© Noémie Leduc, 2019
Évaluation de la biodiversité des invertébrés marins dans les ports commerciaux de l'Arctique grâce à l'ADN
environnemental
Mémoire
Noémie Leduc
Maîtrise en biologie - avec mémoire
Maître ès sciences (M. Sc.)
Québec, Canada
Évaluation de la biodiversité des invertébrés
marins dans les ports commerciaux de l’Arctique
grâce à l’ADN environnemental
Mémoire
Noémie Leduc
Sous la direction de :
Louis Bernatchez, directeur de recherche
Philippe Archambault, codirecteur de recherche
ii
Résumé
D’abord méconnue puis longuement sous-estimée, la biodiversité de l’Arctique fait
maintenant face à d’importantes altérations sous les effets combinés des changements
climatiques ainsi que l’augmentation des activités commerciales dans l’Arctique canadien.
Ces altérations ne sont pas sans risque pour les communautés d’invertébrés marins,
particulièrement dans les zones sensibles telles que les ports commerciaux. La protection de
la biodiversité représente un enjeu majeur, nécessitant une bonne compréhension de
l’organisation spatiale des espèces au moyen d’indices de biodiversité tels que les indices
alpha, beta et gamma, de même que le développement de méthodes de détection efficace.
Dans le cadre de ce projet de maîtrise, la biodiversité obtenue à l’aide de méthodes
d’échantillonnage traditionnelle fut comparée à la biodiversité détectée par l’ADN
environnementale (ADNe), grâce au metabarcoding des gènes COI et 18S, afin de
documenter les patrons de biodiversité des communautés d’invertébrés marins à différentes
échelles spatiales. À partir d’échantillons d’eau de 250 ml récoltés à trois différentes
profondeurs au sein des ports de Churchill, Baie Déception et Iqaluit, il fut possible de déceler
la présence de 202 genres répartis dans plus de 15 phyla. De ces organismes, seulement 9 à
15% furent également collectés par les méthodes traditionnelles, révélant ainsi l’existence de
différences significatives au niveau de la richesse et de la composition des communautés
entre ces différentes approches d’échantillonnage. Outre ces différences majeures, cette étude
a permis de démontrer une réduction de la biodiversité beta dans les communautés détecter
à l’aide de l’ADNe comparativement aux communautés identifiées par la collecte de
spécimens. Cette homogénéisation de la biodiversité souligne le rôle non négligeable de la
dispersion de l’ADNe ainsi que l’influence notoire des stades de vie pélagique dans sa
détection. Les résultats obtenus dans le cadre de cette étude mettent bien en évidence le
potentiel du metabarcoding d’ADNe tout en insistant sur son caractère complémentaire face
aux méthodes traditionnelles pour d’éventuelles applications en gestion et conservation des
communautés d’invertébrés marins de l’Arctique.
iii
Abstract
Arctic biodiversity has long been underestimated and is now facing rapid transformations
due to ongoing climate change and other impacts including shipping activities. These changes
are placing marine coastal invertebrate communities at greater risk, especially in sensitive
areas such as commercial ports. Preserving biodiversity is a significant challenge, going far
beyond the protection of charismatic species and involving suitable knowledge of the
organization of species in space. Therefore, knowledge of alpha, beta and gamma
biodiversity indices are of great importance in achieving this objective together with new
cost-effective approaches to monitor changes in biodiversity. This study compares
metabarcoding of COI mitochondrial genes and 18S rRNA genes from environmental DNA
(eDNA) water samples with standard species collection methods to document patterns of
invertebrate communities at various spatial scales. Water samples (250 mL) were collected
at three different depths within three Canadian Arctic ports; Churchill, MB, Iqaluit, NU and
Deception Bay, QC. From these samples, 202 genera distributed across more than 15 phyla
were detected using eDNA metabarcoding, of which only 9% to 15% were also identified
through species collection at the same sites. Significant differences in taxonomic richness
and community composition were observed between eDNA and species collections, both on
local and regional scales. This study shows that eDNA dispersion in the Arctic Ocean reduces
beta diversity in comparison to species collection while emphasizing the importance of
pelagic life stages for eDNA detection. This study highlights the potential of eDNA
metabarcoding to assess large-scale arctic marine invertebrate diversity while emphasizing
that eDNA and species collection should be considered as complementary tools for providing
a more holistic picture of the marine invertebrate communities living in coastal areas.
iv
Table des matières
Résumé ............................................................................................................................................... ii
Abstract ............................................................................................................................................. iii
Table des matières ............................................................................................................................ iv
Liste des tableaux ............................................................................................................................. vi
Liste des figures ............................................................................................................................... vii
Remerciements ................................................................................................................................. ix
Avant-propos .................................................................................................................................... xi
Introduction ....................................................................................................................................... 1
Problématique générale ................................................................................................................... 1
L’Arctique dans un monde en changement ..................................................................................... 2
La biodiversité comme unité de gestion pour la conservation ........................................................ 3
Indices de biodiversité ................................................................................................................. 3
Méthodes pour recenser la biodiversité ...................................................................................... 5
L’utilisation d’outils moléculaires .................................................................................................. 5
Métabarcoding ............................................................................................................................ 5
ADN environnementale (ADNe) .................................................................................................. 6
Les invertébrés marins .................................................................................................................... 8
Contexte du projet ........................................................................................................................... 9
Objectifs ........................................................................................................................................ 10
Chapter I: Comparing eDNA metabarcoding and species collection for documenting Arctic
metazoan biodiversity ..................................................................................................................... 11
Résumé .......................................................................................................................................... 12
Abstract ......................................................................................................................................... 13
Introduction ................................................................................................................................... 14
Methods ......................................................................................................................................... 17
Sample collection ...................................................................................................................... 17
Species collection .................................................................................................................. 18
Environmental DNA samples ................................................................................................ 19
Metabarcoding .......................................................................................................................... 20
Environmental DNA extraction, amplification and sequencing ............................................ 20
Bioinformatics ........................................................................................................................... 21
Data analysis ............................................................................................................................. 22
Results ........................................................................................................................................... 23
Sequencing quality .................................................................................................................... 23
Arctic coastal gamma diversity ................................................................................................. 25
Alpha biodiversity ..................................................................................................................... 29
Beta diversity ............................................................................................................................. 32
Origin of coastal eDNA ............................................................................................................. 33
Discussion ..................................................................................................................................... 34
v
Overall biodiversity and community structure .......................................................................... 35
Transport and homogenization of eDNA................................................................................... 38
Origins of eDNA ........................................................................................................................ 40
Role of eDNA in Arctic conservation ........................................................................................ 41
Acknowledgement ......................................................................................................................... 43
Annexe (Supporting information) ................................................................................................. 44
Conclusion ........................................................................................................................................ 55
ADNe vs. collecte d’espèces ......................................................................................................... 56
Origine, détection et transport de l’ADNe .................................................................................... 57
Faiblesses et limites du métabarcoding d’ADN environnemental ................................................ 58
Applications et perspectives futures pour la protection de la biodiversité .................................... 59
Bibliographie ................................................................................................................................... 61
vi
Liste des tableaux
Table 1. Summary of the numbers of reads, the proportion of species and genera present in
the Arctic historic database and the mean number of OTUs for the COI primers set and the
18S primers set assigned and non assigned on BOLD and SILVA for each port.
Table 2. Summary of richness, alpha and beta biodiversity indices for the eDNA and species
collection of marine invertebrates communities on abundance data after Hellinger
transformation (Shannon and Pielou indices) and presence/absence transformation (Beta
index). COI primer sets and 18S primer sets are added together.
Table S1. Barque 1.5.1 specific commands for preparation and analysis of pair-end reads.
The sequences from the different COI and 18S primers set were added together.
Table S2. Chordata taxa present in the eDNA data set (COI and 18S primers set added
together) and the appropriate action taken.
Table S3. Genera found in the negative field controls and the appropriate action taken
against the contamination according to COI and 18S primers set.
Table S4. Summary of PERMANOVA statistics tests on marine invertebrates communities
for the phylum relative abundance (number of taxa), Pielou evenness index and alpha
richness. The analyses were performed with method = "bray" for phylum relative abundance
while it was performed with method = "euclidian" for Pielou evenness and alpha richness.
Table S5. Summary of the correlation between dissimilarity and distance across the sites
within Churchill, Deception Bay and Iqaluit ports based on incidence data for the different
sampling methods.
Table S6. Summary of the main phyla identified by sampling collection sampling methods
among Churchill, Deception Bay and Iqaluit ports and their respective presence in BOLD
and SILVA public genetic databases.
vii
Liste des figures
Figure 1. Geographical location of Churchill, Deception Bay and Iqaluit ports in the
Canadian Arctic (map A) and distribution of stations within Churchill (map B), Deception
Bay (map C) and Iqaluit (map D).
Figure 2. Individual-based rarefaction curves of eDNA, benthos and zooplankton genera for
Churchill (blue), Deception Bay (yellow) and Iqaluit ports (magenta) based on incidence
data.
Figure 3. a) Barplots of the number of taxa found in Churchill (blue), Deception Bay
(yellow) and Iqaluit (red). Darker bands represent species collection methods while paler
bands with dashed outline represent eDNA and black bands represent the number of genera
in common between eDNA and species collection. b) Phylum relative proportion of commun
genera between eDNA and species collection based on incidence data. COI and 18S primer
sets are added together for both a) and b).
Figure 4. Marine invertebrate taxonomic composition at the phylum level for eDNA and
species collection methods, respectively, within Churchill, Deception Bay and Iqaluit ports,
based on incidence data. The COI and 18S regions are added together for the eDNA barplot
while the benthic trawl, core, grab and net tows samples are added together for the species
collection barplot.
Figure 5. Biodiversity differences A) among ports based on eDNA only and B) among
sampling methods within ports. Ordination of taxonomic composition (genera) calculated
using Sorensen index (incidence based) with each data point representing a specific sample;
blue squares represent Churchill, yellow dots represent Deception Bay and magenta triangles
represent Iqaluit. Filled symbols are associated with eDNA while empty symbols are
associated with species collections.
Figure 6. Boxplot on alpha diversity for the genera richness and Pielou evenness index in
Churchill, Deception Bay and Iqaluit ports for eDNA (A, C) and species collection (B, D).
These analyses were performed on abundance data with Hellinger transformation, COI and
18S primer sets are added together for the eDNA boxplot.
Figure 7. Sorensen dissimilarity index between pairs of stations as a function of distance
between the stations based on incidence data (presence/absence transformation on
abundance) for different sampling methods (eDNA and species collections of benthos and
zooplankton) in Churchill (blue), Deception Bay (yellow) and Iqaluit (magenta).
Figure 8. Relative abundance of organisms obtained with eDNA and species collection
within Churchill, Deception Bay and Iqaluit ports according to their life history; taxa with a
meroplanktonic life stage, holoplanktonic taxa life stage. Species collection for benthos
include benthic trawl, Van Veen grab, cores and for zooplankton includes pelagic plankton
viii
net tows. The sum of the detection of each taxa (i.e. presence/absence) have been combined
for all primer sets.
Figure S1. The number of Operational taxonomic units (OTUs) assigned and not assigned
on NCBI to the genus level for the COI primers set (COI1 and COI2) and the 18S primers
set (Tareuk and 18S). The assigned OTUs are represented by the grey section of the barplot
while the not assigned OTUs are represented by the red section of the barplot. The
pourcentage written in red represented the mean of the 4 primers for not assigned OTUs in
each port.
Figure S2. Boxplot on alpha diversity for Shannon biodiversity index in Churchill,
Deception Bay and Iqaluit ports. These analyses were performed on abundance data with
Hellinger transformation, COI and 18S primer sets are added together for the eDNA boxplot.
ix
Remerciements
Je voudrais avant tout remercier mon directeur de recherche, le Dr Louis Bernatchez, que ce
soit des plongées en apnée dans les îles Galápagos, aux salles de cours puis finalement
comme directeur de recherche, ta passion pour la science et la génétique est contagieuse et
ne cessera de m’impressionner. Merci de m’avoir accueillie dans ton laboratoire de même
que la grande famille qui le compose ainsi que pour ta confiance en moi depuis mes tout
début dans le domaine de la biologie moléculaire. Ces trois dernières années furent
extrêmement riches en expériences, tant sur le plan personnel que professionnel en grande
partie grâce aux opportunités que tu m’as offertes dans le cadre de ma maîtrise.
J’aimerais également remercier mon codirecteur de recherche, le Dr Philippe Archambault,
pour sa patience, son expertise incroyable ainsi que son émerveillement devant chacun de
mes petits résultats. Merci Phil pour ton écoute et tes précieux conseils, qui ont su me
remonter le moral dans les moments de doutes.
La réussite de ce projet repose également sur la participation de plusieurs personnes. Je tiens
à remercier Anaïs Lacoursière-Roussel pour m’avoir guidée dès mes premiers pas en tant
qu’étudiante à la maîtrise et pour avoir investi beaucoup de temps et d’énergie dans
l’ensemble du projet, de son élaboration jusqu’à la collecte des précieux échantillons et
finalement son aboutissement avec le mémoire ci-présent. Je voudrais également remercier
Kimberly Howland, sans qui ce projet n’aurait pas pu voir le jour tel qu’il est aujourd’hui de
même que Maelle Sevellec, qui malgré le décalage horaire a toujours été présente avec le
sourire à nos rencontres Skype et a su m’encourager et me fournir de précieux conseils pour
la conception de ce manuscript.
Je voudrais ensuite remercier les membres du laboratoire Bernatchez pour leur agréable
compagnie ainsi que leur complicité au cours de ces dernières années. Merci plus
particulièrement aux professionnels de recherche Alysse Perreault-Payette, Cécilia
Hernandez et Bérénice Bougas pour avoir partagé avec moi leurs connaissances, leur
x
expertise en laboratoire de même que pour leurs nombreux conseils. Merci à notre
bioinformaticien Éric Normandeau pour son souci du détail et sa disponibilité pour répondre
aux nombreuses embuches qu’accompagne souvent la bioinformatique. Merci à Damien
Boivin-Delisle, mon "eDNA partner", pour nos discussions diverses mais toujours
enrichissantes de même que nos divagations par certains moments. Finalement merci à
Valérie Cypihot du laboratoire Archambault, pour son amitié, son aide sur le terrain, sa
compagnie lors de nos congrès ainsi que ses conseils.
Pour finir, un grand merci aux membres de ma famille et de ma belle-famille pour leur
encouragement, et par-dessus tout mes parents pour leur soutien et leur confiance en moi
depuis le début de mes études. Je remercie mes amies de la boite de céréales, plus
spécialement Justine Létourneau pour son aide et sa compagnie de grande qualité au sein du
laboratoire. Finalement je remercie tout spécialement mon copain, Gabriel, pour son support
inconditionnel depuis le début de ma maîtrise, son sens de l’humour et son optimisme au
quotidien.
xi
Avant-propos
Ce mémoire est principalement composé d’un article intitulé « Comparing eDNA
metabarcoding and species collection for documenting Arctic metazoan biodiversity ». Cet
article a été soumis à la revue « Environmental DNA » et est actuellement en processus de
révision.
Les co-auteurs sont mesdames Kimberly Howland, Maelle Sevellec, Anaïs Lacoursière-
Roussel, monsieur Éric Normandeau, mon co-directeur Philippe Archambault et mon
directeur Louis Bernatchez.
Anaïs Lacoursière-Roussel, Kimberly Howland, ont élaboré l’étude globale dans lequel
s’inscrit mon projet de maîtrise, organisé les campagnes d’échantillonnage de même que
participé à la collecte d’échantillons sur le terrain. Éric Normandeau a conçu le pipeline
bioinformatique ayant servi aux traitements des données d’ADNe suite au séquençage.
Maelle Sevellec a contribuer aux analyses statistiques de même qu’à la rédaction de l’article
dans son ensemble. Pour ma part, j’ai réalisé le travail sur le terrain et au laboratoire, rédigé
l’article (auteure principale) et effectué l’analyse et l’interprétation des résultats avec l’aide
de mes co-auteurs. Tous les co-auteurs ont également collaboré à la révision du manuscript.
Ce projet a été financé par le réseau de centres d’excellence ArcticNet, le programme fédéral
Savoir polaire Canada (POLAIRE) ainsi que le regroupement Québec-Océan.
1
Introduction
Problématique générale
Depuis le Sommet de la Terre à Rio de Janeiro en 1992, l’intérêt porté à la biodiversité de
même qu’aux conséquences possibles que celle-ci peut engendrer en cas de perte, n’a cessé
de croître et de stimuler la recherche scientifique (Cardinale et al. 2012). Malgré cet intérêt
notoire pour la biodiversité, l’anthropocène plonge la planète dans une importante extinction
de masse, où la surexploitation, la destruction des habitats et les changements climatiques
entraînent chaque jour la disparition de plusieurs espèces (Young et al. 2016). Le maintien
de la biodiversité est nécessaire afin de soutenir la stabilité des processus écosystémiques
dans des environnement en perpétuels changement (Loreau et de Mazancourt 2013) de ce
fait, son déclin actuel représente une crise majeure ainsi qu’un défi considérable pour le 21ème
siècle (Vié et al. 2009).
Parmi les habitats les plus affectés par cette crise on retrouve l’océan Arctique (Jørgensen et
al. 2016). Ce dernier étant le plus petit océan de la planète mais également le moins connu
de tous en raison de son emplacement géographique, ses conditions météorologiques hostiles
ainsi que sa couverture de glace récurrente (National Research Council 1995). Considéré
jusqu’à récemment comme l’un des endroits les plus intact du globe (UNESCO 2010),
l’océan Arctique possède une biodiversité bien supérieure à ce qui était estimé en raison de
connaissances rudimentaires à son sujet (Darnis et al. 2012). Face aux nombreux enjeux
existants, un besoin urgent est né afin de pallier au manque d’informations relatif à la
diversité de même qu’à la distribution des espèces présente dans l’Arctique et ce dans l’espoir
de pouvoir minimiser les impacts en grande partie imposés par l’humain sur la biodiversité
marine de cet océan (Bluhm et al. 2011).
2
L’Arctique dans un monde en changement
Les changements climatiques actuels causent d’importantes altérations aux patrons de
biodiversité à l’échelle de la planète par l’extinction de nombreuses espèces. Globalement,
près de 8% des extinctions d’espèces dans le futur découleraient directement de ces
changements climatiques (Urban 2015). Les habitats marins ne sont pas épargnés par cette
crise, l’augmentation de la température de l’eau, l’acidification des océans, ainsi que les
changements physicochimiques reliés à ces phénomènes causent la dégradation des habitats
(Hoegh-Guldberg et Bruno 2010). De ces changements climatiques résulte également une
importante fonte des glaces qui façonne de nouveaux paysages (Kintisch 2015) et entraine
par le fait même une augmentation considérable des activités humaines dans certains secteurs
maritimes (Ruffilli 2011) .
Lors de la saison estivale, alors que la couverture de glace est à son minimum, c’est 75% de
son volume qui est perdu depuis 1979 (Kintisch 2015). Cette perte importante de glace dans
l’océan Arctique permet l’ouverture de nouveaux chemins pour la navigation des bateaux
commerciaux (Vermeij et Roopnarine 2008) en plus d’augmenter la durée où la navigation
est praticable à certains endroits. Dans le cadre de sa Stratégie pour le Nord, le gouvernement
du Canada compte promouvoir le développement social et économique des communautés
nordiques, ce qui combiné à la popularité grandissante de l’Arctique pour le tourisme et
l’industrie des croisières augmentera considérablement la circulation maritime au cours des
prochaines décennies (Gavrilchuk et al. 2013).
Les effets associés du réchauffement climatique ainsi que de l’augmentation des activités
anthropiques auront des répercussions directes sur la biodiversité ainsi que la distribution des
communautés marines. Ces conséquences seront d’autant plus considérables pour les régions
côtières, qui sont sujettes à de plus grandes pertes de diversité en raison des utilisations
conflictuelles des habitats qui les composent (Gray 1997). De plus, l’augmentation du trafic
maritime devrait accentuer les risques d’introduction d’espèces exotiques (Goldsmit et al.
2017 et 2019). Les principaux vecteurs d’introduction d’espèces exotiques étant bien souvent
l’eau de lest ainsi que les salissures biologiques des coques des bateaux (Piola et al. 2009)
3
qui transportent et déchargent une quantité non négligeable d’organismes à des distances qui
dépassent largement leur capacité naturelle de dispersion (Casas-Monroy et al. 2014), une
augmentation significative du risque d’introduction d’espèces exotiques est attendu dans
l’Arctique canadien au cours des prochaines années (Mooney et Cleland 2001; Casas-
Monroy et al. 2014). Dans certains cas, ces espèces exotiques peuvent devenir des espèces
envahissantes, ayant la capacité d’engendrer de profonds changements écologiques (Carlton
1993) et causer des pertes majeures de biodiversité (Williamson 1999).
Dans un tel contexte, où la biodiversité de même que la composition des communautés
marines sont appelés à connaître d’importantes altérations liées à divers facteurs, des
méthodes permettant une détection rapide et efficiente des changements de biodiversité sont
plus que nécessaires.
La biodiversité comme unité de gestion pour la conservation
Indices de biodiversité
La biodiversité, concept ayant vu le jour dans les années 1980 (Pimm 2001), décrit la variété
structurelle et fonctionnelle de l’ensemble des formes de vie tant au niveau de la génétique,
des populations que des écosystèmes (Sandlund et al. 1992). Depuis quelques années, la
conservation de la biodiversité marine devient progressivement un objectif capital en gestion
environnementale (Spalding et al. 2007), nécessitant ainsi suffisamment d’informations et de
bases de données sur les organismes concernés (Laurila-Pant et al. 2015). Les indices de
biodiversité constituent une approche écologique classique afin d’évaluer la biodiversité,
ceux-ci prenant en considération deux aspects majeurs des communautés soit la richesse
spécifique ainsi que l’abondance relative des individus au sein de chaque espèce (Hamilton
2005, Laurila-Pant et al. 2015). Plusieurs indices existent afin de pouvoir quantifier le rythme
auquel la biodiversité varie dans le temps et l’espace (Gaston 2000), que ce soit au niveau de
l’espèce ou de la communauté et selon différentes échelles spatiales et/ou organisationnelles
à travers les valeurs descriptives alpha, beta et gamma.
4
La biodiversité alpha représente l’assemblage d’une communauté à basse échelle et constitue
la plus petite mesure de la biodiversité (MacArthur 1965), elle reflète habituellement la
richesse spécifique d’un lieu. C’est également l’indice de biodiversité le plus utilisé et revêt
d’une importance cruciale pour la conservation (Socolar et al. 2015). Comme cet indice est
fortement influencé par la taille des échantillons ainsi que par l’envergure des lieux
échantillonnés et peut être biaisée par ces paramètres il est important d’utiliser des techniques
statistiques qui tiennent comptent de ces derniers, telle que la standardisation des données
(Roff et Mark 2011).
La diversité beta, aussi appelée « turnover diversity » fait référence au degré de changement
dans la composition des espèces le long d’un certain gradient (Gray 2000). Cet indice peut
être interprété comme le taux de variation dans la composition des espèces par rapport à
l’environnement ou au sein de communautés (Roff et Mark 2011; Van Dyke 2010), en
fonction de processus menant à l’homogénéisation ou l’hétérogénéisation des communautés.
De ce fait, une bonne compréhension de la diversité beta est essentielle afin de protéger la
biodiversité régionale ainsi que dans l’optique de contribuer directement aux plans de
conservation (Socolar et al. 2015). La diversité beta est une mesure d’autant plus importante
pour les communautés sujettes à d’importantes perturbations (Mori et al. 2018).
Finalement, la diversité gamma constitue quant à elle la diversité totale d’une région donnée,
représentant ainsi la richesse spécifique de chacun des différents habitats au sein de celle-ci.
Puisque l’indice de biodiversité gamma désigne le regroupement d’indice de biodiversité
alpha pour une région, celle-ci est exprimé dans les mêmes unités que cette dernière (Laurila-
Pant et al. 2015).
5
Méthodes pour recenser la biodiversité
Actuellement, et depuis plusieurs décennies, la majorité des suivis de biodiversité en milieu
aquatique sont basés sur l’identification morphologique des organismes, que ce soit au moyen
de sondages visuels ou du dénombrement des individus directement sur le terrain. Ces
méthodes requièrent toutefois des notions de taxonomie avancées, dans un contexte où les
spécialistes pratiquant cette discipline se font de plus en plus rares (Archambault et al., 2010),
créant ainsi une demande pour des approches d’identifications alternatives (Radulovici et al.
2010 ; Thomsen et Willerslev 2015). De plus, certaines techniques d’inventaires
traditionnelles comme le chalutage de fond peuvent se révéler invasives et dommageables
pour les espèces et les écosystèmes (Baldwin et al. 1996; Jones 1992; Robertson et Smith-
Vaniz 2008). À cela s’ajoute la logistique souvent complexe associée aux instruments
d’échantillonnage traditionnel (chalut, benne Van Veen, filets) dans des régions éloignées
et/ou difficiles d’accès tel que l’Arctique (Jorgensen et al. 2016). Le développement de
nouveaux outils non-invasifs permettrait de remédier à certain de ces désavantages et
améliorerait potentiellement la capacité d’échantillonner la biodiversité marine à large
échelle tout en ayant un minimum d’impacts négatifs sur celle-ci.
L’utilisation d’outils moléculaires
Métabarcoding
L’identification simultanée de plusieurs taxons par code-barres ADN (métabarcoding) se
veut être une méthode fort attrayante afin de décrire la composition de communautés
biologiques complexes ainsi que pour en estimer la biodiversité (Brown et al. 2015). Cette
technique est utilisée depuis plusieurs décennies chez les microbiologistes (Coissac et al.
2012) et exploite la diversité génétique présente au sein de séquences particulières d’ADN
chez les organismes afin de les distinguer les uns des autres et de les identifier. Ces séquences
représentent des identifiants génétiques se trouvant à l’intérieur de chaque cellule (Hebert et
al. 2003). Ces marqueurs génétiques doivent être le plus universel possible tout en ayant juste
assez de dissimilitudes pour pouvoir différencier les espèces ou les groupes d’intérêt ainsi, le
6
choix des marqueurs d’ADN appropriés dépend fortement de la résolution désirée pour
l’identification (Drummond et al., 2015).
Le gène de la sous-unité I du cytochrome c (COI), un gène codant une protéine de l’ADN
mitochondrial (ADNmt), est souvent défini comme le marqueur de référence pour le
métabarcoding des métazoaires, car celui-ci permet d’identifier un nombre abondant de
phylum en plus d’avoir une variation génétique permettant d’effectuer le discernement au
niveau des espèces (Hebert et al. 2003; Deagle et al., 2014). De plus, l’ADNmt est davantage
présent que l’ADN nucléaire au sein des cellules, et ce même dans des milieux qui
contiennent de faibles concentrations d’ADN et lorsque celui-ci est dégradé (Rees et al.
2014). Le marqueur 18S, gène présent sur l’ARN ribosomique (ARNr) est aussi largement
utilisé en raison de sa taille et de son abondance dans le génome, en plus de sa caractéristique
à être peu sujet aux mutations, ce qui fait de lui un bon candidat pour distinguer les différents
phyla entre eux (Carugati et al. 2015). Les marqueurs COI et 18S sont souvent ciblés comme
code-barres génétiques en raison de leur accessibilité dans les bases de données publiques
telles que BOLD, SILVA et GenBank.
ADN environnementale (ADNe)
Depuis que Ficetola et al. (2008) ont démontré qu’il était possible de détecter la présence de
vertébrés en utilisant l’ADNe contenu dans des échantillons d’eau, l’intérêt pour l’ADNe
comme outil pour la conservation des poissons, des invertébrés aquatiques et des amphibiens
s’est rapidement intensifié (Goldberg et al. 2015). Comme toutes espèces interagissant avec
leur environnement, celles-ci rejettent constamment de leur ADN dans les endroits qu’elles
fréquentent (Thomsen et Willerslev 2015). Cet ADN représente du matériel génétique
pouvant provenir de gamètes, d’œufs, de fèces, d’urine, de mucus, de salive, ou encore de
sang (Bohmann et al. 2014; Keskin 2014; Rees et al. 2014). Une fois relâché dans
l’environnement, la détection de l’ADNe dépend de trois processus distincts soient sa
production, son transport ainsi que sa dégradation (Goldberg et al., 2015). Différents taux de
production d’ADNe chez de multiples organismes peuvent être attribuables à la taille, l’état
7
de santé, le sexe et la densité de ceux-ci, sans oublier la température du milieu puisque cette
dernière possède une influence majeure sur le métabolisme des organismes (Goldberg et al.
2015; Klymus et al. 2015; Lacoursière-Roussel et al. 2016). La préservation de l’ADN varie
grandement selon le milieu allant de quelques semaines dans les eaux tempérées à plusieurs
milliers d’années dans le permafrost, tout dépendamment de certains facteurs. Les
endonucléases présentes dans le milieu, l’eau, les radiations UV ainsi que l’action des
bactéries ou champignons représentent tous des conditions environnementales qui
contribuent à la dégradation de l’ADN dans différents écosystèmes (Shapiro 2008).
L’utilisation de l’ADNe constitue non seulement une méthode surpassant certains défis
techniques, mais représente également une méthode non invasive qui contribue à diminuer
grandement les perturbations dans les habitats des espèces recensées (Bohmann et al. 2014).
L’utilisation de l’ADNe s’est également révélée être plus efficiente pour détecter les espèces
rares ainsi que les espèces difficiles d’approche comparativement aux méthodes
d’échantillonnage traditionnelle (Goldberg et al. 2015; Smart et al. 2015). Combiné au
metabarcoding, l’ADNe peut fournir un portrait instantané des communautés locales sans
avoir à échantillonner directement chaque organisme présent. Ces nouveaux outils
moléculaires représentent une nouvelle méthode de caractérisation de la biodiversité remplie
de potentiel, constituant une opportunité d’augmenter considérablement l’efficacité des
relevés de biodiversité. Toutefois, il demeure important de souligner que comme toute
approche d’échantillonnage, l’ADNe possède également ces faiblesses. Les principales
lacunes reposent sur son lien étroit avec la qualité des bases de données publiques (Kwong
et al. 2012; Elbrecht et al. 2017) de même que sur une detection grandement influencée par
les amorces utilisées et leurs sources de biais respectives (Elbretch et Leese 2015).
De récentes études portant sur les écosystèmes marins côtiers ont démontrées l’efficacité du
métabarcoding d’ADNe pour en décrire la biodiversité (Deiner et al. 2017 ; Lacoursière-
Roussel et al. 2018). Bien que plusieurs études soient parvenues à détecter une plus grande
richesse à l’aide de l’ADNe comparativement aux méthodes traditionnelles en ce qui a trait
aux communautés de poissons (Thomsen et al. 2012 ; Yamamoto et al. 2017), peu d’études
on fait l’analyse de communautés à l’aide d’indices de biodiversité à plusieurs échelles
8
spatiales entre l’ADNe et les méthodes conventionnelles de collecte de spécimens, qui plus
est dans un milieu extrême comme l’Arctique.
Les invertébrés marins
Bien qu’en écologie la tendance générale prévoit une diminution de la biodiversité face à une
augmentation du gradient latitudinal, les invertébrés marins font potentiellement exception à
cette règle (Kendall 1996). En effet, Wei et al. (en révision) ont récemment montré que la
diversité d’organismes benthiques était plus grande dans l’Arctique Canadien que dans les
eaux canadiennes de l’Atlantique. Certaines études suggèrent qu’il y aurait plus de 4000
espèces d’invertébrés habitant l’océan Arctique (Gradinger et al. 2010 ; Piepenburg et al.
2011 ; Jorgensen et al. 2013) dont plus de 90% au niveau benthique (Josefson et al. 2013),
mais qu’entre 20 et 30% demeurent à découvrir (Pienpenburg et al. 2011, Snelgrove 2010).
Parmi ces organismes, certains n’effectueront que certains stades de leur cycle de vie parmi
le plancton et seront qualifiés de méroplancton alors que d’autres organismes passeront quant
à eux l’entièreté de leur cycle de vie en mode pélagique et seront qualifiés d’holoplancton
(Marcus et Boero 1998).
L’assemblage des organismes benthiques et planctoniques joue un rôle de premier plan dans
la productivité ainsi que dans la structure biologique des écosystèmes marins (Marcus et
Boero 1998). Leur présence dans la diète de poissons, d’oiseaux et de mammifères souligne
bien leur importance dans le réseau trophique (Gajbhiye 2002, Bluhm et al. 2008, CAFF
International Secreteriat 2010). Par conséquent, des changements dans les communautés
d’invertébrés marins pourraient affecter non seulement la stabilité des écosystèmes mais
perturber également les habitudes alimentaires des communautés humaines (Ruiz et al. 1997,
Guyot et al. 2006), qui sont encore grandement dépendante de l’accessibilité de la nourriture
sauvage (Van Oostdam et al. 2005 ; McNeely et Shulski 2011).
9
La distribution hétérogène de même que la difficulté d’identifier correctement les invertébrés
marins rend toutefois plus ardu le suivi de leur biodiversité (Ministère de l’Environnement
2006 ; Jaroslaw et al. 2016). De plus, le nombre d’espèces exotiques envahissantes a plus
que triplé au cours des dernières décennies en Amérique du Nord ainsi que dans les
environnements marins nordiques chez les invertébrés. Jumelé au nombre considérable
d’espèces cryptiques parmi de ces organismes (Knowlton 1993), des méthodes de suivi
fiables sont plus que nécessaire en prévision des changements majeurs que s’apprêtent à vivre
les communautés marine de l’Arctique (Millenium Ecosystem Assessment 2005 ; UNEP
2006).
Contexte du projet
Ce projet de maîtrise s’inscrit dans l’un des quatre volets d’un programme de recherche
fédéral de grande envergure ayant comme objectif général le recensement des communautés
d’invertébrés marins benthiques et pélagiques des principaux ports commerciaux de
l’Arctique canadien. Le but est avant tout d’établir les meilleures banques de données
possibles sur la biodiversité actuelle de ces lieux afin d’être en mesure de pouvoir développer
un suivi efficace de la composition des communautés de même qu’une surveillance accrue
des espèces aquatiques envahissantes susceptibles de s’introduire dans ces communautés.
Cette étude inclut la participation active de l’organisme gouvernemental Pêches et Océans
Canada, de l’organisation Savoir polaire Canada (POLAIRE) ainsi que du réseau de centres
d’excellence du Canada ArcticNet. En prévision d’une augmentation de la navigation en lien
avec le développement économique de la région et des changements climatiques, combinés
aux impacts des changements climatiques eux-mêmes, de nouvelles méthodes
d’échantillonnage alliant faibles coûts et peu de matériel à transporter en région éloignée
suscitent beaucoup d’intérêt. Ce projet de maîtrise contribuera à approfondir nos
connaissances sur le metabarcoding de l’ADN environnemental comme outil de détection de
la biodiversité des invertébrés marins dans des écosystèmes fragiles et productifs tels que les
ports commerciaux afin de mieux en évaluer le potentiel pour de futurs plans de gestion et
de conservation.
10
Objectifs
Dans le cadre de ce projet de maîtrise, il fut question de mettre en relation diverses notions
faisant appel à l’écologie tel que les indices de biodiversité ainsi que des outils moléculaires
tel que le metabarcoding afin d’améliorer nos connaissances sur l’efficacité et l’écologie de
l’ADN environnemental (ADNe) tout en réalisant un bilan de la biodiversité des invertébrés
marins dans trois écosystèmes portuaires de l’Arctique. L’objectif global étant de comparer
les patrons de biodiversité obtenu à différentes échelles spatiales entre l’ADNe et les
méthodes conventionnelles de collecte de spécimens. Pour ce faire, les buts de cette étude
étaient 1) de comparer la biodiversité gamma (richesse spécifique de chacun des ports)
obtenue avec l’ADNe et la collecte de spécimens (chalut, benne Van Veen, carottes de
sédiments et filets) ; 2) obtenir une meilleure compréhension sur des processus tels que la
dispersion, l’hétérogénéisation ou l’homogénéisation biotique des communautés grâce aux
indices de biodiversité alpha et beta et ; 3) examiner comment les cycles de vie des
organismes contribuent et influencent la détection de l’ADNe des invertébrés marins en
milieux côtiers. Ce projet permettra d’acquérir de plus amples notions sur l’ADNe
permettant ainsi d’orienter ce nouvel outil vers une utilisation plus efficiente, notamment
dans les domaines de la gestion et de la conservation.
11
Chapter I: Comparing eDNA metabarcoding and species
collection for documenting Arctic metazoan biodiversity
Chapitre I : Comparaison du métabarcoding d’ADNe et de la collecte
d’espèces afin de documenter la biodiversité des métazoaires de l’Arctique
12
Résumé
Méconnue puis sous-estimée, la biodiversité de l’océan Arctique fait maintenant face à de
nombreux changements en lien avec diverses perturbations d’origine environnemental et
anthropique. Ces transformations menacent plusieurs communautés marines,
particulièrement dans les écosystèmes sensibles tels que les ports commerciaux. La
protection de la biodiversité représente un défi considérable, allant bien au-delà de l’intérêt
porté à quelques espèces charismatiques et nécessitant avant tout de bonnes connaissances
sur l’organisation spatiale des espèces. Dans ce contexte, l’utilisation d’indices de
biodiversité de même que l’acquisition de nouvelles techniques permettant d’évaluer
efficacement la biodiversité se révèlent d’une importance capitale. Cette étude vise à
comparer la biodiversité de même que la composition des communautés d’invertébrés marins
collectées à l’aide de méthodes traditionnelles à celles obtenues grâce à une nouvelle
approche moléculaire ; le metabarcoding d’ADN environnemental (ADNe). Pour ce faire,
des échantillons d’eau de 250 ml furent récoltés à trois différentes profondeurs au sein de 3
ports commerciaux de l’Arctique, soit les ports de Churchill, Baie Déception et Iqaluit. À
partir de ces échantillons, il fut possible d’identifier 202 genres répartis dans plus de 15 phyla
grâce à l’ADNe. De ces organismes, seulement 9 à 15% furent également collectés par les
méthodes traditionnelles, révélant ainsi l’existence de différences significatives au niveau de
la richesse et de la composition des communautés entre ces différentes approches
d’échantillonnage. Parmi les facteurs responsables de ces différences figurent la dispersion
des molécules d’ADNe ainsi que les différences inter-spécifiques dans le cycle vital. Les
résultats obtenus dans le cadre de cette étude mettent bien en évidence la capacité de ces
nouveaux outils moléculaires tout en soulignant leurs caractères complémentaires aux
méthodes traditionnelles pour d’éventuelles applications en gestion et conservation de la
faune marine de l’Arctique.
13
Abstract
Arctic biodiversity has long been underestimated and is now facing rapid transformations
due to ongoing climate change and other impacts including shipping activities. These changes
are placing marine coastal invertebrate communities at greater risk, especially in sensitive
areas such as commercial ports. Preserving biodiversity is a significant challenge, going far
beyond the protection of charismatic species and involving suitable knowledge of the
organization of species in space. Therefore, knowledge of alpha, beta and gamma
biodiversity indices are of great importance in achieving this objective together with new
cost-effective approaches to monitor changes in biodiversity. This study compares
metabarcoding of COI mitochondrial genes and 18S rRNA genes from environmental DNA
(eDNA) water samples with standard species collection methods to document patterns of
invertebrate communities at various spatial scales. Water samples (250 mL) were collected
at three different depths within three Canadian Arctic ports; Churchill, MB, Iqaluit, NU and
Deception Bay, QC. From these samples, 202 genera distributed across more than 15 phyla
were detected using eDNA metabarcoding, of which only 9% to 15% were also identified
through species collection at the same sites. Significant differences in taxonomic richness
and community composition were observed between eDNA and species collections, both on
local and regional scales. This study shows that eDNA dispersion in the Arctic Ocean reduces
beta diversity in comparison to species collection while emphasizing the importance of
pelagic life stages for eDNA detection. This study highlights the potential of eDNA
metabarcoding to assess large-scale arctic marine invertebrate diversity while emphasizing
that eDNA and species collection should be considered as complementary tools for providing
a more holistic picture of the marine invertebrate communities living in coastal areas.
14
Introduction
The Arctic Ocean has been poorly surveyed, but likely harbors a much larger proportion of
undetected biodiversity than previously thought due to a lack of monitoring (Archambault et
al. 2010, Darnis et al. 2012). Recent estimates suggest that there are more than 4000
invertebrate’s species inhabiting the Arctic Ocean (Gradinger et al. 2010, Piepenburg et al.
2011, Jorgensen et al. 2016) with more than 90% being benthic organisms (Josefson et al.
2013). The general pattern of biodiversity decline with increasing latitude may not apply to
marine invertebrates (Kendall 1996) in which a great diversity is found and many species
await discovery (Archambault et al. 2010, Piepenburg et al. 2011). Recently, Wei et al (in
revision) showed that benthic diversity was more diverse in the Canadian Arctic than in
Canadian Atlantic waters. Previously, considered as the second most pristine oceans on earth
(UNESCO 2010), this ecosystem has experienced extensive environmental change since the
1950s (IPCC 2018). In addition to warmer temperatures, increased acidification and greater
freshwater inputs (Arctic Climate Impact Assessment [ACIA] 2004), other activities such as
marine shipping (ACIA 2004, Chan et al. 2012) and the associated risk of introduction of
non-indigenous species (NIS), are increasing (Casas-Monroy et al. 2013, Chan et al. 2013,
Goldsmit et al. 2017, 2019). The number of invasive species has more than tripled since the
beginning of the century in North America and in northern environments (Millennium
Ecosystem Assessment 2005; UNEP 2006). Comprehensive baseline surveys and ongoing
monitoring are thus essential in the Arctic, especially due to the large number of cryptic and
cryptogenic species (Knowlton 1993, Carlton 1996, Goldsmit et al. 2014). However, a better
understanding the invertebrate community structure and its temporal change is challenged by
their heterogeneous distribution, taxonomy and limitations of sampling under ice cover
(Ministry of Environment 2006; Jaroslaw et al. 2016).
The design of a robust monitoring approach aimed at evaluating biodiversity changes such
as species losses and processes that maintain species diversity over longer time frames must
accurately consider the spatial and temporal organization of biodiversity. Diversity can be
measured using different taxonomy-based metrics and at various scales through alpha, beta
and gamma diversity indices. Although alpha biodiversity, which represents the species
15
assemblage of a relatively small area, termed “within-habitat diversity” (MacArthur 1965),
is the most commonly studied biodiversity scale. Beta diversity, often referred as “turnover
diversity”, is the variation in species composition (as a function of presence absence or
relative abundance) among local species assemblages which depends on the balance of biotic
heterogenization and homogenization, and therefore, essential for evaluating responses of
communities to significant disturbances (Mori et al. 2018). Lastly, gamma diversity refers to
the species assemblage of large areas. e.g., regional diversity (Socolar et al. 2015) and is
expressed in the same units as alpha diversity (Laurila-Pant et al. 2015). Large-scale
biodiversity monitoring is essential for understanding more extensive changes in coastal
community composition, but this is logistically challenging and costly in remote areas such
as the Arctic. Coastal metazoan collections are generally invasive (e.g. trawling, grab
sampling), selective, frequently limited to the summer open water period, and rely on some
degree of subjectivity with respect to taxonomic expertise (Jones 1992; Jorgensen et al.
2016).
Ten years after the pioneering study by Ficetola et al. (2008), the environmental DNA
(eDNA) approach offers major advantages over conventional monitoring methods and is
perceived as a game changer for ecological research (Creer et al. 2016). This approach
involves collection and detection of DNA that has been excreted by organisms into the
surrounding environment through metabolic waste products, gametes or decomposition
(Taberlet et al. 2018; Hansen et al. 2018). Analysis of eDNA with metabarcoding, which is
a rapid method of biodiversity assessment that links taxonomy with high-throughput DNA
sequencing (Ji et al. 2013), can provide a snapshot of local species composition without the
need for sampling of individual organisms. Recent studies in coastal marine ecosystems have
demonstrated the feasibility of eDNA metabarcoding to document marine metazoan
biodiversity in the Arctic (Lacoursière-Roussel et al. 2018; Grey et al. 2018). Despite limited
knowledge of the ecology of eDNA (i.e. origin, fate, state and transport; Barnes and Turner
2016), eDNA is increasingly being incorporated within monitoring toolboxes for a large
variety of aquatic organisms and ecosystems (Roussel et al. 2015, Deiner et al. 2017).
16
However, like any sampling approach, eDNA metabarcoding also has its weaknesses which
must be considered to avoid misinterpretation of the results. Although this molecular tool
allows rapid assessment of biodiversity, database gaps hamper the use of eDNA as sequence
assignments are highly dependent on their presence in public databases (Kwong et al. 2012;
Elbrecht et al. 2017). Organism detection is also restricted by the primers used and their
respective biases (Elbretch and Leese 2015). Furthermore, eDNA does not provide any
physiological information or even health information on the organisms detected unlike direct
species collection (Thomsen and Willerslev 2015).
While many studies have compared species composition measured by eDNA with
conventional methods in fish (e.g. Thomsen et al. 2012, Yamamoto et al. 2017), few such
comparative studies have been performed on invertebrates, and even less have considered
the spatial scales of observation. Among marine invertebrate species, meroplankton
(organisms having planktonic larval life stages) and holoplankton (organisms spending their
entire life in planktonic state) represent key components of the food web and ecosystem
stability (Marcus and Boero 1998; Gajbhiye 2002). A better understanding how complex
planktonic life-stages of invertebrates affect the origin and transport of eDNA in coastal
environments is essential to developing genomics-based biodiversity indices aimed at
informing conservation plans.
The main objective of this study is to compare patterns of biodiversity at different spatial
scales revealed by eDNA metabarcoding and conventional species collection within and
among three ports in the Canadian Arctic Ocean. More specifically, gamma biodiversity
(species richness between ports), was compared based on results from eDNA and
conventional collecting methods, namely benthic trawl, Van Veen grab, cores and plankton
net tows. Secondly, alpha (species richness within ports) and beta (similarity of species
between sites within ports) biodiversity indices were contrasted for results based on eDNA
and species collections, to better understand how eDNA may inform species distributions
and ecological processes such as dispersion and biotic heterogenization or homogenization.
Finally, the life histories of organisms were considered in order to interpret how this may
17
affect the origins and detection of eDNA from coastal invertebrates and contribute to
observed discrepancies between eDNA detection and conventional species collections.
Methods
Sample collection
Specimens and eDNA were collected at 13 subtidal stations (≤ 20 m at low tide) in three
commercial ports of the Canadian Arctic during the summer period (Figure 1). Churchill was
surveyed August 11-14, 2015 and Iqaluit was surveyed between August 17-22, 2015 and
between July 24-26, 2016 respectively while Deception Bay was surveyed between August
19-27, 2016. These three Arctic ports were selected because of their vulnerability to potential
changes in the coastal marine invertebrate communities not only to factors such as climate
change, but because of their relatively high levels of shipping activity which place them at
greater risk for introduction of non-indigenous species (Chan et al. 2012, 2013, Goldsmit et
al. 2019).
18
Figure 1. Geographical location of Churchill, Deception Bay and Iqaluit ports in the
Canadian Arctic (map A) and distribution of stations within Churchill (map B), Deception
Bay (map C) and Iqaluit (map D).
Species collection
Throughout the paper, we use specimens collected or species collection to refer to the
following collecting methods; benthic trawls, Van Veen grabs, sediment cores and plankton
tows. We use the term benthic communities to refer to organisms collected through benthic
trawls, Van Veen grabs and sediment cores while we use the term zooplankton communities
to refer to organisms collected through net tows. Benthic invertebrates living on the sea floor
substrate (epifauna) were collected using the benthic trawl with a 500 µm mesh net while
benthic invertebrates living in soft sea bottom (infauna) were collected using a Van Veen
grab (0.1 m2 sample area; Deception Bay and Iqaluit) and then sieved to a minimum of 500
µm. Zooplankton was collected using two net tows of 0.5 m diameter, one vertical Nitex®
80 µm and one oblique Nitex® 250 µm. Zooplankton samples were taken at ten of the 13
stations that were common with the eDNA sampling, whereas benthic trawl and Van Veen
grab samples were taken at the same 13 stations as the eDNA sampling. Trawling and oblique
19
net tows were carried out for 3 minutes at a speed of 1-2 knots to ensure similarity between
the samples of each station within each port. Due to logistical constraints, samples were
collected with the trawl in 2016 instead of 2015 in Iqaluit and infauna samples were not
collected using the Van Veen grab in Churchill. Instead, subtidal core data (15 cm high x 10
cm diameter) from the same areas were used from Goldsmit 2016. Since the sediment volume
accumulated by these subtidal sediment cores was less than that of the Van Veen grab, the
replicates of a given site for the sediment cores were combined together such that the final
volume included for analyses was similar to the volume of site-specific Van Veen grab
samples from the other ports. With the exception of common easily identifiable macro
invertebrates, which were enumerated, recorded and released, all specimens were preserved
in 95% ethanol and later identified by trained taxonomists to the lowest taxonomic level
possible.
Environmental DNA samples
Water samples were collected and filtered following methods outlined in Lacoursière-
Roussel et al. (2018). A 250 ml water sample was taken at each of three different depths
(surface, mid-depth and deep water (i.e., 50 cm from the bottom)) for each station and port
using a 5 L oceanographic Niskin water sampling bottle. Each sample was filtered in the field
using a 0.7 μm glass microfiber filter (Whatman GF/F, 25 mm) and syringes (BD 60 mL,
Kranklin Lakes, NJ, USA). Negative field controls were made by filtering 250 ml of
autoclaved distilled water for every 10 collected samples. All filters were preserved in 2 ml
microtubes containing 700 µl of Longmire’s lysis/preservation buffer, kept at 4°C until the
end of the sampling campaign and then frozen at -20°C until their extraction (at most 4
months). Risks of cross-contamination during the field sampling process were reduced by
using a separate sterile kit for each sample. Sampling kits included bottles and a filter housing
sterilized with a 10% bleach solution and new sterilized gloves, syringes and tweezers sealed
in a transparent plastic bag. Each sampling kit was exposed to UV light for 30 minutes
following assembly.
20
Metabarcoding
Environmental DNA extraction, amplification and sequencing
To avoid risk of laboratory cross-contamination, eDNA extraction, PCR preparation and post
PCR steps were done in three separate rooms. All PCR manipulations were done in a
decontaminated UV hood. All laboratory benches surfaces were cleaned with DNA AWAY
® and all laboratory tools were sterilized with 10% bleach solution and exposed to UV light
for 30 minutes before any manipulations were carried out. DNA was extracted from filters
following a QIAshredder and phenol/chloroform protocol (Lacourssière-Roussel et al.,
2018). Negative control extractions (consisting of 950 µl of distilled water) were done for
each sample batch (i.e., one for every 23 samples) and were treated as normal samples for
the remaining manipulations until sequencing.
To maximize biodiversity detection and reduce the bias of eDNA dominance among species
groups, two pairs of primers from two different genes (COI and 18S) were used. These have
been shown to work well for detecting a wide variety of taxa including invertebrates and have
reasonably comprehensive databases of reference sequences. Following Lacoursière-Roussel
et al. (2018), we used the forward mlCOIintF (Leray et al., 2013) and reverse jgHCO2198
(Geller et al., 2013) (hereafter called COI1) and the forward LCO1490 (Folmer et al., 1994)
and reverse ill_C_R (Shokralla et al., 2015) (hereafter called COI2). Two additional
universal 18S primer pairs were also used, the forward F-574 and reverse R-952 (Hadziavdic
et al.,2014) (hereafter called 18S1) and the forward TAReuk454FWD1 and reverse
TAReukREV3 (Stoeck et al., 2010) (hereafter called 18S2). Three PCR replicates were done
for each sample of each primer set and were then pooled together after amplification and
purification procedures (see Annex for more details). Sequencing was carried out using an
Illumina MiSeq (Illumina, San Diego, USA) with a paired-end MiSeq Reagent Kit V3
(Illumina, San Diego, USA) at the Plateforme d’Analyses Génomiques (IBIS, Université
Laval, Québec, Canada). Each port was analyzed on a separate run to ensure independency,
but the samples within a port were pooled within a single Illumina MiSeq run to ensure the
equality of sequencing depth among samples. Raw sequences reads were deposited in
21
NCBI’s Sequence Read Archive (SRA, http://www.ncbi.nlm.nih.gov/sra) under Bioprojects
PRJNA388333 and PRJNA521343.
Bioinformatics
Adaptor and primer sequences were removed and the raw sequencing reads were
demultiplexed into individual samples files using the MiSeq Control software v2.3. Raw
reads were analyzed using Barque version 1.5.1, an eDNA metabarcoding pipeline
(www.github.com/enormandeau/barque). Forward and reverse sequences were trimmed and
filtered using Trimmomatic v 0.30 with the following parameters: (TrimmomaticPE, -
phred33, LEADING:20, TRAILING:20, SLIDINGWINDOW:20:20, MINLEN:200)
(Bolger et al. 2014). Pairs of reads were merged with FLASh v1.2.11 (Fast Length
Adjustment of Short reads) with the following options: (-t 1 -z -m 30 -M 280) (Magoč &
Salzberg 2011). The amplicons were split using their primer pairs (COI1, COI2, 18S1 and
18S2) and sequences that were either too short or too long were removed. Chimeric
sequences were removed using VSEARCH v 2.5.1 (uchime_denovo command with default
parameters) (Rognes et al. 2016). COI sequences were blasted on the BOLD database and
18S sequences against the SILVA database. Sequences from most terrestrial species (insects,
human, birds and mammals) and sequences that had no taxonomic match were also removed
from the reference databases. Finally, following these steps, chordates others than tunicates
(Table S1) were removed from the results since they were not targeted in this study and would
therefore blur the analyses and subsequent interpretations regarding invertebrate
communities. The Barque pipeline (https://github.com/enormandeau/barque) was then used
to create operational taxonomic units (OTU). The OTUs were generated using vsearch 2.5.1
(id 0.97) (https://github.com/torognes/vsearch) using only reads present more than 20 times
in the full dataset. For each station, sequences collected at the different depths and for all
primers were pooled in order to obtain an overall representation of potential biodiversity.
22
Data analysis
All analyses were performed at the genus level to facilitate comparisons between the
approaches since only ~60 and 80% of the invertebrate taxa could be identified to species
level with species collections and the eDNA approach, respectively. All analyses were
conducted in R version 3.4.3 (R Core team 2017) except for the SIMPER analyses which
were conducted in PRIMER 6 & PERMANOVA+ (Clarke and Gorley, 2006).
In order to determine the effect of sampling effort on overall richness being detected, genera
rarefaction curves were created for each port and data collection type using the "specaccum"
function (number of permutations = 100) in the vegan package of R (Oksanen et al. 2018).
Variation in taxonomic composition detected with eDNA and species collection within ports
was depicted using a barplot generated in R from the raw relative abundance of the genus
taxonomy matrices assigned to the corresponding phylum. PERMANOVAs (number of
permutations = 10, 000) were performed using the vegan package to test the effect of port
and sampling method on taxonomic composition, while nonmetric multidimensional scaling
(nMDS) was used to visualize differences in taxonomic composition among ports and
sampling methods.
Alpha diversity indices (Richness, Shannon diversity H’ and Pielou evenness J) were
calculated with the vegan package (Oksanen et al. 2018) following Hellinger standardization.
Two-way ANOVAs followed by Tukey Honest Significant Difference (Tukey HSD) tests
were performed to evaluate if diversity indices differed among ports and sampling methods.
When standard ANOVA assumptions of normality were not met, PERMANOVAs were done
based on Euclidean distances, thereby ensuring approximate multivariate normality (Clarke
and Warwick 2001), and followed by use of the "pairwise. adonis" function in R to test
variation in sampling approaches among ports.
Beta diversity was estimated from the Sorensen distance using the "vegdist" function in the
vegan package (Oksanen et al. 2018) computed after a presence-absence transformation.
23
Geographic distance matrices between stations within ports were calculated using the
"spDistsN1" function in the sp package of R (Bivand et al. 2008) for Deception Bay and
Iqaluit while distance between Churchill stations were determined using ArcGIS version 10.4
due to some peculiarities of the geographic layout of this port (this port has a large peninsula
separating some sample stations (Figure 1b); since sp simply calculates the straight line
distance between two points, the distances between stations on either side of this peninsula
are underestimated using this package, while ArcGIS allows for calculation of the true
distance by water). The dispersion of eDNA within ports was evaluated from correlations
between beta-diversity and spatial distance matrices using Mantel tests (Li et al. 2018) in the
ade4 package of R (Dray and Dufour 2007) except for Churchill for which the correlation
was calculated using the "cor.test" function (method=Spearman) in the stats package of R
since ArcGIS does not provide a suitable distance matrix format for the Mantel test.
Finally, we investigated the probability of detecting different marine invertebrate taxa
according to their life history, paying particular attention to life histories having pelagic
stages (holoplankton and meroplankton) due to their potential presence in the water column.
To contrast the proportion of species with an entirely pelagic (i.e. holoplankton) versus
benthic-pelagic (i.e. meroplankton phase) life histories, a barplot was constructed in R from
a presence/absence data list with the lowest taxonomic resolution for each organism and the
associated life history category. A PERMANOVA analysis was performed using the vegan
package to test the effect of port on taxonomic composition within each life history type
(holoplankton versus taxa with meroplanktonic stages). Similarity percentage analysis
(SIMPER) in PRIMER 6 & PERMANOVA+ was used to determine which taxa contributed
the most to explaining differences among the groups.
Results
Sequencing quality
A total of 478,046 aquatic metazoan reads were obtained in Churchill, 95,658 reads in
Deception Bay and 203, 245 reads in Iqaluit (see Table S1 for more details about the pipeline
24
processes). The 18S markers generally generated more sequences than COI markers (with a
total of 568 892 and 208 067 sequences respectively), except for in Iqaluit where the reverse
situation was observed. In addition, both genetic markers provided distinctive taxonomic
resolution according to our knowledge of organisms living in Arctic waters1 (Table 1). The
genus taxonomic level provided a satisfying description of biodiversity given that only 10,
17 and 18% of sequences were not assigned at that this taxonomic level in Churchill,
Deception Bay and Iqaluit, respectively (Figure S1). Thus, 2682, 1413 and 1056 operational
taxonomic units (OTUs) where identified at the genus level in the ports of Churchill,
Deception Bay and Iqaluit, respectively.
Table 1. Summary of the numbers of reads, the proportion of species and genera present in
the Arctic historic database and the mean number of OTUs for the COI primers set and the
18S primers set assigned and non assigned on BOLD and SILVA for each port.
Port Number of reads
Proportion of
species known in
Arctic (%)
Proportion of genus
known in Arctic (%)
Mean nb. of
assigned OTUs
(genus)
Mean nb. of
non-assigned
OTUs (genus)
COI 18S COI 18S COI 18S COI 18S COI 18S
Churchill 52 749 425 297 52.3 18.7 61.7 45.9 633 708 39 100
Deception
Bay 30 214 65 454 62.9 18.3 74.3 52.6 348 359 16 105
Iqaluit 125 104 78 141 69.4 15.4 77.6 46.3 238 291 4 92
No amplification was observed on agarose gels for the negative PCR controls, but a small
number of sequences were present in our laboratory and field negative controls. Two
correction factors were applied to ensure the reliability of the data and quality of the resulting
analyses. First, the few sequences present in the laboratory negative controls were subtracted
1 List of known species in the Arctic obtained by pooling various species databases (N = 1054 species,
Howland et al., unpublished data) and published information (Cusson et al., 2007; Piepenburg et al., 2011;
Link et al., 2013; Olivier et al., 2013; Goldsmit et al., 2014; Roy et al., 2015; López et al., 2016; Young et
al., 2016)
25
from the samples from the same extraction batch. These sequences represent 0.003%, 0.1%
and 0.06% of Churchill, Deception Bay and Iqaluit total number of sequences respectively.
Secondly, for the field negative controls, a genus was removed if its abundance in all the field
controls was higher than 2% of the total number of sequences for all samples combined for
that genus. Following application of correction factors for background contamination, 0.1%
and 1.4% of all COI and 18S sequences were removed, respectively (Table S3). An exception
to applying correction was made in the case of 18S Pseudocalanus sequences for which 96%
of all the field contamination occurred in only one field negative control. Given that
Pseudocalanus in real samples represented nearly half of all 18S sequences and this genus is
known to be a dominant part of the Arctic zooplankton community (Dispas 2019), removing
it would cause significant bias to the analyses. When read abundance of a given genus in
field controls was lower than 2% of the total number of sequences for that genus, it was
retained because contamination was considered low enough that it would not lead to false
interpretations. On the contrary, discarding those genera could cause bias in the analyses due
to their high number of sequences in real samples.
Arctic coastal gamma diversity
With the exception of benthos communities sampled using trawls, grabs and cores, genera
rarefaction curves of marine invertebrates were close to saturation, both for zooplankton
surveyed using net tows and eDNA (Figure 2).
26
Figure 2. Individual-based rarefaction curves of eDNA, benthos and zooplankton genera for
Churchill (blue), Deception Bay (yellow) and Iqaluit ports (magenta) based on incidence data.
Vertical bars represented 95% confidence intervals.
When eDNA and species collection datasets were combined, a total of 634 marine
invertebrate genera from 23 phyla were recorded. Gamma richness was consistently higher
for species collections methods than for eDNA, however variable patterns across ports were
observed between these sampling approaches. Gamma richness detected with eDNA was
higher for Churchill and Deception Bay and lower for Iqaluit port, but the opposite pattern
was observed for the gamma richness of communities detected with species collection
(Figure 3a). Although a substantial collective number of organisms were detected, few genera
where shared between eDNA and species collections (Churchill 15%, Deception Bay 15%
and Iqaluit 9%). Of the organisms found with both approaches, annelids accounted for almost
half (42.7%), followed by arthropods and molluscs with 20.2% and 11.2% respectively of
the common genera obtained within all ports (Figure 3b).
27
Figure 3. a) Barplots of the number of taxa found in Churchill (blue), Deception Bay
(yellow) and Iqaluit (red). Darker bands represent species collection methods while paler
bands with dashed outline represent eDNA and black bands represent the number of genera
in common between eDNA and species collection. b) Phylum relative proportion of commun
genera between eDNA and species collection based on incidence data. COI and 18S primer
sets are added together for both a) and b).
The same phyla were generally present among the three ports (Figure 4), with Annelida and
Arthropoda consistently being the most abundant phyla for both eDNA and species
collections. However, for most taxa the relative abundance was significantly different
between eDNA and species collection (PERMANOVA, P < 0.001; Table S4; Figure 4).
Community composition of eDNA clearly differed among ports (PERMANOVA, P < 0.001;
28
Table S4; Figure 5a), and this distinction remained significant (PERMANOVA, P < 0.001;
Table S4; Figure 5b) for the species collections. Differences in community structure with
eDNA versus species collection were mainly driven by annelids and arthropods taxa
(SIMPER analysis; 30% and 23% respectively) followed by molluscs, echinoderms,
cnidarians and bryozoans (SIMPER analysis; 11%, 6%, 5% and 4% respectively). The
remaining differences between eDNA and species collection community compositions may
be partly driven by taxa-specific differences in detectability by these approaches. For
example, some taxa such as Brachiopoda, Foraminifera, Cephalorhyncha and Chaetognatha
(grouped in the Others category with additional phyla of low relative abundance) were only
found using species collection, while others such as Bryozoa were only rarely detected using
eDNA. In contrast, other taxa such as Porifera, Nemertea, Cnidaria and Echinodermata were
more frequently detected with higher read abundances in eDNA samples than in species
collections.
Figure 4. Marine invertebrate taxonomic composition at the phylum level for eDNA and
species collection methods, respectively, within Churchill, Deception Bay and Iqaluit ports,
based on incidence data. The COI and 18S regions are added together for the eDNA barplot
while the benthic trawl, core, grab and net tows samples are added together for the species
collection barplot.
29
Figure 5. Biodiversity differences A) among ports based on eDNA only and B) among
sampling methods within ports. Ordination of taxonomic composition (genera) calculated
using Sorensen index (incidence based) with each data point representing a specific sample;
blue squares represent Churchill, yellow dots represent Deception Bay and magenta triangles
represent Iqaluit. Filled symbols are associated with eDNA while empty symbols are
associated with species collections.
Alpha biodiversity
Similar to gamma diversity, alpha richness for eDNA samples was significantly
higher in Churchill and Deception Bay than Iqaluit (Tukey HSD, p < 0.01), with the number
of genera per station ranging from 49 to 75 (mean = 63 ± 2) in Churchill, 45 to 93 (mean =
70 ± 4) in Deception Bay and from 34 to 53 genera (mean = 41 ± 2) in Iqaluit (Figure 6a).
30
In contrast Churchill had the lowest alpha richness for samples collected using species
collection (Tukey HSD, p < 0.01, Figure 6b) with only 8 to 58 genera per station (mean = 27
± 3) as compared to 30 to 142 (mean = 78 ± 9) and 59 to 151 (mean = 100 ± 8) genera per
station in Deception Bay and Iqaluit, respectively. Overall differences between sampling
approaches varied between ports, with eDNA-based alpha richness being higher than species
collection sample-based richness in Churchill (PERMANOVA, P < 0.001; Table S4), similar
in Deception Bay (PERMANOVA, P = 0.4; Table S4), and lower in Iqaluit (PERMANOVA,
P < 0.001; Table S4). A similar pattern was also observed with the Shannon biodiversity
index (Figure S2).
Figure 6. Boxplot on alpha diversity for the genera richness and Pielou evenness index in
Churchill, Deception Bay and Iqaluit ports for eDNA (A, C) and species collection (B, D).
31
These analyses were performed on abundance data with Hellinger transformation, COI and
18S primer sets are added together for the eDNA boxplot.
Despite the contrasting alpha richness between sampling approaches within each port, the
generally high values of Pielou’s evenness indices revealed a pronounced taxonomic
evenness with little indication of particular genera being over-represented in communities
detected by either eDNA or species collection within the ecosystems (Table 2). Evenness of
communities detected with eDNA was similar across ports except between Deception Bay
and Iqaluit, where a lower or a greater dominance by some taxa was observed in Iqaluit
(PERMANOVA, p < 0.05; Table S4; Figure 6c). This is consistent with the SIMPER
analyses where, for Iqaluit, 19 genera explained 90% of the similarity among stations in
contrast to 30 and 42 genera for Churchill and Deception Bay, respectively. There were no
differences in evenness of communities detected in species collections among the three ports
(PERMANOVA, p = 0.2; Table S4; Figure 6d).
Table 2. Summary of richness, alpha and beta biodiversity indices for the eDNA and species
collection marine invertebrates communities on abundance data after Hellinger
transformation (Shannon and Pielou indices) and presence/absence transformation (Beta
index). COI primer sets and 18S primer sets are added together.
Method Port
Gamma
richness
(Sγ)
Mean
alpha
richness
(Sα) SE
Mean
Pielou
(J) SE
Mean
Shannon
(H') SE
Beta
index
SE
eDNA
Churchill 138 63 ± 2 0.75 ± 0.02 3.12 ± 0.1 0.31 ± 0.004
Deception
Bay 145 70 ± 4 0.82 ± 0.02 3.48 ± 0.1
0.33 ± 0.005
Iqaluit 101 41 ± 2 0.67 ± 0.03 2.50 ± 0.1 0.37 ± 0.005
Species
collection
Churchill 193 27 ± 3 0.79 ± 0.02 2.50 ± 0.1 0.84 ± 0.008
Deception
Bay 292 78 ± 9 0.75 ± 0.04 3.17 ± 0.1 0.62 ± 0.01
Iqaluit 365 100 ± 8 0.84 ± 0.02 3.84 ± 0.1 0.58 ± 0.007
32
Beta diversity
Community structure between stations within ports was significantly different for both
eDNA and species collection but higher indices of beta diversity were observed for the
species collections than for eDNA (Table 2). For eDNA beta indices, higher dissimilarity
among stations was found in Iqaluit (0.37 ± 0.005), followed by Deception Bay (0.33 ± 0.005)
and Churchill (0.31 ± 0.004), while the opposite trend was observed with species collections
(Churchill: 0.84 ± 0.008; Deception Bay: 0.62 ± 0.01; Iqaluit: 0.58 ± 0.007).
Positive correlations between beta diversity and geographic distance between stations were
observed for both eDNA and species collection across all ports. For eDNA, positive
correlations between distance and beta diversity were significant and stronger in Churchill
and Deception Bay (R2 = 0.13 and 0.23 respectively; P < 0.05; Figure 7; Table S5) whereas
a significant albeit weaker correlation was found in Iqaluit (R2 = 0.09; P = 0.02; Table S5;
Figure 7). For species collections, the correlation between beta diversity and geographic
distance varied depending on the port and collection method (zooplankton tow nets versus
benthos sampling methods). In Churchill, none of the correlations were significant
(zooplankton R2 = 0.014; P = 0.2, benthos R2 = 0.004; P = 0.5; Table S5; Figure 7). For
Deception Bay, a lower positive and significant correlation was found for the benthos (R2 =
0.12, P = 0.02; Table S5; Figure 7) than for eDNA (R2 = 0.23; P = 0.01; Table S5; Figure 7)
while a stronger positive and significant correlation was found for the zooplankton (R2 =
0.26; P = 0.01; Table S5; Figure 7). For Iqaluit, a stronger positive and significant correlation
was found for the benthos (R2 = 0.14, P = 0.01; Table S5; Figure 7) than for eDNA (R2 =
0.09; P = 0.02; Table S5; Figure 7) while a negative and non-significant correlation was
found for the zooplankton (R2 = -0.16; P > 0.05; Table S5; Figure 7).
33
Figure 7. Sorensen dissimilarity index between pairs of stations as a function of distance
between the stations based on incidence data (presence/absence transformation on
abundance) for different sampling methods (eDNA and species collections of benthos and
zooplankton) in Churchill (blue), Deception Bay (yellow) and Iqaluit (magenta).
Origin of coastal eDNA
Taxa with the meroplanktonic life history type were the most common life history type
observed by eDNA across ports (≥ 70% of observed taxa; Figure 8). Although the relative
abundance of taxa by life history type varied among ports (PERMANOVA, P < 0.001), the
proportions of taxa with meroplanktonic or holoplankton (taxa with only pelagic stage) life
history types detected by eDNA were similar (Churchill: 69% meroplankton, 14%
holoplankton; Deception Bay: 72% meroplankton, 17% holoplankton; Iqaluit: 80%
meroplankton, 12% holoplankton; Figure 8). Annelida was the most dominant phylum
detected with the meroplankton life history type, followed by Mollusca and Echinodermata
(SIMPER analysis; 45.8% and both 15.7% respectively) whereas Arthropoda (copepods) was
the most dominant phylum in the holoplankton across the three ports (SIMPER analysis;
81.1%). Interestingly, the dominant taxa were similar for the meroplanktonic component of
34
communities detected from both eDNA and species collection approaches, with the exception
of Echinodermata which was replaced by Arthropoda (mostly amphipods) in species
collection samples (SIMPER analysis; Annelida 45.6%, Arthropoda 24.0% and Mollusca
16.5%). In the case of the holoplanktonic component of communities, Arthropoda
(copepods) was the most dominant phylum for both eDNA and zooplankton tows (SIMPER
analysis; 81.1% and 96.1% respectively).
Figure 8. Relative abundance of organisms obtained with eDNA and species collection
within Churchill, Deception Bay and Iqaluit ports according to their life history; taxa with a
meroplanktonic life stage, holoplanktonic taxa life stage. Species collection for benthos
include benthic trawl, Van Veen grab, cores and for zooplankton includes pelagic plankton
net tows. The sum of the detection of each taxa (i.e. presence/absence) have been combined
for all primer sets.
Discussion
Arctic coastal regions are subject to harsh conditions, a wide range of temperatures and
photoperiod, and they support various forms of life under long periods of sea ice cover (Payne
et al. 2012, PAME 2016). Nevertheless, the Arctic Ocean is home to a great diversity of
35
organisms which deserve attention, especially lower trophic taxa, including invertebrates,
which make up the base of ecosystem (Archambault et al. 2010, Piepenburg et al. 2011, Wei
et al. submitted). The presence of marine invertebrates in the diets of arctic fishes, birds and
mammals highlight their trophic relevance (Gajbhiye 2002, Bluhm et al. 2008, CAFF
International Secretariat 2010). Thus, significant changes in their communities could affect
ecosystem stability and disturb the availability of food resources for coastal human
communities (Ruiz et al. 1997, Guyot et al. 2006). Marine biodiversity conservation is
progressively becoming a crucial aim of environmental management (Spalding et al. 2007)
but requires sufficient spatial data on biodiversity (Laurila-Pant et al. 2015). Despite
substantial research efforts in recent years (Piepenburg et al. 2011, Goldsmit et al. 2014, Wei
et al. submitted), there is limited knowledge about the diversity of many invertebrate groups
(Archambault et al. 2010), including distributions and how this is influenced by their life
stage transitions. Indeed, many species new to science still await discovery with species
collection methods (López et al. 2016, Jabr et al. 2018).
To our knowledge, this study is the first to compare eDNA, benthos and zooplankton
community patterns in the Arctic. Here, we used eDNA sampling in parallel with species
collection at Arctic ports to better understand ecological properties of eDNA in relation to
distribution and life stages of invertebrates in the coastal marine environment. While
differing from observations made using species collection approaches, eDNA metabarcoding
of Arctic coastal zone taxa provided relevant, complementary biodiversity information at
various spatial scales using alpha, beta and gamma indices.
Overall biodiversity and community structure
Despite limited sample volumes (only 30 L water in total) and sequencing depth, eDNA
metabarcoding identified 202 marine genera, covering 17 phyla which complemented the
biodiversity information obtained from species collection using traditional benthic trawl,
cores, grab and net tows, representing a combined total of 634 genera, covering 23 phyla for
eDNA and species collection. Bryozoa, Arthropoda and Mollusca were more commonly
36
encountered with species collections of coastal marine communities while Echinodermata,
Porifera, Nemertea and Cnidaria were more frequently detected in eDNA samples. Several
physical and biological factors might explain the differences in detectability of taxa between
approaches. For example, echinoderms and sponges (Porifera) are often attached to large
boulders in the seabed (Bell and Barnes 2003, Chapman 2003) and are difficult to collect
using trawls or a grab, which may negatively bias their detectability in species-based
collections. Identification issues, or a combination with biases in detectability may also
explain differences in community assemblages identified through eDNA and species
collection. For instance, ribbon worms often lack easily diagnosable external body features
making identification challenging in addition to frequently being found under rocks where
they can be difficult to access (Thiel and Norenburg 2009). eDNA metabarcoding may thus
be particularly useful in such cases where taxa are more difficult to sample or identify
morphologically. It is also important to note the considerable variation in previous
sequencing effort deployed among phyla which influences whether or not eDNA from taxa
in these groups can be matched to sequences of morphologically identified organisms. For
example, 54.5–56.3% of the arthropods, cnidarians, and molluscs identified by our traditional
collection sampling methods were present in the sequence databases while only 28.6% of the
bryozoans had been previously sequenced for the barcoding regions used in this study (Table
S6). This clearly limits the ability of eDNA metabarcoding to fully document community
composition in the Arctic, highlighting the importance of improving sequencing effort for
particular taxa to fill the taxonomic gaps in available data bases.
Another salient observation of this study is that detected community structure differed
substantially between sampling methods with benthic communities being more variable
within and between ports and zooplankton communities being more similar within and
between ports. The broader range of biodiversity dissimilarities present among benthic
communities might be explained by the highly variable seabed characteristics, which has an
important role in megafaunal distribution as they impact several factors like larvae
settlement, anchorage and shelter (Kedra et al. 2013, Preez et al. 2016). In contrast,
zooplankton experience less variation in their habitat, due to the greater homogeneity of the
water column relative to benthic substrates (Angel 1993, Gray 1997). The structure of
37
communities detected with eDNA was intermediate between the patterns of communities
detected by the different traditional sampling approaches, representing more community
dissimilarities than plankton communities revealed by net tows but less community
dissimilarities than benthic communities detected by trawls, grabs and cores. This pattern is
likely due to the origin of eDNA, transport and degradation processes. The high prevalence
of meroplanktonic organisms (reflective of benthic communities) detected within eDNA
communities may explain why they display greater dissimilarity than plankton communities
depicted by species collections. On the other hand, eDNA communities likely display less
dissimilarity than benthic communities depicted by species collections due to the
homogenization of eDNA particles through movement in the water as well as their
degradation whereas, living specimens remain in the seabed or seafloor and are less affected
by water movement.
Our observations of distinct patterns of community structure depicted using either COI and
18S primer sets are consistent with several studies that have shown an effect of markers on
the detection rate of marine invertebrates (Drummond et al. 2015, Kelly et al. 2017, Elbrecht
et al. 2017, Djurhuus et al. 2018). Until a more universal primer set is available, this
highlights the importance of using a combination of different primer sets covering different
genomic regions for the time being. Here, our results suggested a greater affinity of COI
primers for annelids, arthropods and echinoderms relative to 18S primers, as previously
reported by Drummond et al. (2015). These affinities could potentially explain why Iqaluit
community composition detected using both COI and 18S clearly differed from Churchill
and Deception Bay communities, mainly because more annelids and echinoderms and less
arthropods taxa were detected in Iqaluit relative to the other two locations.
Despite the large number of taxa observed in this study, many marine invertebrates were
likely missed, as revealed by the rarefaction curves. This is especially true for the benthic
communities for which no plateau was reached. Coastal areas represent complex mosaics of
benthic habitats which contribute to diversifying the epifauna and possibly increasing the
possibility of missing taxa (Gray 1997). For eDNA sampling, the number of genera detected
can be influenced by the number of samples and their vertical and horizontal distributions
38
(Lacoursière-Roussel et al. 2018), type of filters, volume of filtered water, extraction method
(Deiner et al. 2018), sequencing depth and bioinformatics pipeline. Thus, a larger volume of
filtered seawater for each sample and a deeper sequencing depth would likely have improved
the detection rate in this study (Mächler et al. 2016). Similarly, a greater detection rate could
have been achieved by sampling a higher number of stations within each port. Although
eDNA rarefaction curves were very similar between Churchill and Deception Bay ports,
Iqaluit grew less rapidly at first and appeared closer to reaching a plateau than Churchill and
Deception Bay due to the lower alpha and gamma biodiversity measured within this port.
Furthermore, alpha and gamma biodiversity was greater within Iqaluit for the species
collection. This suggests that the inverse trend observed between the two approaches might
reflect decreased previous monitoring effort towards more northern regions which
consequently results in more incomplete sequence reference databases rather than a truly
lower biodiversity. Sequence reference data bases are estimated to contain only 13% of
marine species inhabiting the Arctic Ocean (Hardy et al. 2011), and a latitudinal gradient of
sequencing effort might exist within the Arctic itself. Indeed, we observed an increase in the
percentage of unknown OTUs from Churchill north to Iqaluit.
Transport and homogenization of eDNA
Knowledge pertaining to the spatial patterns of biodiversity is crucial for protecting regional
diversity and supporting conservation planning. The complex mosaic of benthic habitats
existing in Arctic coastal areas makes it more difficult to obtain a representative sampling of
this component of biodiversity. Our results revealed much lower beta indices for eDNA
communities compared to species collection communities which suggests that eDNA
depicted a more homogeneous pattern of species distribution than species collection in
coastal zones, as proposed in several studies of freshwater systems (Ficetola et al. 2008,
Dejean et al. 2011, Thomsen et a. 2012, Li et al. 2018).
Arctic coastal eDNA showed a more homogeneous community structure compared to species
collection but this pattern was affected by the spatial scale. Indeed, our results revealed a
39
significant relationship between the dissimilarities within eDNA communities as a function
of geographic distance which spans between 4 km to nearly 20 km. This is consistent with
classical spatial ecology processes whereby communities close to one another are in turn
more similar than those that are farther apart (Nekola and White 1999) and in line with the
observations of O’Donnell et al. (2016) of greater eDNA dispersion in nearshore marine
habitats. Several studies have also revealed patterns of extensive eDNA dispersion over
considerable distances within river systems (Deiner et al. 2016, Deiner and Altermatt 2014)
which could influence community structure in estuarine settings such as the port of Churchill.
In the case of our study, the very cold Arctic waters may contribute to reducing DNA
degradation, thus providing more time for dispersion over larger distances. This raises the
hypothesis that spatial eDNA homogenization should be more important in the Arctic Ocean
than more southern regions. It will therefore be of interest to evaluate in future studies how
latitude may influence patterns of eDNA biodiversity indices.
The weak correlation between dissimilarity and geographic distance in Iqaluit is in sharp
contrast with the other two ports in this study. This could potentially be explained by the
important tidal range occurring in this region; 7.5-11.7 m, which is over twice as great as in
Churchill (3.3-5.1 m) and in Deception Bay (3.6-5.7 m) (Fisheries and Oceans Canada,
2017). Interestingly, Churchill and Deception Bay ports showed significant distance
differences between their stations (Churchill: 0.2-7km, Deception Bay: 0.3-19km),
suggesting that the correlation between dissimilarity and distance might be consistent at
various spatial scales for marine invertebrates in Arctic coastal environments with similar
tidal conditions. Tidal effects on community composition were previously documented by
Kelly et al. (2018), but no studies have tested the effect of tides on eDNA homogenisation in
coastal environments. In contrast with eDNA results, where dissimilarity increased as a
function of geographic distance between stations, a decay in dissimilarity of communities
with distance was not systematically observed in species collections, which again may reflect
the fact that marine invertebrate communities are often characterised by a pronounced
patchiness distribution (Ministry of Environment 2006). Thus, the homogeneity of eDNA
distribution due to dispersion could potentially improve estimations of biodiversity at small
spatial scales. On the other hand, the dispersion and persistence of eDNA in coastal
40
environments also increases the risk of detecting organisms that are not actually present
locally (Deiner & Altermatt 2014; Jane and al. 2015). Further studies comparing the spatial
distribution of eDNA communities and corresponding species collection communities (either
benthos or plankton) in dynamic systems such as complex coastal areas are needed to
improve our knowledge about how the multiple physical and biological factors influence
eDNA distance decay. Such information will help to better inform eDNA sampling design
for monitoring and management issues.
Origins of eDNA
Benthic species with meroplanktonic life history type were in greater eDNA proportion than
species with strict benthic or pelagic life history. Our results sustain that coastal water eDNA
is a mixture of organic material released in the environment (e.g. feces, skin, mucus) and
plankton degradation and thus underlines the influence of variation in the life histories on
species detection probability. For instance, the fact that the discriminating taxa within the
eDNA and species collection approaches were not the same for the holoplankton and
meroplankton communities suggests that the different reproductive periods of the organisms,
as well as the associated planktonic larval stages, may influence the detection of certain taxa.
As a case in point, the daisy brittle star (Ophiopholis aculeata), the brittle star Ophiura
robusta and the green sea urchin (Strongylocentrotus droebachiensis) were discriminant
echinoderm species detected by eDNA and not by benthic species collection. Interestingly,
these three species are known to synchronise their spawning periods with sharp increases in
sea temperature (Himmelman et al. 2008), which typically occur during July within the
sampled ports (Prinsenberg 1984, Galbraith and Larouche 2011) suggesting that the high
number of sequences observed for those species could reflect the occurrence of these species
in their pelagic phase.
The importance of the planktonic stage in increasing eDNA detection is also supported by
the absence of DNA from amphipods, which were discriminant taxa in species collections
for meroplankton. The few studies on amphipod reproductive biology revealed that breeding
occurs during the spring in most species (Weslawski and Legezynska 2002). However,
41
amphipods represent a complex case as some species are benthic while other species are
planktonic and the two life history types coexist in the same environment. Sampling outside
breeding periods as well as the lack of planktonic stage could explain the lower detectability
of these organisms with eDNA. It is difficult to draw general patterns based on the life
histories of organisms since each species or genera in question might differ substantially, in
addition to the lack of knowledge for life histories of many marine invertebrates inhabiting
the Arctic and their reproduction period. O’Donnell et al. (2016) also concluded that
planktonic larval stages or release of pelagic eggs might play an important role in the eDNA
detection of some organisms. However, given the fact that seasonal factors greatly influence
the proportion of meroplankton or holoplankton organisms (Highfield et al. 2010, Lindeque
et al. 2013) and the ecology of eDNA (e.g. water temperature, UV exposition), further studies
on the detection of various marine invertebrate taxa at different times of the year would be
of great interest toward documenting how life histories of different organisms impact eDNA
detection.
Role of eDNA in Arctic conservation
Under the multiple environmental and anthropogenic factors currently threatening Arctic
coastal biodiversity and facing international objectives such as the protection of 10% of
coastal and marine areas by 2020 (SCBD, 2014), the development of rapid and efficient tools
for monitoring changes in biodiversity is essential. eDNA metabarcoding provides valuable
information towards a broader view of the taxonomic diversity which may help in developing
more rigorous conservation plans, particularly in the Arctic. In addition, this approach
provides numerous advantages due to its time-efficient and non-invasive manner (Deiner et
al. 2017). The simplicity of the sampling protocol for coastal water makes the method easy
to learn, which constitutes a major asset especially in remotes regions such as the Arctic
where it can be easily incorporated into existing sampling or community-based monitoring
programs (e.g., Lacoursière-Roussel et al. 2018). By combining the study of invertebrate
communities at different spatial scales detected by eDNA and species collection, this study
highlights important features related to the ecology of eDNA biodiversity indices such as the
origin of eDNA (i.e., planktonic phases of benthic taxa) and the effect of spatial
42
homogenisation. Taken together, our results suggest that eDNA diversity reflects complex
interactions between life histories of organisms and their spatial species. As public sequence
databases become more complete over time, species detection using eDNA metabarcoding
will improve, and is likely to help in better understanding a wide range of ecological
processes (plankton daily migration, fish seasonal migration, food web interactions, etc.)
where many elements remain undiscovered. Our results highlight that eDNA should be used
as a complementary approach for improving characterization of coastal biodiversity from
species collections as each method yielded distinct information on taxonomic composition
of the invertebrates inhabiting coastal areas.
43
Acknowledgement
This project was funded by ArcticNet, POLAR knowledge, DFO Aquatic Invasive Species
Monitoring Program, Nunavut Wildlife Management Board, Nunavik Marine Region
Wildlife Board. We thank Brian Boyle from IBIS for his expertise at the sequencing platform,
David Lodge, Kristy Deiner and Erin K Grey for their help and advice in terms of eDNA
metabarcoding methodologies and analyses, Jésica Goldsmit for her essential knowledge on
shipping activities and NIS in the arctic and sharing benthic core data from the Port of
Churchill. We thank Melania Cristescu and Guang Zhang for their help in primer selection
and Frederic Chain and Yiyuan Li for the pipeline development, Cecilia Hernandez for
laboratory assistance, and Jérôme Laroche from IBIS for the development of bioinformatics
pipelines. We would like to thank Laure de Montety, taxonomists and members of the
Archambault laboratory for the identification of the benthic organisms as well as for the
knowledge about them. We gratefully acknowledge the Churchill Northern Studies Centre
and Glencore Mine Raglan for providing access to their building facilities. We also thank the
following individuals for field assistance and participating in training: Frédéric Hartog,
Valérie Cypihot, LeeAnn Fishback, Daniel Gibson, Dick Hunter, Austin MacLeod, Thomas
Whittle, Rory McDonald, Frederic Lemire, Adamie Keatanik, Willie Keatanik, Willie Alaku,
Markusie Jaaka.
44
Annexe (Supporting information)
Table S1. Barque 1.5.1 specific commands for preparation and analysis of pair-end reads. The
sequences from the different COI and 18S primers set were added together.
Main step of filtration
Number of
eDNA reads in
Churchill port
Number of
eDNA reads in
Deception Bay
port
Number of
eDNA reads in
Iqaluit port
Number of raw forward and reverse reads 10 553 694 5 767 153 7 960 264
Remove raw reads with low quality 10 171 078 5 547 150 7 679 365
Remaining reads after merging forward and reverse
reads 9 635 596 5 136 684 7 255 383
Remove reads with incorrect length 8 535 538 4 600 181 6 405 199
Remove chimeric reads 8 170 548 4 340 911 6 008 858
Number of final reads 7 262 004 3 578 713 4 880 549
Number of reads with successful BLAST at genus
level 574 581 113 236 206 269
Number of reads with successful BLAST at species
level 478 046 95 651 203 245
45
Table S2. Chordata taxa present in the eDNA data set (COI and 18S primers set added together)
and the appropriate action taken.
Scientific_name Phylum Genus Subphylum Action
Ammodytes_marinus Chordata Ammodytes Fish Removed
Aplousobranchia Chordata Tunicate Remain
Ascidia Chordata Tunicate Remain
Ascidiacea Chordata Tunicate Remain
Ascidiacea_colonial Chordata Ascidiacea Tunicate Remain
Ascidiella Chordata Tunicate Remain
Boltenia_echinata Chordata Boltenia Tunicate Remain
Boltenia_villosa Chordata Boltenia Tunicate Remain
Boreogadus_saida Chordata Boreogadus Fish Removed
Brama_japonica Chordata Brama Fish Removed
Catostomus_catostomus Chordata Catostomus Fish Removed
Catostomus_commersonii Chordata Catostomus Fish Removed
Chelyosoma_macleayanum Chordata Chelyosoma Tunicate Remain
Chen_canagica Chordata Chen Bird Removed
Coregonus_artedi Chordata Coregonus Fish Removed
Coregonus_clupeaformis Chordata Coregonus Fish Removed
Cottidae Chordata Fish Removed
Delphinapterus_leucas Chordata Delphinapterus Beluga Removed
Didemnum_albidum Chordata Didemnum Tunicate Remain
Dendrodoa_grossularia Chordata Dendrodoa Tunicate Remain
Esox_lucius Chordata Esox Fish Removed
Eumesogrammus Chordata Eumesogrammus Fish Removed
Eumicrotremus Chordata Eumicrotremus Fish Removed
Fritillaria_borealis Chordata Fritillaria Tunicate Remain
Gadus_macrocephalus Chordata Gadus Fish Removed
Glossina_pallidipes Chordata Glossina Fly Removed
Gymnelus_viridis Chordata Gymnelus Fish Removed
Gymnocanthus_tricuspis Chordata Gymnocanthus Fish Removed
Halocynthia_pyriformis Chordata Halocynthia Tunicate Remain
Icelus Chordata Icelus Fish Removed
Icelus_bicornis Chordata Icelus Fish Removed
Larus_californicus Chordata Larus Bird Removed
Lateolabrax_maculatus Chordata Lateolabrax Fish Removed
Liparidae Chordata Fish Removed
Liparis_inquilinus Chordata Liparis Fish Removed
Lota_lota Chordata Lota Fish Removed
Lumpenus_lampretaeformis Chordata Lumpenus Fish Removed
Lycodes_mucosus Chordata Lycodes Fish Removed
Maccullochella_macquariensis Chordata Maccullochella Fish Removed
46
Mallatus_villosus Chordata Mallatus Fish Removed
Mallotus_catervarius Chordata Mallotus Fish Removed
Mallotus_villosus Chordata Mallotus Fish Removed
Molgula Chordata Molgula Tunicate Remain
Molgula_retortiformis Chordata Molgula Tunicate Remain
Molgulidae Chordata Tunicate Remain
Myoxocephalus Chordata Myoxocephalus Fish Removed
Myoxocephalus_quadricornis Chordata Myoxocephalus Fish Removed
Myoxocephalus_scorpius Chordata Myoxocephalus Fish Removed
Oikopleura Chordata Tunicate Remain
Osmerus_mordax Chordata Osmerus Fish Removed
Ovis_aries Chordata Ovis Sheep Removed
Pagophilus_groenlandicus Chordata Pagophilus Seal Removed
Pelonaia_corrugata Chordata Pelonaia Tunicate Remain
Perca_flavescens Chordata Perca Fish Removed
Pisces Chordata Fish Removed
Pungitius_pungitius Chordata Pungitius Fish Removed
Pusa_hispida Chordata Pusa Seal Removed
Rangifer_tarandus Chordata Rangifer Reindeer Removed
Rhinichthys_cataractae Chordata Rhinichthys Fish Removed
Salvelinus_fontinalis Chordata Salvelinus Fish Removed
Stichaeidae Chordata Fish Removed
Stichaeus_punctatus Chordata Stichaeus Fish Removed
Stichaeus_punctatus_punctatus Chordata Stichaeus Fish Removed
Styela_rustica Chordata Styela Tunicate Remain
Styela_gibbsii Chordata Styela Tunicate Remain
Styelidae Chordata Tunicate Remain
Triglops Chordata Fish Removed
47
Table S3. Genera found in the negative field controls and the appropriate action taken against the
contamination according to COI and 18S primers set.
Genus
Number
of all
field
control
sequences
Number
of all
sample
sequences
Percentage of
contamination
(%)
Contamination
approaches
Primers
set
Type of
animal/distribution
Alcyonidioides 11 158 7 removed 18S Bryozoan - marine
Alcyonidium 43 587 7 removed 18S Bryozoan - marine
Apistobranchus 1 9 11 removed 18S Worms - arctic marine
waters
Aurelia 232 295 79 removed 18S Jellyfish - marine
Calanus 3 24 123 removed 18S Zooplankton - marine
Eurytemora 443 5 8860 removed 18S Zooplankton - marine,
brackish, fresh waters
Halicephalobus 70 1 7000 removed 18S Nematode - parasite
Halichaetonotus 24 2 1200 removed 18S Gastrotrich - marine
Lampocteis 1 36 3 removed 18S
Comb jelly -
subtropical marine
waters
Larochella 36 49 734 removed 18S Gastropod - marine
Lumbricillus 65 47 138 removed 18S Worms - marine
Mysis 159 157 101 removed 18S Crustacean - marine
Mytilus 25 516 5 removed 18S Bivalve mollusks -
marine
Pseudocalanus 33696 87731 38 remain 18S Zooplankton - arctic
marine waters
Scalibregma 2 72 3 removed 18S Worms - arctic marine
waters
Zaus 18 734 3 removed 18S Crustacean - marine
Aphelochaeta 1 4226 0.02 remain 18S Worms - marine
Boroe 3 216 1 remain 18S Comb jelly - marine
Calycella 1 71 1 remain 18S Cnidarian -marine
Halichondria 1 705 0.1 remain 18S Sponge - marine
Mertensia 9 12352 0.1 remain 18S Comb jelly - marine
Nereis 1 318 0.3 remain 18S Worms - marine
Opercularella 1 806 0.1 remain 18S Cnidaria - marine
Pectinaria 1 30881 0.001 remain 18S Worms - marine
Phyllodoce 2 21422 0.01 remain 18S Worms - marine
Aurelia 1 41 2 removed COI Jellyfish - marine
Cerebratulus 8 6 133 removed COI Worms -arctic marine
waters
Gammarus 68 518 13 removed COI Crustacean - marine
48
Macrocyclops 34 0 removed COI Zooplankton -
freshwater
Macrophiothrix 1 39 3 removed COI Brittle star - marine
Tisbe 6 58 10 removed COI Crustacean - marine
Acartia 10 134600 0.01 remain COI Zooplankton - marine
Balanus 9 104141 0.01 remain COI Crustacean - marine
Clione 1 1638 0.1 remain COI Sea angel - cold marine
waters
Littorina 4 3693 0.1 remain COI Sea snails - marine
Microsetella 1 683 0.1 remain COI Zooplankton - marine
Mysis 1 80 1 remain COI Crustacean - marine
Mytilus 5 754 0.7 remain COI Bivalve mollusks -
marine
Nais 1 1839 0.1 remain COI Worms - brackish and
freshwater
Ophiura 1 10314 0.01 remain COI Brittle star - marine
Pectinaria 1 117866 0.001 remain COI Worms - marine
Phyllodoce 3 35675 0.01 remain COI Worms - marine
Pseudocalanus 70 131765 0.1 remain COI Zooplankton - arctic
marine waters
Scalibregma 2 296 0.7 remain COI Worms - arctic marine
waters
Thelepus 1 6861 0.01 remain COI Worms - arctic marine
waters
49
Table S4. Summary of PERMANOVA statistics tests on marine invertebrates communities for
the phylum relative abundance (number of taxa), Pielou evenness index and alpha richness. The
analyses were performed with method = "bray" for phylum relative abundance while it was
performed with method = "euclidian" for Pielou evenness and alpha richness.
Evaluated parameter Source of variation PERMANOVA
F-value R2 Pr (> f)
eDNA communities
composition Port 40.177 0.416 < 0.001
Churchill vs. Iqaluit 46.785 0.384 0.003
Churchill vs. Deception
Bay 34.180 0.313 0.003
Iqaluit vs. Deception Bay 38.371 0.335 0.003
Species collection
communities composition Port 6.706 0.078 < 0.001
Churchill vs. Iqaluit 8.258 0.067 0.003
Churchill vs. Deception
Bay 6.780 0.056 0.003
Iqaluit vs. Deception Bay 4.449 0.048 0.015
Phylum relative
abundance Method 43.708 0.337 < 0.001
Churchill Sp. collection vs eDNA 12.243 0.265 < 0.001
Deception Bay Sp. collection vs eDNA 38.305 0.615 < 0.001
Iqaluit Sp. collection vs eDNA 64.523 0.723 < 0.001
eDNA Pielou evenness
index Port 10.663 0.372 < 0.001
Churchill vs. Iqaluit 5.55 0.19 0.09
Churchill vs. Deception
Bay 5.356 0.182 0.087
Iqaluit vs. Deception Bay 20.431 0.460 0.003
Species collection Pielou
evenness index Port 1.9 0.08 0.2
Alpha richness Method
Churchill Sp. collection vs eDNA 77.471 0.695 < 0.001
Deception Bay Sp. collection vs eDNA 0.868 0.035 0.365
Iqaluit Sp. collection vs eDNA 53.641 0.691 < 0.001
50
Table S5. Summary of the correlation between dissimilarity and distance across the sites within
Churchill, Deception Bay and Iqaluit ports based on incidence data for the different sampling
methods.
Method Port Correlation test Simulated p-
value R2
eDNA
Churchill Cor.test < 0.001 0.13
Deception Bay Mantel.rtest 0.01 0.23
Iqaluit Mantel.rtest 0.02 0.094
Trawl, grab, cores Churchill Cor.test 0.5 0.004
Deception Bay Mantel.rtest 0.02 0.12 Iqaluit Mantel.rtest 0.01 0.14
Net tows
Churchill Cor.test 0.2 0.014
Deception Bay Mantel.rtest 0.01 0.26
Iqaluit Mantel.rtest 1.0 0.16
51
Table S6. Summary of the main phyla identified by sampling collection sampling methods
among Churchill, Deception Bay and Iqaluit ports and their respective presence in BOLD and
SILVA public genetic databases.
Phylum
Number of
organisms
identified with
trad. methods
Number of
identified
organisms
present in
BOLD
database
Number of
identified
organisms
present in
SILVA
database
Total number of
identified
organisms present
in genetic database
(BOLD or
SILVA)
Percentage of the
identified organisms
collected with trad.
methods available in
genetic database (BOLD
or SILVA)
Annelida 140 80 65 95 67.9
Arthropoda 167 89 43 94 56.3
Brachiopoda 1 0 1 1 100.0
Bryozoa 42 5 10 12 28.6
Cephalorhyncha 2 0 0 0 0.0
Chordata 19 13 6 14 73.7
Cnidaria 16 9 0 9 56.3
Echinodermata 9 7 2 7 77.8
Mollusca 77 37 28 42 54.5
Myzozoa 1 0 0 0 0.0
Porifera 2 0 0 0 0.0
52
Figure S1. The number of Operational taxonomic units (OTUs) assigned
and not assigned on NCBI to the genus level for the COI primers set
(COI1 and COI2) and the 18S primers set (Tareuk and 18S). The
assigned OTUs are represented by the grey section of the barplot while
the not assigned OTUs are represented by the red section of the barplot.
The pourcentage written in red represented the mean of the 4 primers for
not assigned OTUs in each port.
53
Figure S2. Boxplot on alpha diversity for Shannon biodiversity
index in Churchill, Deception Bay and Iqaluit ports. These analyses
were performed on abundance data with Hellinger transformation,
COI and 18S primer sets are added together for the eDNA boxplot.
54
eDNA amplification
Three PCR replicates were done for each sample and each primer set. In brief, the final PCR
mix of each sample replicate was composed of 12.5 µl Qiagen Multiplex Mastermix, 6.5 µl
diH2O, 1.0 µl of each primer (10 µM) and 3.0 µl of DNA. For all samples, the PCR mixture
underwent an initial denaturation step at 95°C for 15 minutes, followed by 35 cycles at 94°C
for 30 seconds, 54°C for 90 seconds (except for the COI2 primer set, which was at 52°C) and
72°C for 60 seconds) and a final elongation at 72°C for 10 minutes. Because barcodes were
different for each sample, a negative PCR control was done for each sample and primer to
ensure no false positive could happen. Following the PCR, each replicate and negative PCR
control were visualized on a 1.5% agarose gel electrophoresis. If no contamination was
visible, the three replicates of each sample were then pooled together. Pooled samples were
purified using Axygen PCR clean up kit following the manufacture’s recommended protocol
and libraries were quantified in equal molar concentrations using AccuClear Ultra High
Sensitivity dsDNA Quantitation Kit using the TECAN Spark 10M Reader.
55
Conclusion
À l’aube de cette 6ème extinction de masse, les diverses perturbations d’origine naturelle et
anthropique sévissant dans l’océan Arctique entraineront des changements majeurs au sein
de la biodiversité, particulièrement dans les régions sensibles telles que les zones côtières.
De bonnes banques de données sur la composition de même que la distribution actuelle des
espèces et communautés sont essentielles afin de pouvoir mieux orienter les programmes de
surveillance ainsi que les gestionnaires dans la création et la délimitation d’éventuelles aires
marines protégées.
L’étude présentée dans ce mémoire avait comme principal objectif d’acquérir de meilleures
connaissances sur l’utilisation du metabarcoding d’ADNe quant à la distribution des
invertébrés marins à différentes échelles spatiales dans les régions côtières de l’Arctique.
Bien que l’ADNe soit de plus en plus utilisé afin d’effectuer des relevés de biodiversité, peu
d’études ont été réalisées dans un environnement aussi extrême que l’Arctique, un milieu où
la basse température de l’eau est susceptible de diminuer la dégradation des molécules
d’ADN et ainsi jouer un rôle non négligeable dans le transport de l’ADNe. Ce rôle pourrait
entre autres être relié à une plus grande dispersion des molécules dans leur environnement.
De plus, à notre connaissance, aucune étude jusqu’à ce jour n’avait comparé la distribution
d’organismes documenté par collectes de spécimens à la distribution des organismes détectés
à l’aide de l’ADNe.
Le développement du séquençage nouvelle génération a permis de révolutionner les
domaines d’études reliés à la génomique ainsi que la biologie moléculaire en permettant le
séquençage rapide de l’ADN. Ainsi des approches telles que le métabarcoding ont pu voir le
jour et ont permis dans le cadre de ce projet la détection de plus d’une centaine d’organismes
marins vivant dans trois ports commerciaux de l’Arctique à partir de moins de 30 litres d’eau
de mer. Ces organismes ont révélé des communautés fort différentes tant au niveau local que
régional entre les ports à l’étude, mais ont aussi dévoilé des distinctions importantes entre les
approches d’échantillonnage. Entre autres, on observe que la dispersion de l’ADNe dans les
56
environnements à l’étude entraîne une réduction de la biodiversité beta obtenu
comparativement à la collecte d’espèces à l’aide de méthodes traditionnelles. Les résultats
obtenus dans cette étude permettent également d’observer le rôle capital des stades de vie
pélagiques dans l’origine de même que la détection de certains taxons avec l’ADNe.
ADNe vs. collecte d’espèces
Au-delà du nombre important d’organismes recensés dans le cadre de ce projet de maîtrise,
l’un des éléments capitaux ressortant de cette étude est la différence considérable entre les
communautés d’invertébrés marins détectées par l’ADNe et celles obtenues par la collecte
d’espèces. Ces résultats vont dans le même sens que ceux rapportés par Kelly et al. (2017),
appuyant la nature complémentaire de ces différentes approches afin d’effectuer des suivis
de biodiversité dans le cas des invertébrés.
Bien qu’au niveau de la diversité gamma, seulement entre 9 et 15% des organismes présent
dans cette étude furent obtenus à la fois avec l’ADNe ainsi que la collecte d’espèces, de
manière plus générale les mêmes phyla furent détectés par les deux approches. Toutefois,
l’abondance relative des phyla au sein de chacune des approches est tant qu’à elle grandement
variable, principalement pour les taxons les moins abondants. Globalement, les organismes
vivants accrochés aux composantes des fonds marins, tel que les échinodermes ou les
porifères, ou encore les organismes possédant des caractéristiques physiques difficilement
identifiable, tel que les némertes, ont davantage été recensés par l’ADNe que la collecte de
spécimens. D’un autre côté, certains taxons furent uniquement retrouvés par la collecte
d’espèces comme ce fut le cas pour les brachiopodes, les foraminifères ainsi que les
Cephalorhyncha et les Chaetognatha. Un second élément témoignant d’une différence non
négligeable entre les deux approches, est le contraste obtenu par rapport à la biodiversité beta
des communautés, révélant une homogénéisation plus grande au sein des communautés
d’ADNe. De par leur dispersion dans l’eau, les molécules d’ADNe permettent fort
probablement de diminuer les taux de changements entre les communautés se trouvant à
proximité, contrastant ainsi avec l’hétérogénéité de la distribution des organismes benthiques
57
grandement influencé par les caractéristiques environnementales telle que les types de fond
marin (Kedra et al. 2013, Preez et al. 2016). Il aurait été très intéressant de posséder des
données visuelles en ce qui attrait aux caractéristiques physiques des fonds marins des divers
sites échantillonnés afin de pouvoir mieux documenter les variations de diversité obtenues
au sein d’un port.
Origine, détection et transport de l’ADNe
Les résultats obtenus dans le cadre de ce projet de maîtrise abondent dans la même direction
que ceux obtenus par O’Donnell et al (2016), suggérant un rôle crucial des phases de vie
pélagiques dans l’origine de l’ADNe. De ce fait, la période de reproduction se trouve à être
un facteur important à considérer lors de campagnes d’échantillonnage avec l’ADNe, puisque
l’abondance de gamètes, d’œufs ou de larves dans la colonne d’eau affectera grandement les
probabilités de détections de certains organismes. Dans un contexte où les changements
climatiques sont susceptibles de faire varier les périodes de reproductions de plusieurs
espèces benthiques (Birchenough et al. 2011), l’utilisation de l’ADNe pourrait devenir un
atout considérable dans le domaine de la phénologie.
Le transport des molécules d’ADNe de même que leur persistance sur de longues distances
est un phénomène bien documenté dans le cadre d’étude portant sur les systèmes lotiques
(Deiner et al. 2016, Deiner and Altermatt 2014). Bien qu’en milieu côtier, une étude de Kelly
et al. (2018) c’est attardé aux effets possibles des marées sur la composition des
communautés, les conséquences plausibles de différents marnages sur la
dispersion/l’homogénéisation de l’ADNe demeurent relativement inexplorées. Alors que la
distance entre les divers sites présents au sein des ports de Churchill et Iqaluit varie
considérablement (Churchill : 0.2-7km et Baie Déception : 0.3-19km), les coefficients de
corrélations obtenus entre la dissimilarité des paires de sites en fonction de la distance entre
ceux-ci sont relativement similaires de même que très faibles. Il est d’autant plus intéressant
de constater que le marnage de ces deux ports est sensiblement le même (Churchill : 3.3-
5.1m et Baie Déception : 3.6-5.7 m) alors que dans le cas d’Iqaluit, où la corrélation entre la
58
dissimilarité et la distance est extrêmement faible, on retrouve des marées avec un marnage
presque doublement supérieur avec une amplitude allant de 7.5 à 11.7 m. La dispersion et
l’homogénéisation de l’ADNe dans ces milieux n’est peut-être donc pas seulement influencer
par la température de l’eau qui diminue la dégradation des molécules, mais aussi par des
phénomènes liés aux environnements côtiers tel que le marnage. D’autres études avec un plus
grand nombre de milieux côtiers seraient dès lors nécessaires afin d’en apprendre davantage
sur ce phénomène.
Faiblesses et limites du métabarcoding d’ADN environnemental
Bien que le metabarcoding d’ADNe donne lieu à l’identification rapide de centaines
d’organismes et représente ainsi une approche fort prometteuse, il n’en demeure pas moins
que certaines limitations existent au sein de cette approche moléculaire et il est essentiel de
saisir la portée de celles-ci sur les résultats obtenus et leur interprétation.
Une limite importante du métabarcoding d’ADNe est son lien étroit avec les bases de données
génétiques publiques, dont l’identification des séquences obtenues suite au séquençage est
entièrement dépendant de leur présence parmi ces bases de données. De ce fait, on estime
que seulement un très faible pourcentage des organismes de l’Arctique y sont répertoriés,
dont environ 50% des organismes benthiques connu des milieux côtiers (Lacoursière-Roussel
et al. 2018). Encore plus critique, seulement 13% des espèces marines de l’Arctique seraient
présentes dans BOLD (Hardy et al. 2011). Toutefois, bien que cette dépendance avec l’état
des bases de données publiques représente actuellement une limite, la disponibilité des
séquences obtenues dans le cadre d’une étude constitue un atout considérable. En effet, de
nouvelles correspondances entre les séquences d’une étude et les bases de données publiques
nécessite peu de temps et peuvent continuellement être mis à jour au fur et à mesure que les
bases de données s’enrichiront.
Pouvoir détecter des organismes sans avoir à les observer ou les manipuler directement
représente un atout majeur, mais possède également ces désavantages puisque les risques de
59
contamination sont omniprésents, et ce jusqu’à la toute fin des manipulations. En effet, outre
les contaminations pouvant survenir lors de l’échantillonnage sur le terrain de même que lors
des manipulations en laboratoire, certaines erreurs peuvent également survenir lors du
séquençage des échantillons. Ce phénomène, appelé "tag jumps" est apte à survenir lors du
séquençage sur les plateformes Illumina en associant des séquences d’ADN aux mauvais
échantillons en raison d’un décalage dans les lectures des barcodes. Ceci peut mener à
l’introduction de mauvaises identifications et par le fait même augmenter faussement la
diversité présente dans les échantillons concernés (Schnell et al. 2015). Bien que ce
phénomène soit généralement peu fréquent, certaines précautions furent prises afin de limiter
au maximum ces mauvais assignements entre les séquences et les échantillons. Entre autres,
les échantillons furent séquencés séparément en fonction de leur provenance respective
(Churchill, Baie Déception et Iqaluit) en addition à l’utilisation d’une plus grande quantité
du contrôle de calibration d’Illumina (PhiX Control) afin d’augmenter la qualité et le niveau
de confiance du séquençage.
Applications et perspectives futures pour la protection de la biodiversité
Le projet de recherche mis de l’avant dans ce mémoire, bien que présentant des notions de
nature davantage théorique qu’appliqué, sera d’une grande utilité pour les plans de gestion et
de conservation futurs. Il serait d’ailleurs très pertinent de réaliser une étude similaire avec
un plus grands nombres d’écosystèmes portuaires arctiques afin d’approfondir le rôle de
différentes variables environnementales telles que la température de l’eau et le marnage sur
la dispersion de l’ADNe. Étant donné l’importance que joue la sensibilisation dans la
protection de la biodiversité, la participation de plusieurs habitants de Churchill, Baie
Déception et Iqaluit lors des campagnes d’échantillonnage de cette étude, est d’un grand
intérêt. La simplicité de la méthode d’échantillonnage utilisée sur le terrain dans le cadre de
ce projet constitue d’ailleurs l’une des forces majeures de l’approche puisqu’elle peut être
effectuée par un large éventail de personnes, ne nécessitant que quelques connaissances
rudimentaires sur le sujet. Outre les efforts de sensibilisation, l’utilisation de divers indices
de biodiversité dans le cadre de cette étude, a permis de soulever de nombreuses pistes de
60
réflexion sur la dispersion de l’ADNe ainsi que sur les facteurs abiotiques et biotiques
influençant sa détection, éléments à ne pas négliger avec l’utilisation de l’ADNe. De cette
façon, des programmes d’échantillonnage plus adaptés pourront être mis sur pied par les
gestionnaires de la faune dans le but de répondre à des problématiques concrètes, que ce soit
pour la détection d’espèce précises ou encore afin d’évaluer la biodiversité d’un secteur pour
une éventuelle délimitation d’une aire protégée. Alors qu’un nombre important d’espèces
exotiques envahissantes se propagent lors de leur phase pélagique, qui sont souvent
taxonomiquement difficile à identifier à ces stades de développement, l’utilisation de l’ADNe
représente une approche fort attrayante. Cette étude met également bien en perspective
l’intérêt qu’il sera nécessaire de porter à la variation temporelle étant donné l’importance que
pourrait jouer les stades de vie pélagique dans l’origine et la détection de l’ADNe chez
certains invertébrés marins et puisque ces stades de vie sont la plupart du temps intimement
lié aux saisons. Pour conclure, alors que la diversité fonctionnelle prend de plus en plus
d’ampleur dans le domaine de l’écologie, le metabarcoding d’ADNe représente une avenue
intéressante et jusqu’ici inexploitée, qui permettrait d’obtenir des informations cruciales sur
les conséquences possibles quant à l’ajout ou la suppression de certains organismes dans les
réseaux trophiques et à plus large échelle dans les écosystèmes.
61
Bibliographie
ACIA. (2004). Impacts of a warming arctic; Arctic Climate Impact Assessment. Cambridge
University Press.
Angel, M. V. (1993). Biodiversity of the Pelagic Ocean. Conservation Biology, 7(4), 760–
772.
Archambault, P., Snelgrove, P. V. R., Fisher, J. A. D., Gagnon, J., David, J., Harvey, M., …
Pepin, P. (2010). From Sea to Sea : Canada ’ s Three Oceans of Biodiversity. PLoS
ONE, 5(8). http://doi.org/10.1371/journal.pone.0012182
Baldwin, C. C., Collette, B. B., Smith, D. G., & Springer, V. G. (1996). Collecting Fishes. In
Methods and Techniques of Underwater Research.
Barnes, M. A., & Turner, C. R. (2016). The ecology of environmental DNA and implications
for conservation genetics. Conservation Genetics, 17(1), 1–17.
http://doi.org/10.1007/s10592-015-0775-4
Bell, J. J., & Barnes, D. K. A. (2003). Effect of Disturbance on Assemblages: An Example
Using Porifera. Biological Bulletin, 205(2), 144–159. http://doi.org/10.2307/1543235
Bivand, R. S., Pebesma, E. J., & Gomez-Rubio, V. (2008). Applied Spatial Data Analysis
with R. Springer. http://doi.org/10.1007/978-0-387-78171-6
Birchenough, S. N. R., Degraer, S., Reiss, H., Borja, Á., Braeckman, U., Craeymeersch, J.,
… Wätjen, K. (2011). 8 Responses of marine benthos to climate change. ICES status
report on climate change in the North Atlantic. p.123–146
Bluhm, B. A., & Gradinger, R. (2008). Regional variability in food availability for arctic
marine mammals. Ecological Applications, 18(2) Supp(2), S77–S96.
http://doi.org/http://dx.doi.org/10.1890/06-0562.1
Bohmann, K., Evans, A., Gilbert, M. T. P., Carvalho, G. R., Creer, S., Knapp, M., … Bruyn,
M. De. (2014). Environmental DNA for wildlife biology and biodiversity monitoring.
Trends in Ecology & Evolution, 29(6), 358–367.
http://doi.org/10.1016/j.tree.2014.04.003
Bolger, A. M., Lohse, M., & Usadel, B. (2014). Trimmomatic: A flexible trimmer for
Illumina sequence data. Bioinformatics, 30(15), 2114–2120.
http://doi.org/10.1093/bioinformatics/btu170
Brown, E. A., Chain, F. J. J., Craese, T. J., MacIsaac, H. J., & Cristescu, M. E. (2015).
Divergence thresholds and divergent biodiversity estimates : can metabarcoding reliably
62
describe zooplankton communities ? Ecology and Evolution, 5(11), 2234 – 2251.
http://doi.org/10.1002/ece3.1485
CAFF International Secreteriat (2010). Arctic Biodiversity Trends 2010 – Selected indicators
of change. Islande.
Carlton, J. T. (1996). Biological Invasions and Cryptogenic Species. Ecology, 77(6), 1653–
1655.
Cardinale, B. J., Duffy, J. E., Gonzalez, A., Hooper, D. U., Perrings, C., Venail, P., …
Naeem, S. (2012). Biodiversity loss and its impact on humanity. Nature, 486(7401), 59–
67. http://doi.org/10.1038/nature11148
Carugati, L., Corinaldesi, C., Dell’Anno, A., & Danovaro, R. (2015). Metagenetic tools for
the census of marine meiofaunal biodiversity: An overview. Marine Genomics, 24, 11–
20. http://doi.org/10.1016/j.margen.2015.04.010
Casas-Monroy, O., Linley, R.D., Adams, J.K., Chan, F.T., Drake, D.A.R. & Bailey, S.A.
(2014). National Risk Assessment for Introduction of Aquatic Nonindigenous Species
to Canada by Ballast Water. In, p. vi + 73 p. Canadian science Advisory Secretariat
(CSAS)
Chan, F. T., Bailey, S. A., Wiley, C. J., & MacIsaac, H. J. (2013). Relative risk assessment
for ballast-mediated invasions at Canadian Arctic ports. Biological Invasions, 15(2),
295–308. http://doi.org/10.1007/s10530-012-0284-z
Chan, F.T., Bronnenhuber, J.E., Bradie, J.N., Howland, K.L., Simard, N. & Bailey, S.A.
(2012). Risk Assessment for ship-mediated introductions of aquatic nonindigenous
species to the Canadian Arctic. In, p. 93. Fisheries and Oceans Canada, Canada.
Chapman, M. G. (2003). The use of sandstone blocks to test hypotheses about colonization
of intertidal boulders. Journal of the Marine Biological Association of the United
Kingdom, 83(2), 415–423. http://doi.org/10.1017/S0025315403007276h
Clarke, K.R., Warwick, R.M. (2001). Changes in communities: an approach to statistical
analysis and interpretation, 2nd edition. PRIMER-E: Plymouth
Chiarucci, A., Bacaro, G., & Scheiner, S. M. (2011). Old and new challenges in using species
diversity for assessing biodiversity. Philosophical Transactions of the Royal Society B:
Biological Sciences, 366(1576), 2426–2437. http://doi.org/10.1098/rstb.2011.0065
Coissac, E., Riaz, T., & Puillandre, N. (2012). Bioinformatic challenges for DNA
metabarcoding of plants and animals. Molecular Ecology, 21(8), 1834–1847.
http://doi.org/10.1111/j.1365-294X.2012.05550.x
Creer, S., Deiner, K., Frey, S., Porazinska, D., Taberlet, P., Thomas, W. K., … Bik, H. M.
(2016). The ecologist’s field guide to sequence-based identification of biodiversity.
63
Methods in Ecology and Evolution, 7(9), 1008–1018. http://doi.org/10.1111/2041-
210X.12574
Cusson, M., Archambault, P., & Aitken, A. (2007). Biodiversity of benthic assemblages on
the Arctic continental shelf : historical data from Canada. Marine Ecology Progress
Series, 331, 291–304.
Darnis, G., Robert, D., Pomerleau, C., Link, H., Archambault, P., Nelson, R. J., … Fortier,
L. (2012). Current state and trends in Canadian Arctic marine ecosystems: II.
Heterotrophic food web, pelagic-benthic coupling, and biodiversity. Climatic Change,
115(1), 179–205. http://doi.org/10.1007/s10584-012-0483-8
Deagle, B. E., Jarman, S. N., Coissac, E., Pompanon, F., Taberlet, P., Deagle, B. E., …
Coissac, E. (2014). DNA metabarcoding and the cytochrome c oxidase subunit I
marker : not a perfect match. Biol. Lett., 10, 2–5.
Deiner, K., & Altermatt, F. (2014). Transport Distance of Invertebrate Environmental DNA
in a Natural River. PLoS ONE, 9(2). http://doi.org/10.1371/journal.pone.0088786
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. http://doi.org/10.1038/ncomms12544
Deiner, K., Bik, H. M., Mächler, E., Seymour, M., Lacoursière-Roussel, A., Altermatt, F.,
… Bernatchez, L. (2017). Environmental DNA metabarcoding: Transforming how we
survey animal and plant communities. Molecular Ecology, 26(21), 5872–5895.
http://doi.org/10.1111/mec.14350
Deiner, K., Lopez, J., Bourne, S., Holman, L. E., Seymour, M., Grey, E. K., … Lodge, D. M.
(2018). Optimising the detection of marine taxonomic richness using environmental
DNA metabarcoding : the effects of filter material , pore size and extraction method.
Metabarcoding and Metagenomics, 1–15. http://doi.org/10.3897/mbmg.2.28963
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(8), 8–11. http://doi.org/10.1371/journal.pone.0023398
Dispas, A. (2019). Étude de référence sur la biodiversité du mésozooplancton dans quatre
ports de l’Arctique canadien en vue d’une augmentation de l’activité maritime, de
l’exploitation des ressources et du réchauffement climatique. Université du Québec à
Rimouski, Québec (Qc), Canada. 146 p.
Djurhuus, A., Pitz, K., Sawaya, N. A., Rojas-Márquez, J., Michaud, B., Montes, E., …
Breitbart, M. (2018). Evaluation of marine zooplankton community structure through
environmental DNA metabarcoding. Limnology and Oceanography: Methods.
http://doi.org/10.1002/lom3.10237
64
Dray, S., & Dufour, A.-B. (2007). The ade4 Package: Implementing the Duality Diagram for
Ecologists. Journal of Statistical Software. Retrieved from
file:///C:/Users/Utilisateur/Downloads/v22i04.pdf
Drummond, A. J., Newcomb, R. D., Buckley, T. R., Xie, D., Dopheide, A., Potter, B. C. M.,
… Nelson, N. (2015). Evaluating a multigene environmental DNA approach for
biodiversity assessment. GigaScience. http://doi.org/10.1186/s13742-015-0086-1
Elbrecht, V., & Leese, F. (2015). Can DNA-based ecosystem assessments quantify species
abundance? Testing primer bias and biomass-sequence relationships with an innovative
metabarcoding protocol. PLoS ONE, 10(7), 1–16.
http://doi.org/10.1371/journal.pone.0130324
Elbrecht, V., Vamos, E. E., Meissner, K., Aroviita, J., & Leese, F. (2017). Assessing
strengths and weaknesses of DNA metabarcoding-based macroinvertebrate
identification for routine stream monitoring. Methods in Ecology and Evolution, 8(10),
1265–1275. http://doi.org/10.1111/2041-210X.12789
Ficetola, G. F., Miaud, C., Pompanon, F., & Taberlet, P. (2008). Species detection using
environmental DNA from water samples. Biology Letters, 4(4), 423–425.
http://doi.org/10.1098/rsbl.2008.0118
Fisheries and Oceans Canada. (2018). Canadian Tide and Current Tables. Retrieved
November 30, 2018, from http://www.charts.gc.ca/publications/tables-eng.asp
Folmer, O., Black, M., Hoeh, W., Lutz, R., & Vrijenhoek, R. (1994). DNA primers for
amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan
invertebrates. Molecular Marine Biology and Biotechnology, 3(5), 294–299.
http://doi.org/10.1371/journal.pone.0013102
Gajbhiye, S. N. (2002). Zooplankton study methods,Importance and significant observations.
In Proc. The National Seminar on Creeks, Estuaries and Mangroves - Pollution and
Conservation (pp. 21–27).
Galbraith, P. S., & Larouche, P. (2011). Sea-surface temperature in Hudson Bay and Hudson
Strait in relation to air temperature and ice cover breakup , 1985 – 2009. Journal of
Marine Systems, 87(1), 66–78. http://doi.org/10.1016/j.jmarsys.2011.03.002
Gaston, K. J. (2000). Global patterns in biodiversity. Nature, 405(6783), 220–7.
http://doi.org/10.1038/35012228
Gavrilchuk, K., and Lesage, V. (2014). Large-scale marine development projects (mineral,
oil and gas, infrastructure) proposed for Canada’s North. Can. Tech. Rep. Fish. Aquat.
Sci. 3069: viii + 84 pp.
Geller, J., Meyer, C., Parker, M., & Hawk, H. (2013). Redesign of PCR primers for
mitochondrial cytochrome c oxidase subunit I for marine invertebrates and application
65
in all-taxa biotic surveys. Molecular Ecology Resources, 13(5), 851–861.
http://doi.org/10.1111/1755-0998.12138
Goldberg, C. S., Strickler, K. M., & Pilliod, D. S. (2015). Moving environmental DNA
methods from concept to practice for monitoring aquatic macroorganisms. Biological
Conservation, 183, 1–3. http://doi.org/10.1016/j.biocon.2014.11.040
Goldsmit, J., Archambault, P., & Howland, K. L. (2014). Establishing a baseline for early
detection of non-indigenous species in ports of the Canadian Arctic. Aquatic Invasions,
9(3), 327–342. http://doi.org/10.3391/ai.2014.9.3.08
Goldsmit, J. (2016). Benthic Non-Indigenous Species in Ports of the Canadian Arctic:
Identification, Biodiversity and Relationships with Global Warming and Shipping
Activity. Doctoral Thesis. Université du Québec à Rimouski, Québec (Qc), Canada. 251
p.
Goldsmit, J., Archambault, P., Chust, G., Villarino, E., Liu, G., Lukovich, J. V, Howland, K.
L. (2017). Projecting present and future habitat suitability of ship- mediated aquatic
invasive species in the Canadian Arctic. Biological Invasions, 20(2), 501–517.
http://doi.org/10.1007/s10530-017-1553-7
Goldsmit, J., McKindsey, C. W., Archambault P., Howland, K. L. (2019) Ecological risk
assessment of predicted marine invasions in the Canadian Arctic. PLoS ONE
Gradinger, R., Bluhm, B. a, Hopcroft, R. R., Andrey, V., Sirenko, B., & W, J. M. (2010). 10.
Marine Life in the Arctic. Life in the World’s Oceans: Diversity, Distribution, and
Abundance, 183–202. http://doi.org/10.1002/9781444325508.ch10
Gray, J. (2000). The measurement of marine species diversity, with an application to the
benthic fauna of the Norwegian continental shelf. Journal of Experimental Marine
Biology and Ecology, 250(1-2), 23–49. http://doi.org/10.1016/S0022-0981(00)00178-7
Gray, J. S. (1997). Marine biodiversity: Patterns, threats and conservation needs. Biodiversity
and Conservation, 6(1), 153–175. http://doi.org/10.1023/A:1018335901847
Grey, E. K., Bernatchez, L., Cassey, P., Deiner, K., Deveney, M., Howland, K. L., … Lodge,
D. M. (2018). Effects of sampling effort on biodiversity patterns estimated from
environmental DNA metabarcoding surveys. Scientific Reports, 8(1), 2–11.
http://doi.org/10.1038/s41598-018-27048-2
Guyot, M., Dickson, C., Paci, C., Furgal, C., & Chan, H. M. (2006). Local observations of
climate change and impacts on traditional food security in two northern Aboriginal
communities. International Journal of Circumpolar Health, 65(5), 403–415.
http://doi.org/10.3402/ijch.v65i5.18135
Hadziavdic, K., Lekang, K., Lanzen, A., Jonassen, I., Thompson, E. M., & Troedsson, C.
(2014). Characterization of the 18s rRNA gene for designing universal eukaryote
66
specific primers. PLoS ONE, 9(2). http://doi.org/10.1371/journal.pone.0087624
Hamilton, A. J. (2005). Species diversity or biodiversity ?, 75, 89–92.
http://doi.org/10.1016/j.jenvman.2004.11.012
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, (August 2017), 1–18.
http://doi.org/10.1111/faf.12286
Hardy, S. M., Carr, C. M., Hardman, M., Steinke, D., Corstorphine, E., & Mah, C. (2011).
Biodiversity and phylogeography of Arctic marine fauna: Insights from molecular tools.
Marine Biodiversity, 41(1), 195–210. http://doi.org/10.1007/s12526-010-0056-x
Hebert, P. D. N., Cywinska, A., Ball, S. L., & DeWaard, J. R. (2003). Biological
identifications through DNA barcodes. Proc. R. Soc. Lond. B, 270, 313–321.
http://doi.org/10.1098/rspb.2002.2218
Highfield, J. M., Eloire, D., Conway, D. V. P., Lindeque, P. K., Attrill, M. J., & Somerfield,
P. J. (2010). Seasonal dynamics of meroplankton assemblages at station L4. Journal of
Plankton Research, 32(5), 681–691. http://doi.org/10.1093/plankt/fbp139
Himmelman, J. H., Dumont, C. P., Gaymer, C. F., Vallières, C., & Drolet, D. (2008).
Spawning synchrony and aggregative behaviour of cold-water echinoderms during
multi-species mass spawnings. Marine Ecology Progress Series, 361, 161–168.
http://doi.org/10.3354/meps07415
Hoegh-Guldberg, O., & Bruno, J. F. (2010). The Impact of Climate Change on the. Science,
328, 1523–1529. http://doi.org/10.1126/science.1189930
Intergovernmental Panel on Climate Change (IPCC) (2018). Summary for Policymakers.
Global warming of 1.5°C. An IPCC Special Report on the impacts of global warming
of 1.5°C above pre-industrial levels and related global greenhouse gas emission
pathways, in the context of strengthening the global response to the threat of climate
change, sustainable development, and efforts to eradicate poverty. World
Meteorological Organization, Geneva, Switzerland, 32 pp.
IOC/UNESCO. (2010). Why monitor the Arctic Ocean? Services to society from sustained
ocean observing system. Retrieved from:
http://unesdoc.unesco.org/images/0018/001898/189843e.pdf
Jabr, N., Archambault, P., Cameron, C. (2018). Biogeography and adaptations of
torquaratorid acorn worms (Hemichordata: Enteropneusta) including two new species
from the Canadian Arctic. Canadian Journal of Zoology. 96 (11), 1221-1229
James T. Carlton, J. B. G. (1993). Ecological Roulette : The Global Transport of
Nonindigenous Marine Organisms. Science, 261, 78–82.
67
http://doi.org/10.1126/science.261.5117.78
Jane, S. F., Wilcox, T. M., Mckelvey, K. S., Young, M. K., Schwartz, M. K., Lowe, W. H.,
… Whiteley, A. R. (2015). Distance, flow and PCR inhibition: EDNA dynamics in two
headwater streams. Molecular Ecology Resources, 15(1), 216–227.
http://doi.org/10.1111/1755-0998.12285
Jarosław, K., Mioduchowska, M., & Petković, M. (2016). Trying to solve current issues with
invertebrate taxonomy – the conceptual web-based application. World Scientific News,
57, 664–673.
Ji, Y., Ashton, L., Pedley, S. M., Edwards, D. P., Tang, Y., Nakamura, A., … Yu, D. W.
(2013). Reliable, verifiable and efficient monitoring of biodiversity via metabarcoding.
Ecology Letters, 16(10), 1245–1257. http://doi.org/10.1111/ele.12162
Jones, J. B. (1992). Environmental impact of trawling on the seabed: A review. New Zealand
Journal of Marine and Freshwater Research, 26(1), 59–67.
http://doi.org/10.1080/00288330.1992.9516500
Jørgensen, L. L., Archambault, P., Piepenburg, D., & Rice, J. (2016). Arctic marine
biodiversity. Oceans and Law of the Sea (Vol. 24).
http://doi.org/10.5670/oceanog.2011.75
Josefson, A. B., Mokievsky, V., Bergmann, M., Blicher, M. E., Bluhm, B., … Włodarska-
Kowalczuk, M. (2013). Marine invertebrates. Arctic Biodiversity Assessment. Status
and Trends in Arctic Biodiversity, 152–185.
Kedra, M., Renaud, P. E., Andrade, H., Goszczko, I., & Ambrose, W. G. (2013). Benthic
community structure, diversity, and productivity in the shallow Barents Sea bank
(Svalbard Bank). Marine Biology, 160(4), 805–819. http://doi.org/10.1007/s00227-012-
2135-y
Kelly, R. P., Closek, C. J., O’Donnell, J. L., Kralj, J. E., Shelton, A. O., & Samhouri, J. F.
(2017). Genetic and Manual Survey Methods Yield Different and Complementary
Views of an Ecosystem. Frontiers in Marine Science, 3(January), 1–11.
http://doi.org/10.3389/fmars.2016.00283
Kelly, R. P., Gallego, R., & Jacobs-Palmer, E. (2018). The effect of tides on nearshore
environmental DNA. PeerJ. http://doi.org/10.7287/peerj.preprints.3436v1
Kendall, M. A. (1996). Are Arctic soft sediment macrobenthic communities impoverished?
Polar Biology, 16(6), 393–399.
Keskin, E. (2014). Detection of invasive freshwater fish species using environmental DNA
survey. Biochemical Systematics and Ecology, 56, 68–74.
http://doi.org/10.1016/j.bse.2014.05.003
68
Kintisch, E. (2015). BOOM & BUST in the Great White North. Science, 349(6248), 578–
581.
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.
http://doi.org/10.1016/j.biocon.2014.11.020
Knowlton, N. (1993). Sibling species in the sea *. Annual Review of Ecology and Systematics,
24(1), 189–216. http://doi.org/10.1146/annurev.ecolsys.24.1.189
Kwong, S., Srivathsan, A., & Meier, R. (2012). An update on DNA barcoding : low species
coverage and numerous. Cladistics, 28, 639–644.
Lacoursière-Roussel, A., Howland, K., Normandeau, E., Grey, E. K., Archambault, P.,
Deiner, K., … Bernatchez, L. (2018). eDNA metabarcoding as a new surveillance
approach for coastal Arctic biodiversity. Ecology and Evolution, 1–15.
http://doi.org/10.1002/ece3.4213
Lacoursière-Roussel, A., Rosabal, M., & Bernatchez, L. (2016). Estimating fish abundance
and biomass from eDNA concentrations: variability among capture methods and
environmental conditions. Molecular Ecology Resources, n/a–n/a.
http://doi.org/10.1111/1755-0998.12522
Laurila-Pant, M., Lehikoinen, A., Uusitalo, L., & Venesjärvi, R. (2015). How to value
biodiversity in environmental management? Ecological Indicators, 55, 1–11.
http://doi.org/10.1016/j.ecolind.2015.02.034
Leray, M., Yang, J. Y., Meyer, C. P., Mills, S. C., Agudelo, N., Ranwez, V., … Machida, R.
J. (2013). A new versatile primer set targeting a short fragment of the mitochondrial
COI region for metabarcoding metazoan diversity: application for characterizing coral
reef fish gut contents. Frontiers in Zoology, 10(1), 34. http://doi.org/10.1186/1742-
9994-10-34
Li et al. (2018). Limited dispersion and quick degradation of environmental DNA in fish
ponds inferred by metabarcoding. BioRxiv. https://doi.org/10.1101/459321
Lindeque, P. K., Parry, H. E., Harmer, R. A., Somerfield, P. J., & Atkinson, A. (2013). Next
generation sequencing reveals the hidden diversity of zooplankton assemblages. PLoS
ONE, 8(11), 1–14. http://doi.org/10.1371/journal.pone.0081327
Link, H., Chaillou, G., Forest, A., Piepenburg, D., & Archambault, P. (2013). Multivariate
benthic ecosystem functioning in the Arctic-benthic fluxes explained by environmental
parameters in the southeastern Beaufort Sea. Biogeosciences, 10(9), 5911–5929.
http://doi.org/10.5194/bg-10-5911-2013
69
López, E., Olivier, F., Grant, C., Archambault, P. (2016). Four new records and a new species
of sedentary polychaetes from the Canadian High Arctic. JMBA UK.
dx.doi.org/10.1017/S0025315416000953
Loreau, M., & de Mazancourt, C. (2013). Biodiversity and ecosystem stability: A synthesis
of underlying mechanisms. Ecology Letters, 16(SUPPL.1), 106–115.
http://doi.org/10.1111/ele.12073
MacArthur, R. H. (1965). Patterns of species diversity. Biol. Rev, 40(4), 510–533.
Mächler, E., Deiner, K., Spahn, F., & Altermatt, F. (2016). Fishing in the Water: E ff ect of
Sampled Water Volume on Environmental DNA-Based Detection of
Macroinvertebrates. Environmental Science and Technology, 50(1), 305–312.
http://doi.org/10.1021/acs.est.5b04188
Magoč, T., & Salzberg, S. L. (2011). FLASH: Fast length adjustment of short reads to
improve genome assemblies. Bioinformatics, 27(21), 2957–2963.
http://doi.org/10.1093/bioinformatics/btr507
Marcus, N. H., & Boero, F. (1998). Minireview: The importance of benthic-pelagic coupling
and the forgotten role of life cycles in coastal aquatic systems. Limnology and
Oceanography, 43(5), 763–768. http://doi.org/10.4319/lo.1998.43.5.0763
McNeeley, S. M., & Shulski, M. D. (2011). Anatomy of a closing window: Vulnerability to
changing seasonality in Interior Alaska. Global Environmental Change, 21(2), 464–
473. http://doi.org/10.1016/j.gloenvcha.2011.02.003
Millennium Ecosystem Assessment. (2005). Ecosystems and Human Well-being: Synthesis.
Retrieved from www.islandpress.org
Ministère de l'Environnement. (2006). Guidelines for Sampling Benthic Invertabrates in
British Columbia Streams. Colombie Britanique.
Ministry of Environment. (2006). Guidelines for Sampling Benthic Invertabrates in British
Columbia Streams. British Columbia.
Mooney, H. A., & Cleland, E. E. (2001). The evolutionary impact of invasive species.
Proceedings of the National Academy of Sciences of the United States of America,
98(10), 5446–5451. http://doi.org/10.1073/pnas.091093398
Mori, A. S., Isbell, F., & Seidl, R. (2018). β-Diversity, Community Assembly, and Ecosystem
Functioning. Trends in Ecology and Evolution, 33(7), 549–564.
http://doi.org/10.1016/j.tree.2018.04.012
National Research Council. (1995). Arctic Ocean Research and Supporting Facilities:
National Needs and Goals. Washington, D.C.: National Academies Press. Retrieved
70
from http://www.nap.edu/catalog/9215
Nekola, J. C., & White, P. S. (1999). The distance decay of similarity in biogeography and
ecology. Journal of Biogeography, 26(4), 867–878. http://doi.org/10.1046/j.1365-
2699.1999.00305.x
O’Donnell, J.L., Kelly, R. P., Andrew, O. S., Jameal F., S., Lowell, N. C., & Gregory D., W.
(2016). Spatial distribution of environmental DNA in a nearshore marine habitat. PeerJ,
0–38. http://doi.org/10.7287/PEERJ.PREPRINTS.2608V1
Olivier, F., San Martín, G., & Archambault, P. (2013). A new species of Streptospinigera
Kudenov, 1983 (Polychaeta, Syllidae, Anoplosyllinae) from the Arctic and north-
western Atlantic with a key to all species of the genus. Polar Biology, 36(10), 1499–
1507. http://doi.org/10.1007/s00300-013-1369-6
PAME. (2016). Hudson Bay Complex LME [Fact sheet]. Retrieved from
https://www.pame.is/index.php/document-library/ecosystem-approach-to-
management-documents/large-marine-ecosystems/402-17-hudson-bay-lme/file
Payne, M.C., Reusser, D.A., and Lee, Henry, II. (2012). Moderate-resolution sea surface
temperature data and seasonal pattern analysis for the Arctic Ocean ecoregions: U.S.
Geological Survey Open-File Report 2011-1246, 20 p., available at
http://pubs.usgs.gov/of/2011/1246/.
Piepenburg, D., Archambault, P., Ambrose, W. G., Blanchard, A. L., Bluhm, B. A., Carroll,
M. L., … Włodarska-Kowalczuk, M. (2011). Towards a pan-Arctic inventory of the
species diversity of the macro- and megabenthic fauna of the Arctic shelf seas. Marine
Biodiversity, 41(1), 51–70. http://doi.org/10.1007/s12526-010-0059-7
Pimm, S. L. (2001). The World According to Pimm: a Scientist Audits the Earth. McGraw‐
Hill, New York.
Piola, R. F., Dafforn, K. a, & Johnston, E. L. (2009). The influence of antifouling practices
on marine invasions. Biofouling, 25(7), 633–644.
http://doi.org/10.1080/08927010903063065
Preez, C. Du, Curtis, J. M. R., & Clarke, M. E. (2016). The structure and distribution of
benthic communities on a shallow seamount (Cobb Seamount, Northeast Pacific
Ocean). PLoS ONE, 11(10), 1–29. http://doi.org/10.1371/journal.pone.0165513
Prinsenberg, S. J. (1984). Freshwater contents and heat budgets of James Bay and Hudson
Bay. Continental Shelf Research, 3(2), 191–200.
Radulovici, A. E., Archambault, P., & Dufresne, F. (2010). DNA barcodes for marine
biodiversity: Moving fast forward? Diversity, 2(4), 450–472.
http://doi.org/10.3390/d2040450
Rees, H. C., Maddison, B. C., Middleditch, D. J., Patmore, J. R. M., & Gough, K. C. (2014).
71
The detection of aquatic animal species using environmental DNA – a review of eDNA
as a survey tool in ecology. Journal of Applied Ecology, 51, 1450–1459.
http://doi.org/10.1111/1365-2664.12306
Robertson, D. R., & Smith-Vaniz, W. F. (2008). Rotenone: An Essential but Demonized Tool
for Assessing Marine Fish Diversity. BioScience, 58(2), 165.
http://doi.org/10.1641/B580211
Roff, J., & Mark, Z. (2011). Marine Conservation Ecology. (Earthscan, Ed.). Earthscan.
Retrieved from http://www.publish.csiro.au/pid/6958.htm
Rognes, T., Flouri, T., Nichols, B., Quince, C., & Mahé, F. (2016). VSEARCH: a versatile
open source tool for metagenomics. PeerJ, 4, e2584. http://doi.org/10.7717/peerj.2584
Roussel, J., Paillisson, J., Tréguier, A., & Petit, E. (2015). The downside of eDNA as a survey
tool in water bodies. Journal of Applied Ecology, (52), 823–826.
http://doi.org/10.1111/1365-2664.12428
Roy, V., Iken, K., & Archambault, P. (2015). Regional variability of megabenthic
community structure across the canadian arctic. Arctic, 68(2), 180–192.
http://doi.org/10.14430/arctic4486
Ruffilli, D.C. (2011). Arctic Marine and Intermodal Infrastructure: Challenges and the
Government of Canada's Response. Library of Parliament.
Ruiz, G. M., Carlton, J. T., Grosholz, E. D., & Hines, A. H. (1997). Global Invasions of
Marine and Estuarine Habitats by Non-Indigenous Species: Mechanisms, Extent, and
Consequences. American Zoologist, 37(6), 621–632.
http://doi.org/10.1093/icb/37.6.621
Sandlund, O. ., Hindar, K., & Brown, A. (1992). Conservation of biodiversity for sustainable
development. Scandinavian University Press. http://doi.org/10.1016/0921-
8009(93)90032-2
Schnell, I. B., Bohmann, K., & Gilbert, M. T. P. (2015). Tag jumps illuminated - reducing
sequence-to-sample misidentifications in metabarcoding studies. Molecular Ecology
Resources, 15(6), 1289–1303. http://doi.org/10.1111/1755-0998.12402
Settles, I. H., Cortina, L. M., Malley, J., & Stewart, A. J. (2006). The climate for women in
academic science: The good, the bad, and the changeable. Psychology of Women
Quarterly, 30(1), 47–58. http://doi.org/10.1111/j.1471-6402.2006.00261.x
Shapiro, B. (2008). Engineered polymerases amplify the potential of ancient DNA. Trends
in Biotechnology, 26(6), 285–287. http://doi.org/10.1016/j.tibtech.2008.03.005
Shokralla, S., Porter, T. M., Gibson, J. F., Dobosz, R., Janzen, D. H., Hallwachs, W., …
Hajibabaei, M. (2015). Massively parallel multiplex DNA sequencing for specimen
identification using an Illumina MiSeq platform. Scientific Reports, 5, 9687.
72
http://doi.org/10.1038/srep09687
Smart, A. S., Tingley, R., R. Weeks, A., R. Van Rooyen, A., & A. McCarthy, M. (2015).
Environmental DNA sampling is more sensitive than a traditional survey technique for
detecting an aquatic invader. Ecological Applications, 25(7), 1944–1952.
Smoot, C. A., & Hopcroft, R. R. (2017). Depth-stratified community structure of Beaufort
Sea slope zooplankton and its relations to water masses. Journal of Plankton Research,
39(1), 79–91. http://doi.org/10.1093/plankt/fbw087
Snelgrove, P.V.R. (2010) Discoveries of the Census of Marine Life: Making Ocean count.
Cambridge University Press. 270p.
Socolar, J. B., Gilroy, J. J., Kunin, W. E., & Edwards, D. P. (2015). How Should Beta-
Diversity Inform Biodiversity Conservation? Trends in Ecology & Evolution, 31(1), 67–
80. http://doi.org/10.1016/j.tree.2015.11.005
Spalding, M. D., Fox, H. E., Allen, G. R., Davidson, N., Ferdaña, Z. a., Finlayson, M., …
Robertson, J. (2007). Marine Ecoregions of the World: A Bioregionalization of Coastal
and Shelf Areas. BioScience, 57(7), 573. http://doi.org/10.1641/B570707
Stoeck, T., Bass, D., Nebel, M., Christen, R., Jones, M. D. M., Breiner, H.-W., & Richards,
T. A. (2010). Multiple marker parallel tag environmental DNA sequencing reveals a
highly complex eukaryotic community in marine anoxic water. Molecular Ecology, 19
(Suppl., 21–31.
Taberlet, P., Bonin, A., Zinger, L., Coissac, E. (2018). Environmental DNA for biodiversity
and monitoring. Oxford university press ISBN: 978-0-19-876722-0. 253 s.
Thiel, M., & Norenburg, J. (2009). Nemertea – Ribbon Worms. Bedim publications.
Retrieved from http://dx.doi.org/10.1038/npg.els.0001586
Thomsen, P. F., Kielgast, J., Iversen, L. L., Møller, P. R., Rasmussen, M., & Willerslev, E.
(2012). Detection of a Diverse Marine Fish Fauna Using Environmental DNA from
Seawater Samples. PLoS ONE, 7(8), 1–9. http://doi.org/10.1371/journal.pone.0041732
Thomsen, P. F., & Willerslev, E. (2015). Environmental DNA – An emerging tool in
conservation for monitoring past and present biodiversity. Biological Conservation,
183, 4–18. http://doi.org/10.1016/j.biocon.2014.11.019
UNEP (The United Nations Convention on Biological Diversity). (2006). Global
Biodiversity Outlook 2. http://doi.org/10.1093/aje/kwq338
Urban, M. C. (2015). Accelerating extinction risk from climate change. Science, 348(6234),
571–573. Retrieved from http://unfccc.int/2860.php
73
Van Dyke, F. (2010). Conservation Biology - Foundations, Concepts, Applications.
(Springer, Ed.) (Second Edi). Illinois. Retrieved from
http://www.springer.com/us/book/9781402068904
Van Oostdam, J., Donaldson, S. G., Feeley, M., Arnold, D., Ayotte, P., Bondy, G., … Kalhok,
S. (2005). Human health implications of environmental contaminants in Arctic Canada:
A review. Science of the Total Environment, 351-352, 165–246.
http://doi.org/10.1016/j.scitotenv.2005.03.034
Vermeij, G. J., & Roopnarine, P. D. (2008). Symmetric transporters for asymmetric transport.
Science, 321(5890), 781–782. http://doi.org/10.1126/science.1161495.Symmetric
Vié, J.-C., Hilton-taylor, C., & Stuart, S. N. (2009). Wildlife in a changing world. IUCN
Gland Switzerland. Retrieved from http://www.iucn.org/dbtw-wpd/html/RL-2009-
001/cover.html
Wei C-L, Cusson M, Archambault P, Belley R, Brown T, Burd BJ, Edinger E, Kenchington
E, Gilkinson K, Lawton P, Link H, Ramey-Balci PA, Scrosati RA, Snelgrove PVR
(submitted) Seafloor Biodiversity of Canada’s Three Oceans: patterns, hotspots, and
potential drivers. Diversity and Distributions
Weslawski, J. M., & Legezynska, J. (2002). Life cycles of some Arctic amphipods. Polish
Polar Research, 23(3), 253–264. http://doi.org/10.1016/j.tourman.2010.03.016
Williamson, M. (1999). Invasions. Ecography, 22(1), 5–12. http://doi.org/10.1111/j.1600-
0587.1999.tb00449.x
Yamamoto, S., Masuda, R., Sato, Y., Sado, T., Araki, H., Kondoh, M., … Miya, M. (2017).
Environmental DNA metabarcoding reveals local fish communities in a species-rich
coastal sea. Scientific Reports, 7(January), 40368. http://doi.org/10.1038/srep40368
Young, H. S., McCauley, D. J., Galetti, M., & Dirzo, R. (2016). Patterns, Causes, and
Consequences of Anthropocene Defaunation. Annual Review of Ecology, Evolution, and
Systematics, 47(1), 333–358. http://doi.org/10.1146/annurev-ecolsys-112414-054142