2020c), where it was reported that the genomic

information obtained from eDNA samples can differ depending on the filter pore size,

can further support the link between eDNA state and persistence. Experimental

verification of our findings and implications will highlight the importance of clarifying

157

the characteristics and dynamics of aqueous eDNA, and will contribute substantially to

the development of eDNA analysis in the future.

158

7.5. Tables

Table 7-1. Published literature on the estimation of first-order eDNA decay rate constants included in the present study.

Study # Decay rate

constant Filter pore size

[µm] Fragment size

[bp] Target gene

Temperature [°C]

Water source Target taxa

Thomsen et al. (2012) 2 0.45 101 to 104 mt 15 Seawater Fish Barnes et al. (2014) 1 1.2 146 mt 25 Freshwater Fish Maruyama et al. (2014) 1 0 100 mt 20 Artificial Fish Strickler et al. (2015) 3 0.45 84 mt 5 to 35 Artificial Amphibian Eichmiller et al. (2016) 4 0.2 149 mt 5 to 35 Freshwater Fish Forsström & Vasemägi (2016) 1 0 75 mt 17 Artificial Crustacean Sassoubre et al. (2016) 5 0.2 107 to 133 mt 19 to 22 Seawater Fish Andruszkiewicz et al. (2017) 2 0.22 107 mt 17 Seawater Fish Jo et al. (2017) 2 0.7 127 to 719 mt 26 Seawater Fish Lance et al. (2017) 4 0.22 190 mt 4 to 30 Artificial Fish Minamoto et al. (2017a) 1 0.7 151 mt 19 Seawater Invertebrate Sansom & Sassoubre (2017) 6 0.4 147 mt 22 Artificial Invertebrate Sigsgaard et al. (2017) 2 0.22 105 mt 35 to 36 Seawater Fish Tsuji et al. (2017) 6 0.7 78 to 131 mt 10 to 30 Freshwater Fish Weltz et al. (2017) 2 0.45 331 mt 4 Seawater Fish Bylemans et al. (2018a) 12 1.2 95 to 515 mt & nu 20 Artificial Fish Collins et al. (2018) 8 0.22 132 to 153 mt 10 to 15 Seawater Fish & Crustacean Cowart et al. (2018) 1 0.45 70 mt -1 Seawater Fish Nevers et al. (2018) 2 1.5 150 mt 12 to 19 Seawater Fish Nukazawa et al. (2018) 2 0.7 149 mt 21 to 22 Freshwater Fish

159

Note: Abbreviations ‘mt’ and ‘nu’ indicate mitochondrial and nuclear DNA, respectively. Filter pore size in studies collecting eDNA via ethanol precipitation or centrifugation was regarded as 0 µm.

Jo et al. (2019) 12 0.7 127 mt 13 to 28 Seawater Fish Moushomi et al. (2019) 4 0 to 0.2 101 to 128 mt & nu 20 Artificial Invertebrate Sengupta et al. (2019) 1 0 86 mt 23 Artificial Invertebrate Jo et al. (2020) 12 0.7 164 nu 13 to 28 Seawater Fish Kasai et al. (2020) 5 0.7 138 mt 10 to 30 Seawater Fish Sakata et al. (2020) 1 0.7 132 mt 17 Freshwater Fish Wood et al. (2020) 4 3 90 to 150 mt 19 Seawater Invertebrate

160

Table 7-2. Results of model selection for the effects of filter pore size, DNA fragment size, target gene, temperature, and water source on

the first-order eDNA decay rates.

Variable GVIF Full model Model_1 Model_2 Model_3

Coeff. SE P value Coeff. SE Coeff. SE Coeff. SE

Intercept 0.0506 0.0975 0.6050 0.0358 0.0552 0.0506 0.0563 0.0709 0.0582 Filter pore size 1.0308 -0.2269 0.1341 0.0942 -0.2058 0.0993 -0.2933 0.1099 -0.2911 0.1095 Fragment size 1.0440 0.0004 0.0005 0.3889 -0.0002 0.0001 -0.0001 0.0001 Gene (nu) 1.0472 -0.3073 2.5630 0.9048 -0.3591 0.1010 -0.3268 0.1012 -0.3365 0.1011 Temperature 1.0281 -0.0043 0.0038 0.2612 -0.0008 0.0026 -0.0012 0.0026 -0.0016 0.0026 Water source (fre)

1.0955 0.1909 0.1567 0.2266 0.0571 0.0272 0.0525 0.0555 0.0573 0.0554

Water source (sea) 0.0308 0.0791 0.6982 0.0858 0.0207 0.0452 0.0295 0.0491 0.0295 Filter pore size: Fragment size -0.0004 0.0004 0.2547 Filter pore size: Gene (nu) 0.0034 0.5853 0.9953 Filter pore size: Temperature 0.0138 0.0056 0.0151 0.0130 0.0052 0.0142 0.0053 0.0149 0.0053 Filter pore size: Water source (fre) -0.0164 0.0948 0.8632 0.0238 0.0841 0.0031 0.0853 Filter pore size: Water source (sea) 0.0709 0.0466 0.1318 0.0783 0.0351 0.0631 0.0368 Fragment size: Gene (nu) -0.0001 0.0196 0.9969 Fragment size: Temperature 0.0000 0.0000 0.7966 Fragment size: Water source (fre) -0.0015 0.0009 0.0796 Fragment size: Water source (sea) -0.0004 0.0003 0.1526 Gene (nu): Temperature 0.0149 0.0047 0.0023 0.0162 0.0046 0.0158 0.0046 0.0156 0.0046 Gene (nu): Water source (fre) n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Gene (nu): Water source (sea) 0.3064 1.0600 0.7731 0.3239 0.0491 0.2966 0.0484 0.3110 0.0495 Temperature: Water source (fre) 0.0041 0.0039 0.2964 Temperature: Water source (sea) 0.0036 0.0033 0.2786

AIC -208.16 -217.38 -217.08 -217.00 ⊿AIC 9.22 0.00 0.30 0.38

161

(Table 7-2 continued)

Variable Model_4 Model_5 Model_6 Model_7

Coeff. SE Coeff. SE Coeff. SE Coeff. SE

Intercept 0.0647 0.0585 -0.0048 0.0838 0.0388 0.0553 0.0387 0.0553 Filter pore size -0.3266 0.1109 -0.2579 0.1257 -0.2103 0.0996 -0.2103 0.0996 Fragment size 0.0000 0.0001 0.0005 0.0004 -0.0002 0.0001 -0.0002 0.0001 Gene (nu) -0.3173 0.1013 -0.3193 0.1011 -0.3116 0.1145 -0.5802 0.2718 Temperature -0.0019 0.0026 -0.0014 0.0026 -0.0010 0.0026 -0.0010 0.0026 Water source (fre) 0.2557 0.1359 0.2711 0.1363 0.0567 0.0272 0.0567 0.0272 Water source (sea) 0.0725 0.0379 0.0982 0.0439 0.0852 0.0207 0.0852 0.0207 Filter pore size: Fragment size -0.0004 0.0004 Filter pore size: Gene (nu) -0.0615 0.0697 Filter pore size: Temperature 0.0159 0.0053 0.0152 0.0053 0.0134 0.0052 0.0134 0.0052 Filter pore size: Water source (fre) -0.0238 0.0894 -0.0164 0.0895 Filter pore size: Water source (sea) 0.0789 0.0388 0.0738 0.0390 Fragment size: Gene (nu) 0.0021 0.0023 Fragment size: Temperature Fragment size: Water source (fre) -0.0014 0.0008 -0.0015 0.0009 Fragment size: Water source (sea) -0.0002 0.0002 -0.0004 0.0002 Gene (nu): Temperature 0.0153 0.0046 0.0153 0.0046 0.0161 0.0046 0.0161 0.0046 Gene (nu): Water source (fre) n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Gene (nu): Water source (sea) 0.2998 0.0500 0.3029 0.0500 0.3213 0.0493 0.2106 0.1383 Temperature: Water source (fre) Temperature: Water source (sea)

AIC -216.71 -216.25 -216.25 -216.23 ⊿AIC 0.67 1.13 1.13 1.15

162


Variable Model_8 Model_9 Model_10 Model_11

Coeff. SE Coeff. SE Coeff. SE Coeff. SE

Intercept 0.0387 0.0553 0.0142 0.0636 0.0383 0.0557 0.0932 0.0685 Filter pore size -0.2103 0.0996 -0.1781 0.1074 -0.2174 0.1001 -0.1777 0.1013 Fragment size -0.0002 0.0001 0.0000 0.0002 -0.0002 0.0001 -0.0002 0.0001 Gene (nu) -1.0870 0.1886 -0.3626 0.1014 -0.9222 0.1542 -0.3484 0.1026 Temperature -0.0010 0.0026 -0.0006 0.0026 -0.0012 0.0026 -0.0036 0.0033 Water source (fre) 0.0568 0.0272 0.0535 0.0278 0.0594 0.0274 -0.0314 0.0825 Water source (sea) 0.0853 0.0207 0.0818 0.0215 0.0892 0.0207 0.0001 0.0660 Filter pore size: Fragment size -0.0002 0.0003 Filter pore size: Gene (nu) 0.1140 0.0762 Filter pore size: Temperature 0.0134 0.0052 0.0128 0.0052 0.0139 0.0052 0.0115 0.0053 Filter pore size: Water source (fre) Filter pore size: Water source (sea) Fragment size: Gene (nu) 0.0059 0.0009 0.0054 0.0008 Fragment size: Temperature Fragment size: Water source (fre) Fragment size: Water source (sea) Gene (nu): Temperature 0.0161 0.0046 0.0161 0.0046 0.0160 0.0047 0.0157 0.0047 Gene (nu): Water source (fre) n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Gene (nu): Water source (sea) 0.3276 0.0496 0.3213 0.0492 Temperature: Water source (fre) 0.0044 0.0039 Temperature: Water source (sea) 0.0044 0.0032

AIC -216.14 -215.91 -215.68 -215.61 ⊿AIC 1.24 1.47 1.70 1.77

163


Variable Model_12 Model_13

Coeff. SE Coeff. SE

Intercept -0.0051 0.0514 -0.0558 0.0809 Filter pore size -0.1851 0.1000 -0.1340 0.1117 Fragment size 0.0005 0.0004 Gene (nu) -0.3504 0.1022 -0.3527 0.1010 Temperature -0.0001 0.0026 -0.0002 0.0026 Water source (fre) 0.0670 0.0270 0.2148 0.1109 Water source (sea) 0.0932 0.0205 0.1282 0.0418 Filter pore size: Fragment size -0.0005 0.0004 Filter pore size: Gene (nu) Filter pore size: Temperature 0.0117 0.0052 0.0125 0.0052 Filter pore size: Water source (fre) Filter pore size: Water source (sea) Fragment size: Gene (nu) Fragment size: Temperature Fragment size: Water source (fre) -0.0012 0.0008 Fragment size: Water source (sea) -0.0003 0.0002 Gene (nu): Temperature 0.0164 0.0047 0.0160 0.0046 Gene (nu): Water source (fre) n.a. n.a. n.a. n.a. Gene (nu): Water source (sea) 0.3069 0.0490 0.3224 0.0494 Temperature: Water source (fre) Temperature: Water source (sea)

AIC -215.50 -215.39 ⊿AIC 1.88 1.99

164

Note: Abbreviation ‘Coeff.’ indicates the coefficient of each variable in GLM. Positive values for the coefficient of the variable ‘Gene

(nu)’ indicate higher eDNA decay rate constant for nuclear than mitochondrial DNA. Positive values for the coefficient of the variable

‘Water source (fre/sea)’ indicate higher eDNA decay rate constant for freshwater or seawater than artificial water samples. The

coefficient of the interaction ‘Gene (nu): Water source (fre)’ was not analysed because no study described eDNA decay rate constants

using a nuclear DNA marker and freshwater samples. P values of each parameter are not shown in the model, except for the full model.

Coefficients of each parameter are shown in bold.

165

Table 7-3. The result of the ANOVA test for the effects of eDNA particle size, target

gene, and water temperature on eDNA decay rate constants.

Response Factor F value P value

Decay rate constant Filter pore size 39.2770 *** Gene 45.8534 *** Temperature 27.3524 *** Filter pore size: Gene 0.2535 0.8570 Filter pore size: Temperature 5.9051 ** Gene: Temperature 2.9600 0.0902

Note: Asterisks indicate the statistical significance of the factor (**, P < 0.01; ***, P <

0.001).

166

7.6. Figures

Figure 7-1. The effects of water temperature and filter pore size on eDNA decay rate

constants. Left, middle, and right graphs show the linear relationships between decay

rate constants and temperature targeting all filter pore sizes (circle), <0.45 µm pore

sizes (square), and >0.7 µm pore sizes (triangle), respectively. Bold and dotted lines

indicate the regression line and the corresponding 95% confidence intervals (CI)

estimated by lm and confint functions in R, respectively. R2 values of the linear

regressions are shown in the top-left corner of each figure, and the asterisks indicate the

statistical significance of the linear regressions (∗∗, P < 0.01).

0 10 20 30 40

0.0

0.2

0.4

0.6

0.8

1.0 R2 = 0.097 **

Overall (n = 106)

0 10 20 30 40

0.0

0.2

0.4

0.6

0.8

1.0 R2 = 0.141 **

< 0.45 µm (n = 46)

0 10 20 30 40

0.0

0.2

0.4

0.6

0.8

1.0 R2 = 0.147 **

> 0.7 µm (n = 60)

Temperature [°C]

Dec

ay ra

te c

onst

ant [

/hou

r]

167

Figure 7-2. The effects of water temperature and target gene on eDNA decay rate

constants. Left, middle, and right graphs show the linear relationships between decay

rate constants and temperature targeting all genes (circle), mitochondrial DNA (square),

and nuclear DNA (triangle), respectively. Bold and dotted lines indicate the regression

line and the corresponding 95% CI estimated by lm and confint functions in R,

respectively. R2 values of the linear regressions are shown in the top-left corner of each

figure, and the asterisks indicate the statistical significance of the linear regressions (*,

P < 0.05; **, P < 0.01).

0 10 20 30 40

0.0

0.2

0.4

0.6

0.8

1.0 R2 = 0.097 **

Overall (n = 106)

0 10 20 30 40

0.0

0.2

0.4

0.6

0.8

1.0 R2 = 0.097 **

Mitochondrial (n = 89)

0 10 20 30 40

0.0

0.2

0.4

0.6

0.8

1.0 R2 = 0.296 *

Nuclear (n = 17)

Temperature [°C]

Dec

ay ra

te c

onst

ant [

/hou

r]

168

Figure 7-3. The effects of water source and target gene on eDNA decay rate constants.

Left, middle, and right graphs show the boxplots of eDNA decay rate constants

targeting all genes, mitochondrial DNA, and nuclear DNA, respectively. In each graph,

decay rate constants derived from artificial water, freshwater, and seawater are shown in

white, bright grey, and dark grey, respectively. Note that no study described eDNA

decay rate constants using a nuclear DNA marker and freshwater samples.

Artificial Freshwater Seawater

0.0

0.2

0.4

0.6

0.8

1.0

(n = 31) (n = 15) (n = 60)

Overall


0.0

0.2

0.4

0.6

0.8

1.0

(n = 26) (n = 15) (n = 48)

Mitochondrial


0.0

0.2

0.4

0.6

0.8

1.0

(n = 5) (n = 0) (n = 12)

Nuclear

Water source

Dec

ay ra

te c

onst

ant [

/hou

r]

169

Figure 7-4. The effects of water source and filter pore size on eDNA decay rate

constant. Upper-left, upper-right, lower-left, and lower-right graphs show the linear

relationships between decay rate constants and filter pore sizes targeting all water

sources (circle), targeting artificial water (square), targeting freshwater (triangle), and

targeting seawater (rectangle). Bold and dotted lines show the regression lines and their

95 % CI, which were estimated by lm and confint functions in R. R2 values of linear

regressions were shown in top-left of each figure. Dots show the marginally statistical

significance of linear regressions (P < 0.1).

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0

0.2

0.4

0.6

0.8

1.0

R2 = 0.034 .

Overall (n = 106)

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0

0.2

0.4

0.6

0.8

1.0

R2 = 0.055 (P > 0.1)

Artificial (n = 31)

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0

0.2

0.4

0.6

0.8

1.0

R2 = 0.046 (P > 0.1)

Freshwater (n = 15)

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0

0.2

0.4

0.6

0.8

1.0

R2 = 0.037 (P > 0.1)

Seawater (n = 60)

Filter pore size [µm]

Dec

ay ra

te c

onst

ant [

/hou

r]

170

Figure 7-5. The effects of DNA fragment size on eDNA decay rate constant. A graph

shows the linear relationship between decay rate constants and fragment sizes within

200 bp (in black, n = 97). Bold lines show regression lines and dotted lines show their

95 % confidence intervals (CI). R2 value of a linear regression was shown in top-left.

Asterisks show the statistical significance of linear regressions (**, P < 0.01). Decay

rate constants derived from longer DNA fragments (>200 bp) are shown in gray but are

excluded from a linear regression.

0 200 400 600 800 1000

0.0

0.2

0.4

0.6

0.8

1.0

Fragment size [bp]

Dec

ay ra

te c

onst

ant [

/hou

r]

R2 = 0.081 ** (n = 97)

171

Figure 7-6. The effects of eDNA particle size, water temperature, and target gene on

eDNA decay rate constants. Upper and lower graphs show the results for mitochondrial

(bright grey) and nuclear (dark grey) eDNA, respectively. Medians and 95% CI of

eDNA decay rate constants are indicated by circles and bars, respectively. Each filter

pore size (10, 3, 0.8, and 0.2 µm) corresponded to a size fraction (>10, 3-10, 0.8-3, and

0.2-0.8 µm).

-0.050.000.050.100.150.20

mitochondrial

-0.050.000.050.100.150.20

nuclear

Dec

ay ra

te c

onst

ant [

/hou

r (×

-1)]

Filter pore size [µm]

10 3 0.8 0.2

13 °C

10 3 0.8 0.2

18 °C

10 3 0.8 0.2

23 °C

10 3 0.8 0.2

28 °C

10 3 0.8 0.2

13 °C

10 3 0.8 0.2

18 °C

10 3 0.8 0.2

23 °C

10 3 0.8 0.2

28 °C

172

Chapter 8. General Discussion

There has been a blossoming of the eDNA application to biological monitoring

targeting various species and environments in this decade (Taberlet et al., 2012;

Takahara et al., 2016; Deiner et al., 2017a); nevertheless, false-positive/-negative

detections and various errors in experimental procedures can confound the reliability of

eDNA detection in the field (Darling & Mahon, 2011; Furlan et al., 2016; Dorazio &

Erickson, 2018; Doi et al., 2019), and quantified eDNA values can measurably vary

even among samples derived from same individuals (Takahara et al., 2012; Klymus et

al., 2015). The former errors relating to eDNA detection may mislead our inferences on

species presence/absence in the field. The latter ones relating to eDNA quantification

may weaken correlations between eDNA concentration and species biomass/abundance

and thus make the estimation uncertain (Yates et al., 2019). To overcome such

uncertainties on eDNA analyses, it would be necessary to better know the characteristics

and dynamics of eDNA including its physiochemical and molecular states and processes

of production, transport, and degradation (Strickler et al., 2015; Barnes & Turner, 2016;

Hansen et al., 2018). Throughout the thesis, especially focusing on the state of eDNA,

which has much less been studied despite of its potential on eDNA transport and

persistence as described in Chapter 1 (General Introduction), I multifacetedly studied

eDNA characteristics, dynamics, and their interactions. The findings refined basic

understandings on spatiotemporal inferences of eDNA, as well as brought novel clues to

solve the uncertainties relating to eDNA detection and quantification.

173

8.1. Nuclear and mitochondrial eDNA

In Chapters 2 to 4, targeting mitochondrial and nuclear eDNA, I examined eDNA

shedding and degradation from Japanese jack mackerel (Trachurus japonicus) in

multiple temperature and species biomass conditions, and found that there were some

differences in eDNA production and degradation between nuclear and mitochondrial

DNA. First, eDNA decay rates were generally higher in nuclear than mitochondrial

DNA, which could depend on the differences in cellular and molecular structures

between nuclei and mitochondria and between their DNA. In a eukaryotic cell, a

mitochondrion is generally 0.5 to 2 µm in diameter (Wrigglesworth et al., 1970; Ernster

& Schatz, 1981), which can greatly vary due to its fission and fusion for the

maintenance of its integrity (Koshiba et al., 2004; Suen et al., 2008), has its own

circular DNA (mitogenome) in the matrix, and is covered with an outer membrane with

relatively small channels (about 2 nm in diameter) (Künkele et al., 1998). In contrast, a

nucleus is generally 5 to 10 µm in diameter (Kornberg, 1974; Lloyd et al., 1979), has its

own linear DNA (nuclear genome) which is generally retracted in a chromatin structure

(Kornberg, 1974), and is covered with an outer membrane with relatively large pores

(up to 40 nm in diameter) (Panté & Kann, 2002; Fahrenkrog & Aebi, 2003). Microbes

and their extra-cellular enzymes are likely to attack mitochondrial DNA more frequently

than chromatin-retracted nuclear DNA, whereas the porous nuclear membrane is likely

to make nuclear DNA susceptible to extra-cellular enzymes. In addition, exonucleases

might not degrade circular mitochondrial genome as long as it is intact (Hosfield et al.,

1998). After released into water, a chromatin in a cell may rapidly lose its structure and

174

function, and nuclear eDNA may be degraded faster than mitochondrial eDNA.

Otherwise, considering that the extent of nuclear DNA retraction in a chromatin and the

frequency of DNase degradation are different depending on the frequency of gene

expressions (Stalder et al., 1980), it might be possible that decay rates of nuclear DNA

in water depends on target genetic regions.

Moreover, the differences in such cellular and molecular states could associate

the effects of environmental conditions on eDNA degradation, which might contribute

to the difference in nuclear and mitochondrial eDNA degradation. In Chapter 3,

nonlinear least-squares regression of eDNA decay curves showed that the coefficients of

water temperature were much larger in nuclear than mitochondrial eDNA degradation,

while the coefficients of fish biomass density were similar between target genes. It

could indicate the difference in water temperature-dependence of eDNA degradation

between nuclear and mitochondria, which is supported by the interaction between target

gene and water temperature on eDNA decay rate constants in Chapter 7. Increase of

microbial activity in moderately higher temperature may accelerate the degradation of

nuclear DNA more than that of mitochondrial DNA because of the relatively larger

pores on nuclear membrane. Besides, the relationship between the ratio of nuclear to

mitochondrial eDNA yields and filter pore sizes in Chapter 6 could also reflect the

degradation processes of nuclear and mitochondrial eDNA and their vulnerabilities

against environmental abiotic/biotic factors; mitochondrial DNA is generally degraded

slower in intra-membrane environments than nuclear DNA whereas the relationship

could be reversed in extra-membrane environments, implying the importance to

175

understand the effects of cellular and molecular states on mitochondrial and nuclear

DNA degradation (Foran, 2006).

Second, eDNA shedding rates of larger fish biomass levels were higher in

nuclear than mitochondrial DNA, and the ratio of mitochondrial to nuclear eDNA

shedding rates were lower in larger fish body sizes and older fish individuals. As

mentioned in Chapter 3, we suggested the possibility that, although further verifications

are needed for other species and wider range of body sizes and age structures, a

decrease in available mitochondrial DNA copy number in a cell owing to the growth

and aging (Clay Montier et al., 2009; Hartmann et al., 2011) influenced relative

concentrations and shedding rates of nuclear and mitochondrial eDNA from Japanese

jack mackerels. Such relationships of eDNA production with metabolism and

development are reasonable given that the production of mitochondrial eDNA from

brook trout (Salvelinus fontinalis) does not scale linearly but allometrically with

individual biomass (Yates et al., 2020a; 2020b). As it is unlikely that either nuclear or

mitochondrial eDNA is only released into external environments, the ratio of nuclear

and mitochondrial eDNA production could be determined by the ratio of nuclear and

mitochondrial DNA copy numbers in a cell before their release from organisms, which

could depend on the cell biology. In Chapter 3, despite a bit longer PCR amplification

length in nuclear (164 bp) than mitochondrial (127 bp) eDNA, eDNA quantifications

were similar among them, or higher in nuclear than mitochondrial eDNA depending on

fish biomass levels. Eukaryotic cells have a larger number of mitochondria which have

a number of mitochondrial DNA (hundreds to thousands of molecules per cell; Robin &

176

Wong, 1988) while ribosomal RNA (rRNA) genes including ITS1 regions in nuclear

DNA have tandem repeats, which can vary depending on taxa (tens to tens of thousands

per nuclear genome; Prokopowich et al., 2003). Any of previous studies reporting

similar or higher detectability of nuclear eDNA relative to mitochondrial one (Bylemans

et al., 2017; Minamoto et al., 2017b; Bylemans et al., 2018a; Dysthe et al., 2018)

targeted ITS regions in rRNA genes, and thus it can be likely that, as with eDNA

degradation, the relative production of nuclear and mitochondrial eDNA depends on

target genetic regions. Moreover, the ratio could differ among cell types. For example,

contrary to somatic cells, sperm cells have highly condensed and protected nuclear

DNA while the number of mitochondrial genomes is relatively low (Coward et al.,

2002). Focusing on the relationship, it was reported that the relative abundance of

nuclear and mitochondrial eDNA can be indicative of recent reproductive activity of

freshwater fishes (Bylemans et al., 2017).

Multi-copy nuclear eDNA analyses such as ribosomal RNA genes achieve

higher detectability and larger yields of target eDNA than mitochondrial eDNA

analyses, and may enable to mitigate the error of eDNA-based estimation of species

biomass/abundance associated with age and developmental stage. Previous studies

suggested that a decrease of mitochondrial eDNA concentrations per wet weight in

larger body size and older fishes could result from an ontogenetic reduction in

metabolic activity (Maruyama et al., 2014; Takeuchi et al., 2019), which is thus likely to

bias eDNA-based estimation of species biomass/abundance in the field when targeting

different body size but similar biomass populations. The findings in my thesis is the first

177

to suggest the possibility to mitigate such errors and to perform more accurate

biomass/abundance estimation by targeting nuclear eDNA. In addition, as nuclear

eDNA was degraded faster than mitochondrial eDNA, the spatiotemporal scale of

nuclear eDNA signals might be possibly more limited than mitochondrial ones, meaning

that nuclear eDNA detection and quantification could provide the finer spatiotemporal

biological information in the field. However, to validate these merits targeting nuclear

eDNA, it would be required (i) to verify that nuclear eDNA represents shorter

persistence time and transport distance than mitochondrial eDNA because both

persistence time and transport distance of eDNA can be a function of its shedding and

decay rates, (ii) to compare variabilities of eDNA quantifications among sampling

replicates between target genes, which would be involved with the reliability of eDNA

quantification, and (iii) to generalize the characteristics and dynamics of nuclear eDNA

by targeting other genetic regions in nuclear genome.

Moreover, particularly with regards to eDNA metabarcoding, sequencing the

region with high genetic variations such as ITS region in nuclear DNA may allow to

obtain inter-/intra-specific inferences with higher taxonomic resolution (e.g.,

distinguishing closely related species) than sequencing mitochondrial eDNA. In

addition, although not being analyzed in my thesis, sequencing single-copy nuclear

eDNA could provide more various population genetic information such as sex ratios and

individual identifications than sequencing mitochondrial eDNA (Uchii et al., 2016;

Sigsgaard et al., 2020; Tsuji et al., 2020) (details are described as below). The shortage

of nuclear DNA sequences from macro-organisms, especially from vertebrates, in

178

database contrary to bacteria and fungi is also the drawback for the further application

of nuclear eDNA analysis (Handelsman, 2004; Toju et al., 2012; Minamoto et al.,

2017b).

Furthermore, in Chapter 3, I showed the possibility that the combined eDNA

applications targeting nuclear and mitochondrial DNA could provide more detailed

biological information such as body size and age structures of a population contrary to

typical eDNA applications targeting a single gene. This study and Bylemans et al.

(2017), focusing on the relative abundance of nuclear and mitochondrial eDNA as an

index of a fish reproduction activity, are among the first showing that analyses of

multiple type of eDNA could be useful approaches to estimate the developmental stage

of the species other than its presence/absence and relative abundance. These approaches

would enable more accurate and detailed ecological monitoring via eDNA than typical

eDNA analyses. Although it appears to be a bit confusing that the relative abundance of

nuclear and mitochondrial eDNA is used as both indices of reproduction activity and

age/body size, water sampling on non-breeding season for target species can mitigate

the problem. With regards to my research, further verifications will be needed (i)

whether the relationship between the ratio of mitochondrial to nuclear eDNA

abundances and body size/age is established for more extensive ranges of fish body size

and age structures and other species, (ii) how environmental conditions such as water

temperature affect the relationship, and (iii) whether the relationship is general and

practical in natural environments.

179

8.2. Long and short eDNA fragments

In Chapter 5, in addition to the comparison between nuclear and mitochondrial eDNA, I

focused on another molecular characteristics of eDNA, DNA fragment size (i.e., PCR

amplification length), and obtained insights on eDNA characteristics and dynamics

between different DNA fragment sizes. First, eDNA was degraded faster in longer DNA

fragments than shorter ones, which could explain why the detection rate and the amount

of target DNA in fecal and environmental samples were lower in longer DNA fragment

sizes (Deagle et al., 2006; Hänfling et al., 2016; Bista et al., 2017; Bylemans et al.,

2018a; Kamenova et al., 2018; Wei et al., 2018). My finding was later supported by the

comparison of DNA degradation between different fragment sizes using artificial

double-stranded DNA (Mikutis et al., 2018).

Ultimately, eDNA degradation can be attributed to (i) damage:

physiochemical and enzymatic loss and cutting of bases (e.g., oxidation, UV radiation,

hydrolysis, deamination, and break down by endo-/exo-nucleases) (Lindahl, 1993;

Ravanat et al., 2001; Arnosti, 2014; Torti et al., 2015) and (ii) uptake: intake of

fragmented DNA molecules by microbes such as bacteria and archaea, which becomes

their nutritional resources of carbon, nitrogen, and phosphorous, and/or is internalized

by their genome (i.e., natural transformation) (Arnosti, 2014; Torti et al., 2015). In

addition, the timing when these factors occur in the eDNA degradation process can be

different; uptake could occur after eDNA release into environments, while damage

could occur both before and after eDNA release into environment (i.e., damaging of

DNA fragment in water would have already begun at the time when it is released from

180

the individual). Moreover, as long as DNA molecules are randomly damaged, contrary

to uptake, only damage could directly be involved with fragment size-dependent

degradation of eDNA (Mikutis et al., 2018). Therefore, higher eDNA decay rates in a

longer DNA fragment size observed in Chapter 5 can indicate that the DNA

fragmentation due to its physiochemical and enzymatic damages is substantial for

eDNA degradation even after its release into environments. In this point, the finding

would be substantial to understand the process and mechanism of eDNA persistence.

On the other hand, relative to the simulation assuming artificial DNA

degradation (Mikutis et al., 2018), the difference in eDNA decay rates among DNA

fragment sizes in Chapter 5 appeared to be smaller; in Mikutis et al. (2018), a decay rate

of 113 bp DNA fragments was twice as high as that of 53 bp fragments, while a decay

rate of 719 bp DNA fragments was about twice as high as that of 127 bp fragments in

my tank experiment. Although the simulation in Mikutis et al. (2018) did not assume the

effects of uptake of DNA molecules, the gap can partly be explained by the difference in

DNA sources. In the water, eDNA can be present with various states including not only

extra-membrane and dissolved DNA but also intra-membrane DNA such as cells and

organelles (Turner et al., 2014), which possibly results in the multi-phasic degradation

of eDNA (Eichmiller et al., 2016). Contrary to extra-membrane eDNA, DNA

fragmentation may occur less frequently in intra-membrane eDNA, which could likely

make it uncertain that aqueous eDNA degrades faster in longer DNA fragments (the

detailed relationship between eDNA state and persistence is discussed below).

In addition, Bylemans et al. (2018a) reported that, although concentrations of

181

goldfish (Carassius auratus) eDNA in experimental tanks negatively correlated with the

length of DNA fragments, eDNA decay rates were not different among fragment sizes.

The discrepancy might also be explained by the difference in DNA sources. My tank

experiment in Chapter 5 monitored eDNA degradation in rearing water by transferring it

to other tanks without stimulating the fish, while Bylemans et al. (2018a) monitored it

in rearing water by removing the fish from the tanks. Thus, in Bylemans et al. (2018a),

the proportion of intra-membrane eDNA in rearing water, which could derive from

slough cells from skin and scale (Pilliod et al., 2014; Sassoubre et al., 2016), would be

likely higher than my tank experiment. Although further studies are necessary to reveal

the detailed mechanism of eDNA degradation in relation to DNA fragment size, such a

discrepancy might result from an abundance of intra-membrane eDNA in water which

could be more robust against DNA fragmentation than extra-membrane eDNA.

Second, eDNA concentrations were correlated more clearly with fish biomass

in longer DNA fragments. As a higher eDNA decay rate can mean shorter persistence

time and transport distance of eDNA, the result would at first be attributed to the finer

spatiotemporal signal (i.e., eDNA more recently released and closer to the individual) of

a longer eDNA fragment in the water. In addition, notably, longer eDNA fragments

were not detected at all nearby a fish market contrary to the highest concentration of

shorter eDNA fragments. Thus, a clearer correlation between longer eDNA fragments

and fish biomass would also result from the removal of the detection of highly degraded

and fragmented eDNA, which could likely derive from carcasses and resuspension from

substrates, by targeting a longer eDNA fragment.

182

The sources of eDNA can include not only living individuals but also dead

individuals (Merkes et al., 2014; Dunker et al., 2016; Yamamoto et al., 2016) and

resuspension from substrates (Turner et al., 2015; Fremier et al., 2019). To allow the

accurate estimation of species distribution and abundance using eDNA, discriminating

these eDNA sources and selectively detecting the eDNA derived from living individuals

would be required. For example, the false-positive detection of carcasses-derived eDNA

could lead the overestimation of species abundance when targeting the fish species most

of which are killed after reproduction and incubation (e.g., salmonid; Caswell et al.,

1984; Tillotson et al., 2018), or some environmental accidents (e.g., eutrophication and

fish diseases; San Diego-McGlone et al., 2008; Uchii et al., 2011; Gomes et al., 2017).

In addition, bottom trawling or harsh weather conditions can disturb bottom sediments,

where a plenty of eDNA particles are adsorbed (Turner et al., 2015; Sakata et al., 2020),

and potentially bias the estimation of species biomass/abundance (Hansen et al., 2018).

These issues have so far been discussed for eDNA monitoring targeting aqueous macro-

organisms (Darling & Mahon, 2011; Hansen et al., 2018), and such eDNA not derived

from living individuals (relic DNA) can actually cause the estimation biases in

abundance and composition of soil microbes (Calini et al., 2016). Therefore, if the

detection of longer eDNA fragments inferred the finer spatiotemporally and fresher

signal of eDNA recently released from individuals than shorter eDNA fragments, eDNA

application using a longer eDNA fragment would selectively detect the eDNA derived

from living individuals and enable more accurate ecological monitoring via eDNA.

For the practical use of a longer eDNA fragment in eDNA-based ecological

183

monitoring, future studies should investigate following points. Firstly, because eDNA

yields per filtration water volume decrease in a longer eDNA fragment, a collection

efficiency of longer eDNA fragments from the field must be increased. As one of the

solutions, tank experiment in Chapter 6 showed that water filtration using a larger pore

size filter increased not only filtration efficiency but also the relative yields of longer to

shorter eDNA fragments, and thus the validity of this finding will be strengthened by

the experiment in natural environments. Secondly, it should be confirmed that the

eDNA derived from dead individuals and resuspension is actually more degraded and

fragmented than that derived from living individuals. By comparing DNA fragment

sizes of aqueous eDNA released from living and dead individuals, the fragment size

enough to discriminate living individuals or not might be identified, which could

directly be used as the index to discriminate living and dead individuals. Even if it is not

the case, the effect of eDNA not derived from living individuals might be removed by

combining other molecular approaches. For example, viability PCR, using fluorescence

dye not penetrating intact cell membrane, can selectively detect viable cells

(Rawsthorne et al., 2009; Fittipaldi et al., 2011; Carini et al., 2016). Alternatively, the

detection of RNA molecules, which is generally less stable than DNA, might provide

finer spatiotemporally and fresher inferences of target organisms like a longer eDNA

fragment (described below; Barnes & Turner, 2016). Thirdly, the effects of various

biotic/abiotic factors on the relationship between eDNA persistence and DNA fragment

size should also be assessed. A fragment size-dependent degradation of eDNA may be

accelerated or suppressed depending on environmental conditions. Answering them

184

would allow the practical application of a longer eDNA fragment in natural

environment in the future.

8.3. Linking eDNA characteristics to its dynamics

In Chapters 2 and 4, I investigated the particle size distribution of nuclear and

mitochondrial eDNA from Japanese jack mackerels and obtained the inferences on

cellular characteristics of eDNA from macro-organisms. First, both nuclear and

mitochondrial eDNA was detected frequently at 3-10 µm size fraction. This implies that

much of fish eDNA is present as intra-cellular DNA but not dissolved DNA (i.e., DNA

molecules in the filtrate passing through a 0.2 µm pore size filter) in water (Turner et

al., 2014; Wilcox et al., 2015; Barnes et al., 2020). Second, just after the removal of

fish, eDNA concentrations increased more at larger size fractions. A physical stress

against the individuals might accelerate exfoliation of scales and mucus from their body

surfaces, which could shift the particle size distribution of target eDNA in experimental

tanks toward a larger size fraction. Third, time-series decreases in eDNA concentrations

were suppressed at smaller size fractions, which led to an increase in relative abundance

of smaller-sized eDNA in experimental tanks after fish removal. As mentioned in

Chapter 4, an apparent persistence of smaller-sized eDNA could be increased by the

inflow of degraded eDNA from larger to smaller size fractions. Further comparison of

eDNA particle size distribution and its time-series changes between artificial and natural

environments would generalize the understanding of the state and persistence of

aqueous eDNA which is summarized in Chapter 4 (Figure 4-6).

185

Through the studies in Chapters 2 to 5, the characteristics and dynamics of

eDNA were individually examined from the perspectives of molecular and cellular

states of eDNA; genetic region (nuclear or mitochondrial), DNA fragment size, and

particle size. However, considering that most of eDNA are released into environments

as intra-cellular DNA such as cell and tissue fragments, and various particle size and

state of eDNA are present in water, these eDNA characteristics should influence the

dynamics of eDNA. For example, with regard to the persistence of eDNA, if an intact

cell membrane protects DNA molecules from enzymatic degradation in aquatic

environment, such intra-cellular eDNA at larger size fractions might possess less-

degraded DNA molecules (i.e., longer DNA fragments) than extra-cellular eDNA at

smaller size fractions. Moreover, the effects of environmental factors on eDNA

degradation is likely to depend on its molecular and cellular states. Nevertheless, it has

so far remained unknown how molecular and cellular states of eDNA interact

environmental factors and how the interaction influences the dynamics of eDNA

including its production, transport, and persistence.

In these points, the findings in Chapters 6 and 7 provided the clues to

comprehensively understand the relationship between eDNA characteristics and

dynamics. Chapter 6 described that the relative yield of long to short eDNA increased

and the relative yield of nuclear to mitochondrial eDNA fragments decreased by

selective collection of larger-sized eDNA (i.e., eDNA detected at larger size fractions)

via a larger pore size filter. These results suggested that DNA fragmentation could be

suppressed in intra-cellular eDNA relative to extra-cellular eDNA due to the cellular

186

membrane, and implied that the persistence of nuclear and mitochondrial DNA could be

reversed between cellular and aquatic environments. In addition, Chapter 7 described

the various interactions between filter pore size, target gene, water temperature, and

water source on first-order eDNA decay rate constants based on a meta-analysis of

previous eDNA papers. The result in a meta-analysis was generally consistent with the

result of a re-analysis of a previous tank experiment which measured the time-series

changes in marine fish eDNA concentrations in multiple size fractions after fish

removal (Chapter 4). These results suggested that the mechanism of eDNA persistence

and degradation could be fully understood by knowing both environmental factors and

cellular and molecular states of eDNA in water.

In both chapters, it was particularly indicated that eDNA characteristics and

dynamics could substantially vary depending on its particle size and filter pore size. A

larger-sized eDNA possessed less degraded DNA fragments (Figure 6-2) but showed

higher CV values among sampling replicates (Table 6-4) than a smaller-sized eDNA

(i.e., eDNA detected at smaller size fractions). In contrast, likely due to the inflow of

degraded eDNA from larger to smaller size fractions, a smaller-sized eDNA showed a

lower decay rates and a smaller dependence of its degradation on water temperature and

possibly water sources than a larger-sized eDNA. The discrepancy between a higher

relative abundance of longer eDNA fragments but a higher eDNA decay rate in larger

size fractions may imply a temporal gap in the process of eDNA degradation between

degradation of its cellular structure and fragmentation of DNA molecules; DNA

fragmentation can occur following a certain amount of reduction in eDNA particle size

187

(i.e., degradation of its cellular structure) rather than reduction in eDNA particle size

and DNA fragmentation simultaneously occur in the eDNA degradation process. The

hypothesis may be reasonable considering that DNA fragmentation might occur less

frequently in intra-cellular than extra-cellular DNA, which is suggested in the section

8.2, while future works will be needed to further investigate when and how degradation

of cellular structure and DNA fragmentation occur in the process of eDNA degradation

in detail. The process of eDNA degradation similar to the hypothesis has been

considered for the eDNA derived from plants and microbes in soil and sediment (Poté et

al., 2005; Levy-Booth et al., 2007; Nielsen et al., 2007; Corinaldesi et al., 2008; Torti et

al., 2015), while there is some cytologic differences of eDNA sources between them and

macro-organisms such as vertebrates (e.g., the structure of nuclear DNA and the

presence of cell wall).

The difference in eDNA characteristics and dynamics between its particle

size may significantly enable to expand the new applicability of eDNA analysis as well

as to understand “the ecology of eDNA (Barnes & Turner, 2016)”. On the basis of all

the knowledge throughout the thesis, I summarize and discuss the potential for the

qualitative difference between larger-sized and smaller-sized eDNA. A selective

collection of larger-sized eDNA increases the capture efficiency of longer DNA

fragments, and likely less-degraded DNA, in water, which may enable to detect

haplotype diversity and SNPs more sensitively and improve the resolution of eDNA-

based taxonomic identification. Besides, considering its apparent shorter persistence and

particle size, larger-sized eDNA may show spatiotemporally finer signals in the field. In

188

contrast, smaller-sized eDNA is likely to possess more degraded DNA molecules and to

show broader spatiotemporal signals. However, it might not necessarily be a shortage

depending on the purpose of eDNA application; if researchers had an interest in among-

site comparison of representative species communities, smaller-sized eDNA might

reflect broader spatiotemporal scales of biodiversity information and allow to decrease

the sampling replicates and survey costs. With regard to estimation of species

biomass/abundance, because eDNA can be heterogeneously distributed due to

aggregation of cell and tissue fragments in environments (Furlan et al., 2016; Song et

al., 2017), a selective collection of larger-sized eDNA can cause to increase the

variation of eDNA quantification among replicates. On the other hand, quantification of

smaller-sized eDNA may be less dispersed among replicates and, owing to the inflow of

degraded eDNA from larger size fractions, be less dependent on environmental

parameters affecting eDNA degradation. It would allow to simplify the procedure of

modelling for biomass/abundance estimation accounting eDNA degradation with its

accuracy and reliability retained (as discussed in Chapter 7). Therefore, by

differentiating the particle size of target eDNA according to the purpose, it would be

possible to develop and establish the strategy of eDNA application more appropriate for

different purposes of studies. The qualitative trade-off of eDNA information between

larger-sized and smaller-sized eDNA could have the potential to refine our

understanding on the spatiotemporal range of eDNA signal, to reduce the uncertainties

relating to eDNA detection and quantification, and to substantially improve the

performance of eDNA detection and quantification. It will allow to fill a gap between

189

eDNA detection/quantification and species presence/abundance in the field and to

establish more accurate and detailed monitoring of biodiversity and fishery resources in

the future.

8.4. Further perspectives for the innovation of eDNA applications

I suggest the qualitative trade-off of eDNA information between its particle size in my

thesis, which is one of the major significances of my thesis. How can it contribute to the

innovation of eDNA-based biological monitoring and ecological understanding? As

mentioned above, larger-sized eDNA is likely to have less-degraded genomic

information as well as narrower spatiotemporal biological signals. Thus, the use of

larger-sized eDNA could be effective for the studies of population genetics and

population ecology, which is often destructive and difficult because of physical

collection of tissue samples from individuals. Some studies have assessed genetic

diversity of fish species using eDNA analysis (Uchii et al., 2016; Sigsgaard et al., 2017;

Uchii et al., 2017; Sigsgaard et al., 2020; Tsuji et al., 2020), in which mitochondrial

DNA fragments (100-500 bp) in water was mainly targeted. By selectively collecting

larger-sized and less-degraded eDNA, longer mitochondrial DNA fragments, and

possibly mitochondrial genome (Deiner et al., 2017b) can be efficiently captured and

sequenced from water samples. It may enable the estimation of intraspecific genetic

diversity and species identification with higher accuracy and resolution.

Moreover, selective collection of larger-sized eDNA using a larger pore size

filter may allow the enrichment of nuclear DNA from water samples and the

190

optimization of eDNA metabarcoding not based on PCR such as shotgun sequencing. In

the former, nuclear genome, particularly single-copy nuclear gene, has been used to

calculate effective population size more precisely than maternally-inherited

mitogenome (Birky et al., 1983; Moore, 1995) and to estimate sex ratio and kinship of

the population (Devlin & Nagahama, 2002; Dallas et al., 2003; Iacchei et al., 2013).

However, its copy number is much fewer than those of mitochondrial or multi-copy

nuclear (e.g., rRNA) genes, and thus its detectability from water samples is also

expected to be much lower. Water filtration via a larger pore size filter for selective

collection of larger-sized eDNA can accordingly increase the filtration volume by

preventing filter clogging, and accordingly may enhance the detectability of single-copy

gene in water samples. In the latter, although read abundance based on shotgun

sequencing can be used as the index of initial DNA amount than that based on PCR, it

remains an inefficient approach for aquatic macro-organisms due to the large amount of

non-target DNA derived from bacteria, algae, and fungi (Stat et al., 2017; Bovo et al.,

2020). Even in the pond with high density of common carp (Cyprinus carpio), carp

eDNA accounted for only 0.0004 % of total DNA (<0.01 ng mitochondrial DNA per 1 L

pond water) (Turner et al., 2014). Hybridization-based sequence capturing can be

effective for capture enrichment of target eDNA, while additional development of

hybridization probes is much costly (Creer et al., 2016; Wilcox et al., 2018; Giebner et

al., 2020). In contrast, using a larger pore size filter may decrease the capture efficiency

of such non-target eDNA and relatively increase the capture efficiency of target eDNA,

which would be much more cost-effective. Similarly, regarding amplicon sequencing

191

such as eDNA metabarcoding, enrichment of target eDNA in water samples using a

larger pore size filter may increase the relative abundance of target eDNA reads

available to analyze in a sample.

On the other hand, how should smaller-sized eDNA be used for ecological

studies? Larger-sized eDNA can be collected by using a larger pore size filter, while

collection of smaller-sized eDNA needs pre-filtering, which rather increases the effort

of water filtration and the risk of contamination in filtration process. Therefore, smaller-

sized eDNA itself would have limited uses contrary to larger-sized eDNA. Rather, we

should discuss in the future how the selective collection of larger-sized eDNA via a

larger pore size filter could be used differently from the inclusive collection of various

sizes of eDNA via a smaller pore size filter.

Another significance of my thesis is to expand the applicability of eDNA

analysis by utilizing nuclear and longer eDNA fragments as well as mitochondrial

shorter eDNA, which has been used in most of eDNA studies, suggesting the possibility

to extract more detailed ecological information from eDNA signals than species

presence and relative abundance. I showed that the combination of nuclear and

mitochondrial eDNA quantification could lead to estimate the age and developmental

stage of fish in Chapter 3, and the detection of longer eDNA fragments could remove

the effect of older eDNA derived from carcasses and resuspension from sediment in

Chapter 5, respectively. These inferences would substantially contribute to the

expansion of eDNA basic information such as its characteristics and dynamics, and

could be a clue against the uncertainties relating to eDNA detection and quantification.

192

To further progress in such an innovation of eDNA application, in addition to further

studies of basic information on multiple eDNA including nuclear and longer eDNA

fragments, I focus on the applicability of RNA molecules in environments

(environmental RNA; eRNA) and the combined use of eDNA and eRNA analyses.

Although all somatic cells in the individual generally have the same genomic

information, the pattern of their gene expressions can greatly vary depending on cell

type, developmental stage, and life history of the species (Budovskaya et al., 2008;

Cristescu, 2019). Thus, eRNA might be an index to estimate the age structure and

developmental stage of a population as well as the combination of nuclear and

mitochondrial eDNA. In addition, because RNA molecules are physiochemically

unstable and degradable rapidly contrary to DNA (van Hoof & Parker, 2002; Cristescu,

2019), eRNA might be able to provide finer spatiotemporal information on species

presence and might be an index to identify whether the individual is alive or dead. A

few studies targeting eRNA including eukaryotes have previously reported that species

composition based on eRNA metabarcoding was different from that based on eDNA

metabarcoding, possibly due to discrimination of living organisms (Laroche et al., 2016;

Pochon et al., 2017). On the other hand, eRNA release from fanworms (Sabella

spallanzanii and Styela clava) was much fewer than eDNA release, while eRNA

degradation was not significantly different from eDNA degradation (Wood et al., 2019),

which might vary depending on target genetic region, taxa, and environment.

The ‘extended environmental nucleic acid (e-eNA)’ analysis, which includes

mitochondrial and nuclear eDNA, shorter and longer eDNA fragments, larger-/smaller-

193

sized eDNA, and even eRNA, will substantially update the current eDNA analysis

(Figure 8-1). It will be applicable to the studies of physiological ecology including

metabolism, trophic state, and stress response of a population, as well as population

ecology as mentioned above. In Chapters 2 and 3, I found that shedding rates of

Japanese jack mackerel eDNA were promoted in warmer temperatures, implying that

even eDNA production can reflect the physiology of species such as their metabolism

rates and responses to environmental stress. In addition, the combined use of

mitochondrial and nuclear DNA can infer the age structure and/or spawning activity of

fish (Bylemans et al., 2017; Chapter 3 in the thesis). Nevertheless, it would be

insufficient to estimate the physiological state of species by only using DNA

information in environmental samples. Therefore, it would be better to utilize RNA

information reflecting specific states of physiology, which may be able to estimate

physiological information on a population more directly from environmental samples.

For example, ribosomal RNA (rRNA) concentration, or the ratio of rRNA to DNA

concentrations, has frequently been used to evaluate the metabolic state and growth rate

of microbial community (Muttray & Mohn, 1999; Blazewicz te al., 2013) and even of

fish species (Buckley, 1984; Chícharo & Chícharo, 2008), in the latter of which blood

samples are mainly used. The metabolism of fish population and community could be

also inferred non-invasively and cost-effectively by evaluating rRNA in water samples.

Moreover, targeting messenger RNA (mRNA) gene which expresses in the specific cell

type and/or tissue could directly indicate other physiological and ecological

information; spawning and reproduction activities of a population could directly be

194

indicated by the detection of the reproduction cell-specific gene (Tsuri et al., 2020).

Utilizing eRNA will also enable to directly identify physiological sources of

eDNA from macro-organisms, as is the case of forensic studies which have used mRNA

to identify different body fluids left at crime scenes (Vennemann & Koppelkamm,

2010). Recently, Tsuri et al. (2020) succeeded in detecting tissue-specific mRNA from

macro-organisms (Zebrafish [Danio rerio]) in water samples, indicating that feces and

epidermal cells are likely to be major sources of fish eDNA except for the reproduction

period. However, as far as I know, there is no study to reveal the source of eDNA from

other taxa. Considering the potential differences of species ecology and morphology, the

extent of production of eDNA derived from epidermis and mucus can vary depending

on taxa; contrary to fish and amphibians, mussels, crustaceans, and insects are likely to

produce less eDNA from their body surface (Ficetola et al., 2008; Mächler et al., 2014;

Klymus et al., 2015; Dougherty et al., 2016). In addition, it is likely that feces-derived

eDNA is more degraded than epidermal cells given the potential process of eDNA

production. By detecting the gene expressing in specific cell and/or tissue type, new

insights on developmental state, stress, and likely other aspects of demography may be

provided from environmental samples (Deiner et al., 2020).

In the summary, I revealed the characteristics and dynamics of marine fish

eDNA targeting multi-copy nuclear and longer DNA fragments as well as mitochondrial

shorter DNA fragment, and suggested that the combined use of various type of eDNA

including mitochondrial eDNA, nuclear eDNA, and longer eDNA fragments could

discriminate living and dead organisms and estimate the age and developmental stage of

195

individuals from water samples, and aqueous eDNA with different cellular and

molecular states could show different spatiotemporal inferences of the species and

qualitative information. Understanding the knowledge on the characteristics and

dynamics of multiple eDNA in the thesis would contribute to the further enrichment of

basic information of eDNA analysis. Moreover, new applicability of the analyses of

multiple eDNA would play important roles in the development of future eDNA analysis

for the study of population ecology. Similarly, the enrichment of less-degraded DNA

and single-copy gene in water by selective collection of larger-sized eDNA would

indicate the possibility of future development of eDNA analysis for the study of

population genetics. The feasibility of these approaches could be enhanced by e-eNA

analysis targeting both eDNA and eRNA. It would provide more precisely and

comprehensively ecological information than current eDNA analysis, and create the

novel research area of meta-genomics and meta-transcriptomics for macro-organisms.

The findings in my thesis is the important groundwork to innovate eDNA analysis for

biodiversity monitoring, ecological assessment, and fishery resource management in the

future.

196

8.5. Figure

Figure 8-1. Future prospects of ecosystem monitoring brought by the extended

environmental nucleic acid (e-eNA) analysis. As a summary of the thesis, I proposed e-

eNA analysis, which targets mitochondrial and nuclear eDNA, shorter and longer eDNA

fragments, larger-/smaller-sized eDNA, and even eRNA, for better understanding of the

ecology and ecosystems based on environmental samples. The novel approach will

allow to collect the information on population ecology, population genetics, and

physiological ecology non-destructively and cost-effectively, and accordingly to

promote further understanding of these ecological research areas.

Extended environmental nucleic acid (e-eNA) analysis

Environmental DNA (meta-genomics)

mitochondrial / nuclear gene longer / shorter DNA fragment

Environmental RNA (meta-transcriptomics)

Population ecology・biomass/abundance density・age/body size structure・developmental stage・mortality・life history

larger- / smaller-sized particle

circular mtDNAchromatin &histone modification

messenger RNA(mRNA)

transfer RNA(tRNA)

ribosomal RNA(rRNA)

A U C G

A U C G

A G C

Driving the understanding of

multiple ecological research areas

Physiological ecologyPopulation genetics・genetic diversity・(intra-/inter-specific)・functional diversity・individual identification・phylogeography

・growth rate・metabolism・stress response・immunology・trophic stateetc. etc. etc.

197

References throughout the thesis

Ahrenholtz, I., Lorenz, M. G., & Wackernagel, W. (1994). The extracellular nuclease of

Serratia marcescens: studies on the activity in vitro and effect on transforming DNA in a

groundwater aquifer microcosm. Archives of Microbiology, 161(2), 176-183.

Alaeddini, R., Walsh, S. J., & Abbas, A. (2010). Forensic implications of genetic

analyses from degraded DNA — a review. Forensic Science International: Genetics,

4(3), 148-157.

Alfreider, A., Pernthaler, J., Amann, R., Sattler, B., Glockner, F., Wille, A., & Psenner,

R. (1996). Community analysis of the bacterial assemblages in the winter cover and

pelagic layers of a high mountain lake by in situ hybridization. Applied and

Environmental Microbiology, 62(6), 2138-2144.

Amann, R. & Fuchs, B. M. (2008). Single-cell identification in microbial communities

by improved fluorescence in situ hybridization techniques. Nature Reviews

Microbiology, 6(5), 339-348.

Andruszkiewicz, A. E., Zhang, W. G., Lavery, A. C., & Govindarajan, A. F. (2020).

Environmental DNA shedding and decay rates from diverse animal forms and thermal

regimes. Environmental DNA, in press. DOI: 10.1002/edn3.141

Andruszkiewicz, E. A., Koseff, J. R., Fringer, O. B., Ouellette, N. T., Lowe, A. B.,

Edwards, C. A., & Boehm, A. B. (2019). Modeling environmental DNA transport in the

coastal ocean using Lagrangian particle tracking. Frontiers in Marine Science, 6, 477.

Arai, T., Aoyama, J., Limbong, D., & Tsukamoto, K. (1999). Species composition and

inshore migration of the tropical eels Anguilla spp. recruiting to the estuary of the

Poigar River, Sulawesi Island. Marine Ecology Progress Series, 188, 299-303.

Arnosti, C. (2014). Patterns of microbially driven carbon cycling in the ocean: links

between extracellular enzymes and microbial communities. Advances in Oceanography,

2014, 706082.

198

Balasingham, K. D., Walter, R. P., Mandrak, N. E., & Heath, D. D. (2017).

Environmental DNA detection of rare and invasive fish species in two great lakes

tributaries. Molecular Ecology, 27(1), 112-127.

Barnes, M. A., Chadderton, W. L., Jerde, C. L., Mahon, A. R., Turner, C. R., & Lodge,

D. M. (2020). Environmental conditions influence eDNA particle size distribution in

aquatic systems. Environmental DNA, in press. DOI: 10.1002/edn3.160

Barnes, M. A., & Turner, C. R. (2016). The ecology of environmental DNA and

implications for conservation genetics. Conservation Genetics, 17(1), 1-17.

Barnes, M. A., Turner, C. R., Jerde, C. L., Renshaw, M. A., Chadderton, W. L., &

Lodge, D. M. (2014). Environmental conditions influence eDNA persistence in aquatic

systems. Environmental Science & Technology, 48(3), 1819-1827.

Barton, B. A. (2002). Stress in fishes: A diversity of responses with particular reference

to changes in circulating corticosteroids. Integrative and Comparative Biology, 42(3),

517-525.

Bartoń, K. (2019). MuMIn: Multi-Model Inference. R package version 1.43.6.

https://CRAN.R-project.org/package=MuMIn

Bates, D., Maechler, M., Bolker, B., & Walker, S. (2015). Fitting linear mixed-effects

models using lme4. Journal of Statistical Software, 67, 1-48.

Belgrader, P., Benett, W., Hadley, D., Richards, J., Stratton, P., Mariella, R., &

Milanovich, F. (1999). PCR detection of bacteria in seven minutes. Science, 284(5413),

449-450.

Beng, K. C., & Corlett, R. T. (2020). Applications of environmental DNA (eDNA) in

ecology and conservation: opportunities, challenges and prospects. Biodiversity and

Conservation, 29, 2089-2121.

199

Bergmann, C. (1847). Über die verhältnisse der wärmeökonomie der thiere zu ihrer

grösse. Göttinger Studien, 5, 595-708.

Biggs, J., Ewald, N., Valentini, A., Gaboriaud, C., Dejean, T., Griffiths, R. A., Foster, J.,

Wilkinson, J. W., Arnell, A., Brotherton, P., Williams, P., & Dunn, F. (2015). Using

eDNA to develop a national citizen science-based monitoring programme for the great

crested newt (Triturus cristatus). Biological Conservation, 183, 19-28.

Birky, C. W., Maruyama, T., & Fuerst, P. (1983). An approach to population and

evolutionary genetic theory for genes in mitochondria and chloroplasts, and some

results. Genetics, 103(3), 513-527.

Bista, I., Carvalho, G. R., Walsh, K., Seymour, M., Hajibabaei, M., Lallias, D.,

Christmas, M., & Creer, S. (2017). Annual time-series analysis of aqueous eDNA

reveals ecologically relevant dynamics of lake ecosystem biodiversity. Nature

Communications, 8, 14087.

Blazewicz, S. J., Barnard, R. L., Daly, R. A., & Firestone, M. K. (2013). Evaluating

rRNA as an indicator of microbial activity in environmental communities: limitations

and uses. The ISME Journal, 7(11), 2061-2068.

Bohmann, K., Evans, A., Gilbert, M. T. P., Carvalho, G. R., Creer, S., Knapp, M., Yu, D.

W., & De Bruyn, M. (2014). Environmental DNA for wildlife biology and biodiversity

monitoring. Trends in Ecology & Evolution, 29(6), 358-367.

Booton, G. C., Kaufman, L., Chandler, M., Oguto-Ohwayo, R., Duan, W., & Fuerst, P.

A. (1999). Evolution of the ribosomal RNA internal transcribed spacer one (ITS-1) in

cichlid fishes of the Lake Victoria region. Molecular Phylogenetics and Evolution,

11(2), 273-282.

Boussarie, G., Bakker, J., Wangensteen, O. S., Mariani, S., Bonnin, L., Juhel, J. B.,

Kiszka, J. J., Kulbicki, M., Manel, S., Robbins, W. D., Vigliola, L., & Mouillot, D.

200

(2018). Environmental DNA illuminates the dark diversity of sharks. Science Advances,

4(5), eaap9661.

Bovo, S., Utzeri, V. J., Ribani, A., Cabbri, R., & Fontanesi, L. (2020). Shotgun

sequencing of honey DNA can describe honey bee derived environmental signatures

and the honey bee hologenome complexity. Scientific Reports, 10, 9279.

Brill, R. W., Holts, D. B., Chang, R. K. C., Sullivan, S., Dewar, H., & Carey, F. G.

(1993). Vertical and horizontal movements of striped marlin (Tetrapturus audax) near

the Hawaiian Islands, determined by ultrasonic telemetry, with simultaneous

measurement of oceanic currents. Marine Biology, 117, 567-574.

Buckley, L. J. (1984). RNA-DNA ratio: an index of larval fish growth in the sea. Marine

Biology, 80(3), 291-298.

Budovskaya, Y. V., Wu, K., Southworth, L. K., Jiang, M., Tedesco, P., Johnson, T. E., &

Kim, S. K. (2008). An elt-3/elt-5/elt-6 GATA transcription circuit guides aging in C. elegans. Cell, 134(2), 291-303.

Burnham, K. P., & Anderson, D. R. (2002). Model selection and inference: a practical

information–theoretic approach, 2nd edn. Springer, Berlin.

Butchart, S. H. M., Walpole, M., Collen, B., Van Strien, A., Scharlemann, J. P. W.,

Almond, R. E. A., Baillie, J. E. M., Bomhard, B., Brown, C., Bruno, J., Carpenter, K.

E., Carr, G. M., Chanson, J., Chenery, A. M., Csirke, J., Davidson, N. C., Dentener, F.,

Foster, M., Galli, A., Galloway, J. N., Genovesi, P., Gregory, R. D., Hockings, M.,

Kapos, V., Lamarque, J., Leverington, F., Loh, J., McGeoch, M. A., McRae, L.,

Minasyan, A., Morcillo, M. H., Oldfield, T. E. E., Pauly, D., Quader, S., Revenga, C.,

Sauer, J. R., Skolnik, B., Spear, D., Stanwell-Smith, D., Stuart, S. N., Symes, A.,

Tierney, M., Tyrrell, T. D., Vié, J., & Watson, R. (2010). Global biodiversity: indicators

of recent declines. Science, 328(5982), 1164-1168.

Buxton, A. S., Groombridge, J. J., & Griffiths, R. A. (2017). Is the detection of aquatic

201

environmental DNA influenced by substrate type? PLoS ONE, 12(8), e0183371.

Bylemans, J., Furlan, E. M., Gleeson, D. M., Hardy, C. M., & Duncan, R. P. (2018a).

Does size matter? An experimental evaluation of the relative abundance and decay rates

of aquatic environmental DNA. Environmental Science & Technology, 52(11), 6408-

6416.

Bylemans, J., Furlan, E. M., Hardy, C. M., McGuffie, P., Lintermans, M., & Gleeson, D.

M. (2017). An environmental DNA-based method for monitoring spawning activity: A

case study, using the endangered Macquarie perch (Macquaria australasica). Methods

in Ecology and Evolution, 8(5), 646-655.

Bylemans, J., Gleeson, D. M., Hardy, C. M., & Furlan, E. (2018b). Toward an ecoregion

scale evaluation of eDNA metabarcoding primers: A case study for the freshwater fish

biodiversity of the Murray-Darling Basin (Australia). Ecology and Evolution, 8(17),

8697-8712.

Carini, P., Marsden, P. J., Leff, J. W., Morgan, E. E., Strickland, M. S., & Fierer, N.

(2016). Relic DNA is abundant in soil and obscures estimates of soil microbial diversity.

Nature Microbiology, 2(3), 1-6.

Carraro, L., Hartikainen, H., Jokela, J., Bertuzzo, E., & Rinaldo, A. (2018). Estimating

species distribution and abundance in river networks using environmental DNA.

Proceedings of the National Academy of Sciences, 115(46), 11724-11729.

Caswell, H., Naiman, R. J., & Morin, R. (1984). Evaluating the consequences of

reproduction in complex salmonid life cycles. Aquaculture, 43(1-3), 123-134.

Ceballos, G., Ehrlich, P. R., Barnosky, A. D., García, A., Pringle, R. M., & Palmer, T.

M. (2015). Accelerated modern human–induced species losses: Entering the sixth mass

extinction. Science Advances, 1(5), e1400253.

Chen, W. & Ficetola, G. F. (2019). Conditionally autoregressive models improve

202

occupancy analyses of autocorrelated data: An example with environmental DNA.

Molecular Ecology Resources, 19(1), 163-175.

Chícharo, M. A. & Chícharo, L. (2008). RNA: DNA ratio and other nucleic acid derived

indices in marine ecology. International Journal of Molecular Sciences, 9(8), 1453-

1471.

Chow, S., Ueno, Y., Toyokawa, M., Oohara, I., & Takeyama, H. (2009). Preliminary

analysis of length and GC content variation in the ribosomal first internal transcribed

spacer (ITS1) of marine animals. Marine Biotechnology, 11(3), 301-306.

Clarke, A. & Johnston, N. M. (1999). Scaling of metabolic rate with body mass and

temperature in teleost fish. Journal of Animal Ecology, 68(5), 893-905.

Clay Montier, L. L. C., Deng, J. J., & Bai, Y. (2009). Number matters: control of

mammalian mitochondrial DNA copy number. Journal of Genetics and Genomics,

36(3), 125-131.

Collins, R. A., Wangensteen, O. S., O’Gorman, E. J., Mariani, S., Sims, D. W., &

Genner, M. J. (2018). Persistence of environmental DNA in marine systems.

Communications Biology, 1, 185.

Cook, J. R. & Heyse, J. F. (2000). Use of an angular transformation for ratio estimation

in cost-effectiveness analysis. Statistics in Medicine, 19(21), 2989-3003.

Corinaldesi, C., Beolchini, F., & Dell’Anno, A. (2008). Damage and degradation rates

of extracellular DNA in marine sediments: Implications for the preservation of gene

sequences. Molecular Ecology, 17(17), 3939-3951.

Coward, K., Bromage, N. R., Hibbitt, O., & Parrington, J. (2002). Gamete physiology,

fertilization and egg activation in teleost fish. Reviews in Fish Biology and Fisheries,

12(1), 33-58.

203

Creer, S., Deiner, K., Frey, S., Porazinska, D., Taberlet, P., Thomas, W. K., Potter, C., &

Bik, H. M. (2016). The ecologist's field guide to sequence-based identification of

biodiversity. Methods in Ecology and Evolution, 7(9), 1008-1018.

Cristescu, M. E. (2019). Can environmental RNA revolutionize biodiversity science?

Trends in Ecology & Evolution, 34(8), 694-697.

Dallas, J. F., Coxon, K. E., Sykes, T., Chanin, P. R., Marshall, F., Carss, D. N., Bacon, P.

J., Piertney, S. B., & Racey, P. A. (2003). Similar estimates of population genetic

composition and sex ratio derived from carcasses and faeces of Eurasian otter Lutra lutra. Molecular Ecology, 12(1), 275-282.

D’Anjou, R. M., Bradley, R. S., Balascio, N. L., & Finkelstein, D. B. (2012). Climate

impacts on human settlement and agricultural activities in northern Norway revealed

through sediment biogeochemistry. Proceedings of the National Academy of Sciences,

109(50), 20332-20337.

Darling, J. A. & Mahon, A. R. (2011). From molecules to management: adopting DNA-

based methods for monitoring biological invasions in aquatic environments.

Environmental Research, 111(7), 978-988.

Deagle, B. E., Eveson, J. P., & Jarman, S. N. (2006). Quantification of damage in DNA

recovered from highly degraded samples – a case study on DNA in faeces. Frontiers in

Zoology, 3(1), 11.

Deiner, K. & Altermatt, F. (2014). Transport distance of invertebrate environmental

DNA in a natural river. PLoS ONE, 9(2), e88786.

Deiner, K., Bik, H. M., Mächler, E., Seymour, M., Lacoursière-Roussel, A., Altermatt,

F., Creer, S., Bista, I., Lodge, D. M., De Vere, N., Pfrender, M. E., & Bernatchez, L.

(2017a). Environmental DNA metabarcoding: Transforming how we survey animal and

plant communities. Molecular Ecology, 26(21), 5872-5895.

204

Deiner, K., Fronhofer, E. A., Mächler, E., Walser, J. C., & Altermatt, F. (2016).

Environmental DNA reveals that rivers are conveyer belts of biodiversity information.

Nature Communications, 7, 12544.

Deiner, K., Renshaw, M. A., Li, Y., Olds, B. P., Lodge, D. M., & Pfrender, M. E.

(2017b). Long-range PCR allows sequencing of mitochondrial genomes from

environmental DNA. Methods in Ecology and Evolution, 8(12), 1888-1898.

Deiner, K., Yamanaka, H., & Bernatchez, L. (2020). The future of biodiversity

monitoring and conservation utilizing environmental DNA. Environmental DNA, in

press. DOI: 10.1002/edn3.178

Dejean, T., Valentini, A., Duparc, A., Pellier-Cuit, S., Pompanon, F., Taberlet, P., &

Miaud, C. (2011). Persistence of environmental DNA in freshwater ecosystems. PLoS

ONE, 6, e23398.

Dell'Anno, A. & Corinaldesi, C. (2004). Degradation and turnover of extracellular DNA

in marine sediments: Ecological and methodological considerations. Applied and

Environmental Microbiology, 70(7), 4384-4386.

Devlin, R. H. & Nagahama, Y. (2002). Sex determination and sex differentiation in fish:

an overview of genetic, physiological, and environmental influences. Aquaculture,

208(3-4), 191-364.

Díaz-Ferguson, E. E. & Moyer, G. R. (2014). History, applications, methodological

issues and perspectives for the use environmental DNA (eDNA) in marine and

freshwater environments. Revista De Biologia Tropical, 62, 1273-1284.

Ding, A. A. & Wu, H. (1999). Relationships between antiviral treatment effects and

biphasic viral decay rates in modeling HIV dynamics. Mathematical Biosciences,

160(1), 63-82.

Djurhuus, A., Closek, C. J., Kelly, R. P., Pitz, K. J., Michisaki, R. P., Starks, H. A.,

205

Walz, K. R., Andruszkiewicz, E. A., Olesin, E., Hubbard, K., Montes, E., Otis, D.,

Muller-Karger, F. E., Chavez, F. P., Boehm, A. B., & Breitbart, M. (2020).

Environmental DNA reveals seasonal shifts and potential interactions in a marine

community. Nature Communications, 11, 254.

Doi, H., Fukaya, K., Oka, S. I., Sato, K., Kondoh, M., & Miya, M. (2019). Evaluation

of detection probabilities at the water-filtering and initial PCR steps in environmental

DNA metabarcoding using a multispecies site occupancy model. Scientific Reports, 9,

3581.

Doi, H., Inui, R., Akamatsu, Y., Kanno, K., Yamanaka, H., Takahara, T., & Minamoto,

T. (2017). Environmental DNA analysis for estimating the abundance and biomass of

stream fish. Freshwater Biology, 62(1), 30-39.

Doi, H., Uchii, K., Takahara, T., Matsuhashi, S., Yamanaka, H., & Minamoto, T. (2015).

Use of droplet digital PCR for estimation of fish abundance and biomass in

environmental DNA surveys. PLoS ONE, 10(3), e0122763.

Dorazio, R. M. & Erickson, R. A. (2018). EDNAOCCUPANCY: An R package for

multiscale occupancy modelling of environmental DNA data. Molecular Ecology

Resources, 18(2), 368-380.

Dougherty, M. M., Larson, E. R., Renshaw, M. A., Gantz, C. A., Egan, S. P., Erickson,

D. M., & Lodge, D. M. (2016). Environmental DNA (eDNA) detects the invasive rusty

crayfish Orconectes rusticus at low abundances. Journal of Applied Ecology, 53(3),

722-732.

Dudgeon, D., Arthington, A. H., Gessner, M. O., Kawabata, Z. I., Knowler, D. J.,

Lévêque, C., Naiman, R. J., Prieur-Richard, A., Soto, D., Stiassny, M. L. J., & Sullivan,

C. A. (2006). Freshwater biodiversity: importance, threats, status and conservation

challenges. Biological Reviews, 81(2), 163-182.

Dunker, K. J., Sepulveda, A. J., Massengill, R. L., Olsen, J. B., Russ, O. L., Wenburg, J.

206

K., & Antonovich, A. (2016). Potential of environmental DNA to evaluate northern pike

(Esox lucius) eradication efforts: an experimental test and case study. PLoS ONE, 11(9),

e0162277.

Dunn, N., Priestley, V., Herraiz, A., Arnold, R., & Savolainen, V. (2017). Behavior and

season affect crayfish detection and density inference using environmental DNA.

Ecology and Evolution, 7(19), 7777-7785.

Dulvy, N. K., Sadovy, Y., & Reynolds, J. D. (2003). Extinction vulnerability in marine

populations. Fish and Fisheries, 4, 25-64.

Dysthe, J. C., Franklin, T. W., McKelvey, K. S., Young, M. K., & Schwartz, M. K.

(2018). An improved environmental DNA assay for bull trout (Salvelinus confluentus)

based on the ribosomal internal transcribed spacer I. PLoS ONE, 13(11).

Easton, J. H., Gauthier, J. J., Lalor, M. M., & Pitt, R. E. (2005). Die-off of pathogenic E. coli O157:H7 in sewage contaminated waters. Journal of the American Water Resources

Association, 41(5), 1187-1193.

Eichmiller, J. J., Best, S. E., & Sorensen, P. W. (2016). Effects of temperature and

trophic state on degradation of environmental DNA in lake water. Environmental

Science & Technology, 50(4), 1859-1867.

Ellison, S. L., English, C. A., Burns, M. J., & Keer, J. T. (2006). Routes to improving

the reliability of low level DNA analysis using real-time PCR. BMC Biotechnology,

6(1), 33.

Ellenberg, J., Siggia, E. D., Moreira, J. E., Smith, C. L., Presley, J. F., Worman, H. J., &

Lippincott-Schwartz, J. (1997). Nuclear membrane dynamics and reassembly in living

cells: targeting of an inner nuclear membrane protein in interphase and mitosis. The

Journal of Cell Biology, 138(6), 1193-1206.

Ernster, L. & Schatz, G. (1981). Mitochondria: A historical review. The Journal of Cell

207

Biology, 91(3), 227s-255s.

Evans, N. T. & Lamberti, G. A. (2018). Freshwater fisheries assessment using

environmental DNA: A primer on the method, its potential, and shortcomings as a

conservation tool. Fisheries Research, 197, 60-66.

Fahrenkrog, B. & Aebi, U. (2003). The nuclear pore complex: nucleocytoplasmic

transport and beyond. Nature Reviews Molecular Cell Biology, 4(10), 757-766.

Ficetola, G. F., Miaud, C., Pompanon, F., & Taberlet, P. (2008). Species detection using

environmental DNA from water samples. Biology Letters, 4(4), 423-425.

Fish, L. J. (1988). Why multivariate methods are usually vital. Measurement and

Evaluation in Counseling and Development, 21(3), 130-137.

Fittipaldi, M., Codony, F., Adrados, B., Camper, A. K., & Morató, J. (2011). Viable real-

time PCR in environmental samples: can all data be interpreted directly? Microbial

Ecology, 61(1), 7-12.

Foote, A. D., Thomsen, P. F., Sveegaard, S., Wahlberg, M., Kielgast, J., Kyhn, L.,

Salling, A. B., Galatius, A., Orlando, L., & Gilbert, M. T. P. (2012). Investigating the

potential use of environmental DNA (eDNA) for genetic monitoring of marine

mammals. PLoS ONE, 7, e41781.

Foran, D. R. (2006). Relative degradation of nuclear and mitochondrial DNA: An

experimental approach. Journal of Forensic Sciences, 51(4), 766-770.

Fremier, A. K., Strickler, K. M., Parzych, J., Powers, S., & Goldberg, C. S. (2019).

Stream transport and retention of environmental DNA pulse releases in relation to

hydrogeomorphic scaling factors. Environmental Science & Technology, 53(12), 6640-

6649.

208

Fukaya, K., Murakami, H., Yoon, S., Minami, K., Osada, Y., Yamamoto, S., Masuda, R.,

Kasai, A., Miyashita, K., Minamoto, T., & Kondoh, M. (2020). Estimating fish

population abundance by integrating quantitative data on environmental DNA and

hydrodynamic modelling. Molecular Ecology, in press. DOI: 10.1111/mec.15530

Fukumoto, S., Ushimaru, A., & Minamoto, T. (2015). A basin-scale application of

environmental DNA assessment for rare endemic species and closely related exotic

species in rivers: A case study of giant salamanders in Japan. Journal of Applied

Ecology, 52(2), 358-365.

Furlan, E. M., Gleeson, D., Hardy, C. M., & Duncan, R. P. (2016). A framework for

estimating the sensitivity of eDNA surveys. Molecular Ecology Resources, 16(3), 641-

654.

Gantz, C. A., Renshaw, M. A., Erickson, D., Lodge, D. M., & Egan, S. P. (2018).

Environmental DNA detection of aquatic invasive plants in lab mesocosm and natural

field conditions. Biological Invasions, 20(9), 2535-2552.

Giebner, H., Langen, K., Bourlat, S. J., Kukowka, S., Mayer, C., Astrin, J. J., Misof, B.,

& Fonseca, V. G. (2020). Comparing diversity levels in environmental samples: DNA

sequence capture and metabarcoding approaches using 18S and COI genes. Molecular

Ecology Resources, 20(5), 1333-1345.

Giguet-Covex, C., Pansu, J., Arnaud, F., Rey, P. J., Griggo, C., Gielly, L., Domaizon, I.,

Coissac, E., David, F., Choler, P., Poulenard, J., & Taberlet, P. (2014). Long livestock

farming history and human landscape shaping revealed by lake sediment DNA. Nature

Communications, 5, 3211.

Goldberg, C. S., Sepulveda, A., Ray, A., Baumgardt, J., & Waits, L. P. (2013).

Environmental DNA as a new method for early detection of New Zealand mudsnails

(Potamopyrgus antipodarum). Freshwater Science, 32(3), 792-800.

Goldberg, C. S., Strickler, K. M., & Pilliod, D. S. (2015). Moving environmental DNA

209

methods from concept to practice for monitoring aquatic macroorganisms. Biological


Gomes, G. B., Hutson, K. S., Domingos, J. A., Chung, C., Hayward, S., Miller, T. L., &

Jerry, D. R. (2017). Use of environmental DNA (eDNA) and water quality data to

predict protozoan parasites outbreaks in fish farms. Aquaculture, 479, 467-473.

Grunstein, M. (1997). Histone acetylation in chromatin structure and transcription.

Nature, 389(6649), 349.

Handelsman, J. (2004). Metagenomics: application of genomics to uncultured

microorganisms. Microbiology and Molecular Biology Reviews, 68(4), 669-685.

Hänfling, B., Lawson Handley, L., Read, D. S., Hahn, C., Li, J., Nichols, P., Blackman,

R. C., Oliver, A., & Winfield, I. J. (2016). Environmental DNA metabarcoding of lake

fish communities reflects long-term data from established survey methods. Molecular

Ecology, 25(13), 3101-3119.

Hansen, B. K., Bekkevold, D., Clausen, L. W., & Nielsen, E. E. (2018). The sceptical

optimist: challenges and perspectives for the application of environmental DNA in

marine fisheries. Fish and Fisheries, 19(5), 751-768.

Hartmann, N., Reichwald, K., Wittig, I., Dröse, S., Schmeisser, S., Lück, C., Hahn, C.,

Graf, M., Gausmann, U., Terzibasi, E., Cellerino, A., Ristow, M., Brandt, U., Platzer,

M., & Englert, C. (2011). Mitochondrial DNA copy number and function decrease with

age in the short-lived fish Nothobranchius furzeri. Aging Cell, 10(5), 824-831.

Hayakawa, M., Torii, K., Sugiyama, S., Tanaka, M., & Ozawa, T. (1991). Age-associated accumulation of 8-hydroxydeoxyguanosine in mitochondrial DNA of human

diaphragm. Biochemical and Biophysical Research Communications, 179(2), 1023-

1029.

Heard, W. R. (1991). Life history of pink salmon (Oncorhynchus gorbuscha). In C.

210

Groot & L. Margolis (Eds.), Pacific Salmon Life Histories, (pp. 121-234). Vancouver,

Canada: UBC press.

Henderson, H. F., Hasler, A. D., & Chipman, G. G. (1966). An ultrasonic transmitter for

use in studies of movements of fishes. Transactions of the American Fisheries Society,

95, 350-356.

Hillis, D. M. & Dixon, M. T. (1991). Ribosomal DNA: Molecular evolution and

phylogenetic inference. The Quarterly Review of Biology, 66(4), 411-453.

Hosfield, D. J., Mol, C. D., Shen, B., & Tainer, J. A. (1998). Structure of the DNA

repair and replication endonuclease and exonuclease FEN-1: Coupling DNA and PCNA

binding to FEN-1 activity. Cell, 95(1), 135-146.

Iacchei, M., Ben-Horin, T., Selkoe, K. A., Bird, C. E., García-Rodríguez, F. J., &

Toonen, R. J. (2013). Combined analyses of kinship and FST suggest potential drivers of

chaotic genetic patchiness in high gene-flow populations. Molecular Ecology, 22(13),

3476-3494.

Iwai, N., Yasumiba, K., & Takahara, T. (2019). Efficacy of environmental DNA to

detect and quantify stream tadpoles of Odorrana splendida. Royal Society Open

Science, 6(1), 181798.

Jane, S. F., Wilcox, T. M., McKelvey, K. S., Young, M. K., Schwartz, M. K., Lowe, W.

H., Letcher, B. H., & Whiteley, A. R. (2015). Distance, flow and PCR inhibition: e DNA

dynamics in two headwater streams. Molecular Ecology Resources, 15(1), 216-227.

Jerde, C. L., Olds, B. P., Shogren, A. J., Andruszkiewicz, E. A., Mahon, A. R., Bolster,

D., & Tank, J. L. (2016). Influence of stream bottom substrate on retention and transport

of vertebrate environmental DNA. Environmental Science & Technology, 50(16), 8770-

8779.

Jo, T., Arimoto, M., Murakami, H., Masuda, R., & Minamoto, T. (2020b). Estimating

211

shedding and decay rates of environmental nuclear DNA with relation to water

temperature and biomass. Environmental DNA, 2(2), 140-151.

Jo, T., Arimoto, M., Murakami, H., Masuda, R., & Minamoto, T. (2019b). Particle size

distribution of environmental DNA from the nuclei of marine fish. Environmental


Jo, T., Fukuoka, A., Uchida, K., Ushimaru, A., & Minamoto, T. (2020a). Multiplex real-

time PCR enables the simultaneous detection of environmental DNA from freshwater

fishes: a case study of three exotic and three threatened native fishes in Japan.

Biological Invasions, 22(2), 455-471.

Jo, T., Murakami, H., Masuda, R., & Minamoto, T. (2020c). Selective collection of long

fragments of environmental DNA using larger pore size filter. Science of the Total

Environment, 139462.

Jo, T., Murakami, H., Masuda, R., Sakata, M., Yamamoto, S., & Minamoto, T. (2017).

Rapid degradation of longer DNA fragments enables the improved estimation of

distribution and biomass using environmental DNA. Molecular Ecology Resources,

17(6), e25-e33.

Jo, T., Murakami, H., Yamamoto, S., Masuda, R., & Minamoto, T. (2019a). Effect of

water temperature and fish biomass on environmental DNA shedding, degradation, and

size distribution. Ecology and Evolution, 9(3), 1135-1146.

Kamenova, S., Mayer, R., Rubbmark, O. R., Coissac, E., Plantegenest, M., & Traugott,

M. (2018). Comparing three types of dietary samples for prey DNA decay in an insect

generalist predator. Molecular Ecology Resources, 18(5), 966-973.

Katano, I., Harada, K., Doi, H., Souma, R., & Minamoto, T. (2017). Environmental

DNA method for estimating salamander distribution in headwater streams, and a

comparison of water sampling methods. PLoS ONE, 12(5), e0176541.

212

Klymus, K. E., Richter, C. A., Chapman, D. C., & Paukert, C. (2015). Quantification of

eDNA shedding rates from invasive bighead carp Hypophthalmichthys nobilis and silver

carp Hypophthalmichthys molitrix. Biological Conservation, 183, 77-84.

Kornberg, R. D. (1974). Chromatin structure: A repeating unit of histones and DNA.

Science, 184(4139), 868-871.

Koshiba, T., Detmer, S. A., Kaiser, J. T., Chen, H., McCaffery, J. M., & Chan, D. C.

(2004). Structural basis of mitochondrial tethering by mitofusin complexes. Science,

305(5685), 858-862.

Künkele, K. P., Heins, S., Dembowski, M., Nargang, F. E., Benz, R., Thieffry, M., Walz,

J., Lill, R., Nussberger, S., & Neupert, W. (1998). The preprotein translocation channel

of the outer membrane of mitochondria. Cell, 93(6), 1009-1019.

Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. B. (2017). lmerTest package:

tests in linear mixed effects models. Journal of Statistical Software, 82(13), 1-26.

Lacoursière-Roussel, A., Howland, K., Normandeau, E., Grey, E. K., Archambault, P.,

Deiner, K., Lodge, D. M., Hernandez, C., Leduc, N., & Bernatchez, L. (2018). eDNA

metabarcoding as a new surveillance approach for coastal Arctic biodiversity. Ecology

& Evolution, 8(16), 7763-7777.

Lance, R. F., Klymus, K. E., Richter, C. A., Guan, X., Farrington, H. L., Carr, M. R.,

Thompson, N., Chapman, D. C., & Baerwaldt, K. L. (2017). Experimental observations

on the decay of environmental DNA from bighead and silver carps. Management of

Biological Invasions, 8(3), 343.

Laroche, O., Wood, S. A., Tremblay, L. A., Ellis, J. I., Lejzerowicz, F., Pawlowski, J.,

Lear, G., Atalah, J., & Pochon, X. (2016). First evaluation of foraminiferal

metabarcoding for monitoring environmental impact from an offshore oil drilling site.

Marine Environmental Research, 120, 225-235.

213

Levy-Booth, D. J., Campbell, R. G., Gulden, R. H., Hart, M. M., Powell, J. R.,

Klironomos, J. N., Paulsb, K. P., Swantonb, C. J., Trevorsa, J. T., & Dunfield, K. E.

(2007). Cycling of extracellular DNA in the soil environment. Soil Biology and

Biochemistry, 39(12), 2977-2991.

Li, Y., Evans, N. T., Renshaw, M. A., Jerde, C. L., Olds, B. P., Shogren, A. J., Deiner,

K., Lodge, D. M., Lamberti, G. A., & Pfrender, M. E. (2018). Estimating fish alpha- and

beta-diversity along a small stream with environmental DNA metabarcoding.

Metabarcoding and Metagenomics, 2, e24262.

Lindahl, T. (1993). Instability and decay of the primary structure of DNA. Nature,

362(6422), 709-715.

Lloyd, H. M., Jacobi, J. M., & Cooke, R. A. (1979). Nuclear diameter in parathyroid

adenomas. Journal of Clinical Pathology, 32(12), 1278-1281.

Lodge, D. M., Turner, C. R., Jerde, C. L., Barnes, M. A., Chadderton, L., Egan, S. P.,

Feder, J. L., Mahon, A. R., & Pfrender, M. E. (2012). Conservation in a cup of water:

Estimating biodiversity and population abundance from environmental DNA. Molecular

Ecology, 21(11), 2555-2558.

Long, E. O. & Dawid, I. B. (1980). Repeated genes in eukaryotes. Annual Review of

Biochemistry, 49(1), 727-764.

Mächler, E., Deiner, K., Steinmann, P., & Altermatt, F. (2014). Utility of environmental

DNA for monitoring rare and indicator macroinvertebrate species. Freshwater Science,

33(4), 1174-1183.

MacKenzie, D. I., Nichols, J. D., Hines, J. E., Knutson, M. G., & Franklin, A. B. (2003).

Estimating site occupancy, colonization, and local extinction when a species is detected

imperfectly. Ecology, 84(8), 2200-2207.

Martellini, A., Payment, P., & Villemur, R. (2005). Use of eukaryotic mitochondrial

214

DNA to differentiate human, bovine, porcine and ovine sources in fecally contaminated

surface water. Water Research, 39(4), 541-548.

Maruyama, A., Nakamura, K., Yamanaka, H., Kondoh, M., & Minamoto, T. (2014). The

release rate of environmental DNA from juvenile and adult fish. PLoS ONE, 9(12),

e114639.

Masuda, R. (2008). Seasonal and interannual variation of subtidal fish assemblages in

Wakasa Bay with reference to the warming trend in the Sea of Japan. Environmental

Biology of Fishes, 82, 387-399.

Mauvisseau, Q., Burian, A., Gibson, C., Brys, R., Ramsey, A., & Sweet, M. (2019).

Influence of accuracy, repeatability and detection probability in the reliability of

species-specific eDNA based approaches. Scientific Reports, 9, 580.

Matsui, K., Honjo, M., & Kawabata, Z. (2001). Estimation of the fate of dissolved DNA

in thermally stratified lake water from the stability of exogenous plasmid DNA. Aquatic

Microbial Ecology, 26(1), 95-102.

Merkes, C. M., McCalla, S. G., Jensen, N. R., Gaikowski, M. P., & Amberg, J. J.

(2014). Persistence of DNA in carcasses, slime and avian feces may affect interpretation

of environmental DNA data. PLoS ONE, 9(11), e113346.

Mikutis, G., Schmid, L., Stark, W. J., & Grass, R. N. (2019). Length-dependent DNA

degradation kinetic model: Decay compensation in DNA tracer concentration

measurements. AIChE Journal, 65(1), 40-48.

Minamoto, T., Fukuda, M., Katsuhara, K. R., Fujiwara, A., Hidaka, S., Yamamoto, S.,

Takahashi, K., & Masuda, R. (2017a). Environmental DNA reflects spatial and temporal

jellyfish distribution. PLoS ONE, 12(2), e0173073.

Minamoto, T., Naka, T., Moji, K., & Maruyama, A. (2016). Techniques for the practical

collection of environmental DNA: filter selection, preservation, and extraction.

215

Limnology, 17(1), 23-32.

Minamoto, T., Uchii, K., Takahara, T., Kitayoshi, T., Tsuji, S., Yamanaka, H., & Doi, H.

(2017b). Nuclear internal transcribed spacer-1 as a sensitive genetic marker for

environmental DNA studies in common carp Cyprinus carpio. Molecular Ecology

Resources, 17(2), 324-333.

Minamoto, T., Yamanaka, H., Takahara, T., Honjo, M. N., & Kawabata, Z. (2012).

Surveillance of fish species composition using environmental DNA. Limnology, 13(2),

193-197.

Mitani, F. & Ida, E. (1964). A study on the growth and age of the jack mackerel in the

East China Sea. Bulletin of the Japanese Society of Scientific Fisheries, 30(12), 968-

977. (in Japanese).

Mizumoto, H., Urabe, H., Kanbe, T., Fukushima, M., & Araki, H. (2018). Establishing

an environmental DNA method to detect and estimate the biomass of Sakhalin taimen, a

critically endangered Asian salmonid. Limnology, 19(2), 219-227.

Moore, W. S. (1995). Inferring phylogenies from mtDNA variation: mitochondrial-gene

trees versus nuclear-gene trees. Evolution, 49(4), 718-726.

Morita, K., Fukuwaka, M. A., Tanimata, N., & Yamamura, O. (2010). Size-dependent

thermal preferences in a pelagic fish. Oikos, 119(8), 1265-1272.

Moushomi, R., Wilgar, G., Carvalho, G., Creer, S., & Seymour, M. (2019).

Environmental DNA size sorting and degradation experiment indicates the state of

Daphnia magna mitochondrial and nuclear eDNA is subcellular. Scientific Reports, 9,

12500.

Murakami, H., Yoon, S., Kasai, A., Minamoto, T., Yamamoto, S., Sakata, M. K.,

Horiuchi, T., Sawada, H., Kondoh, M., Yamashita, Y., & Masuda, R. (2019). Dispersion

and degradation of environmental DNA from caged fish in a marine environment.

216

Fisheries Science, 85(2), 327-337.

Muttray, A. F. & Mohn, W. W. (1999). Quantitation of the population size and metabolic

activity of a resin acid degrading bacterium in activated sludge using slot-blot

hybridization to measure the rRNA: rDNA ratio. Microbial Ecology, 38(4), 348-357.

Nakamura, T. & Hamano, A. (2009). Seasonal differences in the vertical distribution

pattern of Japanese jack mackerel, Trachurus japonicus: Changes according to age?

ICES Journal of Marine Science, 66(6), 1289-1295.

Nichols, P. K. & Marko, P. B. (2019). Rapid assessment of coral cover from

environmental DNA in Hawai'i. Environmental DNA, 1(1), 40-53.

Nielsen, K. M., Johnsen, P. J., Bensasson, D., & Daffonchio, D. (2007). Release and

persistence of extracellular DNA in the environment. Environmental Biosafety

Research, 6(1-2), 37-53.

Nukazawa, K., Hamasuna, Y., & Suzuki, Y. (2018). Simulating the advection and

degradation of the environmental DNA of common carp along a river. Environmental


Ogram, A., Sayler, G. S., & Barkay, T. (1987). The extraction and purification of

microbial DNA from sediments. Journal of Microbiological Methods, 7(2-3), 57-66.

Okabe, S. & Shimazu, Y. (2007). Persistence of host-specific Bacteroides–Prevotella 16S rRNA genetic markers in environmental waters: effects of temperature and salinity.

Applied Microbiology and Biotechnology, 76(4), 935-944.

Okazaki, Y., Hodoki, Y., & Nakano, S. I. (2013). Seasonal dominance of CL500-11

bacterioplankton (phylum Chloroflexi) in the oxygenated hypolimnion of Lake Biwa,

Japan. FEMS Microbiology Ecology, 83(1), 82-92.

Panté, N. & Kann, M. (2002). Nuclear pore complex is able to transport

217

macromolecules with diameters of ∼39 nm. Molecular Biology of the Cell, 13(2), 425-

434.

Person-Le Ruyet, J., Mahe, K., Le Bayon, N., & Le Delliou, H. (2004). Effects of

temperature on growth and metabolism in a Mediterranean population of European sea

bass, Dicentrarchus labrax. Aquaculture, 237(1), 269-280.

Pilliod, D. S., Goldberg, C. S., Arkle, R. S., & Waits, L. P. (2014). Factors influencing

detection of eDNA from a stream-dwelling amphibian. Molecular Ecology Resources,

14(1), 109-116.

Pochon, X., Zaiko, A., Fletcher, L. M., Laroche, O., & Wood, S. A. (2017). Wanted dead

or alive? Using metabarcoding of environmental DNA and RNA to distinguish living

assemblages for biosecurity applications. PLoS ONE, 12(11), e0187636.

Port, J. A., O’Donnell, J. L., Romero-Maraccini, O. C., Leary, P. R., Litvin, S. Y.,

Nichols, K. J., Yamahara, K. M., & Kelly, R. P. (2016). Assessing vertebrate

biodiversity in a kelp forest ecosystem using environmental DNA. Molecular Ecology,

25, 527-541.

Poté, J., Ackermann, R., & Wildi, W. (2009). Plant leaf mass loss and DNA release in

freshwater sediments. Ecotoxicology and Environmental Safety, 72(5), 1378-1383.

Poté, J., Rossé, P., Rosselli, W., & Wildi, W. (2005). Kinetics of mass and DNA

decomposition in tomato leaves. Chemosphere, 61(5), 677-684.

Price, P. B. & Sowers, T. (2004). Temperature dependence of metabolic rates for

microbial growth, maintenance, and survival. Proceedings of the National Academy of

Sciences, 101(13), 4631-4636.

Prokopowich, C. D., Gregory, T. R., & Crease, T. J. (2003). The correlation between

rDNA copy number and genome size in eukaryotes. Genome, 46(1), 48-50.

218

R Core Team. (2016). R: A Language and Environment for Statistical Computing. R

Foundation for Statistical Computing, Vienna, Austria.

R Core Team. (2019). R: A Language and Environment for Statistical Computing. R

Foundation for Statistical Computing, Vienna, Austria.

Ravanat, J. L., Douki, T., & Cadet, J. (2001). Direct and indirect effects of UV radiation

on DNA and its components. Journal of Photochemistry and Photobiology B: Biology,

63(1-3), 88-102.

Rawsthorne, H., Dock, C. N., & Jaykus, L. A. (2009). PCR-based method using

propidium monoazide to distinguish viable from nonviable Bacillus subtilis spores.

Applied and Environmental Microbiology, 75(9), 2936-2939.

Robin, E. D. & Wong, R. (1988). Mitochondrial DNA molecules and virtual number of

mitochondria per cell in mammalian cells. Journal of Cellular Physiology, 136(3), 507-

513.

Robinson, B. H. & Bailey, T. G. (1981). Sinking rates and dissolution of midwater fish

fecal matter. Marine Biology, 65(2), 135-142.

Robson, H. L., Noble, T. H., Saunders, R. J., Robson, S. K., Burrows, D. W., & Jerry, D.

R. (2016). Fine-tuning for the tropics: application of eDNA technology for invasive fish

detection in tropical freshwater ecosystems. Molecular Ecology Resources, 16(4), 922-

932.

Rockström, J., Steffen, W., Noone, K., Persson, Å., Chapin, F. S., Lambin, E. F., Lenton,

T. M., Scheffer, M., Folke, C., Schellnhuber, H. J., Nykvist, B., De Wit, C. A., Hughes,

T., Van der Leeuw, S., Rodhe, H., Sörlin, S., Snyder, P. K., Costanza, R., Svedin, U.,

Falkenmark, M., Karlberg, L., Corell, R. W., Fabry, V. J., Hansen, J., Walker, B.,

Liverman, D., Richardson, K., Crutzen, P., & Foley, J. A. (2009). A safe operating space

for humanity. Nature, 461(7263), 472-475.

219

Rogers, S. W., Donnelly, M., Peed, L., Kelty, C. A., Mondal, S., Zhong, Z., & Shanks,

O. C. (2011). Decay of bacterial pathogens, fecal indicators, and real-time quantitative

PCR genetic markers in manure-amended soils. Applied and Environmental

Microbiology, 77(14), 4839-4848.

Roussel, J. M., Paillisson, J. M., Tréguier, A., & Petit, E. (2015). The downside of

eDNA as a survey tool in water bodies. Journal of Applied Ecology, 52(4), 823-826.

Sakata, M. K., Yamamoto, S., Gotoh, R. O., Miya, M., Yamanaka, H., & Minamoto, T.

(2020). Sedimentary eDNA provides different information on timescale and fish species

composition compared with aqueous eDNA. Environmental DNA, 2(4), 505-518.

Sandersfeld, T., Mark, F. C., & Knust, R. (2017). Temperature-dependent metabolism in

Antarctic fish: Do habitat temperature conditions affect thermal tolerance ranges? Polar

Biology, 40(1), 141-149.

San Diego-McGlone, M. L., Azanza, R. V., Villanoy, C. L., & Jacinto, G. S. (2008).

Eutrophic waters, algal bloom and fish kill in fish farming areas in Bolinao, Pangasinan,

Philippines. Marine Pollution Bulletin, 57(6-12), 295-301.

Sansom, B. J. & Sassoubre, L. M. (2017). Environmental DNA (eDNA) shedding and

decay rates to model freshwater mussel eDNA transport in a river. Environmental


Sassa, C. & Konishi, Y. (2006). Vertical distribution of jack mackerel Trachurus japonicus larvae in the southern part of the East China Sea. Fisheries Science, 72(3),

612-619.

Sassoubre, L. M., Yamahara, K. M., Gardner, L. D., Block, B. A., & Boehm, A. B.

(2016). Quantification of environmental DNA (eDNA) shedding and decay rates for

three marine fish. Environmental Science & Technology, 50(19), 10456-10464.

220

Schulz, C. J. & Childers, G. W. (2011). Fecal Bacteroidales diversity and decay in

response to variations in temperature and salinity. Applied and Environmental

Microbiology, 77(8), 2563-2572.

Sengupta, M. E., Hellström, M., Kariuki, H. C., Olsen, A., Thomsen, P. F., Mejer, H.,

Willerslev, E., Mwanjei, M. T., Madsena, H., Kristensena, T. K., Stensgaard, A., &

Vennervald, B. J. (2019). Environmental DNA for improved detection and

environmental surveillance of schistosomiasis. Proceedings of the National Academy of

Sciences, 116(18), 8931-8940.

Seymour, M., Durance, I., Cosby, B. J., Ransom-Jones, E., Deiner, K., Ormerod, S. J.,

Colbourne, J. K., Wilgar, G., Carvalho, G. R., De Bruyn, M., Edwards, F., Emmett, B.

A., Bik, H. M., & Edwards, F. (2018). Acidity promotes degradation of multi-species

environmental DNA in lotic mesocosms. Communications Biology, 1, 4.

Shadel, G. S. & Clayton, D. A. (1997). Mitochondrial DNA maintenance in vertebrates.

Annual Review of Biochemistry, 66(1), 409-435.

Shogren, A. J., Tank, J. L., Andruszkiewicz, E. A., Olds, B., Jerde, C., & Bolster, D.

(2016). Modelling the transport of environmental DNA through a porous substrate using

continuous flow-through column experiments. Journal of the Royal Society Interface,

13(119), 20160290.

Shogren, A. J., Tank, J. L., Andruszkiewicz, E., Olds, B., Mahon, A. R., Jerde, C. L., &

Bolster, D. (2017). Controls on eDNA movement in streams: Transport, retention, and

resuspension. Scientific Reports, 7, 5065.

Shogren, A. J., Tank, J. L., Egan, S. P., August, O., Rosi, E. J., Hanrahan, B. R.,

Renshaw, M. A., Gantz, C. A., & Bolster, D. (2018). Water flow and biofilm cover

influence environmental DNA detection in recirculating streams. Environmental


Shogren, A. J., Tank, J. L., Egan, S. P., Bolster, D., & Riis, T. (2019). Riverine

221

distribution of mussel environmental DNA reflects a balance among density, transport,

and removal processes. Freshwater Biology, 64(8), 1467-1479.

Sigsgaard, E. E., Jensen, M. R., Winkelmann, I. E., Møller, P. R., Hansen, M. M., &

Thomsen, P. F. (2020). Population-level inferences from environmental DNA—Current

status and future perspectives. Evolutionary Applications, 13(2), 245-262.

Sigsgaard, E. E., Nielsen, I. B., Bach, S. S., Lorenzen, E. D., Robinson, D. P., Knudsen,

S. W., Pedersen, M. W., Al Jaidah, M., Orlando, L., Willerslev, E., Møller, P. R., &

Thomsen, P. F. (2016). Population characteristics of a large whale shark aggregation

inferred from seawater environmental DNA. Nature Ecology & Evolution, 1, 0004.

Song, J. W., Small, M. J., & Casman, E. A. (2017). Making sense of the noise: The

effect of hydrology on silver carp eDNA detection in the Chicago area waterway

system. Science of the Total Environment, 605, 713-720.

Stalder, J., Larsen, A., Engel, J. D., Dolan, M., Groudine, M., & Weintraub, H. (1980).

Tissue-specific DNA cleavages in the globin chromatin domain introduced by DNAase

I. Cell, 20(2), 451-460.

Stat, M., Huggett, M. J., Bernasconi, R., DiBattista, J. D., Berry, T. E., Newman, S. J.,

Harvey, E. S., & Bunce, M. (2017). Ecosystem biomonitoring with eDNA:

metabarcoding across the tree of life in a tropical marine environment. Scientific

Reports, 7, 12240.

Stewart, K. A. (2019). Understanding the effects of biotic and abiotic factors on sources

of aquatic environmental DNA. Biodiversity and Conservation, 28(5), 983-1001.

Strickler, K. M., Fremier, A. K., & Goldberg, C. S. (2015). Quantifying effects of UV-B,

temperature, and pH on eDNA degradation in aquatic microcosms. Biological


Suen, D. F., Norris, K. L., & Youle, R. J. (2008). Mitochondrial dynamics and

222

apoptosis. Genes & Development, 22(12), 1577-1590.

Taberlet, P., Coissac, E., Hajibabaei, M., & Rieseberg, L. H. (2012). Environmental

DNA. Molecular Ecology, 21(8), 1789-1793.

Takahara, T., Honjo, M. N., Uchii, K., Minamoto, T., Doi, H., Ito, T., & Kawabata, Z.

(2014). Effects of daily temperature fluctuation on the survival of carp infected with

Cyprinid herpesvirus 3. Aquaculture, 433, 208-213.

Takahara, T., Minamoto, T., & Doi, H. (2013). Using environmental DNA to estimate

the distribution of an invasive fish species in ponds. PLoS ONE, 8(2), e56584.

Takahara, T., Minamoto, T., Yamanaka, H., Doi, H., & Kawabata, Z. (2012). Estimation

of fish biomass using environmental DNA. PLoS ONE, 7(4), e35868.

Takahara, T., Yamanaka, H., Minamoto, T., Doi, H., & Uchii, K. (2016). Current state of

biomonitoring method using environmental DNA analysis. Japanese Journal of Ecology,

66(3), 583-599. (in Japanese)

Takeuchi, A., Iijima, T., Kakuzen, W., Watanabe, S., Yamada, Y., Okamura, A., Horie,

N., Mikawa, N., Miller, M. J., Kojima, T., & Tsukamoto, K. (2019). Release of eDNA

by different life history stages and during spawning activities of laboratory-reared

Japanese eels for interpretation of oceanic survey data. Scientific Reports, 9, 6074.

Thomsen, P. F., Kielgast, J., Iversen, L. L., Møller, P. R., Rasmussen, M., & Willerslev,

E. (2012b). Detection of a diverse marine fish fauna using environmental DNA from

seawater samples. PLoS ONE, 7(8), e41732.

Thomsen, P. F., Kielgast, J. O. S., Iversen, L. L., Wiuf, C., Rasmussen, M., Gilbert, M.

T. P., Orlando, L., & Willerslev, E. (2012a). Monitoring endangered freshwater

biodiversity using environmental DNA. Molecular Ecology, 21(11), 2565-2573.

Thomsen, P. F., Møller, P. R., Sigsgaard, E. E., Knudsen, S. W., Jørgensen, O. A., &

223

Willerslev, E. (2016). Environmental DNA from seawater samples correlate with trawl

catches of subarctic, deepwater fishes. PLoS ONE, 11(11), e0165252.

Thomsen, P. F. & Willerslev, E. (2015). Environmental DNA–An emerging tool in

conservation for monitoring past and present biodiversity. Biological Conservation, 183,

4-18.

Tillotson, M. D., Kelly, R. P., Duda, J. J., Hoy, M., Kralj, J., & Quinn, T. P. (2018).

Concentrations of environmental DNA (eDNA) reflect spawning salmon abundance at

fine spatial and temporal scales. Biological Conservation, 220, 1-11.

Toju, H., Tanabe, A. S., Yamamoto, S., & Sato, H. (2012). High-coverage ITS primers

for the DNA-based identification of ascomycetes and basidiomycetes in environmental

samples. PLoS ONE, 7(7), e40863.

Torti, A., Lever, M. A., & Jørgensen, B. B. (2015). Origin, dynamics, and implications

of extracellular DNA pools in marine sediments. Marine Genomics, 24, 185-196.

Tréguier, A., Paillisson, J. M., Dejean, T., Valentini, A., Schlaepfer, M. A., & Roussel, J.

M. (2014). Environmental DNA surveillance for invertebrate species: advantages and

technical limitations to detect invasive crayfish Procambarus clarkii in freshwater

ponds. Journal of Applied Ecology, 51(4), 871-879.

Tsuchida, S. (2002). Experimental study on temperature preference of Japanese marine

fish. Report of Marine Ecology Research Institute, 4, 11-66. (in Japanese).

Tsuji, S., Miya, M., Ushio, M., Sato, H., Minamoto, T., & Yamanaka, H. (2020).

Evaluating intraspecific genetic diversity using environmental DNA and denoising

approach: A case study using tank water. Environmental DNA, 2(1), 42-52.

Tsuji, S., Ushio, M., Sakurai, S., Minamoto, T., & Yamanaka, H. (2017). Water

temperature-dependent degradation of environmental DNA and its relation to bacterial

abundance. PLoS ONE, 12(4), e0176608.

224

Tsuji, S., Yamanaka, H., & Minamoto, T. (2016). Effects of water pH and proteinase K

treatment on the yield of environmental DNA from water samples. Limnology, 18(1), 1-

7.

Tsuri, K., Ikeda, S., Hirohara, T., Shimada, Y., Minamoto, T., & Yamanaka, H. (2020).

Messenger RNA typing of environmental RNA (eRNA): A case study on zebrafish tank

water with perspectives for the future development of eRNA analysis on aquatic

vertebrates. Environmental DNA, in press. DOI: 10.1002/edn3.169

Turner, C. R., Barnes, M. A., Xu, C. C., Jones, S. E., Jerde, C. L., & Lodge, D. M.

(2014). Particle size distribution and optimal capture of aqueous macrobial eDNA.

Methods in Ecology and Evolution, 5(7), 676-684.

Turner, C. R., Uy, K. L., & Everhart, R. C. (2015). Fish environmental DNA is more

concentrated in aquatic sediments than surface water. Biological Conservation, 183, 93-

102.

Uchii, K., Doi, H., & Minamoto, T. (2016). A novel environmental DNA approach to

quantify the cryptic invasion of non-native genotypes. Molecular Ecology Resources,

16(2), 415-422.

Uchii, K., Telschow, A., Minamoto, T., Yamanaka, H., Honjo, M. N., Matsui, K., &

Kawabata, Z. (2011). Transmission dynamics of an emerging infectious disease in

wildlife through host reproductive cycles. The ISME Journal, 5(2), 244-251.

Ushio, M., Fukuda, H., Inoue, T., Makoto, K., Kishida, O., Sato, K., Murata, K.,

Nikaido, M., Sado, T., Sato, Y., Takeshita, M., Iwasaki, W., Yamanaka, H., Kondoh, M.,

& Miya, M. (2017). Environmental DNA enables detection of terrestrial mammals from

forest pond water. Molecular Ecology Resources, 17(6), e63-e75.

Ushio, M., Murata, K., Sado, T., Nishiumi, I., Takeshita, M., Iwasaki, W., & Miya, M.

(2018). Demonstration of the potential of environmental DNA as a tool for the detection

225

of avian species. Scientific Reports, 8, 4493.

Van Hoof, A. & Parker, R. (2002). Messenger RNA degradation: beginning at the end.

Current Biology, 12(8), R285-R287.

Vennemann, M. & Koppelkamm, A. (2010). mRNA profiling in forensic genetics I:

Possibilities and limitations. Forensic Science International, 203(1-3), 71-75.

Warne, R. T. (2014). A Primer on Multivariate Analysis of Variance (MANOVA) for

Behavioral Scientists. Practical Assessment, Research & Evaluation, 19(17), 1-10.

Wei, N., Nakajima, F., & Tobino, T. (2018). A microcosm study of surface sediment

environmental DNA: decay observation, abundance estimation, and fragment length

comparison. Environmental Science & Technology, 52(21), 12428-12435.

Weltz, K., Lyle, J. M., Ovenden, J., Morgan, J. A., Moreno, D. A., & Semmens, J. M.

(2017). Application of environmental DNA to detect an endangered marine skate

species in the wild. PLoS ONE, 12(6), e0178124.

Wilcox, T. M., McKelvey, K. S., Young, M. K., Lowe, W. H., & Schwartz, M. K.

(2015). Environmental DNA particle size distribution from Brook Trout (Salvelinus

fontinalis). Conservation Genetics Resources, 7(3), 639-641.

Wilcox, T. M., Zarn, K. E., Piggott, M. P., Young, M. K., McKelvey, K. S., & Schwartz,

M. K. (2018). Capture enrichment of aquatic environmental DNA: A first proof of

concept. Molecular Ecology Resources, 18(6), 1392-1401.

Willerslev, E., Hansen, A. J., Binladen, J., Brand, T. B., Gilbert, M. T. P., Shapiro, B.,

Bunce, M., Wiuf, C., Gilichinsky, D. A., & Cooper, A. (2003). Diverse plant and animal

genetic records from Holocene and Pleistocene sediments. Science, 300(5620), 791-

795.

Williams, M. A., O'Grady, J., Ball, B., Carlsson, J., de Eyto, E., McGinnity, P., Jennings,

226

E., Regan, F., & Parle-McDermott, A. (2019). The application of CRISPR-Cas for

single species identification from environmental DNA. Molecular Ecology Resources,

19(5), 1106-1114.

Wilson, C., Wright, E., Bronnenhuber, J., MacDonald, F., Belore, M., & Locke, B.

(2014). Tracking ghosts: combined electrofishing and environmental DNA surveillance

efforts for Asian carps in Ontario waters of Lake Erie. Management of Biological

Invasions, 5(3), 225.

Wood, S. A., Biessy, L., Latchford, J. L., Zaiko, A., von Ammon, U., Audrezet, F.,

Cristescu, M. E., & Pochon, X. (2020). Release and degradation of environmental DNA

and RNA in a marine system. Science of the Total Environment, 704, 135314.

Worm, B., Barbier, E. B., Beaumont, N., Duffy, J. E., Folke, C., Halpern, B. S., Jackson,

J. B. C., Lotze, H. K., Micheli, F., Palumbi, S. R., Sala, E., Selkoe, K. A., Stachowicz, J.

J., & Watson, R. (2006). Impacts of biodiversity loss on ocean ecosystem services.

Science, 314, 787-790.

Wotton, R. S. & Malmqvist, B. (2001). Feces in aquatic ecosystems: feeding animals

transform organic matter into fecal pellets, which sink or are transported horizontally by

currents; these fluxes relocate organic matter in aquatic ecosystems. BioScience, 51(7),

537-544.

Wrigglesworth, J. M., Packer, L., & Branton, D. (1970). Organization of mitochondrial

structure as revealed by freeze-etching. Biochimica Et Biophysica Acta (BBA) - Bioenergetics, 205(2), 125-135.

Wu, Q., Kawano, K., Uehara, Y., Okuda, N., Hongo, M., Tsuji, S., Yamanaka, H., &

Minamoto, T. (2018). Environmental DNA reveals nonmigratory individuals of

Palaemon paucidens overwintering in Lake Biwa shallow waters. Freshwater Science,

37(2), 307-314.

Yamamoto, S., Masuda, R., Sato, Y., Sado, T., Araki, H., Kondoh, M., Minamoto, T., &

227

Miya, M. (2017). Environmental DNA metabarcoding reveals local fish communities in

a species-rich coastal sea. Scientific Reports, 7, 40368.

Yamamoto, S., Minami, K., Fukaya, K., Takahashi, K., Sawada, H., Murakami, H.,

Tsuji, S., Hashizume, H., Kubonaga, S., Horiuchi, T., Hongo, M., Nishida, J., Okugawa,

Y., Fujiwara, A., Fukuda, M., Hidaka, S., Suzuki, K. W., Miya, M., Araki, H.,

Yamanaka, H., Maruyama, A., Miyashita, K., Masuda, R., Minamoto, T., & Kondoh, M.

(2016). Environmental DNA as a ‘snapshot’of fish distribution: A case study of

Japanese jack mackerel in Maizuru Bay, Sea of Japan. PLoS ONE, 11(3), e0149786.

Yamanaka, H. & Minamoto, T. (2016). The use of environmental DNA of fishes as an

efficient method of determining habitat connectivity. Ecological Indicators, 62, 147-

153.

Yamanaka, H., Minamoto, T., Matsuura, J., Sakurai, S., Tsuji, S., Motozawa, H., Hongo,

M., Sogo, Y., Kakimi, N., Teramura, I., Sugita, M., Baba, M., & Kondo, A. (2017). A

simple method for preserving environmental DNA in water samples at ambient

temperature by addition of cationic surfactant. Limnology, 18(2), 233-241.

Yates, M. C., Fraser, D. J., & Derry, A. M. (2019). Meta-analysis supports further

refinement of eDNA for monitoring aquatic species-specific abundance in nature.

Environmental DNA, 1(1), 5-13.

Yates, M. C., Glaser, D., Post, J., Cristescu, M. E., Fraser, D. J., & Derry, A. M. (2020a).

The relationship between eDNA particle concentration and organism abundance in

nature is strengthened by allometric scaling. Molecular Ecology, in press. DOI:

10.1111/mec.15543

Yates, M. C., Wilcox, T. M., McKelvey, K. S., Young, M. K., Schwartz, M. K., & Derry,

A. M. (2020b). Allometric scaling of eDNA production in stream-dwelling brook trout

(Salvelinus fontinalis) inferred from population size structure. Environmental DNA, in

press. DOI: 10.1002/edn3.150

228

You, Y., Rankin, S. C., Aceto, H. W., Benson, C. E., Toth, J. D., & Dou, Z. (2006).

Survival of Salmonella enterica serovar Newport in manure and manure-amended soils.

Applied and Environmental Microbiology, 72(9), 5777-5783.

Zhang, C. I. & Lee, J. B. (2001). Stock assessment and management implications of

horse mackerel (Trachurus japonicus) in Korean waters, based on the relationships

between recruitment and the ocean environment. Progress in Oceanography, 49(1-4),

513-537.

Zobel, M., Davison, J., Edwards, M. E., Brochmann, C., Coissac, E., Taberlet, P.,

Willerslev, E., & Moora, M. (2018). Ancient environmental DNA reveals shifts in

dominant mutualisms during the late Quaternary. Nature Communications, 9, 139.

229

Appendix

Appendix S1. Detailed information on a literature survey for macrobial eDNA studies.

I conducted a Google Scholar search (https://scholar.google.co.jp/) of literatures relating

to eDNA characteristics and dynamics during 2008 to 2019. The literature search

included the terms “eDNA” or “environmental DNA” in the title and/or text. I filtered

the papers with the following criteria: the papers which (i) targeted eDNA from macro-

organisms (not only from microorganism, fungi, plankton, virus, and bacteria), but did

not target ancient DNA (aDNA) in sediment or ice core samples; (ii) was published in

the international journals, (iii) was peer-reviewed (not preprint server or the paper

before inclusion in an issue), and (iv) was not review papers, news, views,

introductions, opinions, responses, and perspectives.

From the remaining 535 papers, by carefully reading them, I then selected the

papers whose main objectives corresponded the keywords as follows: (a) production:

the study focusing on physiological and ecological sources of eDNA, and biotic/abiotic

factors affecting eDNA production, (b) state: the study focusing on physiochemical and

molecular states of eDNA, (c) transport: the study focusing on vertical/horizontal

movement of eDNA such as transport, diffusion, and retention, and environmental

factors affecting eDNA movement, and (d) persistence: the study focusing on eDNA

persistence and degradation, and environmental factors affecting them (some papers can

be assigned to multiple keywords). As a result, 78, 16, 31, and 54 papers were assigned

to keywords ‘production, ‘state’, ‘transport’, and ‘persistence’.

230

Acknowledgements

First and foremost, I would like to thank my chief-supervisor, Prof. Toshifumi

Minamoto, for giving me this PhD opportunity. It is my pleasure that he has invested

substantial time and effort in my thesis. I am grateful so much for his constructive

assistance and comments on my research plan, laboratory work, and manuscript writing.

Besides, I want to thank the remaining members of my supervisory panel, Profs. Atushi

Ushimaru, Nobuko Ohmido, Yasuoki Takami, and Reiji Masuda. All of them provided

me with helpful comments and suggestions on my research plan, interpretation of the

results, and presentation in the meetings.

I am also grateful for Dr. Hiroaki Murakami in Maizuru Fisheries Research Station,

Kyoto University. Without him and Prof. Masuda, I could not have completed almost all

of my tank experiments in the thesis. They kindly assisted with my tank experiments

including preparation of fishes, set-up of experimental tanks, and water sampling, as

well as provided helpful comments on my research plan and interpretation of the results.

I thank Dr. Satoshi Yamamoto in Kyoto University, who had been postdoctoral fellow in

Prof. Minamoto’s lab. He also assisted with my tank experiment and provided helpful

comments on research plan and manuscript writing in Chapter 2 and 5. I thank Ms. Mio

Arimoto, who took charge of a part of qPCR experiments in Chapters 3 and 4, and Mr.

Masayuki K. Sakata, who is the most intimate contemporary for me and assisted my

tank experiment in Chapters 2 and 5, in Kobe University.

Moreover, I want to thank everyone that has contributed to my thesis. Thanks to other

members of Prof. Minamoto’s lab, Drs. Qianqian Wu, Ryohei Nakao, and Masayoshi

Hiraiwa, and Shunsuke Hidaka, Ayaka Fujiwara, Sei Tomita, Mone Kawata, and Saki

Ikeda for supporting my tank experiments. Thanks to other members of Maizuru

Fisheries Research Station, Takaya Yoden, Mizuki Ogata, Sachia Sasano, and Misaki

Shiomi, for supporting my tank experiments and treating me with fresh and delicious

seafood, which is my vitality in performing the experiment in MFRS. Thanks to all the

past and present members of Prof. Minamoto’s lab, Mushikusa seminar, and Ecological

Joint Seminar in Kobe University for providing great atmosphere and helpful comments

on my research.

231

This work was supported by JST CREST (Grant Number JPMJCR13A2), JSPS

KAKENHI (Grant Number JP19H03031), and Grant-in-Aid for JSPS Research Fellow

(Grant Number JP18J20979). Without these financial contributions, the research in the

thesis would not have been possible.

Lastly, my special thanks are due to my family, Eiraku, Kazumi, and Kyoka, and

partner, Akiho. They have always been supportive of me.

Kobe University Repository : Thesis

Documents