Using Genomics to Restore and Future-Proof Underwater Seaweed Forests

Using genomics to restore and future-proof underwater seaweed forests

Georgina V. Wood

A dissertation submitted in fulfillment of the requirements for the degree of Doctor of Philosophy

Supervised by Ezequiel Marzinelli and Adriana Vergés

Evolution & Ecology Research Centre The School of Biological, Earth and Environmental Sciences Faculty of Science UNSW Sydney

January 2020

Thesis/Dissertation Sheet

Surname/Family Name : WOOD Given Name/s : GEORGINA VALENTINE Abbreviation for degree as give in the University calendar : PhD Faculty : SCIENCE School : BIOLOGICAL, EARTH AND ENVIRONMENTAL SCIENCES USING GENOMICS TO RESTORE AND FUTURE-PROOF UNDERWATER Thesis Title : SEAWEED FORESTS

Abstract 350 words maximum: (PLEASE TYPE) Worldwide, foundation species are declining, leading to significant loss of biodiversity and ecosystem goods and services. With anthropogenic pressures predicted to continue to have major effects on foundation species, restoration is emerging as a key management tool to halt or reverse decline. Planning long-term restoration solutions that include population resilience under extant and future conditions as an explicit objective is of utmost importance; yet, it is still in its infancy for marine systems. In this thesis, I investigated approaches that can be used to develop and improve restoration and future-proofing strategies for declining underwater seaweed forests. A literature review identified the most significant challenges predicted to influence marine macrophytes into the next century. The incorporation of novel tools such as genomics emerged as being essential to combat these challenges. I applied genomics to design and experimentally assess the restoration of Phyllospora comosa, a foundational seaweed that has suffered historical declines along the coast of Sydney. Single Nucleotide Polymorphisms (SNPs) were used to characterise genetic diversity and structure on extant populations surrounding Phyllospora’s gap in distribution and subsequently inform restoration, via transplantation, of five Phyllospora populations. Although donor transplant provenance influenced survival of transplanted donors, recruitment was rapid and SNP analyses showed that genetic diversity and structure of the F1 generation successfully resembled extant populations. Landscape genomic tools were then used to characterise overall and potentially adaptive genetic diversity and structure along Phyllospora’s entire latitudinal range. Genetic diversity was unevenly distributed and putative loci under selection that linked to sea temperature were identified and may be useful to assist adaptation. Finally, the potential to harness important host-microbial interactions was investigated. A combination of interdisciplinary tools was used to demonstrate that the environment and host traits together explain 54% of variation in Phyllospora-associated microbial communities, with host genetics explaining half of this. Key genetic loci and phenotypic traits were strongly related to taxa with known associations to seaweed defense, disease and tissue degradation. Overall, my research demonstrates the capacity of genomic data to optimise intervention strategies for marine forests and paves the way for development of innovative solutions to prevent further degradation.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents a non-exclusive licence to archive and to make available (including to members of the public) my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known. I acknowledge that I retain all intellectual property rights which subsist in my thesis or dissertation, such as copyright and patent rights, subject to applicable law. I also retain the right to use all or part of my thesis or dissertation in future works (such as articles or books).

…………………………………………………………… ……….……………………...…….… Signature Date The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years can be made when submitting the final copies of your thesis to the UNSW Library. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research. ORIGINALITY STATEMENT

‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’

Signed ……………………………………………......

Date ……………………………………………...... INCLUSION OF PUBLICATIONS STATEMENT

UNSW is supportive of candidates publishing their research results during their candidature as detailed in the UNSW Thesis Examination Procedure.

Publications can be used in their thesis in lieu of a Chapter if: • The candidate contributed greater than 50% of the content in the publication and is the “primary author”, ie. the candidate was responsible primarily for the planning, execution and preparation of the work for publication • The candidate has approval to include the publication in their thesis in lieu of a Chapter from their supervisor and Postgraduate Coordinator. • The publication is not subject to any obligations or contractual agreements with a third party that would constrain its inclusion in the thesis

Please indicate whether this thesis contains published material or not:

This thesis contains no publications, either published or submitted for publication ☐ (if this box is checked, you may delete all the material on page 2)

Some of the work described in this thesis has been published and it has been documented in the relevant Chapters with acknowledgement ☒ (if this box is checked, you may delete all the material on page 2)

This thesis has publications (either published or submitted for publication) ☐ incorporated into it in lieu of a chapter and the details are presented below

CANDIDATE’S DECLARATION I declare that: • I have complied with the UNSW Thesis Examination Procedure • where I have used a publication in lieu of a Chapter, the listed publication(s) below meet(s) the requirements to be included in the thesis. Candidate’s Name Signature Date (dd/mm/yy)

‘I hereby grant the University of New South Wales or its agents a non-exclusive licence to archive and to make available (including to members of the public) my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known. I acknowledge that I retain all intellectual property rights which subsist in my thesis or dissertation, such as copyright and patent rights, subject to applicable law. I also retain the right to use all or part of my thesis or dissertation in future works (such as articles or books).’

‘For any substantial portions of copyright material used in this thesis, written permission for use has been obtained, or the copyright material is removed from the final public version of the thesis.’

Signed ……………………………………………......

Date ……………………………………………......

AUTHENTICITY STATEMENT ‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis.’

Signed ……………………………………………......

Date ……………………………………………...... Acknowledgements

All theses are the outcome of multiple architects and mine has had exceptionally many. First and foremost, I am enormously grateful to my wonderful supervisors. To Ziggy Marzinelli: your scientific rigour, good humour and level-headedness have carried me throughout this journey. To Adriana Vergés, I owe you many of the opportunities I have had to communicate my love for research and the ocean; thank you for continually supporting and inspiring me. To Mel Coleman, I will never forget you so graciously took me into your home during those initial months; thank you for teaching me how to unravel the secrets of tricky seaweed DNA extraction and for always being on the ball. Many thanks to Alex Campbell; it is largely due to you that I joined Operation Crayweed in the first place; our early days in the field have stuck in my mind. To Peter Steinberg, you have always stepped in at the right time with your wisdom. Thank you all for welcoming me into the crayweed family; you have motivated me greatly and I look forward to being able to continue working with you in the future.

I have received generous awards from Holsworth Research Fund, Evolution & Ecology Research Centre, School of BEES and UNSW that went towards genotyping costs, fieldwork equipment, conference costs and an international workshop. This thesis has greatly benefitted from their financial contributions.

Over the years I have had the privilege of working and speaking with a fantastic team of scientists. Many thanks are due to Damon Bolton – my crayweed partner in crime. What would this project have been without you! Thank you for all of the early morning coffees, latenight BBQs and laughs on the rock platform. I will be forever reminiscent of our intense fieldwork seasons. Thank you also to Tim Lynch, whose early encouragement sparked my PhD journey, as well as Bill Sherwin, Emma Johnston and Alistair Poore for valuable comments and papers shared. To the SIMS staff and folk at USYD, thank you for welcoming me into your spaces. I have also consistently been surrounded by a large group of hardworking, enthusiastic people. Thank you to the tens of volunteers who have helped me transplant, count fish and measure seaweed along the way. A huge shout out in particular to the Crayweed ‘A- Team’, Sofie Voerman and Gary Truong, as well as Sam Baxter – we shall soon have to get out for another dive! I owe a very special acknowledgement to Ryan and Mali for giving up their afternoons to volunteer every Tuesday; you both have bright futures ahead of you. To all of the awesome people in the marine labs – what a great crew we have! Special mention to Ruby Garthwin, Sophie Powell, Marc Uya, Juliana Ferrari, Sally Bracewell, Rosie Steinberg, Seb Vadillo Gonzalez, Aaron Eger, Janine Ledet, Brendan Lanham, Shannen Smith, Paula Sgarlatta, Giulia Ferretto and so many others. Thank you for being such great company during fieldwork, writing and also in non – PhD life.

Many thanks to my family – Mum, Dad and Florentine for your support and encouragement over the years. Finally, to Kingsley: you did everything I have mentioned above and so much more. Thank you for sharing this journey with me.

I love you all.

George, 2020. Table of Contents

List of tables………………………………….…………………………………………..……...……………………iv

List of figures………………………………….…………………….……….…….…………….…….……….……vii

List of abbreviations……………………………..……………………….…………….………..………..…..…xI

Chapter 1: General Introduction……………..………………………….…………………1

1.1 Restoration and future-proofing: the New Ecological Frontier…………………………….2

1.2 Novel techniques to restore and future-proof…………………………………………..…………3

1.3 Focusing on seaweeds as candidates for novel management efforts………….……….4

1.4 Thesis aims and outline………………………………………………………………….……………………6

1.5 Acknowledgement of co-authored papers…………………….………………………….…………9

Chapter 2: Restoring subtidal marine macrophytes in the Anthropocene: trajectories and future-proofing………………………………………….……….……..10

2.1 Abstract…………………………………………………………………………………..…………..………………11

2.2 Introduction…………………………………………………………….…..…………………….………………..12

2.3 Restoring into the future: for when, what and where do we restore?...... 17

2.4 Future-proofing: challenges ahead and new ecological solutions………….…….……….21

2.5 Moving forward: issues of management, political and public support.…….…………..31

2.6 Concluding remarks………………………………………….……………………………………………….…35

i Chapter 3: Using genetics to optimise and measure success in restoration of underwater forests…………………………….……………………………………………36

3.1 Abstract………………………………………………………………………………………………………………..37

3.2 Introduction…………………………………………………………..…………………………………………….38

3.3 Methods……………………………………………………..…………………………………………...………….40

3.4 Results…………………………………………………..………………………………………………….………….48

3.5 Discussion and conclusion……………………………………………..……………………….…….………57

Chapter 4: Marine forests on the cline: establishing a baseline for assisted evolution………………………………………………………………….…………………………62

4.1 Abstract……………………………………………………………………………………………………………….63

4.2 Introduction………………………………………………………………………..………………………………64

4.3 Methods………………………………………………………………………………………………………………67

4.4 Results…………………………………………………………………………………………………………………72

4.5 Discussion……………………………………………………………………………………………………………80

4.6 Conclusion…………………………………………………………………………………………………………..86

Chapter 5: The influence of host genetics, phenotype and geography on the microbiome of a foundational seaweed…………………………………………88

5.1 Abstract……………………………………………………………………………………………………………….89

5.2 Introduction…………………………………………………………………………………………………………90

5.3 Methods………………………………………………………………………………………….……..……………93

5.4 Results……………………………………………………………………………………………….………….……100

ii 5.5 Discussion………………………………….……………………………………………………………………….111

5.6 Conclusion………………………………………………………………………………….………………………117

Chapter 6: General Discussion……………………………………………………………118

6.1 Incorporating genomic data into restoration and future-proofing strategies for underwater forests.………………………………………….………………………………………………………120

6.2 The value of incorporating ecological knowledge………….…………………………...……..126

6.3 Limitations and future directions…………………………………………….………………….……..127

6.4 Conclusion………………………………………………….……………………………………………….……..133

Literature cited………………………………………………………………………………….134

Appendix A: Supplementary material to accompany Chapter 3………….179

Appendix B: Supplementary material to accompany Chapter 4……….…184

Appendix C: Supplementary material to accompany Chapter 5………….198

iii List of tables

3.1: Genetic diversity of Phyllospora comosa from the six extant sites surrounding Sydney………………………………………………………………………………..………..………..………………….52

3.2: Genetic diversity of Phyllospora from donor populations and restored F1 recruits……………………………………………………………………………………………………...………………55

4.1: Overall population structure (pairwise Fst) among extant Phyllospora comosa populations based on 109 loci * ………………………………………………………..………………………75

4.2: ANOVA output table for final dbRDA model, showing annual maximum sea surface temperature averaged over 26 years (Max_SST), average annual range in sea surface temperature (Range_SST) and average annual standard deviation in sea surface temperatures (SD_SST) experienced at 13 sites is associated with Phyllospora comosa’s genetic structure. The final model was selected via model selection based on adjusted p-values………………………………………………………………………………………………………………..……77

4.3: Overall genetic diversity of 13 Phyllospora comosa populations across its latitudinal range based on 109 loci………………………………………………………………………………………….….79

5.1: Mantel tests results describing relationship between (i) Phyllospora comosa genetic distance (Euclidean) and (ii) phenotypic distance (Euclidean) on surface-associated microbial community dissimilarities calculated using Bray-Curtis on square-root transformed relative abundances.*…………………………………………………………………………108

5.2: ANOVA output table for (i) initial and (ii) final dbRDA models, showing geographic, host phenotypic and host genetic associations with Phyllospora comosa’s associated microbial community structure. The final model was selected via model selection based on p-values, using a stepwise selection procedure……………………………………………………109

S3.1: Population structure (pairwise FST) among extant Phyllospora populations.*….182

iv S3.2: AMOVA between extant Phyllospora sites………………………………………….…………..183

S4.1: Proportions of males and female Phyllsopora comosa adults sampled at 13 sites………………………………………………………………………………………………………………………….185

S4.2: Phyllospora comosa private alleles detected across 13 sites ………………………………………………………………………………………………………………………………..186

S4.3: Observed heterozygosity estimates of Phyllsopora comosa adults sampled at 13 sites…….…………………………………………………………………………………………………………………..187

S4.4: Phyllospora comosa’s putatively adaptive loci identified with different methods…………………………………………………………………………………………………………………..189

S4.5.1: Population structure (pairwise Fst) among extant Phyllospora populations based on outlier loci identified using Bayescan (3)*……………………………………………………………190

S4.5.2: Population structure (pairwise Fst) among extant Phyllospora populations based on outlier loci (8) identified using PCadapt *……………………………………………………………191

S4.5.3: Population structure (pairwise Fst) among extant Phyllospora populations based on SST-associated loci (36) identified using lfmm *…………………………………………………..192

S4.6: Test statistics for pairwise comparisons of observed heterozygosity for Phyllsopora comosa adults sampled at 13 sites…………………………………………………………………………..193

S4.7.1: Genetic diversity of thirteen Phyllospora populations across its latitudinal range, based on outlier loci (3) identified using Bayescan……………………………………………………195

S4.7.2: Genetic diversity of thirteen Phyllospora populations across its latitudinal range, based outlier loci identified using PCAdapt (8)………………………………………………………….196

v S4.7.3: Genetic diversity of thirteen Phyllospora populations across its latitudinal range, based on SST-associated loci (36)……………………………………………………………………………..197

S5.2: Results of pairwise comparisons of dispersion in host genetic data between eight sites. *………………………………………………………………………………………………………………………202

S5.3: Adjusted p-values for pairwise comparisons of host phenotype data between eight sites.* ……………………………………………………………………………………………………………………..203

S5.4: Results of pairwise comparisons of dispersion in host phenotype data between eight sites. *……………………………………………………………………………………………………………..205

S5.5: Results of tests for differences between individual host phenotype data between 8 sites……………………………………………………………………………………………………………………….207

S5.6: Spearman correlation coefficient for all phenotypic traits measured……………………………………………..………………………………………………………………….208

S5.7: List of “core” microbial taxa. Each amplicon sequence variant (ASV) was present in all 156 individuals of the seaweed Phyllospora comosa sampled during a study spanning their entire latitudinal distribution……………………………………………………………………………209

S5.8: Adjusted p-values for pairwise comparisons of microbial community data between 8 sites, based on (a) Bray-Curtis dissimilarity matrix based on overall microbial data and (b) Jaccard distances based on overall microbial data.*……………………………………………210

S5.9: Results of pairwise comparisons of dispersion in microbial community data between 8 sites, based on (a) Bray-Curtis dissimilarity matrix based on overall microbial data and (b) Jaccard distances based on overall microbial data. *………………………………212

S5.10: Microbial taxa significantly related to geography, host phenotypic and genetic traits shown to influence overall microbial community, as identified using DEseq2…..213

vi List of figures

2.1: Incorporating future-proofing principles into restoration of subtidal marine macrophytes.(Conceptual diagram)……………………………………………………………………………16

3.1: (a) Genetic structure of Phyllospora comosa populations. (b) Photograph showing transplanted algae from the northern donor site ...... 42

3.2: Relationship between geographic distance (km) and genetic distance (pairwise FST) between Phyllospora comosa individuals collected at six extant sites………………………………………………………………………………………..……………….…………………53

3.3: Transplant survival and condition. (a) Survival and (b) epibiosis of adult Phyllospora comosa transplants at restoration sites (WH: Whale Beach, FW: Freshwater, SO: South Head, CO: Coogee, MA: Maroubra) after six months……………………………………………………54

3.4: Recruitment at restored sites. (a) Photograph of restored Phyllospora comosa recruits at Freshwater.(b) Scatterplot output from DAPC for Phyllospora donor and restored F1 recruits (boxed in legend) populations. (c) Membership probability plot showing probability of restored F1 recruits genetic assignment to Bateau Bay (BB) or Shark Park (SP) donors at four restored sites. WH: Whale Beach, FW: Freshwater, SO: South Head, CO: Coogee, MA: Maroubra. Horizontal dotted lines represent 95% confidence intervals for assignment to each site (= likely a pure bred cross between donor from the same site) (d) Summary table showing number of recruits significantly assigned to BB, SP or a mix of both (threshold 0.95 or 0.8 as accuracy of model was ~80%, shown in brackets)...... 56

4.1: (a) Map of 13 sites where Phyllospora comosa was sampled; inset depicts Phyllospora’s entire distributional range; ocean coloured with average maximum annual SST from 1992 - 2018. PM: Port Macquarie; FO: Forster; AB: Anna Bay: BB: Bateau Bay: TE: Terrigal: PB: Palm Beach; CR: Cronulla; SP: Shark Park; SH: Shellharbour; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport. (b) Genetic structure plots showing individuals assigned to inferred clusters using all loci or outlier loci detected using Bayescan, PCadapt and LFMM (SST-associated loci). Each row represents an individual; different colours within columns indicate maximum likelihood probability of belonging to different clusters…………………………………………………………………………………………………….68

vii 4.2: Distance-based Redundancy Analysis showing the relationship between environmental variables and genetic structure (SNPs) of Phyllospora comosa populations ...... 76

4.3: Relationship between geographic distance (km) and overall genetic distance (pairwise Fst) between Phyllospora comosa individuals calculated using 109 loci...... 78

5.1: Map of sites where the seaweed Phyllospora comosa was sampled for a study on host genetics, phenotype and surface-associated microbial communities ...... 95

5.2: (a) Principal Component Analysis (PCA) of the seaweed Phyllospora comosa’s genetic structure, based on allele frequencies at 114 SNP loci; (b) PCA of phenotype, based on 11 traits describing morphology and condition; (c) nMDS analysis of Phyllospora-associated microbial communities ...... 103

5.3: Relationship between geographic distance (km) and pairwise distance/dissimilarity between centroids of the eight Phyllospora comosa populations for genetics (Euclidean distances), overall surface-associated microbial communities (Bray-Curtis dissimilarity of square-root transformed relative abundances) and phenotype (Euclidean distances)...... 104

5.4: Relationship between site-level host genetic diversity (HE) and a) species richness and b) Simpson diversity of microbial communities associated with the seaweed Phyllospora comosa’s surface...... 105

5.5: Relationship between microbial community dissimilarity (Bray-Curtis on square- root data) based on all ASVs and genetic distance (Euclidean) between all Phyllospora comosa individuals (a) across and within all sites (b) within each site...... 106

5.6: : Relationship between microbial community dissimilarity (Bray-Curtis on square root relative abundances) based on all ASVs and phenotypic distance (Euclidean) between all Phyllospora comosa individuals (a) across and within all sites (b) within each site ...... 107

viii 5.7: Associations between geography, genetics and phenotype of the Phyllospora comosa host with 156 microbial community samples isolated from the surface of host fronds; as inferred by distance-based Redundancy Analysis (dbRDA)...... 110

S3.1: Average (a) wet weight biomass and (b) density of reproductive conceptacles of individuals………………………………………………………………………………………………………………..181

S4.1: Principal Component Analysis (PCA) of the seaweed Phyllospora comosa’s genetic structure, based on allele frequencies at 109 SNP loci……………………………………………..188

S4.2: Relationship between geographic distance (km) and genetic distance (pairwise Fst) between Phyllospora individuals for selective loci datasets; a) outlier loci identified using Bayescan; b) outlier loci identified using PCAdapt and c) sea surface temperature- associated loci. Data collected at thirteen sites and plots fitted with linear regression (fitted values: blue line; 95% Confidence Intervals: grey shade)...... 194

S5.1: Figure depicting traits measured or characterised for each of 156 Phyllospora comosa individuals……………………………………………………………………………………………………199

S5.2: Dispersion of Phyllospora comosa genetic data for 8 sites. PM: Port Macquarie; FO: Forster; AB: Anna Bay; CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport...... 201

S5.3: Dispersion of Phyllospora comosa phenotype data for 8 sites. PM: Port Macquarie; FO: Forster; AB: Anna Bay; CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport...... …204

S5.4: Plots of all phenotypic traits measured, ordered by site…….……………..………………206

S5.5: Dispersion of Phyllospora comosa microbial community data for 8 sites, based on (a) Bray-Curtis dissimilarity matrix based on overall microbial data and (b) Jaccard distances based on overall microbial data. PM: Port Macquarie; FO: Forster; AB: Anna Bay: CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport...... 211

S5.6 Relationship between (a) geography, (b) host genetics and (c) phenotype on Phyllospora comosa’s overall microbial communities. Taxa shown are the five most

ix abundant ASVs that were significantly with each respective variable of interest. Sites ordered north to south; PM: Port Macquarie; FO: Forster; AB: Anna Bay: CR: Cronulla; MB: Malua Bay; ED: Eden; NT: Bicheno; ST: Southport...... 220

x List of abbreviations

AMOVA Analysis of molecular variance

ANOVA Analysis of variance

AR Allelic richness

ASV Amplicon sequence variant

DNA Deoxyribonucleic acid

FIS Inbreeding coefficient

FST Genetic distance among populations

HE Expected heterozygosity

Ho Observed heterozygosity

HWE Hardy-Weinberg equilibrium

IBD Isolation by distance

LD Linkage disequilibrium

PAM Pulse-amplitude modulated fluorometer

PERMANOVA Permutational multivariate analysis of variance

SD Standard deviation

SE Standard error

SNP Single nucleotide polymorphism

xi Using genomics to restore and future-proof underwater seaweed forests

Chapter 1:

General Introduction

Using genomics to restore and future-proof underwater seaweed forests

Our world is changing due to human activities more rapidly than ever before (Steffen et al., 2015). Approximately one-third of Earth’s arable land surface is degraded (Gibbs et al., 2015) and between 50-90% of marine ecosystems are now in an altered state (Lotze et al., 2011). Anthropogenic impacts have affected biodiversity (Dornelas et al., 2014), ecosystem functions and services (Millennium Ecosystem Assessment, 2005; Pecl et al., 2017) and even threaten our own planet’s stability (Steffen et al., 2018). Yet some shifts in ecosystem dynamics are showing signs of recovery after active solution-based management strategies have been applied (Camill et al., 2004; Lefcheck et al., 2018; Rey Benayas et al., 2009; Warren et al., 2017). Ecological understanding is critical to help guide these actions towards a more stable and sustainable future.

1.1 Restoration and future-proofing: the New Ecological frontier

Against the backdrop of profound and rapid environmental transformation, habitat restoration is emerging as a useful tool to halt or reverse the degradation of natural ecosystems (Rey Benayas et al., 2015). While restoration has traditionally involved active biological intervention that aims to reinstate historical, or match present “healthy” (i.e. not or less impacted) ecosystems, it is now recognised that the advance of pervasive stressors (e.g. those related to climate change) call for the development of novel long-term solutions. Such solutions explicitly require incorporation of the ability for ecosystems or populations to recover from perturbations (i.e. resilience; Holling, 1978) and/or persist in the face of extant stressors and future change (i.e. resistance; Sutherland 1990). One result of this is that ecosystems may necessarily need to be different from how they have previously or presently existed to survive in the future. Increasingly, it is acknowledged that applying such “future-proofing” strategies are not only critical for restoration, but in many cases are also necessary to help buffer natural extant populations that are facing continued rapid environmental change (van Oppen et al., 2015; Wood et al., 2019).

Using genomics to restore and future-proof underwater seaweed forests

Restoration efforts are increasing worldwide and are predicted to continue to rise, with 2020-2030 being named the “Decade of Restoration” (UN, 2019). So far, restoration has had some significant successes, increasing the provision of biodiversity and ecosystem services in previously degraded areas (Rey Benayas et al., 2009). However, current success rates are still nowhere near what is required to reach proposed targets by the UN, which aim to see 350 million hectares of degraded ecosystems restored by 2030. The speed and effectiveness of restoration activities vary depending on the ecosystem type and state at the outset (Johnson et al., 2017; O’Brien et al., 2018; Rey Benayas et al., 2009), as well as the methods of restoration action (Bullock et al., 2011). Further, many projects are plagued by issues with poorly designed evaluation programs - making success difficult to define - and slowing progress by masking the real rates of restoration success or failure (Morandi et al., 2014). Intercepting ecosystem declines as soon as possible and developing the science to maximise the success of both restoration and future- proofing strategies are critical priorities.

1.2 Novel techniques to restore and future-proof

Recent innovations in science include novel techniques to aid or enable the development of restoration and future-proofing strategies and improve their success (Hobbs, 2018; Breed et al., 2019). In particular, the advent of genetic and genomic tools, e.g. rapid DNA sequencing technologies, is now enabling scientists to examine and understand components of ecosystems that have previously been prohibitively expensive to access and therefore were often ignored (Mijangos et al., 2015). For example, the genetic diversity and provenance of donor material for restoration can directly affect the establishment rates and fitness of restored populations and may also enhance their adaptive capacity and resilience (Forsman and Wennersten, 2016; Hughes and Stachowicz, 2004; van Oppen et al., 2015). Understanding the genetic characteristics of target species can empower restorationists to harness desirable traits and to enhance adaptive potential. Other components that may be studied are more exploratory. For example, shotgun sequencing and environmental genomics are greatly facilitating the study of

Using genomics to restore and future-proof underwater seaweed forests

microbes and other organisms in the environment (Handley, 2015; Lacoursière‐ Roussel et al., 2016; Venter et al., 2004). An understanding of these components may also lead to the development of novel and innovative restoration and future- proofing solutions, e.g. that of promoting beneficial interactions between microbes and hosts (e.g. Hong and Lee 2014; van Oppen et al., 2015). Importantly, conducting restoration or future-proofing programs without paying attention to all of these elements may cripple their success and lead to subsequent failure (Granado et al., 2018; Williams, 2001). The incorporation of genomic tools is therefore considered critical for guiding the scaling up of restoration and future-proofing activities (Breed et al., 2019; Mijangos et al., 2015).

1.3 Focusing on seaweeds as candidates for novel management efforts

The increasing focus on restoration development and expansion is largely limited to habitats on land (e.g. see David et al., 2019). Despite widespread stress and degradation, restoration in the marine realm has had relatively few successes (Bayraktarov et al., 2016; Rey Benayas et al., 2009) and remains limited in scope (Eger et al., 2019; Gillies et al., 2015). This is particularly true for temperate rocky reefs, which are undergoing widespread changes and stand to gain much from research focused on restoration and future-proofing. Temperate rocky coastal systems are underpinned by forest-forming seaweeds, which are in decline due to multiple stressors, such as poor water quality and climate change, in many places around the world (Krumhansl et al., 2016; Wernberg et al., 2019a). Underwater seaweed forests have high ecological and economic value, they underpin coastal biodiversity, provide key ecosystem goods and services and are among the most productive ecosystems on earth (Bennett et al., 2016; Krause-Jensen and Duarte, 2016; Steneck et al., 2002). Due to their immense value and the increasing challenges that these ecosystems face, interest in the restoration of seaweed forests is growing such that the number of seaweed restoration projects have doubled in the past decade alone (Eger et al., 2019). Most seaweed restoration projects are still in their initial development stages and are done at relatively small spatial scales (<1000m2; Eger et al., 2019). In some systems,

Using genomics to restore and future-proof underwater seaweed forests

however, we are at the point of being able to restore at scales that match that of degradation, which are generally much larger (Marzinelli et al., in prep). Ensuring that the rapidly advancing restoration and future-proofing theory is transformed into a truly effective practice requires the testing of solutions and ideas in the field. By embedding rigorous scientific practices into active restoration projects, restoration and future-proofing strategies can be refined. This, in turn, can lead to the enhancement of project outcomes as they scale up (Breed et al., 2018; Gellie et al., 2018). Now is therefore the perfect time to refine restoration strategies for these systems and incorporate future-proofing principles into this work.

1.3.1 Phyllospora comosa as a case study

Phyllospora comosa (hereafter, Phyllospora, Order Fucales) is a foundation species of seaweed commonly known as ‘crayweed’ which forms dense, often monospecific forests along the shallow temperate reefs of Eastern Australia (Underwood et al., 1991; Wormersley, 1987). Phyllospora forests provide food and habitat for unique and biologically diverse communities, which include the two most valuable fisheries in Australia: abalone and rock lobster (Bishop et al., 2010; Marzinelli et al., 2014). Phyllospora forests disappeared from 70km of Sydney’s coastline during the 1970’s and 1980’s, likely due to outfalls of raw sewage along the metropolitan coast (Coleman et al., 2008). Although water quality has increased dramatically since the early 1990’s due to changes in waste management (Scanes and Phillip, 1995), Phyllospora has not naturally returned, likely due to recruitment limitation, e.g. due to low supply of propagules or pre-emptive competition with algal mats or other macrophytes. Pilot experiments conducted in 2011 demonstrated that transplanting adult, reproductive Phyllospora from the edges of Sydney to areas where it has disappeared could lead to recruitment and the subsequent re- establishment of populations (Campbell et al., 2014a) and have stimulated plans for a wider-scale restoration program along Sydney’s coastline (known as ‘Operation Crayweed’ – see www.OperationCrayweed.com). The work presented

Using genomics to restore and future-proof underwater seaweed forests

within this thesis, particularly Chapters 3-5, arose from a need to further our understanding of how to maximize restoration success by incorporating knowledge of genetic diversity of donor populations and to develop future- proofing strategies as we attempt to scale activities up.

1.4 Thesis aims and outline

The overarching aim of this thesis is to investigate and develop methodologies that can be used to improve the restoration and future-proofing of declining underwater forests. The thesis has a specific focus on harnessing transdisciplinary approaches, especially genomics. It includes a series of empirical studies to demonstrate how the novel integration of ecological and genomic data can be used to inform the design of future-proof underwater forest restoration programs and measure subsequent success. These studies are incorporated into the ongoing restoration of the declining fucoid Phyllospora, thus linking the conceptual framework of modern ecology and genetics with a practical and significant conservation outcome. Firstly, in Chapter 2 I discuss the theory and conceptual background for designing restoration programs for marine macrophytes (Wood et al., 2019). I examine these ideas within a future-centric context and identify four significant challenges that we predict are likely to influence these ecosystems into the next century. I discuss emerging ecological solutions to combat these threats, and consider key changes to managerial, political and public frameworks to support the scientific solutions proposed and influence the uptake of marine restoration into the future. In Chapter 3 I build upon the framework I developed in Chapter 2 by focusing on seaweed genetic structure and diversity, which are identified as key biological features that can determine both short and long-term restoration success in many systems. Thus, I focus on restoring lost underwater forests which mimic genetic diversity and structure of deemed “healthy” extant populations. In this chapter, I apply population genetics to design and experimentally assess Phyllospora’s restoration program (Wood et al., in review). Single Nucleotide

Using genomics to restore and future-proof underwater seaweed forests

Polymorphisms (SNPs) are employed to characterise genetic diversity, structure and the effect of geographic distance on extant populations surrounding the gap in Phyllospora’s distribution. This information is then used to design the restoration of five new populations across Sydney. Within the restoration activities, I embed a manipulative field experiment to empirically test whether donor provenance influences restoration success. I describe the effects of Phyllospora donor provenance on adult condition and survival, as well as subsequent recruitment and the genetic diversity and structure of the first generation (F1) of restored populations. Chapters 4 and 5 further investigate the potential to future-proof underwater forests, be they “natural” (i.e. have not been the subject of direct and active management interventions) or restored. In Chapter 4, the objective is to provide a platform that can be used to explicitly design strategies to assist adaptation of Phyllospora in the context of a rapidly changing climate. I identify genetically rich and/or depauperate i.e. potentially vulnerable populations and assess signals of natural selection. To do this, I characterised Phyllospora’s overall and adaptive genetic diversity along the species’ entire latitudinal (31-43o latitude) and thermal (~12-26oC) range. I then discuss ways in which we can triage populations in need of genetic management and identify donor populations that may be also be used to provide such solutions. Chapter 5 incorporates novel ideas about future-proofing by drawing on the emerging concept of holobionts and the hologenome theory of evolution (Zilber-Rosenberg and Rosenberg, 2008). As is discussed in Chapter 2, the microbiome is a key biological feature that will likely influence the survival, condition and adaptation of seaweed populations into the decades ahead. In this chapter, I investigate the driving factors of Phyllospora’s surface-associated microbial communities using ecological surveys, amplicon sequencing and host genetics. I decouple the influence of geography/the environment from that of characteristics of the host, particularly host genetics and phenotype, on associated microbial communities and specific amplicon sequence variants (ASVs), with the ultimate objective of determining the degree to which future-proofing techniques should accommodate host-specific traits in their design. 7

Using genomics to restore and future-proof underwater seaweed forests

Finally, Chapter 6 summarises these works and discusses the key lessons learnt, postulating ideas about the future and the relevance of this thesis in a broader ecological context.

Using genomics to restore and future-proof underwater seaweed forests

1.5 Acknowledgement of co-authored papers Chapter 2-5 in this thesis are written in the format of a series of individual papers; a review paper (Chapter 2) and three primary research papers (Chapters 3–5). Chapter 2 had been published at time of submission (Wood et al., 2019); Chapter 3 was in review at time of submission and is now published (Journal of Applied Ecology 2020) and Chapters 4 and 5 were being drafted for submission to scientific peer-reviewed journals (Chapter 4 is now published; Global Change Biology 2021). Although these chapters focus on different aspects of the restoration and future proofing framework and I have made every effort to avoid repetition, there are inevitably some unavoidable overlaps - especially in the Introduction and Methods sections. While the work in these papers was primarily conducted and written by me for my thesis, they are nevertheless collaborative; hence I have used the phrasing “we”, as is conventional for collaborative scientific papers, throughout. Co-authors are acknowledged at the start of chapters. The reference details of the published chapters are:

Wood, G., Marzinelli, E., Coleman, M., Campbell, A., Santini, N., Kajlich, L., Verdura, J., Wodak, J., Steinberg, P., Vergés, A. (2019). Restoring subtidal marine macrophytes in the Anthropocene: trajectories and future-proofing. Marine and Freshwater Research, 70(7), 936-951.

Wood G., Marzinelli E.M., Vergés A., Campbell A.H., Steinberg P.D., Coleman M.A. Using genomics to design and evaluate the performance of underwater forest restoration. J Appl Ecol. 2020;57:1988–1998. https://doi. org/10.1111/1365- 2664.13707

Wood G., Marzinelli E.M., Campbell A.H., Steinberg P.D., Vergés A., Coleman M.A. Genomic vulnerability of a dominant seaweed points to futureproofing pathways for Australia's underwater forests. Glob Change Biol. 2021;00:1–13. https://doi.org/10.1111/ gcb.15534

Using genomics to restore and future-proof underwater seaweed forests

Chapter 2:

Restoring subtidal marine macrophytes in the Anthropocene: trajectories and future-proofing

Published as: Wood, G., Marzinelli, E., Coleman, M., Campbell, A., Santini, N., Kajlich, L., Verdura, J., Wodak, J., Steinberg, P., Vergés, A. (2019). Restoring subtidal marine macrophytes in the Anthropocene: trajectories and future-proofing. Marine and Freshwater Research, 70(7), 936-951. (see Appendix A)

Using genomics to restore and future-proof underwater seaweed forests

2.1 Abstract Anthropogenic activities have caused profound changes globally in biodiversity, species interactions and ecosystem functions and services. In terrestrial systems, restoration has emerged as a useful approach to mitigate these changes and is increasingly recognised as a tool to fortify ecosystems against future disturbances. In marine systems, restoration is also gaining traction as a management tool, but it is still comparatively scant and underdeveloped relative to terrestrial environments. Key coastal habitats, such as seaweed forests and seagrass meadows are showing widespread patterns of decline around the world. As these important ecosystems increasingly become the target of emerging marine restoration campaigns, it is important not only to address current environmental degradation issues, but also to focus on the future. Given the rate at which marine and other environments are changing and given predicted increases in the frequency and magnitude of multiple stressors, we argue for an urgent need for subtidal marine macrophyte restoration efforts that explicitly incorporate future-proofing in their goals. Here we highlight emerging scientific techniques that can help achieve this, and discuss changes to managerial, political and public frameworks that are needed to support scientific innovation and restoration applications at scale.

Using genomics to restore and future-proof underwater seaweed forests

2.2 Introduction In recent decades, humanity’s far-reaching and long-lasting effects on Earth’s natural systems have reached such a magnitude that many now argue we are living in a new geological age: the Anthropocene (Zalasiewicz et al., 2008). Human activities such as agriculture, urbanisation and industrialisation have produced climatic, biological and geochemical signatures that are clearly distinct from the preceding Holocene epoch (Waters et al., 2016). These human effects have led to profound changes in global biodiversity (Dornelas et al., 2014), species interactions (Tylianakis et al., 2008) and ecosystem functions and services in ways that have already affected human health, culture and economy (Millennium Ecosystem Assessment, 2005; Pecl et al., 2017). Against this backdrop of profound and rapid transformation of the Earth’s ecosystems, habitat restoration has become a major focus of conservation and natural resource management (Dobson et al., 1997; Millennium Ecosystem Assessment, 2005) and is emerging as a management strategy that can provide realistic, context-specific pathways to a sustainable future (DeFries et al., 2012). A recent meta-analysis estimated that, globally, restoration practices have already increased the provision of biodiversity and ecosystem services in previously degraded areas by an average of 44 and 25% respectively (Rey Benayas et al., 2009). Nevertheless, the speed and effectiveness of restoration activities still vary widely depending on the ecosystem state at the outset (Johnson et al., 2017; O’Brien et al., 2018), the type of environment being restored (Rey Benayas et al., 2009) and the methods of restoration action (Bullock et al., 2011). Environmental restoration is currently undergoing multiple shifts in conceptual theory and practice (Higgs et al., 2014). Traditionally, restoration comprised simple actions aimed at returning a site’s ecosystem function(s) or species composition back to a historic state. In contrast, modern restoration strategies embrace both social and ecological considerations (Martin, 2017) and are guided by multiple goals including not only environmental ones, but also those that speak to human well-being (e.g. alleviating poverty; Perring et al., 2015). New approaches recognise the non-equilibrium nature of ecosystem dynamics and that restoration may take multiple trajectories, very likely including towards novel 12

Using genomics to restore and future-proof underwater seaweed forests

states (Choi 2004, 2007; Perring et al., 2015). This has been accompanied by an increased focus on restoring ecological processes that generate, maintain and affect biodiversity and ecosystem functioning, rather than restoring particular species per se (Perring et al., 2015). Restoration is also emerging as a strategy to mitigate against future predicted effects, rather than only as a response to extant ones (Menz et al., 2013; Wilson and Forsyth, 2018). Thus, there is a recognised need for restoration efforts not only to focus on matching past (which may be unknown) or current characteristics of the systems, but also to include an explicit consideration of future desirable characteristics (Hobbs and Harris, 2001). Many terms have been suggested for these interventions, such as rehabilitation, repair, reconciliation (Rozenweig, 2003; Goodsell and Chapman, 2009) or even ‘future-proofing’ (i.e. the building of resistant or resilient, self-sustaining populations that can withstand conditions both now and into the future). Here, we use the term ‘restoration’ in the broadest sense (i.e. including future-proofing). Restoring into the future increasingly requires the incorporation of cross- disciplinary scientific and humanitarian fields including ecology, microbiology, population genetics, engineering and social, economic and political sciences to target multiple aspects of socioecological systems and promote resilience for unforeseen changes (Wilson et al., 2011; Menz et al., 2013; Zhang et al., 2018). However, to date the most significant advances in restoration efforts have been made on land, with marine restoration comparatively scant and underdeveloped (Rey Benayas et al., 2009). This is despite the fact that between 50 and 90% of marine ecosystems are already in a degraded state (Lotze et al., 2006). This disparity in progress is probably due, in part, to marine habitats being largely ‘out of sight, out of mind’ and the facts that: (1) marine ecosystem degradation is often only recognised decades after the initial problem (e.g. Coleman et al., 2008); (2) it can be difficult to pinpoint and reverse the source(s) of degradation because of the open nature of marine systems; and (3) many stressors causing marine degradation originate on land and must be removed or mitigated before restoration (Crouzeilles et al., 2016). Addressing these problems

Using genomics to restore and future-proof underwater seaweed forests

is also accentuated by the substantial logistical challenges related to working underwater (Gillies et al., 2015). Subtidal marine macrophytes (i.e. seagrasses and seaweeds) are habitat formers that provide ecosystem services of global importance. They are among the most productive primary producers on the planet (Mann, 1973; Duarte and Chiscano, 1999) and provide habitat and other resources for many species (Heck and Orth, 1980; Teagle et al., 2017). Seagrasses and seaweeds also represent significant marine carbon sinks (Trevathan-Tackett et al., 2015; Krause-Jenson and Duarte, 2016) and are likely natural buffers against coastal erosion likely natural buffers against coastal erosion (Duarte et al 2013) due to their ability to attenuate wave action (Mendez et al., 2004; Bouma et al, 2005; Yang et al., 2012). However, they are declining in many places around the world (Waycott et al., 2009; Vergés et al., 2014; Krumhansl et al., 2016), with significant consequences to biodiversity, ecosystem function and services projected to accelerate in coming years. Despite their importance, systems dominated by seaweeds and seagrasses receive little public recognition and research funding compared with other marine habitats, such as coral reefs, mangroves or salt marshes (Duarte et al., 2008; Bennett et al., 2016), hindering our ability to communicate, understand and mitigate their losses. Thus, a greater awareness and concern of the importance of these macrophyte-dominated systems and the challenges they face is critical to advance marine macrophyte restoration and to maintain the ecosystem functions and services they provide (Bennett et al., 2016). In this paper we discuss the theory and conceptual background for designing restoration programs for seaweed and seagrass habitats and examine these within a future-centric context, including issues related to when (in the future) to restore for, what to restore and where. We identify the four most significant challenges that we predict will influence the restoration of seagrasses and seaweeds into the next century, namely climate change, overfishing, water quality and ocean sprawl (i.e. coastal and offshore development; Firth et al., 2016), and discuss emerging ecological solutions that have arisen in terrestrial systems (and, where possible, marine environments) that may be adapted or applied to subtidal macrophyte 14

Using genomics to restore and future-proof underwater seaweed forests

systems and used to combat these threats (Fig. 2.1). We also consider some of the key changes to managerial, political and public frameworks that are necessary to support the scientific solutions proposed, and that will affect the prevalence and uptake of marine restoration into the future. By doing so, we hope to instigate further research into future-proofing methods and their application in this field.

Using genomics to restore and future-proof underwater seaweed forests

Figure. 2.1: Incorporating future-proofing principles into restoration of subtidal marine macrophytes*. (a) Anthropogenic-induced stressors, notably climate change, overfishing, water quality and ocean sprawl, are having significant effects on subtidal marine macrophyte systems, and these are expected to continue into the future. Future-proofing aims to enable the persistence of ecosystems by enhancing resistance and resilience to both current and future stressors. (b) Although the concept of future-proofing is still novel in these environments, emerging developments and innovations in science may provide solutions to the challenges facing them into the future. However, in many cases multiple solutions may be necessary to combat not only cumulative, but also unpredictable and synergistic effects, and more research is needed before suitable management plans can be decided. (c) Early but recent applications of future-proofing strategies include (photographs from top to bottom): field-trials to determine the suitability of low-latitude seaweed genotypes for restoration in Sydney, Australia; the development of seagrass seeding methods in Western Australia; and eco-engineering seawalls in Sydney, Australia.

*Photographs courtesy of Adriana Vergés (degraded kelp forest), Rebecca Morris (seawall), John Turnbull (urchin barren, crayweed forest and seagrass meadow), Steve Spring (Marine Photobank; sewage outfall) and Katherine Dafforn (engineered seawall). Icons are from Jane Thomas, Kim Kraeer, Lucy van Essen-Fishman, Catherine Collier, Dieter, Tracey and Jane Hawkey, Integration and Application Network, University of Maryland Center for Environmental Science (ian.umces.edu/imagelibrary/). 16

Using genomics to restore and future-proof underwater seaweed forests

2.3 Restoring into the future: for when, what and where do we restore?

2.3.1 When to restore for Most ecological experiments are relatively short term, and a major challenge with restoration efforts is to adequately identify the relevant temporal scale(s) for habitat interventions. If restoration efforts are aimed at the future, when should we restore for? The next 10 years? The next century? The next millennium? Several factors come into play and should be considered when determining this. The first is related to the goals of the restoration and who they are set by. In many cases, restoration may be limited to logistical and managerial time scales that are often shorter than those needed to address societal or ecological needs. For instance, most agencies that fund restoration projects in Australia typically have a time frame of 1–10 years (Bayraktarov et al., 2016), whereas recovery of seaweed and seagrass ecosystems may take from 2 to over 30 years to reach a fully functional state (Campbell et al., 2014a; O’Brien et al., 2017). Second, the temporal scale of stressors that are more likely to affect the systems we are restoring will affect how and if it is technically possible to future- proof populations for specific stressors. For example, choosing donor populations from local environments similar to those of restoration sites may help restored populations survive the initial years of transplantation, but novel stressors, such as increasing sea temperatures over the next 50–100 years, could require specialised genotypes to be sourced from foreign populations, such as those already adapted to warm temperatures at the warm (rear) edge of the distribution (discussed below in the ‘Designing restored populations with genetics in mind’ section). Considering that restored populations will need to survive between the initial restoration effort and when the restoration is aimed at, approaches such as these may also require adaptive monitoring and incremental introductions of resistant or adapted genotypes over long periods of time (Lindenmayer and Likens, 2009). Third, and related to the second point, is that as the restoration goals become longer term (e.g. a century v. a decade) and environmental conditions continue to change, so increases the philosophical dilemma posed as to whether we should attempt to restore populations that we are uncertain will survive in a far-away future. For example, Martínez et al., (2018) recently modelled the 17

Using genomics to restore and future-proof underwater seaweed forests

distribution of 15 habitat-forming seaweeds up to 2100 and predicted that 4 of 15 prominent habitat-forming seaweed species were likely to become extinct due to ocean warming in Australia over the next century, which prompts the question, should we be trying to restore these species? Modelling environmental change and shifting species’ distribution over time, but also integrating settlement (e.g. Cetina- Heredia et al., 2014, 2015) and adaptive potential (discussed for seaweeds and seagrasses in Duarte et al., 2018) into these models, may be used as a tool to aid decisions regarding which species may be most appropriate to restore (i.e. those still within their environmental limits). Similarly, these approaches can be used to define specific temporal scales of interest for restoration and to target restoration at scales that are relevant to management or identified societal and ecological needs. It may also help identify species or target areas that will benefit most if complemented by techniques that improve general resilience, such as increasing genetic, species or functional diversity as insurance policies for unforeseen change.

2.3.2 What to restore Defining what we want to restore is equally critical to the temporal goals of restoration. Motivations for restoring marine macrophytes, as in other systems, have traditionally focused on human use (i.e. ecosystem services) as defined by regulatory requirements or economics (Wiens and Hobbs, 2015). These efforts have typically focused on re-establishing individual foundation species to mitigate past or extant anthropogenic effects, with the assumption that bringing these species back would also reinstate their associated ecological communities and the functions and services they provide (Keenan et al., 1997; Brudvig, 2011). For example, Marzinelli et al., (2016) restored a habitat-forming seaweed with the aim of also restoring associated communities of epifauna. However, if restoration is increasingly recognised as a tool in mitigating or adapting to future threats, such as those derived from climate change, it may be easier or more desirable to restore functions rather than target species if the latter are not predicted to cope with future environmental conditions. Restoring specific ecosystem functions and services that will result in tangible benefits for humans, such as provision of food, improving water quality, preventing erosion, enhancing blue carbon sequestration 18

Using genomics to restore and future-proof underwater seaweed forests

or providing areas for recreation (Abelson et al., 2016), may be achieved by introducing alternative species that may be more successfully restored, or even engineering solutions such as artificial habitats as a replacement (Seaman, 2007). Although this will still require decisions as to what species are required to maintain these services, the focus on function may mean that restored systems in the future may decreasingly resemble historical ones.

2.3.3 Where to restore Equally important to when and what we are trying to restore, is where we restore. Although there is increasing attention in marine environments on where less interventionist approaches to conservation, such as spatial management, should be undertaken (e.g. in the choice and use of ecological bright spots; Cinner et al., 2016), a coherent framework for how to identify appropriate places for restoration is still embryonic (Anthony et al., 2017). Factors to consider when targeting sites for restoration of macrophytes include knowledge of the reproductive biology of the target species and the factors that affect dispersal, recruitment and colonisation at scales relevant for the proposed restoration (Coleman and Kelaher, 2009). For example, understanding ecological issues such as the minimum population or habitat size necessary to sustain the long-term survival (and potentially expansion) of the restored species and the ecological (diversity, productivity etc.) benefits of restoring in a particular place need to be considered. Assessing suitability of sites for restoration by anticipating changes to supply and dispersal of propagules, as well as survival and self-sustainability based not only on current hydrodynamics and existing ecological interactions, but also on likely future changes to these and other factors, such as temperature and acidification, are necessary to achieve desired outcomes at scale. Oceanographic modelling can be used as a tool to select sites where interventions are likely to have the highest probability of success and expansion. For example, modelling future change in hydrodynamics of the dominant current system on the east coast of Australia, the East Australian Current, has shown that peak settlement locations of species may change (Cetina-Heredia et al., 2014, 2015), altering the ability of

Using genomics to restore and future-proof underwater seaweed forests

areas to be self-sustaining or to act as a source to replenish degraded areas (Coleman et al., 2017). Other factors that need to be considered when deciding where to focus restoration efforts involve species interactions. For example, in regions where herbivores such as sea urchins are abundant, tipping points in macrophyte– herbivore dynamics may preclude successful restoration efforts in these systems and lead to continued overgrazing. Thus, except where there is extraordinary ability for continual removal of sea urchin grazers, resources would be better spent elsewhere (Johnson et al., 2017). For restoration to be successful, it must also lead to benefits that match the spatial scale of the degradation. In contrast with terrestrial systems, where restoration is undertaken at large scales (Lamb, 2014), restoration in marine systems generally and of marine macrophytes in particular has historically been done at very small scales (tens to hundreds of metres) with outcomes that rarely match the scale of degradation of the relevant habitat or environment (tens to hundreds of kilometres or more; van Katwijk et al., 2016). However, it is not clear that constraints on the scale of restoration of at least some macrophytes are primarily due to a lack of ecological knowledge or appropriate techniques (Gillies et al., 2015; Lefcheck et al., 2018). For instance, seaweed is farmed in areas of East and South-east Asia at substantial scales of hundreds of hectares or more (McHugh, 2003). The scale of seaweed aquaculture in Asia and the development of new seed-based techniques for large-scale restoration of some seagrasses (Marion and Orth, 2010; Kendrick and Statton, 2019) suggest that methods for upscaling of macrophyte restoration are, in at least some cases, available. Rather, the constraints may often be societal, particularly financial or political (Bayraktarov et al., 2016). Thus, ‘where to restore’ for large-scale restoration of macrophytes may be based on societal considerations as much as ecological, targeting places where there is the potential to develop coalitions of government, communities, private parties and researchers that will broadly support the effort. This is analogous to restoration efforts in some tropical rainforest sites (Brancalion et al., 2012; Chaves et al., 2015). Where resources are limited, restoration should focus on small-scale efforts that leverage large-scale outcomes. For example, macroalgal restoration 20

Using genomics to restore and future-proof underwater seaweed forests

efforts in Sydney performed at a scale of metres are resulting in outcomes at scales of hundreds of metres to kilometres (when done at multiple sites), with self- sustaining populations established via multiple, but very small-scale, interventions (Marzinelli et al., unpubl. data).

2.4 Future-proofing: challenges ahead and new ecological solutions

Identifying the cause of degradation (i.e. why to restore) is arguably the first step in a successful restoration program as it will inform when, what, where and how to restore. The best restoration strategy may simply involve alleviating the stressors that caused degradation. Factors to consider in this regard include those that affect colonization, adult reproduction and survival. Arguably, the most significant challenges to restoration of macrophytes into the future are linked to climate change. Effects of ocean warming, acidification, changes in circulation patterns and the frequency and intensity of storms are predicted to increase in the next decades and century, strongly affecting subtidal macrophytes directly through physiological and physical effects (e.g. Filbee-Dexter and Scheibling, 2012; Harley et al., 2012; Koch et al., 2012), as well as indirectly through changes to key ecological interactions that can cascade through ecosystems (e.g. Vergés et al., 2016; Provost et al., 2017). However, these climatic changes will not occur in isolation, particularly around coastal areas with high human usage, where macrophytes have often historically been found. Such macrophyte-dominated coastal areas are subjected to multiple stressors, of which overfishing, pollution and habitat modification through coastal development, or ‘ocean sprawl’, are among the most important.

Overfishing has been described as one of the primary causes of ecological collapse in the 20th century (Jackson et al., 2001) and can affect marine macrophytes in multiple ways. For example, historical overfishing of herbivores, such as the green turtle and dugong, have been linked to the decline of water quality, soil hypoxia and proliferation of wasting disease that has caused significant declines of seagrass meadows (Jackson et al., 2001) and may reduce seed dispersal and connectivity (Tol et al., 2017). Conversely, population explosions of herbivores

Using genomics to restore and future-proof underwater seaweed forests

following the removal of apex predators by fishing have caused trophic cascades leading to widespread deforestation of kelp forests (Sala et al., 1998; Shears and Babcock, 2003; Johnson et al., 2005; Spyksma et al., 2017). Predicted increases in the human population and its reliance on protein from fish (Delgado et al., 2003; Gerland et al., 2014) are likely to continue to increase fishing pressure and subsequent changes to ecological communities and species interactions (Carney et al., 2005; Yoon et al., 2014). Poor water quality and associated pollution due to poor management of sewage outfalls, stormwater and agricultural run-offs have been among the most significant stressors for macrophytes in coastal systems over the past century (Airoldi and Beck, 2007; Burkholder et al., 2007; Lotze et al., 2011). In past decades, substantial efforts to ameliorate water quality have taken place, particularly in developed countries (Scanes and Phillip, 1995; Lyerly et al., 2014; Sydney Water, 2017), and this, in turn, has made restoration possible in some regions (Campbell et al., 2014a; Bellgrove et al., 2017; Lefcheck et al., 2018). A fundamental challenge in the future will be to maintain and/or improve water quality in coastal areas as urbanisation and agriculture increase as a consequence of human population growth (Burkholder et al., 2007). Continued population growth has also led to increased coastal development, including seawalls, jetties, aquaculture and wind farms, developments on reclaimed land and even entire islands, all of which have ecological effects (Bulleri and Chapman, 2010). Given their current, and ever growing, extent, such novel or modified habitats must be accommodated in any serious consideration of the global rehabilitation of marine macrophytes (Dafforn et al., 2015). Critically, these multiple stressors interact in complex ways, and understanding their interactive effects can lead to better management approaches (Côté et al., 2016). For instance, in addition to having obvious direct positive effects on macrophyte restoration, improvements in water quality can potentially dampen synergistic effects of future climate change by slowing the increase in competing algal turfs (Falkenberg et al., 2013). Rigorous science capable of distinguishing between the effects of climate change, overfishing and degradation in water quality 22

Using genomics to restore and future-proof underwater seaweed forests

on the ecological collapse of coastal ecosystems is critically important for identifying appropriate remedial actions and designing successful restoration programs. Below we describe approaches to restoration designed to ameliorate these stressors into the future.

2.4.1 Designing restored populations with genetics in mind Preserving genetic diversity in habitat-forming species has long been recognised as one of the key pillars of restoration science (Society for Ecological Restoration, 1993; International Union for Conservation of Nature, 2013), and is predicted to become increasingly relevant in highly variable environments or those subject to rapid anthropogenic change (Hughes et al., 2008; Forsman and Wennersten, 2016). Genetic diversity can directly affect establishment rates and fitness, environmental resilience and the adaptive capacity of macrophyte populations (Hughes and Stachowicz 2004; Forsman and Wennersten, 2016; Wernberg et al., 2018). In habitat formers, genetic diversity has been suggested to influence associated biodiversity and biomass, as well as ecosystem functions such as primary productivity, rates of decay and flux of nutrients (Whitham et al., 2006; Hughes et al., 2008; Reynolds et al., 2012; Kettenring et al., 2014). Although recent innovations in next-generation sequencing have provided cost-effective tools to study genes at multiple scales (Morin et al., 2004; Selkoe et al., 2016), the application of genetic and evolutionary theory to restoration design, practice and assessment remains in its infancy (Mijangos et al., 2015). Many restoration genetics studies over the past three decades have focused on determining the minimum amount of donor stock needed, as well as defining the spatial limits of donor provenance needed, to preserve local genetic diversity (Mijangos et al., 2015). Sourcing from too many locations has traditionally been discouraged in order to avoid outbreeding depression and introducing maladapted ecotypes into restored populations (e.g. Calumpong and Fonseca, 2001; McKay et al., 2005; Vander Mijnsbrugge et al., 2010). This may be due to a strong research focus on long-lived clonal macrophytes, such as seagrasses, which generally have higher genetic structure and outbreeding depression over regional scales than other sexually reproducing marine species (Lu and Williams, 1994; McKenzie and

Using genomics to restore and future-proof underwater seaweed forests

Bellgrove, 2006; Coleman and Kelaher, 2009). However, the relevance of these concerns to most restoration cases has recently been challenged (Frankham, 2015). Calls for the standard practice of evaluation of genetic rescue (i.e. mixing of novel genotypes into a recipient population to increase genetic variation of fragmented populations) reflect a shifting paradigm in conservation genetics (Ralls et al., 2018). New challenges to the way we conduct restoration are also emerging as climate change, habitat fragmentation and other anthropogenic pressures threaten to radically alter species distributions, with resulting reductions in gene flow and the loss of rear-edge (low-latitude and warm-adapted, often genetically diverse and distinct) genotypes expected to hamper subtidal macrophyte species’ adaptive potential over the next 50 years (Assis et al., 2017; Martínez et al., 2018; Wernberg et al., 2018). Thus, to avoid losses at all temporal scales, planning long-term solutions that include population resilience (the capacity to recover quickly) as an explicit objective is of utmost importance (Timpane-Padgham et al., 2017). For example, this may include ‘mixing in’ genotypes that have adaptive potential or are already adapted to conditions such as warmer waters or consumer resistance (assisted gene flow). There are some examples of the identification of macrophyte donors that may be suitable for such an approach (Clarke et al., 2013; O’Leary et al., 2017; Zerebecki et al., 2017;), although evidence of success in situ is limited (Wernberg et al., 2010; Bennett et al., 2015; Kleynhans et al., 2016). In cases where species have rapidly shrinking ranges, which is the case for many marine species (Chefaoui et al., 2018; Martínez et al., 2018; Wernberg et al., 2018), they may also be translocated to environments outside their historical range, where they are expected to persist in the future (assisted colonisation), a concept that has a fairly long history in terrestrial and aquatic environments (Gallagher et al., 2014; Armstrong et al., 2015). Another novel approach yet to be explored in macrophytes but one that is beginning to be explored in corals and their endosymbionts is the ‘artificial selection’ of genotypes bred under simulated future conditions (assisted evolution; van Oppen et al., 2015). Given the importance of diversity to resilience to stochastic and unforeseeable events, conserving a mixture of these ‘successful’ and rarer genotypes in a ‘portfolio approach’ is likely the most appropriate approach (Reusch et al., 2005; Webster et al., 2017; but see Houde et al., 2015). 24

Using genomics to restore and future-proof underwater seaweed forests

Although we recognise that it is important to thoroughly investigate the mechanisms and potential unexpected consequences of manipulating population genetics in specific systems before intervening, the time is ripe for a discussion of such techniques in marine macrophyte environments. This kind of thinking and analysis is already happening for terrestrial, and in some cases aquatic, systems. For example, guidelines for assessing vulnerability and evolutionary responses of species to climate change (Williams et al., 2008; Sgrò et al., 2011), determining the suitability of introducing genotypes for genetic translocations (Lesica and Allendorf 1999; Weeks et al., 2011; Hoffmann et al., 2015), including assisted colonisation (Gallagher et al., 2007), and advice on the genetic management of species with different mating systems and ploidy levels are already available (Frankham et al., 2017). Accumulating evidence from terrestrial systems also shows that highly divergent populations can be successfully crossed (Kronenberger et al., 2017) with benefits that persist to at least the F3 generation (Frankham, 2016). Genetic engineering or modification (GM), for example by using the clustered regularly interspaced short palindromic repeat (CRISPR) -CRISPR- associated protein 9 (Cas9) genome editing tool to engineer novel resilient organisms may also become more prevalent in the future. GM has a long history of research funded by agriculture and aquaculture, including for seaweeds (Beardmore and Porter, 2003; Qin et al., 2012; Lin and Qin, 2014). For example, Levin et al., (2017) discussed the potential use of gene editing to enhance the stress tolerance of coral endosymbionts Symbiodinium spp. in order to assist coral restoration. Aside from developing the science of this technology, use of GM in wild populations will require new policies, perhaps based on agricultural genetically modified organism (GMO) guidelines for use (for a project using GMO for terrestrial restoration purposes currently under legal consideration in the US, see www.esf.edu/chestnut/about.asp, accessed 25/02/2019). Efforts to genetically modify species in the wild would also require extensive further consultation among scientists and the public, including ethical considerations, and careful consideration of manifestations into real-world ecological scenarios.

Using genomics to restore and future-proof underwater seaweed forests

2.4.2 Managing species interactions Changes in species interactions need to be fully considered to successfully future- proof restoration efforts. In many instances an important obstacle to restoration success, especially in its initial stages, is herbivory (Carney et al., 2005; Campbell et al., 2014a; Yoon et al., 2014). The magnitude and importance of this problem is likely to increase in the future because climate change is leading to increases in herbivory in temperate reefs, with biodiverse and abundant tropical marine herbivores expanding their distribution into temperate regions worldwide (Vergés et al., 2014; Hyndes et al., 2016). Reducing effects of herbivores may include planning techniques, such as determining optimum patch size and density for recruitment success (in some cases, higher seeding densities can saturate consumers; Grant et al., 1982). Transplant success may also be enhanced by seeding within patches of chemically defended unpalatable macrophytes (Westermeier et al., 2013) or in association with artificial plants that can either deter herbivores because of their unpalatability (Tuya et al., 2017) or remove consumers via a whip effect (Vasquez and McPeak 1998). Finally, herbivory also varies greatly in space and time (Duffy and Hay, 1990), and loss to consumers can be minimised by restoring macrophytes during times of the year when herbivory is lowest (typically in the winter; Carney et al., 2005) or in refuges that are least accessible for herbivores (e.g. wave-exposed shallow sites; Duggins et al., 2001). Recovery of trophic structures within marine reserves (Babcock et al., 2010; Coleman et al., 2015) could also afford restored macrophytes increased protection from climate-induced range expansion of herbivores (Vergés et al., 2014; Hyndes et al., 2016) relative to areas open to extractive activities. Restoring in marine reserves and other types of protected areas (e.g. aquatic reserves, scientific reserves) may afford increased levels of protection from extractive activities including harvesting, new development and associated stressors (e.g. Coleman et al., 2013). These areas are often chosen for their role in connecting habitats (Coleman et al., 2011a, 2011b), but may also provide macrophytes in surrounding areas with enhanced protection from extant and future stress and maximise the chances of long-term success and persistence of restoration efforts.

Using genomics to restore and future-proof underwater seaweed forests

In some systems, the removal of herbivores may in itself be the main method of restoration. On small scales this may include protecting the most vulnerable juvenile transplants from herbivores such as fish using meshes, cages or other devices, particularly during early stages of restoration (Carney et al., 2005; Yoon et al., 2014). The removal of sea urchins has also been used with some success to re-establish underwater seaweed forests in some regions (Ford and Meux, 2010; Watanuki et al., 2010; Sanderson et al., 2016). However, this approach is logistically demanding and expensive, because it generally requires sustained and regular removal of sea urchins. New approaches now seek to enhance the sustainability and alleviate the costs associated with the continued removal of sea urchins by developing a parallel sea urchin fishery (D. Fujita and J. Keane, pers. comm.). Alternatively, focusing herbivore removal efforts to areas where such activity may have the most benefit for macrophytes (e.g. incipient barrens rather than established barrens) has been suggested to be a more effective restoration pathway (Johnson et al., 2017). The management of species interactions to enhance restoration is often biased towards negative interactions such as herbivory or competition, but positive interactions such as facilitation can also play a key role (Halpern et al., 2007). Facilitation is predicted to be more important in stressful warmer conditions (Gilman et al., 2010), and new restoration efforts that incorporate facilitation principles as our oceans warm may enhance restoration outcomes. For example, planting mixtures of seagrass species improves their overall survival and growth (Williams et al., 2017) and restoration of slow-growing seagrasses that typically dominate communities can be facilitated by first transplanting fast-growing opportunistic species (Kenworthy et al., 2018).

2.4.3 Harnessing microbial interactions Emerging evidence shows that host-associated microbial communities often have critically important roles in the normal development (Marshall et al., 2006), functioning (Ruby and Nealson 1976; Dubilier et al., 2008) and defence (Engel et al., 2002; Sharp et al., 2007) of their hosts and in biogeochemical cycling processes

Using genomics to restore and future-proof underwater seaweed forests

in the sediments, oceans and atmosphere (Arrigo, 2004; Azam and Malfatti 2007). As such, an important first step in restoration activities should include an assessment of host-associated microbial communities, particularly focusing on pathogens that may lead to macrophyte diseases at a restoration site before, during or after restoration (Gellie et al., 2017; Marzinelli et al., 2018). For example, transplantation of corals for restoration was shown to have negative effects on their associated microbiomes, leading to dysbiosis, disease and low survivorship of transplanted fragments (Casey et al., 2015). However, only few other marine examples exist, highlighting the need to incorporate an understanding of microbial communities in assessments of the suitability and success of restoration activities in marine ecosystems. One recent study found that microbial communities associated with restored seaweeds generally remained similar to those associated with seaweeds in extant populations (Campbell et al., 2015). There are also reports of seagrass restoration leading to changes in sediment microbiomes after ~12 months (Bourque et al., 2015), but how microbiomes on the surfaces of leaves and rhizoids respond and how this may affect restoration success is unknown. Manipulation of microbiomes may also become a useful tool for enhancing the success of restoration activities in marine ecosystems. Such techniques are reasonably well established in terrestrial restoration projects. For example, the addition of salt-tolerant bacteria (Bacillus spp.) to seedlings of coastal plants requiring restoration significantly increased plant growth (Hong and Lee, 2014). Inoculation of plant growth-promoting bacteria, a practice that has been used in agriculture for centuries, could be applied to many restoration projects, including seagrass habitats, to enhance success in stressful environments into the future (Holguin et al., 2001; Bashan et al., 2014). The development of such tools requires an understanding of which microbial taxa or functions enhance the resilience of the host to extant and future stressors, which can be challenging given the astounding diversity present in these communities (Thompson et al., 2017). Once the key taxa or functions are identified, methods can be developed to promote suitable bacterial strains (prebiotics or probiotics) to the host or the environment. These steps require going beyond basic descriptions of patterns generated from environmental gene sequencing (‘-omics’), towards experimentally showing cause– 28

Using genomics to restore and future-proof underwater seaweed forests

effect via manipulations before interventions. Similar approaches to those already used to inoculate soils in terrestrial systems may be used to inoculate sediments in seagrass systems. Similarly, inoculation methods designed to study diseases in subtidal macrophytes (e.g. Campbell et al., 2014b) may also be used to manipulate surface-associated microbiomes to enhance the resilience of the host (Egan et al., 2013).

2.4.4 Developing sustainable methods for restoration Although developing new methods and approaches to enhance restoration outcomes is important, this needs to be done while also prioritising the conservation of at least some remaining wild populations. Individual transplants harvested from wild populations remain the primary mode of seagrass and macroalgal restoration (Reigersman et al., 1939; Marion and Orth, 2010; Gianni et al., 2013). Although some studies have demonstrated that most seagrass donor populations may recover easily from infrequent (less than once a year), small-scale, dispersed harvesting activities (Fonseca et al., 1994, 1998; Verduin et al., 2012), in general the effects of harvesting for restoration are highly variable, dependent on species, environmental conditions, size of harvesting patches and techniques used and remain largely unexplored in many instances (Marion and Orth, 2010; Verduin et al., 2012; Matheson et al., 2017). Given that restoration often requires multiple transplantation efforts for success (van Katwijk et al., 2009), continuing large-scale restoration practices into the future may also run the risk of jeopardising the viability of donor populations to naturally recover after a perturbation, particularly when coupled with increasing fragmentation and environmental stressors, as has happened in saltmarsh ecosystems (e.g. Laegdsgaard, 2006). To avoid this, it is crucial to determine the optimal intensity, frequency and timing of harvesting of donor populations in order to develop harvesting strategies that are sustainable now and into the future. Alternatively, one solution may be to explant material that has already been dislodged following severe weather events, as is being trialled in eastern Australia (Vergés et al., pers. comm.). In some cases, especially for vulnerable and slow-

Using genomics to restore and future-proof underwater seaweed forests

growing species, it will likely be necessary to culture macrophytes in specialised facilities specifically for restoration purposes. In many cases, the basic ecological knowledge needed to do this with relevant species already exists (Falace et al., 2018; Verdura et al., 2018). However, high mortality of lab-reared seedlings once explanted to the field (Cumming et al., 2020) as well as basic lack of funds, facilities and infrastructure necessary for the establishment of seed banks and nurseries (already used in terrestrial and coral systems; Merritt and Dixon, 2011; Rinkevich, 2014) at the levels required for restoration thus generally result in current harvesting practices. This is changing, as researchers increasingly harness mariculture techniques and facilities to conduct rapid, large-scale fertilization and multiple outplantings of individuals for restoration. For example in Australia, a dedicated culturing and seeding facility was recently set up at the University of Tasmania to facilitate efforts to restore forests of Macrocystis porifera along the Tasmanian Coast in conjunction with marine permaculture activities (C. Layton, pers. Comm). Recent pilot work by Statton and Kendrick. (unpubl. data) to streamline techniques for seed collection and processing has also made improvements to the availability and delivery of large quantities of viable, ‘restoration-ready’ Posidonia australis seagrass seeds and work by this group has already produced new seagrass shoots in situ in Western Australia (see http://seagrassrestoration.net/seed-based-restoration-1, accessed 24/02/2019). Understanding processes that enable the establishment of large-scale nurseries for multiple species, similar to those used in forestry and terrestrial restoration of mine sites, remains a priority.

2.4.5 Integrating restoration into development Restoration can also be integrated into development and management strategies, especially in urban areas. In addition to developing or updating waste disposal methods that improve water quality (e.g. Scanes and Phillip, 1995; Sydney Water, 2017), designing seascapes, in the same way terrestrial land planners design landscapes as interspersed mosaics of natural and artificial habitat, may be used to improve the environmental effects of ocean sprawl. Ecologically appropriate

Using genomics to restore and future-proof underwater seaweed forests

designs for artificial structures have received considerable attention (e.g. Dafforn et al., 2015), but approaches to designing seascapes that combine artificial and natural structures of habitats are only starting to emerge (e.g. Firth et al., 2014), including for macrophytes (Davis et al., 2017; Mamo et al., 2018. Perkol-Finkel et al., (2012) demonstrated successful seeding of macrophytes onto artificial structures (breakwaters), with success dependent on various characteristics of the substratum. This group also highlighted the importance of understanding the ecology of the target species and system, because the prospects for restoration were significantly constrained by herbivory (Ferrario et al., 2016). Although these challenges need to be overcome, such studies clearly raise the potential for targeted design of artificial structures to either facilitate seeding or transplantation of macrophytes onto the structures, or to encourage natural colonisation in both current and future oceans (Davis et al., 2017). Strain et al., (2018) have reviewed features of artificial habitats that facilitate macroalgal success, and there are many design features of artificial structures (e.g. type of substratum, physical complexity or heterogeneity and slope, among others; Dafforn et al., 2015) that are readily manipulable and thus could be brought into play to enhance the restoration of macrophytes. Along these lines, approaches to the design of artificial reefs so as to target specific species (although, to date, typically fish) has been reviewed by Baine (2001), and artificial reef design companies often tout such taxa-specific (including macrophytes) design considerations in their products (http://haejoo.com/purpose-built-artificial-reef/, accessed 24/02/2019).

2.5 Moving forward: issues of management, political and public support

2.5.1 Who does and who pays for restoration? The cost of restoration can vary considerably according to the techniques used and whether further research and development is needed to optimise results. For example, in Australia basic seaweed transplanting efforts are estimated at ~US$6,850 per restoration site plus project management costs of ~US$18,500, whilst assisted recovery techniques such as urchin culling have been estimated at ~US$980,478 per 1 km2 (Layton et al., 2020). Research costs e.g. those associated

Using genomics to restore and future-proof underwater seaweed forests

with genetic sequencing and experimental phases should be considered alongside the often hidden cost of using sub-optimal techniques such as restoring with some maladapted genotypes under the portfolio approach. Of fundamental importance for any restoration project are the related questions of who pays for it (Holl and Howarth, 2000; Richardson and Lefroy, 2016) and who is responsible to see that it is implemented. In an ideal world, the entity that damages or degrades a habitat would pay and be responsible for its restoration or rehabilitation. This is not simple in any system, but it is particularly complicated in marine systems because the degradation is often due to multiple factors derived from multiple entities or activities (Crain et al., 2008). In addition, the open nature of most marine systems means there is often a disjunction between those who cause the degradation and those who bear the costs of the effect. Examples of this include diffuse coastal pollution from terrestrial run-off, or the climate change- driven invasion and removal of temperate kelp forests by immigrant herbivores arriving via warm water currents (Ling et al., 2009; Vergés et al., 2014). Further confounding the issue of ‘who pays’ is the nature of ownership and zoning of marine spaces, particularly relative to those on land. In coastal and national territorial waters, marine spaces below the low tide mark are usually publicly owned and governed, but the benefits of restoration efforts typically flow to a wide range of both private and public stakeholders. This contrasts with many terrestrial situations, where the boundaries of ownership are usually clearly delineated, the direct causes of the degradation often more localised and obvious (e.g. deforestation) and the entities to whom potential costs and benefits would flow also more readily identified. This review focuses on coastal systems and macrophytes, but all these issues of ownership and benefits become even more complicated in open ocean systems. Funding of restoration efforts comes from three sources: government, philanthropic or private (e.g. development mitigation; Richardson and Lefroy, 2016). In Australia, and to a large extent globally, marine restoration efforts are almost entirely funded through government or philanthropic efforts (Richardson and Lefroy, 2016). This makes the selection of where and what to restore problematic, because this selection is often governed by serendipity, the interest 32

Using genomics to restore and future-proof underwater seaweed forests

of local scientists or managers in a particular habitat or species or the public profile or political resonance of that habitat or species (http://www.gbrmpa.gov.au/media-room/latest-news/corporate/2018/$500- million-funding-game-changer-for-the-great-barrier-reef, accessed 24/02/2019). Cunningham (2002) has suggested that we are entering the era of the ‘restoration economy’, where more and more economic activity is derived from existing infrastructure and repairing ecosystems rather than investing in new infrastructure. However, business models, and thus the structure of financing for marine restoration, are still poorly developed. Some positive steps are being made in the development of such models, such as the Reducing Emissions from Deforestation and Forest Degradation (REDD+) initiative, or voluntary markets and other programs from the United Nations Framework Convention on Climate Change (UNFCCC; e.g. Wylie et al., 2016), which are being considered for ‘blue (marine)’ carbon stocks (Lovelock et al., 2017; Macreadie et al., 2017). In general, however, marine macrophyte restoration would greatly benefit from a more formal and concerted effort on development of models that explicitly link economic benefits of restoration (be it carbon stocks, tourism, fishing or other benefits) to their costs, thus hopefully providing a structure for determining who should pay.

2.5.2 Public engagement, social licence and science communication Finally, restoration solutions are heavily dependent on community attitudes and need to reflect society’s common values and beliefs as they change (Martin, 2017). Relying on conjecture or unsubstantiated claims in the literature for the cause and extent of macrophyte loss can be misleading and result in misplaced efforts, failed restoration, and an eventual lack of confidence in the ability of science to serve the public’s needs (e.g., see Foster and Schiel., 2010). Involving communities in restoration projects from the outset and being supported by governance can result in durable community effort that contributes to restoration success and enables restoration at larger scales (Datta et al., 2012; Brown et al., 2014; On-prom, 2014). Community participation can also save money on restoration costs (Watanuki et al., 2010) and is linked to increasing environmental awareness more broadly (Peters

Using genomics to restore and future-proof underwater seaweed forests

et al., 2015), as well as providing a sense of ownership and responsibility (dela Cruz et al., 2014). Although fostering direct participation from the public in restoration activities can be challenging in subtidal environments due to work and safety issues, new modes of engaging the public can be highly successful in raising awareness about local degradation and the role of science and restoration in improving our coastlines. The rise of online modes of participation, whereby participants either upload or help process photographic materials, are becoming increasingly successful (Brancalion and Van Melis, 2017). For example, in the project Explore the Seafloor (http://exploretheseafloor.net.au/, accessed 16/01/2019), more than 8000 citizen scientists processed over 225 000 photographs to map the location of kelp and sea urchin populations around Australia. Films and social media are highly effective in connecting the general public with the underwater world, which is inaccessible to many. Social media in particular is becoming an essential tool to raise awareness about environmental projects generally and has been used successfully in crowd-funding campaigns to support restoration (e.g. www.operationcrayweed.com, accessed 24/02/2019). The arts are also increasingly used to communicate science to the public (Lesen et al., 2016). In Australia recently, coastal sculptural works have showcased seaweed and seagrass restoration (see http://studiotcs.com.au/work/operationcrayweed-art-work-site, accessed 24/02/2019; Pfeiffer et al., 2017) and have provided a platform for science communication that led to a significant increase in respondents’ ‘willingness to pay’ to restore seaweed forests over more distantly located coral reefs (Kajilich et al., unpubl. data). Aside from promoting individual projects, as underwater restoration enters new territory at larger scales the public will need to better understand the science used to assess whether these new techniques align with social ideology, identity and trust (Groffman et al., 2010). The use of artistic platforms may render the ethical considerations of intervention ecology in plainer relief for the general public, enabling broader discussion of the social licence for new technologies and ecological or engineering solutions. For example, in 2017 The Museum of Underwater Art (http://moua.com.au, accessed 24/02/2019) was launched on the 34

Using genomics to restore and future-proof underwater seaweed forests

Great Barrier Reef and aims to stimulate public engagement with existential issues facing marine life, with the artwork also functioning as surrogate ecosystems for the original degraded ones. Ultimately, however, it is up to scientists to bring these solutions to the public’s attention so that non-experts may understand their implications and embrace their potential.

2.6 Concluding remarks Many marine ecosystems worldwide are already drastically different to those that existed 100, or even 20, years ago. The restoration of subtidal marine ecosystems is not only a significant concern for scientists, but it is also an issue of equity and justice for the millions of people who benefit from marine ecosystem goods and services around the world. In many instances, the threats discussed here (i.e. climate change, ocean sprawl, overfishing and water quality) are already upon us, and the effects of some of these will increase in the coming decades. Therefore, a major challenge is to implement restoration solutions, perhaps multiple and in combination, before current effects on ecosystems are too great to rectify. In subtidal macrophyte systems, the technologies discussed in this paper may still be in their infancy but are more advanced and accessible than ever before. Although the nature of the marine environment adds to the challenges involved in conducting this work, increased interest in seaweed and seagrass systems will ultimately lead to the development of solutions that can be applied to both current and future issues. This paper aims to highlight the key elements of emerging research that should be investigated in order to promote the resilience and longevity of these ecosystems and the services they provide into the future.

Using genomics to restore and future-proof underwater seaweed forests

Chapter 3:

Using genetics to optimise and measure success in restoration of underwater forests

Published as: Wood G., Marzinelli E.M., Vergés A., Campbell A.H., Steinberg P.D., Coleman M.A. Using genomics to design and evaluate the performance of underwater forest restoration. J Appl Ecol. 2020;57:1988–1998. https://doi. org/10.1111/1365- 2664.13707

Using genomics to restore and future-proof underwater seaweed forests

3.1 Abstract Restoration is an emerging intervention to reverse the degradation and loss of marine habitat-formers and the ecosystem services they underpin. Current best practice seeks to restore populations by transplanting donor individuals chosen to replicate genetic diversity and structure of extant, nearby populations. However, genetic characteristics are rarely empirically examined across generations, despite their potential role in influencing restoration success. We used genetics to design a restoration program for lost underwater forests of Phyllospora comosa, a dominant forest-forming macroalga that went locally extinct from reefs off Sydney, Australia. Population genetic diversity and structure of nearby extant populations informed choice of donor sites. We further tested whether donor provenance influenced restoration success. Extant populations of Phyllospora within a 100 km radius of Sydney comprised three distinct genetic clusters with similar levels of genetic diversity. We transplanted reproductive adults from two of these donor sites, with the aim of restoring five Phyllospora forests with similar levels of genetic structure and diversity to nearby extant populations. Although donor provenance influenced survival of transplanted adults, recruitment of the first (F1) generation was rapid and genetic diversity and structure of this generation successfully resembled extant populations. This likely occurred because transplanted individuals reproduced synchronously and rapidly post-transplantation, prior to mortality of adult donor transplants, overcoming a major bottleneck in marine forest restoration. As restoration and the need to “future-proof” marine ecosystems increase globally, it will be critical to understand the role of genetic diversity and structure in restoration success. This study demonstrates that evidence-based selection of donor macroalgal forests using knowledge of genetic diversity and structure of extant populations can result in restoration success across generations.

Using genomics to restore and future-proof underwater seaweed forests

3.2 Introduction

Global habitat degradation affects essential ecosystem functions and services and is one of the most serious threats faced by humanity (IPBES, 2018). Recovering degraded or lost habitats and maintaining ecosystem functions through active interventions, such as restoration, is now a major focus for management and conservation (Kovács-Hostyánszki et al., 2017; UN, 2019). Understanding the factors that enhance restoration success is therefore crucial. In particular, decisions regarding the geographic or genetic origin (‘provenance’) of donor biological material can profoundly influence restoration success (Miller et al., 2017) but are only beginning to receive attention (Breed et al., 2018; Mijangos et al., 2015).

Provenance generally defines the genetic diversity, identity and structure of a restored population and thus much of the population’s ability to persist, respond and adapt to subsequent change (Bischoff et al., 2010; Sgrò et al., 2011). Many guidelines for restoration practitioners recommend mimicking genetic characteristics of natural populations by sourcing donors from multiple extant, “local” populations (Bischoff et al., 2010; Bucharova et al., 2019), although alternative strategies involving the use of distant and/or stress-tolerant populations are increasingly being proposed in response to human impacts such as landscape fragmentation and climate change (Broadhurst et al., 2008; Wood et al., 2019).

Most restoration programs, however, are still undertaken without any a priori knowledge of background population genetics (Mijangos et al., 2015). Even when some knowledge of provenance is utilised in restoration design, the maintenance of genetic diversity or structure and its relationship to restoration success is rarely quantified, particularly over more than one generation (Mijangos et al., 2015). This can lead to restored populations with poor genetic characteristics and subsequent failure (Granado et al., 2018; Williams, 2001). To rectify these gaps, the combined use of genetic tools and provenance trials embedded into the design, implementation and assessment of restoration projects are needed (Breed et al., 2018). 38

Using genomics to restore and future-proof underwater seaweed forests

Underwater macroalgal forests underpin ecosystem goods and services along temperate rocky coastlines (Steneck and Johnson, 2013) but they are declining in many places around the world (Krumhansl et al., 2016). Restoration of underwater forests is still in its early stages (e.g Layton et al., 2019) and long-term success is difficult to predict. The youth of such programs creates an opportunity to include provenance and other demographic considerations in their design.

Here, we used population genetics to delineate appropriate provenance and assess effects on restoration success of Phyllospora comosa (hereafter, Phyllospora), a dominant macroalga that forms extensive underwater forests along the south-east coast of Australia (Coleman and Wernberg 2017). Phyllospora underpins coastal biodiversity and valuable ecosystem functions and services (Coleman and Wernberg 2017, Bishop et al., 2010, Marzinelli et al., 2014), but disappeared from 70km of Sydney’s coastline in the 1970-80s (Coleman et al., 2008). While this decline was likely due to poor water quality, which has since improved (Coleman et al., 2008; Scanes and Philip, 1995), Phyllospora has not returned naturally – likely due to recruitment limitation. However, transplanted Phyllospora can survive and reproduce in Sydney (Campbell et al., 2014) demonstrating that active interventions are likely to be effective in re-establishing Phyllospora forests where they have been lost.

To identify appropriate provenance for Phyllospora restoration, we used Single Nucleotide Polymorphisms (SNPs) to characterise genetic diversity, structure and the effect of geographic distance on extant populations surrounding the gap in distribution. Given the absence of historical genetic data from Sydney, we used this genetic information to restore five populations with the aim to mimic genetic diversity and structure of nearby extant populations. We quantified differences in morphology, survival and condition of transplanted individuals of differencing provenance and assessed subsequent impacts on the establishment, genetic diversity and likely origin of first generation (F1) recruits. If provenance influenced restoration success via the survival and condition of transplanted adults, we predicted that this would result in differences in the genetic diversity and structure of the F1 generation compared to extant donor populations.

Using genomics to restore and future-proof underwater seaweed forests

3.3 Methods

3.3.1 Study species Phyllospora is a dioecious perennial macroalga found from Port Macquarie in northern New South Wales (NSW) to Southern Tasmania in Eastern Australia (Underwood et al., 1991; Wormersley, 1987). In NSW, Phyllospora forms dense beds on exposed rocky reefs from the low tide mark to c. 5 m depth (Underwood et al., 1991). It has a lifespan of ~ two to six years and is reproductive year-round (Cumming et al., 2019). Reproduction occurs via spawning of motile sperm into the water column, which fertilise stalked eggs attached to the fronds of female adults and drop to the seafloor (Burridge, 1990). Gas filled vesicles may also facilitate dispersal of fertile wrack and likely result in long distance dispersal and connectivity (Coleman and Kelaher, 2009).

3.3.2 Determination of appropriate provenance for restoration

3.3.2.1 Sample collection To characterise patterns of genetic diversity and structure of extant Phyllospora populations, we sampled three sites north and three sites south of the distributional gap in Sydney (Fig. 3.1), over the Austral summer of 2016 (November - December). At each site, 30 reproductive individuals (>1 m apart) were haphazardly sampled at depths of 1-5 m from an area of 500 m2 of reef. The sex, density of reproductive conceptacles (measured within 3 x 25mm2 quadrats on the thallus frond) and wet weight biomass of each individual was recorded to assess any biases in sex ratios and potential reproductive capacity at each site. Lateral branches were removed from thalli and kept cool until processing (within 48 hours of collection). From each individual, 30 unfouled apical tips were removed, rinsed in fresh water and dried to remove external salt, epiphytes and water (see Coleman and Brawley, 2005). Samples were snap-frozen in liquid nitrogen and then stored at -80°C.

Frozen frond tips (~25 mg) were ground to a powder in a Qiagen Tissuelyser 2000 using stainless steel beads without thawing. DNA was then extracted using

Using genomics to restore and future-proof underwater seaweed forests

the Qiagen DNeasy Plant Mini kit with some modifications (see Appendix B). DNA was extracted over 2-4 reactions and pooled for each algal sample, then cleaned using a Qiagen PowerClean Pro Cleanup kit. Quantity/purity of DNA was measured in Qubit 2.0 (ThermoFisher Scientific Inc) and Nanodrop 2,000 UV-Vis (ThermoFisher Scientific Inc) respectively.

Using genomics to restore and future-proof underwater seaweed forests

Figure 3.1: (a) Genetic structure of Phyllospora comosa populations. Left: STRUCTURE plot showing individuals from extant populations assigned to 3 inferred clusters. Each column represents an individual; different colours within columns indicate maximum likelihood probability of belonging to different clusters. Right: Map of extant and restored sites coloured according to average probability of belonging to each genetic cluster. (Legend continues overleaf)

Using genomics to restore and future-proof underwater seaweed forests

(Legend continued from previous page): Sites from top to bottom are: BB: Bateau Bay, TE: Terrigal, PB: Palm Beach, WH: Whale Beach, FW: Freshwater, SO: South Head, CO: Coogee, MA: Maroubra, CR: Cronulla, SP: Shark Park and SH: Shell Harbour. The three pie charts on either side of the black semicircle represent extant populations and four smaller pie charts on the line represent recruits at restored sites. (b) Photograph showing transplanted algae from the northern donor site (BB: shorter and lighter) and southern donor site (SP: taller and darker).

Using genomics to restore and future-proof underwater seaweed forests

3.3.2.2 Genotyping and bioinformatics 177 samples from across the six sites were genotyped using an Agena Bioscience MassARRAY with iPlex GOLD technology on a custom panel of 354 SNP loci. These had been previously established by genotyping by sequencing (GBS) runs of seven samples at the Australian Genome Research Facility (AGRF, www.agrf.org.au) that were subjected to preliminary assay design (MassARRAY software) to select the top SNPs with a high Minor Allele Frequency (MAF) and reasonable flanking sequence.

Subsequent bioinformatics and data analyses were conducted using the statistical platform R (version 3.6; R Core Team, 2019). The quality of genotyping was then assessed for each locus in the sample collection using the poppr package (Kamvar et al., 2014). SNPs and samples with a call rate below 90% of the total, or with a minor allele frequency (MAF) below 0.05 were excluded from further analysis.

3.3.2.3 Genetic diversity and structure in extant populations Exact tests for Hardy-Weinberg Equilibrium (HWE) deviations were calculated across all loci using hierfstat (Goudet, 2005). We then estimated Linkage disequilibrium (LD) for each locus and across all loci using Fisher’s exact tests in genepop 1.1.2 (Rousset, 2008) with 10,000 dememorization and in 100 batches with 999 iterations per batch. Multiple tests in the detection of HWE and LD were corrected using the false discovery approach (Storey, 2002). None of the pairwise comparisons of loci were found to be significantly in LD. Five loci deviated from HWE across all sites and were removed, leaving 118 loci across 177 samples.

Genetic differentiation and diversity were then evaluated by generating estimates of observed heterozygosity (HO), expected heterozygosity (HE), and heterozygote excess (FIS) for each locus and for each sampling group using the R package diveRsity (Keenan et al., 2013). FIS estimates were assessed for significance using 1000 permutations with 95% confidence intervals. We also calculated allelic richness (with allele counts rarefied by the minimum number of individuals

Using genomics to restore and future-proof underwater seaweed forests

genotyped) and the Shannon-Weiner diversity index using the hierfstat and poppr packages, respectively.

Genetic structure was assessed by estimating Wier and Cockham’s FST (the proportion of genetic variance contained in a subpopulation relative to the total genetic variance). Pairwise comparisons of FST between sampling locations and their significance were assessed using bootstrapping (999) to construct 95% confidence intervals in hierfstat. To identify the number of genetic groups in the dataset, we used STRUCTURE version 2.3.4 (Pritchard et al., 2000). The number of possible genetic clusters (K) varied from one to seven (the number of analysed locations plus one) and was assessed using 20 independent runs with a 10,000 burnin time and a Markov Chain Monte Carlo iteration of 200,000. This was performed on a dataset which allowed admixture and assumed correlated allele frequencies with no prior information about populations. To determine the most probable value of real clusters (K), we used the ad hoc criterion (Evanno et al., 2005). The software CLUMPAK (Kopelman et al., 2015) was then used to find the optimal alignment of multiple replicate analyses of each K and to graphically display them. To calculate the percentage of genetic variation attributed among and within sites, an Analysis of Molecular Variance (AMOVA) was performed using the poppr package, also using the broader genetic clusters identified with the STRUCTURE analysis (“north”, “north-south mixed”, “south” and “Shell Harbour” clusters) as a priori groupings. We also performed a Mantel test using the ade4 package (Dray and Dufour, 2007) to examine the relationship between geographic distance on genetic distance. As the coastline was relatively linear at the broad scale needed for site comparisons, the dist function was used to create the geographic distance matrix by calculating Euclidean distances in geographic space between collection sites based on their coordinates. For the genetic distance matrix, we used the collection site-based pairwise FST values generated from hierfstat.

Using genomics to restore and future-proof underwater seaweed forests

3.3.3 Restoration experiment

Donor male and female Phyllospora thalli were haphazardly collected from Shark Park (SP; predominately characterised by the southern genetic cluster mixture with some northern) and Bateau Bay (BB; predominately northern genetic cluster; see Results). Ninety reproductive-aged, “healthy” (i.e. the thallus had less than <5% percentage cover of grazing, epibiosis and bleaching when visually assessed during collection) individuals from each population were transplanted to each of 5 sites in Sydney that had previously been identified as suitable habitat (moderately exposed and having large, flat boulders at 4-5 m depth) in the austral Spring of 2017 (October-November). Mats were placed between 1- ~5m away from each other. Differences in morphology between algae of each provenance were tested by comparing thallus length and number of branches using t-tests in R. Algae were transplanted over 1-3 days by cable-tying individuals in natural densities (15 algae / m2) to 6 x 2 m2 plastic mats per site that had been attached to the rocky reef (as per Campbell et al., 2014). All transplanted thalli were treated in the same way (e.g. similar collection approach, time out of the water; Campbell et al., 2014). Individuals from the two donor sites were evenly distributed (mixed) across each of the mats at each site, and they were identified using cable tie attachment units of different colours (Fig. 3.1).

3.3.3.1 Survival and condition of transplants The number of adult algae from each donor site with holdfast, stipe and fronds present was quantified six and nine months after transplantation. Measurements of survival were pooled across the six mats per site according to transplant provenance (BB or SP). To assess transplant condition, we also quantified the percentage of epibiosis, which can negatively affect algae and indicate stress. To compare the effects of donor provenance (BB versus SP, as a fixed factor), restoration site (random factor) and their interaction on (i) the percentage of survivors per mat and (ii) percentage of epibiosis per individual, we fitted linear mixed models in the lme4 package followed by F-tests with Kenward-Roger approximation using the KRmodcomp function in the pbkrtest package (Halekoh

Using genomics to restore and future-proof underwater seaweed forests

and Højsgaard, 2014). “Mat” was fitted as a random effect to account for non- independence between individuals of different provenance interspersed on the same mat. This was done for the data collected at six months only, as the numbers of adults remaining nine months following transplantation were very low (see Results). This mortality rate was similar to previous transplantation efforts (Campbell et al 2014) and was not deemed an issue as transplants do not re-attach to the seafloor and were transplanted only to provide a reproductive unit to produce the next generation of recruits. Normality and homogeneity of variance assumptions were checked visually with histograms of the residuals and scatter plots of residuals versus fitted values, respectively.

3.3.3.2 Recruitment and origin of the F1 generation There is no published evidence that Phyllospora recruitment is seasonal, however previous restoration attempts have shown that recruitment density at Phyllospora transplant sites is very high six months after transplantation and then declines to approximately “natural” levels (i.e. relative to reference populations) over the subsequent six months (Campbell et al., 2014). We quantified total numbers of

Phyllospora F1 recruits at each site ten months after transplantation, when recruits were expected to be well-established and were large enough to be non- destructively sampled.. Recruits were defined as Phyllospora individuals that were present at restoration sites but were not transplants. All recruits were still in the early life history stage when they were quantified (i.e. < 30cm tall). Dedicated searches up to 20m from the restoration mats revealed that the majority of recruitment was within 30cm of the mats, as observed in previous work (Campbell et al., 2014) (see Results). As adult survival had declined significantly by 10 months post-transplant and condition/survival at 6 months was deemed more relevant to reproductive output, we tested whether numbers of recruits found in or around each mat was dependant on total donor survival (at six months) using a linear model (data were pooled across sites). We then quantified genetic diversity and the relative contribution of north versus south donors to reproduction by sampling recruits at each restoration site as described above, although only one apical tip was taken from each recruit to 47

Using genomics to restore and future-proof underwater seaweed forests

avoid mortality. Overall, genomic data from a total of 30 recruit samples (20 for population genetic analysis at the most successful site, Coogee, and two-five from the remaining sites for comparison of genetic assignment) were genotyped and re- filtered as described above (extant populations). Only loci that were common to both donor and recruit populations following filtering were used, leaving 102 loci across 30 samples for subsequent comparative analysis. As we had low sample sizes of recruits at most sites (see Results), measures of genetic diversity could only be meaningfully interpreted for the recruit samples collected from Coogee, where n = 20. To check the number of genetic clusters among donor and recruit groups, multivariate analyses were carried out using discriminate analyses of principal components (DAPC) using ADEGENET (Jombart, 2008). The optimal number of clusters was selected based on the lowest Bayesian Information Criterion (BIC). Genetic assignment of recruits to parent populations was then performed using Monte-Carlo and K-fold cross-validation coupled with machine-learning classification algorithms in the assignPOP package (Chen et al., 2018). We used the existing data from the north and south donor populations as reference groups and used the Bayesian model approach to estimate membership probabilities to these. We did not test for differences in genetic assignment of recruits among restored populations as recruit sample sizes were very uneven between sites.

3.4 Results

3.4.1 Determination of appropriate provenance for restoration

Sex ratios were roughly 50:50 at all sites, except at Palm Beach which had 40% less females than males sampled (Fig. S3.1). Wet weight biomass and density of reproductive conceptacles also varied between sites, but individuals from both

Using genomics to restore and future-proof underwater seaweed forests

donor sites had similar densities of reproductive propagules and biomass (Fig. S3.1), suggesting comparable reproductive capacity in their sites of origin.

3.4.1.1 Genetic diversity and structure of extant populations

Mean allelic richness and genetic diversity estimates varied only slightly among the six extant populations sampled (AR: 1.95, SE 0.01, HO: 0.34, SE 0.01, HE: 0.33, SE 0.01, SD: 3.38, SE 0.03), with no difference between observed and expected heterozygosity (Bartlett’s test, p = 0.63; Table 3.1). Northern sites were characterised by low, positive FIS values (indicative of inbreeding trend), and southern sites were characterised by low, negative FIS values (indicative of outbreeding trend); however, none of the sites differed significantly from zero or each other (Table 3.1).

All pairwise FST tests between all pairs of sites confirmed that populations were genetically different (global FST: 0.05; see Table S3.1). STRUCTURE analysis revealed three genetic clusters: Bateau Bay (BB) and Terrigal (TE) formed a “northern” cluster, Shark Park (SP) and Shellharbour (SH) each formed their own cluster with little admixture, while Palm Beach (PB) and Cronulla (CR) (on either side of the gap in distribution) comprised mixes of the “northern” and SP clusters (Fig. 3.1a). A small but significant amount of genetic variation was explained by genetic clusters (AMOVA, 3.15%, p < 0.05, See Table S3.2) and sites (AMOVA, 2.99%, p < 0.05), while the majority of genetic variation occurred among individuals (AMOVA, 94.8%, p = 0.07). Geographic distance and pairwise FST values between each pair of sites were positively correlated (Mantel test, p < 0.05, r = 0.6, Fig. 3.2).

3.4.2 Differences between transplants from each provenance

At the time of transplantation, there were significant differences in the total thallus length (t86 = -3.48, p < 0.001) and number of branches (t66 = 7.58, p < 0.001) between algae from the two donor sites, with algae from the north significantly shorter and

Using genomics to restore and future-proof underwater seaweed forests

more branched (BB: 88.12, SE 3.81 cm and 24.88, SE 1.72, respectively) than algae from the south (SP: 107.02, SE 3.94 cm and 10.88, SE 0.66, respectively). Between 13-40% of transplanted adults remained in restored sites after six months (Fig. 3.3) and c. 8% (SE 2.73) remained across all sites after 9 months. At six months, survival of adults sourced from the north (BB: 31.3, SE 3.56%) was significantly higher than those sourced from the south (SP: 19.1, SE 3.56%; F1,4 = 11.002, p = 0.03). Most of the remaining algae had higher levels of epibiosis than when transplanted (<5%), with significantly higher epibiosis on adults sourced from the south (30.24, SE 3.30%) than from the north (22.92, SE 3.30%; F1, 3.7 = 8.5970, p = 0.047). Both of these patterns were consistent across all sites.

3.4.3 Recruitment and gene flow

Phyllospora recruits were observed at four transplant sites (Fig. 3.4a), with most recruits found on mats (92%) or within 30 cm of these (8%) ten months after adult transplantation, except for two recruits found 10 m away from the mats at Coogee. Total numbers of recruits differed widely between restoration sites (WH: 6, FW:11, SH: 0, CO: 96, MA: 5) and were not significantly related to total adult survival at six months (F1,28 = 1.349, p = 0.26). Allelic richness and genetic diversity estimates varied slightly between the

F1 generation and donor populations (Table 3.1). Observed heterozygosity was larger than or equal to expected heterozygosity in all sites, although this trend was non-significant (Bartlett’s test, p > 0.41). All of the F1 populations were characterised by low, negative FIS values with WH, FW and MA having values that differed significantly from zero, which is indicative of outbreeding, although the number of replicates were low (Table 3.2).

DAPC partitioned the donor and F1 recruit populations into two genetic clusters (Fig. 3.4b). Across all transplant sites, 17 recruits were assigned to Shark Park and 13 were assigned to Bateau Bay (Fig. 3.4c), likely indicating breeding between male and female donor algae from each of these sites, respectively. Probability of assignment varied from 0.503 to 1, with lower probability of recruits assigned to Bateau Bay than Shark Park (BB: 0.80, SE 0.07, SP: 0.89, SE 0.07; Fig.

Using genomics to restore and future-proof underwater seaweed forests

3.4d). Monte-Carlo cross-validation using assignPOP showed the mean self- assignment rates for the Bateau Bay and Shark Park populations following training were 77% and 83%, respectively, indicating there was some uncertainty around delineation between genetic clusters.

Using genomics to restore and future-proof underwater seaweed forests

Table 3.1: Genetic diversity of Phyllospora comosa from the six extant sites surrounding Sydney.

a b c d e f Site n Total AR Ho He SD FIS all loci alleles BB 26 234 1.98 0.337 0.341 3.26 0.011 TE 28 232 1.95 0.351 0.354 3.4 0.006 PB 29 231 1.94 0.309 0.319 3.37 0.032 CR 30 229 1.93 0.337 0.329 3.4 -0.026 SP 31 232 1.96 0.328 0.318 3.43 -0.033 SH 30 233 1.97 0.317 0.314 3.43 -0.012

aBB: Bateau Bay, TE: Terrigal, PB: Palm Beach, CR: Cronulla, SP: Shark Park, SH: Shell Harbour bRarefied allelic richness (AR) c Observed heterozygosity (HO) d Expected heterozygosity (HE) e Shannon-weiner diversity (SD) f Inbreeding coefficient (FIS). None significant.

Using genomics to restore and future-proof underwater seaweed forests

Figure 3.2: Relationship between geographic distance (km) and genetic distance (pairwise FST) between Phyllospora comosa individuals collected at six extant sites, fitted with linear regression. Comparison between the 2 donor sites BB: Bateau Bay and SP: Shark Park shown in red. 95% Confidence Intervals shaded in grey. The minor peak around 60-70 km corresponds to the gap in distribution in Phyllospora around Sydney.

Using genomics to restore and future-proof underwater seaweed forests

Figure 3.3: Transplant survival and condition. (a) Survival and (b) epibiosis of adult Phyllospora comosa transplants at restoration sites (WH: Whale Beach, FW: Freshwater, SO: South Head, CO: Coogee, MA: Maroubra) after six months. Coloured bars depict the provenance. * p<0.05.

Using genomics to restore and future-proof underwater seaweed forests

Table 3.2: Genetic diversity of Phyllospora from donor populations and restored F1 recruits.

a b c d e f Site n Total AR Ho He SD FIS all alleles loci BB 26 204 1.37 0.36 0.36 3.26 -0.01 WH 2 173 1.39 0.38 0.29 0.69 -0.32 a FW 5 188 1.36 0.39 0.32 1.61 -0.23

CO 20 202 1.36 0.36 0.35 3.00 -0.04

MA 3 182 1.37 0.38 0.30 1.10 -0.24 SP 31 202 1.34 0.35 0.33 3.43 -0.05

a Donor sites are in grey: BB: Bateau Bay and SP: Shark Park. F1 of restored populations are in white; WH: Whale Beach, FW: Freshwater, CO: Coogee, MA: Maroubra bRarefied allelic richness (AR) c Observed heterozygosity (HO) d Expected heterozygosity (HE) e Shannon-weiner diversity (SD) f Inbreeding coefficient (FIS). Significant values in bold.

Using genomics to restore and future-proof underwater seaweed forests

Figure 3.4: (a) Photograph of restored Phyllospora comosa recruits at Freshwater. (b) DAPC Scatterplot output donor and restored F1 recruits (boxed in legend) populations. (c) Membership probability plot showing probability of restored F1 recruits genetic assignment to Bateau Bay (BB) or Shark Park (SP) donors at four restored sites. WH: Whale Beach, FW: Freshwater, SO: South Head, CO: Coogee, MA: Maroubra. Horizontal dotted lines represent 95% confidence intervals for assignment to each site (= likely a pure bred cross between donor from the same site) (d) Summary table showing number recruits significantly assigned to BB, SP or a mix (threshold 0.95 or 0.8 as accuracy of model was ~80%; in brackets).

Using genomics to restore and future-proof underwater seaweed forests

3.5 Discussion Provenance is widely considered as vital for restoration success; however, it is rarely empirically examined. Here, we characterised three genetic groups of the dominant, forest-forming seaweed Phyllospora comosa surrounding Sydney and, for the first time, used this genetic information to design a restoration program. Our empirical tests showed that, despite differences in survival and condition of donors of different provenance, the subsequent F1 generation had levels of genetic diversity and structure that resembled a mix of the extant donor populations, likely due to the rapid reproduction of transplanted individuals. This suggests that it is possible to use population mixing to restore desired levels of genetic diversity and structure, even if transplant survival and condition are hampered by the restoration method and/or other factors at their destination. These techniques provide critical information for restoring genetically diverse Phyllospora forests at the large scales of loss. Moreover, these methods could be extended to “future- proof” populations via purposely selecting donors with desirable genetic traits in the context of climate change and other stressors. However, total recruitment varied significantly between sites, and recruitment levels were generally low compared to both natural populations and previous transplantation efforts (Campbell et al 2014). This suggests that whilst it is possible to use population mixing to restore desired levels of genetic diversity and structure, other factors such as donor provenance, transplant numbers, environmental factors and ecological interactions at restoration sites must also be considered during future restoration efforts in order to achieve long-term success (Campbell et al., 2014; Vergés et al. 2020).

3.5.1 Extant population genetics and provenance choices

We found a small amount of genetic variation between extant populations, although they had similar levels of heterozygosity and no evidence of inbreeding overall. There was a trend for allelic richness to be slightly lower closer to the gap in its distribution near Sydney, possibly reflecting the impact of fragmentation of Phyllospora forests on dispersal and gene flow. Overall, our results are largely

Using genomics to restore and future-proof underwater seaweed forests

consistent with prior research indicating that Phyllospora has relatively low genetic diversity across its’ central range and disperses widely, at least on scales that contribute to estimates of gene flow, facilitated by the presence of large, gas- filled vesicles that aid rafting on the ocean surface (Coleman et al., 2008; Coleman et al., 2011; Durrant et al., 2015). This previous work has suggested that dispersal is largely non-linear and likely influenced by local eddies and/or northwardflowing currents that prevail inland off the coast of NSW. However, by using SNPs we were able to detect a weak but significant pattern of isolation by distance and previously uncharacterised genetic structure. This suggests that Phyllospora’s dispersal patterns are also highly influenced by the East Australian Current, which flows in a north-southerly direction.

The lack of strong differentiation in Phyllospora populations around Sydney indicated that restoring with the aim to resemble the genetic structure of surrounding extant populations could be achieved by sourcing donor algae from within a ~60 km radius. Moreover, the genetic structure detected between the extant populations surrounding Sydney - in particular the presence of multiple genetic clusters, including genetic admixture within sites - suggested that populations restored into Sydney should be comprised of a mixture, primarily of genetic clusters found directly north and south of Sydney. Sourcing from two, rather than one, area would lead to a more realistic genetic mix of these clusters while also minimising potential harvesting impacts on more genetically mixed sites bordering of the distributional gap (i.e PB and CR).

3.5.2 Provenance effects on adult survival and condition

Provenance influenced Phyllospora transplant survival and condition, with higher survival and better condition of transplants sourced from the north. This was unlikely due to our transplantation method because algae from both sites were handled similarly. Differences in susceptibility to herbivory may explain some of the observed patterns, as restoration sites with greater fouling and differences in survival also had the highest levels of grazers (Wood et al., in prep.). Donors from

Using genomics to restore and future-proof underwater seaweed forests

the south had fewer branches and were longer than the “bushier” algae from the north, potentially leaving them susceptible to higher stress and biomass loss caused by grazing. Generally, traits that influence herbivory in Phyllospora and other species, e.g. tissue chemistry and morphology, are highly variable and likely shaped by local abiotic conditions (Peters, 2015; Weigner, 2016). Future restoration efforts intending to improve transplant survival and condition may thus benefit from further work to determine the underlying mechanism(s) driving provenance effects and selecting donors based on this.

3.5.3 Provenance effects on recruitment

While Phyllospora disperses widely, over 10 years of monitoring data associated with the current project have revealed that recruitment has not occurred at any of the degraded/restoration sites prior to our intervention (Marzinelli et al., and Wood et al., unpublished). Restoration resulted in rapid reproduction; with 80% of sites containing an F1 generation after six months, largely within 30cm of the transplant mats. These results emphasise that although this species has high capacity for dispersal (via ocean currents) and subsequent multi-generational gene flow, within-generation dispersal remains generally limited. Where the F1 generation had similar recruitment density to natural populations (Coogee), observed heterozygosity and allelic richness was similar to that of donor populations. Interestingly, despite transplants from the north having higher survival, slightly more recruits were attributed to reproduction between donors from the south or were putative crosses between the north and south. This could be due to a slight bias in the validity of assignment test estimates because donor populations were very genetically similar to begin with, or a comparatively lower reproductive output from adults from the north. Additionally, without the ability to detect paternity/maternity or the particular identity of parental individuals (which would require more loci), we are unable to determine if admixture in recruits was a result of true outcrossing or simply the progeny of a few particularly fecund individuals. However, we contend that the F1 populations are likely composed of a genetic mix of contributions of each provenance due to the life 59

Using genomics to restore and future-proof underwater seaweed forests

history of Phyllospora and fucoids in general (Pearson and Serrao, 2006). Synchronous reproduction and gamete release are likely induced by osmotic and/or hydrostatic stress occurring during transplantation. This not only results in restoration achieving its aims of mimicking patterns of genetic diversity and structure of extant populations, but also overcomes a major bottleneck of marine restoration - the need to maintain adult donor plants on engineered structures in situ, that can be rapidly removed due to waves and during storms (Campbell et al., 2014).

While sites with low recruit density (Whale Beach, Freshwater, South Head and Maroubra) exhibited some evidence of outbreeding (negative FIS values) in the

F1 generation, we believe that outbreeding depression or maladaptation was unlikely to be the cause of lower recruitment as the level of genetic differentiation observed in this study is generally considered low. Instead, we contend that the low densities of recruits were likely due to the high levels of grazing at these sites

(pers. obs). Nevertheless, crossing over a genetic distance of GST 0.02 - as we did here – has reportedly caused fitness problems in F2 generations of other marine species, e.g. pink salmon (Beacham et al., 1988). The effects of low recruitment at some sites is also likely to cause genetic bottlenecks as effective population size diminishes due to mortality. Future work monitoring gene flow and diversity in restored populations beyond the F1 generation will be critical in determining such effects.

3.5.4 Implications and future directions for seaweed restoration

As underwater forests are predicted to continue to be affected by environmental stressors into the future, there is an increasing need to apply genetic and genomic techniques to underwater restoration projects (Wood et al., 2019). This study demonstrates that Phyllospora populations can be mixed to achieve desired genetic structure and diversity, at least in the F1 generation. Restoration efforts are continuing at these sites and in some instances, recruitment numbers continued to increase following this study. Subsequent efforts to enhance recruitment at

Using genomics to restore and future-proof underwater seaweed forests

some sites have included supplemental transplantations to “boost” recruitment numbers and in some cases, herbivore exclusion or removal. Although results will vary depending on the algal species being restored, we suggest that mixing populations of species that exhibit synchronous spawning and rapid fertilisation is likely to conserve diversity and heterozygosity, or even increase it. Further work is needed to determine if this is the case for species with more complex life histories, such as the with true kelps, e.g. Ecklonia radiata or Macryocystis spp. Additional positive effects of genetic diversity documented in other systems, e.g. increased establishment rates, fitness and ecosystem functioning (Forsman and Wennersten, 2016) may also be conferred, however further investigation of whether these apply to underwater forests is still needed. Future work in this field will also determine if mixing algae containing desirable genetic characteristics may be used to “future-proof” populations. Determining ideal provenance mixtures and available genetic variation for selection would necessitate sampling populations at larger scales and would be facilitated by a multi-modelling approach which incorporates differences in regional and heritable differentiation (Breed et al., 2019; Rossetto et al., 2019). Experimental trials to determine progeny fitness and expression of desirable traits such as growth and productivity across scales and generations will facilitate this outcome (Breed et al., 2018; Weeks et al., 2011). Overall, the ability to manage genetic diversity, as has been demonstrated here, will likely play a crucial role in maintaining underwater forests into the future.

Using genomics to restore and future-proof underwater seaweed forests

Chapter 4:

Reduced genetic diversity and gene flow in marine forests on the edge: implications for restoration and future-proofing

Published as: Wood G., Marzinelli E.M., Campbell A.H., Steinberg P.D., Vergés A., Coleman M.A. Genomic vulnerability of a dominant seaweed points to futureproofing pathways for Australia's underwater forests. Glob Change Biol. 2021;00:1–13. https://doi.org/10.1111/ gcb.15534

Using genomics to restore and future-proof underwater seaweed forests

4.1 Abstract

Loss and degradation of habitats under climatic stress have prompted calls for management and restoration interventions to boost future resilience through assisted evolution strategies. Such efforts rely on fundamental knowledge of underlying neutral and adaptive genomic patterns across a species’ distribution, but data is scant for most species of conservation concern. Seaweed forests provide significant ecological, economic and social values along temperate coastlines globally, but their widespread declines are prompting calls to restore and/or “future-proof” populations. Here, we used genomics to characterise overall and potentially adaptive genetic diversity and structure of a declining forest-forming seaweed, Phyllospora comosa, along its entire latitudinal (31-43o latitude) and thermal (~12-26oC) range. Phyllospora showed relatively high connectivity throughout its central range, but there was evidence of genetic structure and putative selection, particularly at the range-edges. Genetic diversity was unevenly distributed, with rear and leading-edge populations harbouring only half the diversity of central populations, but also containing loci putatively under selection that linked to sea temperature. Our results suggest that populations at range-edges may be locally adapted to marginal environmental conditions, but reduced gene flow and diversity may compromise overall adaptability under future climates.

This work demonstrates how genomics can be used to explore potential restoration and future-proofing strategies and provide a basis for manipulative experiments which assess true adaptative traits and resilience.

Using genomics to restore and future-proof underwater seaweed forests

4.2 Introduction The impacts of human activities on natural systems are pervasive and significant, having affected biodiversity, species interactions and key ecosystem functions and services around the globe (Halpern et al., 2008; Pecl et al., 2017; Tylianakis et al., 2008). The need to curtail habitat and biodiversity loss is urgent, with the degradation of nature now recognised as one of the greatest threats to humanity and planetary stability (Steffen et al., 2018; IPCC, 2019). Conservation and restoration have emerged as useful tools to slow, mitigate or reverse environmental degradation (McKay et al., 2015; Rinkevich, 2014). However, current rates of environmental change - particularly due to the effects of the climate crisis - are outpacing natural rates of migration, adaptation and evolution (Burrows et al., 2011; Cang et al., 2016; Deutsch et al., 2015; Radchuck et al., 2019). As a result, many species are unable to persist under current conservation and restoration initiatives (Martinez et al., 2018; van Oppen et al., 2015). Further, relying on techniques which aim to preserve or restore ecosystems to historic or current states is unlikely to be effective under future climates (Breed et al., 2018; Perring et al., 2015; van Oppen et al., 2015). Recently, there has been growing interest in using “assisted evolution” as a strategy to increase the resilience of organisms in the face of current and future environmental stress (Aitken and Whitlock, 2013; van Oppen et al., 2015, Anthony et al., 2017). Assisted evolution strategies harness natural genetic variability of wild individuals, and may include one or several of the following approaches: (1) genetic rescue, i.e. the introduction of individuals from genetically diverse populations in similar environments into genetically depauperate populations to enhance genetic diversity, fitness and adaptive potential (e.g Reynolds et al., 2012; Whiteley et al., 2015); (2) assisted gene flow, i.e. the intentional movement of pre-adapted individuals within a species range to facilitate adaptation to anticipated local conditions (Aitken and Whitlock, 2013); and (3) assisted colonisation, i.e. the movement of species to suitable environments outside of their native range to facilitate species’ persistence (Weeks et al., 2011). More transformative strategies to increase resilience to future stress may also include artificial hybridisation, manipulation of host-associated microbial communities (natural or synthetic), 64

Using genomics to restore and future-proof underwater seaweed forests

stress conditioning and gene editing (van Oppen et al., 2015; Coleman and Gould, 2019). Such strategies are being developed to accelerate the rate of naturally occurring evolutionary processes to boost resilience to change. However, manipulating genetic composition also carries genetic risks, including potential maladaptation and outbreeding depression (Aitken and Whitlock, 2013; Weeks et al., 2011). To properly design and assess the feasibility of using assisted evolution solutions, it is critical to understand underlying patterns of both neutral and adaptive (i.e. ecologically relevant) genetic diversity and gene flow across species distributions, and how they link to extant and future environmental conditions. Such data, however, are only beginning to emerge for most species. Underwater seaweed forests underpin temperate coastal ecosystems and hold significant ecological and economic value (Bennett et al., 2016; Coleman and Wernberg 2017; Wernberg et al., 2019b), but they are in decline in many places around the world (Krumhansl et al., 2016). While causes of decline vary, temperature stress associated with ocean warming and marine heatwaves is among the most pervasive causes of loss (Smale, 2020; Wernberg et al., 2011; Wernberg et al., 2016). Heat stress can directly affect seaweed survival, condition and resilience to additional perturbations (Wernberg et al., 2010) and can also make seaweeds more vulnerable to consumers (Simonson et al., 2015; Vergés et al., 2014; Provost et al., 2017) and disease (Qiu et al., 2019). This can also lead to loss of genetic diversity (Gurgel et al., in press). South-eastern Australia harbours the highest number of endemic seaweeds in the world (Phillips, 2001; Bennett et al., 2016), but it is warming three times faster than the global average (Hobday and Pecl, 2014; Ridgeway, 2007). Under best-case carbon emission reduction scenarios, ocean warming alone is predicted to decrease the distribution of over 85% of the dominant forest-forming species in south-east Australia by 75% in 2100 (Martinez et al., 2018). These predictions are based on species-wide thermal tolerance data; however, losses could be even more widespread if intraspecific variation in thermal tolerance exists or if range shifting herbivores enhance decline (Vergés et al., 2016). With such a high risk of loss of habitat forming species and consequent major ecological changes along this coastline, Australian seaweed forests are likely excellent candidates for assisted 65

Using genomics to restore and future-proof underwater seaweed forests

evolution strategies (Wood et al., 2019). However, such efforts are hampered by a lack of knowledge of macroalgal genomics that underpin such strategies, particularly for functional genomics and adaptive potential at relevant scales (Coleman et al., in review a). Understanding the underlying genomics of seaweed forests to inform assisted adaptation strategies is required to pinpoint where desired adaptive diversity occurs. Patterns of genetic diversity in macrophytes are seldom uniformly distributed and variation can reflect both contemporary and historical processes. For example, genetic refugia (areas of unique or high genetic diversity) may occur in rear‐edge populations, deep-water populations or other areas that correspond to past climate refugia that make them an important contributor to overall genetic diversity of a species (Assis et al., 2013; Assis et al., 2016; Diekmann and Serrao, 2012; Lourenco et al., 2016). Further, rear and leading-edge populations are often less genetically diverse and exhibit increased divergence due to habitat fragmentation, increased random genetic drift, bottleneck effects and reduced gene flow (Provan and Maggs, 2011; Assis et al., 2013). Because of their proximity to species thermal limits (especially at the warm rear-edge) and low genetic diversity, range-edge populations are often considered to be particularly vulnerable to stress under climate change scenarios (Wernberg et al., 2018) but may also be locally adapted and thus present ideal targets for assisted adaptation strategies that seek to boost thermal tolerance. When considering species persistence, the conservation of vulnerable and unique genetic lineages is thus often considered key to avoiding the loss of existing thermal adaptations and erosion of the capacity to adapt to a changing climate (Assis et al., 2013; Nicastro et al., 2013). Phyllospora comosa (hereafter Phyllospora) is an endemic forest-forming seaweed along the south-eastern Australian coastline that supports vital ecosystem functions and unique biodiversity, including economically important fish and shellfish (Bishop et al., 2010; Coleman and Wernberg, 2017; Marzinelli et al., 2014). Phyllospora has suffered historical declines in Sydney - the middle of its range, likely due to poor water quality (Coleman et al., 2008), but it is the subject of Australia’s most successful macroalgal restoration program (see www.operationcrayweed.com; Campbell et al., 2014; Layton et al., 2019). However, 66

Using genomics to restore and future-proof underwater seaweed forests

Phyllospora’s long-term persistence in extant and restored areas is significantly threatened by ocean warming, with a projected 87% loss in distribution by 2100 (Martinez et al., 2018). Assisted evolution may play a critical role in boosting resilience in this key species. Characterising spatial patterns in overall and adaptive genetic structure and diversity across the entire species range, and then confirming the ecological effects of such patterns via experimental manipulations is required to underpin such efforts. Here, we provided a genetic baseline for Phyllospora by characterising overall and adaptive genetic diversity and structure across it’s entire latitudinal (31- 43° latitude) and thermal (~12-26oC) range. We identified specific adaptive SNP loci, including those linked to temperature, to assess the potential role of selection in determining patterns of genetic diversity and structure.

4.3 Methods

4.3.1 Field sampling We sampled 13 sites spread across Phyllospora’s latitudinal range over the Austral summers of 2016 and 2018 (five sites in 2016 and nine sites in 2018; Fig. 4.1a). At each site, a minimum of 20 and maximum of 50 (in accordance with current advice on sampling for high-throughput sequencing methods; see Mao et al., 2020; Morin et al., 2009; Nazareno et al., 2017 and Rhode et al., 2017). Reproductive Phyllospora individuals (whole male and female thalli; > 1 m apart) were haphazardly sampled from a 500 m2 area of reef. The sex of each individual was recorded to assess any biases in sex ratios at each site, but ratios were mostly equal. The exception was at the northern-most site, Port Macquarie, where there were many more males than females (Table S4.1). Ten unfouled apical tips were removed from each individual for genotyping. The methods for tissue processing and DNA extraction are described in Chapter 3.

Using genomics to restore and future-proof underwater seaweed forests

Figure 4.1: (a) Map of 13 sites where Phyllospora comosa was sampled; inset depicts Phyllospora’s entire distributional range; ocean coloured with average maximum annual SST from 1992 - 2018. PM: Port Macquarie; FO: Forster; AB: Anna Bay: BB: Bateau Bay: TE: Terrigal: PB: Palm Beach; CR: Cronulla; SP: Shark Park; SH: Shellharbour; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport. (b) Genetic structure plots showing individuals assigned to inferred clusters using all loci or outlier loci detected using Bayescan, PCadapt and LFMM (SST-associated loci). Each row represents an individual; different colours within columns indicate maximum likelihood probability of belonging to different clusters. 68

Using genomics to restore and future-proof underwater seaweed forests

4.3.2 Genotyping and bioinformatics 337 samples from across the 13 sites were genotyped on a custom panel of 354 SNPs as described in Chapter 3. Bioinformatics and data analyses were then conducted using the statistical platform R (version 3.6; R Core Team, 2019). Because they are based on thresholds calculated across populations, quality filtering steps may erroneously remove private alleles found in high frequencies at differentiated populations. We thus first checked for private alleles on the raw data using poppr. As private alleles were found in low frequencies (see Table S4.2), any subsequent removal due to filtering steps was deemed unlikely to affect overall results. We then quality filtered the data, excluding SNPs and samples with a call rate below 90% of the total, or with a minor allele frequency (MAF) below 0.05. We also filtered for Linkage Disequilibrium using SNPrelate (Zheng et al., 2012) with a threshold value of 0.7, which removed two loci from the dataset. Exact tests for Hardy-Weinberg Equilibrium (HWE) deviations were calculated across all samples and loci in the dataset using hierfstat (Goudet, 2005) and corrected for multiple testing using the Benjamini-Hochberg false discovery rate (FDR) procedure (Benjimini and Hochberg, 1995). 45% of loci deviated from the HWE, but only at one or two sites, so these loci were retained in the dataset. One locus was identified as deviating from HWE at ten (77%) sites and exhibited high heterozygote deficiencies (FIS = 0.918). We removed this locus as this is likely due to null alleles or other genotyping errors (Spencer et al., 2014). An initial comparison of observed heterozygosity and sample size for each site established that the use of unbalanced samples did not bias genetic diversity estimates (Table S4.3), so all samples were included for subsequent analysis. This left a total of 109 loci out of the original 354 across 331 of the original 337 samples

4.3.3 Data analyses 4.3.3.1 Tests for adaptation Three different approaches were used to detect loci under putative selection, following recommendations to utilize multiple methods to increase robustness of results (Benestan et al., 2016; Rellstab et al., 2015): (i) BayeScan 2.1 (Foll and 69

Using genomics to restore and future-proof underwater seaweed forests

Gaggiotti, 2008, Foll, 2012), (ii) PCadapt (Luu et al., 2017) and (iii) latent factor mixed models (LFMMs; Frichot et al., 2013). An FDR correction of 0.05 was applied to all methods to avoid the occurrence of false positives.

BayeScan uses a Bayesian approach to detect FST outlier loci by using linear regression to decompose FST coefficients into population‐ and locus‐specific components and estimates the posterior probability of a locus showing deviation from Hardy–Weinberg proportions. We used a total of ten separate runs, from 50,000 to 500,000 iterations with a 10% burn-in period. PCadapt detects outlier loci by first ascertaining population structure using principal component analysis (PCA). It then identifies markers under putative selection as those that are excessively correlated with population structure. This approach has greater power in the presence of admixed individuals and when population structure is continuous (Luu et al., 2017), as is likely with Phyllospora populations (Coleman and Kelaher 2009). PCadapt identified four principal components in the dataset; to identify outlier loci we assessed significance using adjusted p-values. LFMM detects outlier loci associated with environmental variables whilst accounting for the confounding effects of population structure. We first screened for important environmental predictors by fitting the overall genetic data against several variables based on sea surface temperature (SST) using distance-based linear models and redundancy analysis (dbRDA; McArdle and Anderson, 2001) in the vegan package (Oksanen et al., 2018). Missing allele frequencies were imputed using sNMF (Frichot et al., 2014) in the LEA package (Frichot and François, 2015) using the most common allele frequency observed at each site. SST data was downloaded from the Integrated Marine Observing System (IMOS) database using the raster package (Hijmans, 2017). Because acclimation or adaptation may be occurring on multiple timescales, we downloaded monthly SST data over the period of 1992 to 2018 and calculated mean, maximum, minimum, range and variation (standard deviation from the mean: SD) temperature for the periods of 1992-2018 (26 years, the longest period of available data) and 2013-2018 (six years, to use recent data within the lifespan of current Phyllospora individuals). Environmental variables that were strongly correlated (Pearson’s r2 > 0.7) were 70

Using genomics to restore and future-proof underwater seaweed forests

removed, leaving one in the set to represent those removed. The final model included the average of annual maximum, range and standard deviation (SD) in SST occurring at each site over the past 26 years. Analyses used a stepwise procedure in both directions and model selection based on adjusted p-values. Outlier loci associated with SST variables that had been identified as important with dbRDA were then identified using z-score cutoff in the lfmm v2.0 package (Caye et al., 2019). The number of genetic clusters (K = 9, see the section Population Structure below) was fitted as a latent variable in the model. We compiled a set of “key” outlier loci which overlapped with two or more variables to assess genetic diversity and structure associated with SST, as a combination of these is likely important when choosing donors for assisted evolution strategies.

4.3.3.2 Genetic structure and diversity of overall versus adaptive datasets A common approach in selection studies is to run multiple tests and look for duplicate detections across methods, which is effective in reducing the number of false positives and in identifying strong selective sweeps in large SNP datasets (e.g. Feng et al., 2015; Selechnik et al., 2019). However, as our SNP panel was relatively small to start with and the study objective was to provide any possible evidence of selection, including recent selection or selection on standing genetic variation (see Forester et al., 2018), we conducted all further analyses of genetic diversity and structure on four datasets. These included (1) the full SNP dataset (109 loci; hereafter referred to as “overall” genetic diversity and structure) and (2) “adaptive” SNPs, i.e. loci that had been identified as putatively under selection using (i) Bayescan (three loci), (ii) PCadapt (eight loci) or (iii) lfmm (key loci only; 36 loci). Genetic structure was visualised using Principal Component Analysis

(PCA) and statistically assessed using two approaches: Wier and Cockham’s FST (the proportion of genetic variance contained in a subpopulation relative to the total genetic variance) and sNMF (Frichot et al., 2014). Pairwise comparisons of

FST between sampling locations and their significance was assessed using p-values calculated using bootstrapping (999) in the dartR package (Gruber et al., 2018) and corrected using the Benjamini-Hochberg FDR Procedure. sNMF is based on sparse non-negative matrix factorization to estimate the genetic ancestry 71

Using genomics to restore and future-proof underwater seaweed forests

components for each individual. 15 runs were performed with alpha = ten for each K value (one to 14). Cross-entropy was used to guide the choice of the number of ancestral populations and the results from the best run were visualised using the barplots function. To examine the effects of geographic distance on genetic distance across the latitudinal range, we performed Mantel tests on the overall and adaptive datasets using ade4 v1.7-13 (Dray and Dufour, 2007). For the geographic distance matrix, we used the dist function to calculate the Euclidean distances in geographic space between collection sites based on their coordinates. For the genetic distance matrix, we used the collection site-based pairwise FST values generated from hierfstat. Population genetic summary statistics were calculated on the overall dataset and on the adaptive datasets to describe and compare overall and population‐specific genetic diversity. Genetic differentiation and diversity were evaluated by generating estimates of observed heterozygosity (HO) and expected heterozygosity (HE) for each locus and for each sampling group using the diveRsity package (Keenan et al., 2013), and allelic richness with allele counts rarefied by the minimum number of individuals genotyped using hierfstat. We tested for differences in observed heterozygosity between sites using the Hs.test function in adegenet. We also calculated departures from random mating (FIS, i.e. inbreeding/outbreeding estimates) for the overall dataset and assessed significance using 1,000 permutations with 95% confidence intervals.

4.4 Results

4.4.1 Overall population structure While some caution is heeded when unbalanced samples are used across populations (Goudet et al., 1996), exploratory analysis that may reduce statistical power to detect differences in genetic diversity and structure between populations, PCA ordinations showed that sites were clustered in a broad, hierarchical latitudinal pattern (Fig. S4.1). Pairwise FST tests using the overall SNP dataset between all pairs of sites confirmed that all sites were genetically different (all

Using genomics to restore and future-proof underwater seaweed forests

pairwise tests significant; Table 4.1) and were highly differentiated overall (global

FST = 0.225). Using LEA, we identified the most likely number of genetic clusters in the overall dataset as 9 (Fig. 4.1). In the dbRDA, average annual maximum, range and SD of SST explained a significant proportion of overall genetic differentiation in Phyllospora, as measured by adjusted-R2 (adj-R2 = 0.17, p < 0.001; Fig. 4.2 and

Table 4.2). Geographic and genetic distance (pairwise FST values) were significantly associated across all loci (r = 0.82, p = 0.001; Fig. 4.3).

4.4.2 Signals of selection and adaptive genetic structure In total, we detected 47 loci (43%) that were putatively under selection; six of these were identified through two or more methods (Table S4.4). In the BayeScan program, we detected three polymorphic loci with statistically significant patterns of divergent genetic differentiation. PCadapt identified eight outlier SNPs which correlated with four genetic clusters, one of which overlapped with the Bayescan output. Lfmm identified 28 loci that were associated with maximum SST, 39 loci that were associated with the range in SST and 62 loci that were associated with the SD of SST at each site. Overall, there were 36 loci that had overlapping associations with two or more of these temperature variables. As with the overall genetic dataset, each adaptive genetic dataset was significantly correlated with geographic distance (Bayescan: r = 0.78, p = 0.001; PCadapt: r = 0.44; p = 0.048 and SST-associated loci identified with lfmm: r = 0.84; p = 0.001; Fig. S4.2). Pairwise tests on adaptive datasets revealed that most sites were still, on average, highly differentiated (Global FST 0.06-0.41; Tables S4.5.1- 4.5.3). However, not all sites were significantly different to others, with 8.5-24% of pairwise comparisons of FST values between sites not significant. This potentially indicates shared adaptations between sites. The optimum number of genetic clusters for putatively adaptive SNPs was dependant on the dataset used, with K = two-six (Fig. 4.1b). While patterns of genetic structure varied depending on the dataset, all exhibited strong differentiation between mainland and Tasmanian sites, which typically differ in temperature by 3-6 °C. Three of the four structure plots also revealed significant structure between the northern-most sites (Port Macquarie, Forster and Anna Bay)

Using genomics to restore and future-proof underwater seaweed forests

and central sites (Fig.4.1b), which typically vary in temperature by 0.5-3 °C. Almost a quarter of the loci (24%) deviated from Hardy-Weinberg equilibrium at the warm rear-edge population (Port Macquarie), suggesting non-random mating or a small effective population size in this marginal population.

4.4.3 Genetic diversity

Most sites were characterised by small non-significant FIS estimates; however, three sites (Forster, Shark Park and Eden) had significantly large, negative values indicative of outbreeding (Table 4.3). Overall genetic diversity as measured by observed heterozygosity differed significantly between most sites (Table S4.6) and was generally more similar among sites in the central range (i.e. BB, TE, PB, CR, SP, SH and MB) and between range-edge sites (PM, FO and SO). Expected heterozygosity (HO; 0.172-0.336) and allelic richness (AR; 1.523-1.957) also appeared to vary between sites. All measures of genetic diversity demonstrated a clear trend for lower genetic diversity at sites near the rear and leading edges compared to Phyllospora’s central range (Table 4.3). Genetic diversity estimates based on the adaptive datasets had generally similar trends to patterns in overall genetic diversity, with greater adaptive diversity in central populations. This was with the exception of the northern-most site, Port Macquarie, which also had high adaptive diversity in the Bayescan and PCadapt datasets (Tables S4.6.1-4.6.3).

Using genomics to restore and future-proof underwater seaweed forests

Table 4.1: Overall population structure (pairwise Fst) among extant Phyllospora comosa populations based on 109 loci *

Site a PM FO AB BB TE PB CR SP SH MB ED BI SO

PM FO 0.278

AB 0.275 0.288 BB 0.252 0.192 0.115

TE 0.245 0.202 0.132 0.022 PB 0.249 0.212 0.124 0.024 0.051 CR 0.235 0.194 0.120 0.029 0.044 0.022 SP 0.260 0.217 0.146 0.037 0.057 0.035 0.039 SH 0.276 0.251 0.174 0.081 0.098 0.060 0.070 0.069 MB 0.299 0.281 0.192 0.116 0.119 0.126 0.119 0.133 0.142 ED 0.472 0.462 0.391 0.326 0.326 0.344 0.339 0.355 0.341 0.266 BI 0.516 0.491 0.389 0.338 0.327 0.336 0.340 0.362 0.353 0.302 0.303 SO 0.568 0.543 0.430 0.355 0.344 0.355 0.366 0.388 0.385 0.367 0.367 0.255

* Significant values in bold. aPM: Port Macquarie; FO: Forster; AB: Anna Bay: BB: Bateau Bay: TE: Terrigal: PB: Palm Beach; CR: Cronulla; SP: Shark Park; SH: Shellharbour; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

Using genomics to restore and future-proof underwater seaweed forests

PM FO AB BB TE PB CR SP SH MB ED BI SO

Figure 4.2: Distance-based Redundancy Analysis showing the relationship between environmental variables and genetic structure (SNPs) of Phyllospora comosa populations from 13 sites across the species’ latitudinal range. Environmental data include: annual maximum sea surface temperature averaged over 26 years (Max_SST), average annual range in sea surface temperature (Range_SST) and average annual standard deviation in sea surface temperatures (SD_SST) experienced at each site. PM: Port Macquarie; FO: Forster; AB: Anna Bay; BB: Bateau Bay; TE: Terrigal; PB: Palm Beach; CR: Cronulla; SP: Shark Park; SH: Shellharbour; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

Using genomics to restore and future-proof underwater seaweed forests

Table 4.2: ANOVA output table for final dbRDA model, showing annual maximum sea surface temperature averaged over 26 years (Max_SST), average annual range in sea surface temperature (Range_SST) and average annual standard deviation in sea surface temperatures (SD_SST) experienced at 13 sites is associated with Phyllospora comosa’s genetic structure. The final model was selected via model selection based on adjusted p-values.

Model Df Variance F p Adjusted R2

~ Max_SST + 3 8.970 24.246 0.001 0.174 Range_SST + 327 40.308 SD_SST

Using genomics to restore and future-proof underwater seaweed forests

F Pairwise Pairwise

Geographic distance (Km)

Figure 4.3: Relationship between geographic distance (km) and overall genetic distance (pairwise FST) between Phyllospora comosa individuals calculated using 109 loci. Data collected at 13 sites and plots fitted with linear regression (fitted values: blue line; 95% Confidence Intervals: grey shade).

Using genomics to restore and future-proof underwater seaweed forests

Table 4.3: Overall genetic diversity of 13 Phyllospora comosa populations across its latitudinal range based on 109 loci.

a b c d Site n Global FST FIS HO HE AR Port Macquarie 20 0.116 0.175 0.184 1.752 Forster 20 -0.051 0.206 0.197 1.587 Anna Bay 20 0.064 0.256 0.273 1.841 Bateau Bay 26 -0.004 0.332 0.331 1.929 Terrigal 29 0.009 0.338 0.340 1.945 Palm Beach 29 0.012 0.31 0.317 1.947 Cronulla 49 0.225 -0.021 0.335 0.327 1.957 Shark Park 29 -0.042 0.319 0.304 1.896 Shellharbour 30 0.001 0.309 0.304 1.894 Malua Bay 20 0.002 0.336 0.336 1.919 Eden 20 -0.036 0.312 0.296 1.825 Bicheno 20 -0.016 0.234 0.231 1.711 Southport 19 -0.035 0.179 0.172 1.523

a Inbreeding coefficient (FIS). Values that differ significantly from zero shown in bold. b Observed heterozygosity (HO) c Expected heterozygosity (HE) dRarefied allelic richness (AR)

Using genomics to restore and future-proof underwater seaweed forests

4.5 Discussion

There is an urgent need for interventionist approaches such as assisted evolution that enhance ecosystem resilience to environmental changes. However, the genetic information needed to sensibly implement such interventions is critically lacking. Our study investigated overall and adaptive genetic structure and diversity of the declining forest-forming seaweed Phyllospora comosa across its latitudinal and thermal range. Phyllospora had relatively high gene flow, indicating high connectivity over generations at regional scales, however overall genetic diversity was unevenly distributed and was generally greatly reduced at the range-edges. These patterns will likely have severe implications for population and species-wide responses to environmental change, particularly as populations at the warm range- edge are pushed against maximum thermal thresholds as ocean waters warm (Martinez et al., 2018). Range edges also harboured putative adaptive diversity indicative of selection, and many of these loci were linked to SST, which may present targets for assisted adaptation strategies to boost thermal resilience into the future. While further work is needed to confirm the presence of adaptation, genomics is a useful exploratory tool to guide future experimental manipulations. Importantly, our results identify vulnerable populations and enable the development of management strategies to enhance the resilience and resistance of these key underwater forests into the future.

4.5.1 Patterns of overall genetic diversity and structure Overall genetic diversity across Phyllospora’s range was medium to low (average

HE was ~0.2). Genetic differentiation between sites was, however, high, with genetic structure strongly associated with geographic distance (82% variation explained) and weakly (17%) associated with a combination of SST variables. Unlike previous work that used microsatellites, mitochondrial and chloroplast markers and showed weak patterns of genetic and no haplotype or nucleotide diversity across Phyllospora’s central range (Coleman et al., 2008; Coleman et al., 2011a; Durrant et al., 2015), our investigation spanning Phyllospora’s entire latitudinal distribution and including both neutral and adaptive portions of the

Using genomics to restore and future-proof underwater seaweed forests

genome was able to reveal clear genetic clustering and differences in genetic diversity across the entire range. Notably, range-edges were clearly differentiated and only had approximately half the genetic diversity of the central populations. This pattern may be a result of smaller population sizes at range-edges, limited connectivity or selection (Wernberg et al., 2018). These explanations are not mutually exclusive, and all may in part account for the observed patterns. For example, Phyllospora’s connectivity is thought to be facilitated by large and abundant gas filled floats that enable rafting and dispersal (Coleman et al., 2008; Coleman et al., 2011a; although see Cole, 2017); under a scenario of limited connectivity, high differentiation of populations at lower latitudes may be a result of prevailing ocean currents. The East Australian Current separation point is at ~31oS (around Forster, NSW). Further south, eddies may facilitate greater dispersal and mixing, leading to less genetic structure (Coleman et al., 2011a, b). Connectivity patterns linked to the East Australian Current may therefore partly explain the relative distinctiveness of the only population north of this point (Port Macquarie), which also represents the rear range edge for the species. Similarly, the leading- edge Tasmanian populations (Bicheno and Southport) were also highly differentiated from the central range. This pattern may be due to recurrent bottleneck effects from reduced connectivity between mainland Australian and Tasmanian populations, which are separated by ~150 km of open ocean. Such patterns are largely congruent with connectivity patterns of other marine species in the region, which also show genetic structuring corresponding to the poleward flowing boundary currents (Coleman et al., 2011b, Wernberg; Thomsen et al., 2013) and historic biogeographic isolation (Waters, 2008). The high overall genetic differentiation exhibited at Phyllospora’s warm range-edge population (Port Macquarie, NSW), along with significant deviations from HWE in 50% of loci, is suggestive of non-random mating or small population size. This site was unusual in that it was characterised by a very high proportion of males (80% compared to an average of 42% across the other sites) which may account for this pattern. This sex, and therefore reproductive bias may be a result of differential survival of male and female individuals as a result of temperature stress or other stressors associated with their marginal habitat. Under future 81

Using genomics to restore and future-proof underwater seaweed forests

climates, dioecious seaweeds are expected to exhibit sex bias towards males which show less sensitivity to change relative to females, which invest heavily in reproduction (Hultine et al., 2016). Indeed, the pattern for selection which was emphasized in the rear edge populations (see below) may have a sex-bias with potential selection for longer survival of male plants under warmer waters. While high female biased ratios were observed at several colder sites in NSW (Shellharbour, Eden), there was also male bias at Bicheno in Tasmania. Considering that sampling was relatively low at each site (n=20), further sampling is needed to robustly assess sex ratios. Further, manipulative experiments will be needed to thoroughly test any hypotheses regarding sex-based selection.

4.5.2 Adaptive genetic diversity and structure We found evidence of potential selection for adaptive loci across the latitudinal range of Phyllospora that linked to SST and range-edge populations. However, the occurrence of false positives is often high for selection studies (Narum and Hess, 2011; Whitlock and Lotterhos, 2015) and the tests run on the adaptive datasets were based on very low numbers of loci (n = 3 - 36), so these results should be interpreted with caution. Further, as both overall and adaptive datasets exhibited isolation by distance, disentangling the effects of genetic drift and true selection remain a challenge. Clustering analysis and pairwise FST tests of the loci putatively under selection suggest that functional differentiation is relatively high and occurs between both range edge and central populations, as well as on smaller scales (i.e. between individual sites). Interestingly, while the overall genetic diversity of central populations was up to two-fold that of edge populations, adaptive genetic diversity was up to ~1.5 times higher on the rear-edge (Port Macquarie) than on central populations, further indicating the potential harbouring of local adaptation in this location. Genetic differentiation between rear, central and leading-edge populations associated with SST was strongest between mainland and Tasmanian (leading-edge) populations, although there was also evidence for differentiation between rear-edge (Port Macquarie, Forster and Anna Bay) and central populations. Crucially, maximum SST may link to thermal tolerance, which is a

Using genomics to restore and future-proof underwater seaweed forests

hereditary trait that can vary among genotypes on similar spatial scales for other fucoid seaweeds (Clark et al., 2013; Miller at al. in press, although see Bennett et al., 2015). Given that SST varies along a latitudinal gradient, along which many other environmental variables also co-vary (e.g. nutrients, daylength), the driving factor(s) behind these patterns of selection remain to be established. A previous study has shown that at high temperatures Phyllospora from the northern range are able to maintain consistently greater photosynthetic functionality than those from Tasmania, which may indicate a higher upper thermal tolerance (Flukes et al., 2015). However, this was based on a short-term experiment that was not designed to identify the presence of heritable traits. Future experimental evidence will thus be needed to determine if the differences in putatively functional genetic diversity characterised here truly contribute to a genetic basis for adaptations to SST.

4.5.3 Informing assisted evolution These results have clear implications for future management and conservation of this key forest-forming seaweed. Firstly, our results identify Phyllospora’s rear and leading range-edges as potentially the most vulnerable to the multistressors that will characterise our future oceans. However, vulnerability of range-edges depends on whether low diversity is a result of selection and, if so, for which agents of selection or stressors. While the limited genetic diversity and therefore potential overall adaptive capacity of range-edge populations may render them vulnerable to a variety of stressors, these populations may be locally adapted to temperature stress. Indeed, warming waters is a significant threat to Phyllospora populations, with the East Australian Current continuing to penetrate and separate further south (Cetina-Heredia et al., 2014). While we anticipate that range-edge populations may be resilient to warming despite low diversity due to prior selection, more diverse central populations and especially the geographically isolated Tasmanian populations may be yet to experience such selective forces and suffer impacts in the future. Marine heatwaves can have a major impact on forest-forming seaweeds (Wernberg et al.,

Using genomics to restore and future-proof underwater seaweed forests

2016) and can occur anywhere throughout a species’ range and their impact may also depend on prior selection and standing genetic diversity (Wernberg et al., 2018). However, unlike gradual warming, heatwaves may rapidly promote selection to heat tolerance (Coleman et al., in review b, Gurgel et al., in press) provided there is enough standing genetic diversity upon which selection can occur and the temperature rise does not exceed the capacity of seaweeds to respond. Pollution and decreased water quality are also a threat to Phyllospora and were likely responsible for its historical decline around Sydney, Australia’s largest city, located in the central range of the species’ distribution (Coleman et al., 2008). Areas of greatest pollution are likely to occur in the more populated regions near Sydney, where genetic diversity was greatest, suggesting that there may be some adaptive capacity to cope with this stressor – although historical declines suggest this has its limits (Coleman et al., 2008). Based on these observations, we suggest that genetic interventions could be considered via two strategies: (1) increase overall genetic diversity in range-edge populations via genetic rescue – i.e. introduce individuals from genetically diverse populations from the central range - to boost resilience to stressors other than temperature; and/or (2) target potentially thermally adapted individuals (e.g. from the warm rear edge at Port Macquarie to Anna Bay) for assisted gene flow strategies. Central populations with high genetic diversity might be used to supplement diversity in vulnerable range-edge populations. However, this may come at the expense of temperature tolerance in rear-edge populations if there has indeed been prior selection. In turn, this may subsequently dilute the beneficial process of dispersal of temperature tolerant genes into central populations, a natural way to boost resilience to future temperature stress. Alternatively, strategies which target potentially thermally adapted individuals for assisted gene flow may rapidly increase tolerance to thermal stress in central populations. Indeed, evidence across a wide variety of systems suggests that explicitly including pre-existing adaptations can be more successful than simply buffering adaptive potential alone (Houde, 2016). Mixing very genetically differentiated populations (e.g by moving individuals between Phyllospora’s central and range-edge populations) has 84

Using genomics to restore and future-proof underwater seaweed forests

traditionally been discouraged to avoid significant effects such as maladaptation, outbreeding depression and the breakdown of co-adapted gene complexes (Frankham et al., 2011). Recent evidence suggests that such events are actually rarer than initially thought (Ralls et al., 2012), although we emphasize that care should be taken when mixing populations that are very genetically distinct as was observed at large spatial scales in this system. Additionally, adaptation to one stressor (e.g. temperature) may come at the expense to others (Bennett and Lenski, 2007; Foster-Huenneke, 1991; Rodriguez-Verdugo, 2014) and “genetic pollution” effects arising from admixture between very genetically differentiated populations will likely not be observed until the F2 or F3 generation. Further, the underlying genetics of foundation species can have broader ecosystem implications such as affecting associated functioning, habitat and food provisioning in some systems (Whitham et al., 2006). Careful planning and long-term studies of potential genetic effects are therefore vital (Weeks et al., 2011). While these issues should be considered, there are risks with any interventionist strategy, such as potential negative effects on other co-occurring organisms/assemblages, and these must be considered alongside the benefits (Weeks et al., 2011). Local adaptation to higher temperatures may be crucial in determining seaweed survival under stochastic events such as heatwaves, but a mixture of genetic diversity and phenotypic plasticity will likely play a large role in determining which individuals survive under gradual ocean warming (McCoy and Widdicombe, 2019). Phyllospora has a high level of plasticity and northern genotypes can indeed survive in experimental temperature and nitrate conditions vastly different to their original environment (at least on short timescales; Flukes et al., 2015). Similar survival and function of northern genotypes in novel environments may mean that assisted evolution strategies involving translocation of non-local individuals could be carried out with few negative and potentially many positive ecosystem-levels effects.

Using genomics to restore and future-proof underwater seaweed forests

4.5.4 Study limitations and future work The number of investigations that explicitly aim to identify environmental drivers of seaweed genetic patterns is increasing (e.g. Johansson et al., 2015; Vranken et al.,

In Prep). This is one of the first to use FST tests to map potential signatures of selection or adaptation for climate-proof traits in wild populations.While the number of SNPs used here is relatively low compared to some contemporary studies utilising even more powerful sequencing methods, such as genotyping by sequencing (e.g. Fraser et al., 2016), our work provides a strong basis for beginning to triage populations for management interventions, including the identification of which populations are most at risk and may harbour selective traits which may prove useful for assisted evolution. Nevertheless, we are limited in our ability to make causal inferences about genetic-functional interactions. As all loci exhibited patterns of isolation-by-distance, we were unable to tease apart the effect of temperature (or some other correlated driver) versus geographic distance, and further experiments are needed to determine possible mechanistic links between physiological response and presence of loci under selection. If it is established that thermal adaptation / heritable tolerance does exist and is unique to certain populations, breeding experiments to determine whether genetic mixing is feasible will also be needed. The breeding of sufficient individuals to adult or reproductive stages under laboratory conditions, especially over multiple generations, is challenging due to Phyllospora’s large size and requirements for survival. However, new restoration technologies that allow rapid and easy deployment of juvenile stages into populations (Fredriksen et al., in review) may provide a pathway for implementation. Given reductions in regional genetic diversity (Gurgel et al., in press), population resilience (Wernberg et al., 2010, 2018) and loss of rear edge populations of several seaweed species (Vergés et al., 2016, Smale and Wernberg, 2013, Wernberg et al., 2016), the need to initiate such solutions is urgent.

4.6 Conclusion These results have clear implications for Phyllospora’s conservation and, more generally, the future of underwater forests. The underlying patterns of genetic

Using genomics to restore and future-proof underwater seaweed forests

variation we describe here may affect the response of this key forest-forming species to future stressors and lead to loss that is faster or occurs at larger scales than previously predicted. As Phyllospora is a foundation species that underpins entire ecosystems (Coleman and Wernberg, 2017), the loss of populations is likely to have severe ecological effects in south-east Australia, impacting associated goods and services (Bishop et al., 2010, Marzinelli et al., 2014). Accordingly, intercepting the ongoing decline of seaweeds may require novel genetic management strategies (Coleman and Goold, 2019).

Active manipulation of the genetic makeup of populations requires careful science-based planning and bioethical considerations (Filbee-dexter and Smajdor, 2019). However, given the current state of the marine environment and future projections of degradation, strategies such as those discussed in this paper may well be the least controversial available (e.g see Coleman and Gould, 2019). If species are unable to adapt to the pace of climate change, harnessing standing adaptations as discussed here may still not be enough to ensure survival. While further work to determine the feasibility and risks of these solutions is necessary, embedding field trials into existing restoration programs as experimental management strategies might be the most realistic way to pursue these avenues, before we lose the opportunity to do so. This study provides a strong platform upon which to begin.

Using genomics to restore and future-proof underwater seaweed forests

Chapter 5:

The influence of host genetics, phenotype and geography on the microbiome of a foundational seaweed

Publication in preparation:

Wood, G., Campbell, A., Steinberg, P., Vergés, A., Coleman, M. and Marzinelli, E. (in prep.) The influence of host genetics, phenotype and geography on the microbiome of a foundational seaweed.

Using genomics to restore and future-proof underwater seaweed forests

5.1 Abstract

Host-microbial interactions are rapidly emerging as critical to the function, health, survival, resilience and adaptation of eukaryotic organisms. Yet, we have little understanding of the factors that underpin host-associated microbiomes. Such understanding is particularly important for foundation hosts in the wild, e.g. trees or corals, that underpin biodiversity and ecosystem functioning, as impacts can cascade throughout an entire ecosystem. Understanding host-microbe interactions can also enable the development of ecological management interventions that confer resilience and increase adaptive capacity. In this study, we focused on the marine foundation seaweed Phyllospora comosa and examined the influence of host genetics, phenotype and geography/environment on the surface-associated microbial communities associated with this species along its entire latitudinal distribution (1300 km). We characterised individual host genetics using a panel of 354 single nucleotide polymorphisms (SNPs), surface-associated microbial communities using 16S rRNA gene amplicon sequencing and quantified multiple morphological/functional traits. Combined, these factors explained 54% of variation in Phyllospora-associated microbial communities. Microbial communities were strongly related to local (site-level) environmental conditions, while effects of host phenotype and genetics were most apparent at larger, regional spatial scales. We identified several key genetic loci and phenotypic traits in Phyllospora that were strongly related to multiple microbial amplicon sequence variants (ASV), including taxa with known associations to seaweed defence, disease and tissue degradation. Our results indicate that Phyllospora’s surface- associated microbial communities are jointly shaped by local environmental conditions and host-specific differences. The strength of each of these factors is, however, context-specific - and likely dependant on trade-offs between each of these traits as well as additional ecological or environmental influences. Phyllospora management efforts seeking to manipulate or incorporate microbial associations would likely benefit from considering both local environmental conditions and host-specific traits in their design.

Using genomics to restore and future-proof underwater seaweed forests

5.2 Introduction Host-associated microbiomes play a critical role in the functioning, health, survival, resilience and adaptation of eukaryotic organisms (Egan et al., 2013; McFall-Ngai et al., 2013; Rosenberg and Zilber-Rosenberg, 2018). Indeed, it has been suggested that hosts and their microbiome form a coherent biological entity - or “holobiont” (Rohwer, 2002) – which needs to be studied holistically in order to better understand the ecology and evolution of eukaryotic hosts (e.g. the “hologenome” theory of evolution; Zilber-Rosenberg and Rosenberg, 2008; Wilkins et al., 2019). This, in turn, can inform management interventions that confer resilience and increase adaptive capacity of hosts (e.g. Breed et al., 2019; van Oppen et al., 2015; Wood et al., 2019). Free-living microbial communities are strongly driven by environmental and geographic factors such as temperature, light, the chemical environment and dispersal limitation (de Vries et al., 2012; Gusareva et al., 2019; Hellweger et al., 2014; Rusch et al., 2007). While the environment also acts as a source for holobiont microbiota, host-associated microbial communities are likely shaped by strong selective forces driven by host biology and behaviour (Bauer et al., 2018; Coyte et al., 2015). In particular, microbial taxa may be associated with or excluded from a community based on host phenotypic variation in chemical composition, morphology or condition (Arumugam et al., 2011; Neefjes et al., 2011; Srinivas et al., 2013). Microbial communities may also be shaped by underlying host genetics (Benson et al., 2010; Bulgarelli et al., 2013; Rawls et al., 2006), such that in many cases microbes and their eukaryotic hosts have co-evolved (Hacker et al., 2005; O’Brien et al., 2019; Rosenberg and Falkovitz, 2004; Zilber-Rosenberg and Rosenberg, 2008). Much of our understanding of host-microbial interactions is still largely limited to humans or mammalian model systems (e.g. mice), as well as economically important species such as domesticated plants and livestock (Blekhman et al., 2015; Wang et al., 2016), limiting inferences and application to other systems. In contrast, little is known about the mechanisms and strength of host-microbiome interactions in wild (non-model) organisms. This is particularly true for hosts in marine environments, which despite their ecological importance 90

Using genomics to restore and future-proof underwater seaweed forests

have been understudied compared to those in terrestrial systems, with the exception of reef-building coral taxa (e.g. see Bourne et al., 2009; Hernandez- Agreda et al., 2016; McFall-Ngai et al., 2013). Recent studies however demonstrate the presence of large and often highly diverse communities of marine host- associated microorganisms in habitat-formers such as seaweeds and seagrasses (e.g. Fahimipour et al., 2017; Marzinelli et al., 2015). The recent advent of rapid next-generation DNA sequencing technologies is increasingly providing an opportunity to expand our understanding of host-microbiome interactions in these systems, because it enables associations between non-model hosts and their microbiomes to be studied in high resolution at relatively low cost (e.g using host- genome-wide association studies; Awany et al., 2018). In temperate marine rocky reefs, seaweed forests are dominant habitats that underpin coastal biodiversity and functioning (Steneck et al., 2013). Despite being surrounded by the same “microbial soup” within any one location, seaweeds carry diverse biofilms on their surface (Egan et al., 2008; Wahl et al., 2012) that are distinct among different species (Lachnit et al., 2009) and from the surrounding sediment or seawater (Burke et al., 2011a; Roth-Schulze et al., 2016; Thomson, 2017). Seaweed-associated microbial communities play important roles in seaweed development, reproduction, functioning and defence (see Egan et al., 2013; Hollants et al., 2013 and Singh and Reddy, 2016 for reviews). They are also key players in biotransformation and nutrient cycling in the oceans due to their ability to decompose algal cell walls (Goecke et al., 2010; Hollants et al., 2013; Michel et al., 2006). The composition and species’ identity of these microbial communities on single host species often varies significantly in time and space (Bengtsson et al., 2010; Campbell et al., 2015; Lachnit et al., 2011; Staufenberger et al., 2008; Tujula et al., 2010). There is evidence for the role of both host characteristics and the environment underpinning such variation. It was initially suggested that microorganisms colonize algal surfaces using a ‘competitive lottery model,’ in which multiple species can colonize algal surfaces, so long as a core set of functions (rather than specific taxa) is represented (Burke et al., 2011a). Recent evidence suggests that seaweed microbial communities can also be strongly affected by host 91

Using genomics to restore and future-proof underwater seaweed forests

condition (Campbell et al., 2011, 2015; Marzinelli et al., 2015, Qiu et al., 2019). However, environmental changes such as increases in water temperature can lead to disruptions to host-microbiome associations with strong consequences to the host (Campbell et al., 2011; Qiu et al., 2019) and may therefore be an underlying mechanism of seaweed decline in many places around the world. Improving our understanding of the mechanisms and specificity behind seaweed-microbiome interactions may aid in the development of conservation and restoration tools, similar to the bioremediation and soil inoculation tools that have been developed for agriculture and restoration on land (e.g. Bashan et al., 2014; Holguin et al., 2001; Hong and Lee, 2014). Phyllospora comosa (hereafter, Phyllospora) is a dominant marine forest- forming seaweed that is found on shallow subtidal reefs in south-eastern Australia. It is ecologically and economically important but has suffered widespread decline along the metropolitan coastline of Sydney, Australia’s largest city (Coleman et al., 2008). The loss of Phyllospora has been attributed to historically poor water quality, although the potential underlying mechanism may have been host- microbial dysbiosis and/or disease influenced by sewage outfall discharges (Campbell et al., in review). Given that water quality has now improved, Phyllospora is currently being restored onto Sydney’s reefs in one of Australia’s most ambitious marine restoration programs (Layton et al., 2018). Restoration relies on the transplantation of reproductive adults from surrounding populations (Campbell et al., 2014; Marzinelli et al., 2016). Previous sampling of surface- associated microbial communities on transplanted Phyllospora showed that some components of their microbiome remained unchanged, despite hosts being moved to a different location (Campbell et al., 2015), suggesting that characteristics of the host may at least partly influence associated microbial communities. Here, we investigated the influence of Phyllospora’s genetics, morphology, condition and geography on its microbiome. We isolated 156 surface-associated bacterial and archaeal communities from genetically diverse Phyllospora individuals sampled across 1300km of coast, which represents the entire latitudinal distribution of this seaweed (12° of latitude, -31S to -43S). If surface-associated microbial communities are driven by characteristics of the host, we predicted that 92

Using genomics to restore and future-proof underwater seaweed forests

the diversity, abundance and composition of microbes would be more strongly correlated with Phyllospora’s phenotypic and genetic diversity than with geography. We further examined the relationship between microbial taxa and particular combinations of host allelic frequencies, morphology and condition as potential drivers of their abundances.

5.3 Methods

5.3.1 Sampling We sampled Phyllospora individuals and their associated microbial communities at eight rocky reef sites spanning Phyllospora’s latitudinal distribution (~1300 km; Fig. 5.1) along south-eastern Australia. Sites were visited in random order over the Austral summer (January) of 2018, to avoid temporal correlations with geographic distribution. At each site, twenty Phyllospora adults >1 m apart were haphazardly collected from an area of 500m2 at 1-5 m depth. 20 cm sections were cut from the middle of the thallus of each Phyllospora individual and placed into a press seal bag whilst underwater. On the surface, samples were kept in their bags on ice for up to 30 minutes, rinsed with 0.22 m filtered seawater to remove epiphytes and a sterilized cotton swab was used to sample the surface (see Marzinelli et al., 2015). Swabs were swiped firmly over 5 cm2 of healthy frond tissue for 30 seconds before being stored immediately in liquid nitrogen, transported to UNSW Sydney and kept at -80°C until DNA extractions were performed. Phenotypic trait data were recorded for each individual. We characterised seaweed morphology by quantifying wet weight (biomass), length of the primary frond (the length from the top of the stipe to the apical tip), total length (including secondary fronds), thallus circumference, frond width, stipe width and stipe length (see Peters, 2015 for sampling details). We recorded sex (male/female) and the mean number of reproductive conceptacles found within a 25 mm2 quadrat placed over three randomly selected fronds to estimate reproductive capacity. We assessed photosynthetic capacity by quantifying maximum quantum yield of one dark-adapted leaf per individual using a Pulse Amplitude Modulated (PAM) fluorometer (WALZ, Germany), and seaweed condition by visually estimating

Using genomics to restore and future-proof underwater seaweed forests

levels of herbivory, bleaching, biofouling and the presence of putative disease (stipe rot; see Fig. S5.1 and Campbell et al., 2014a for further sampling details). Ten unfouled apical tips were then removed from each individual for genotyping (see below). The methods for tissue processing, DNA extraction and genotyping of Phyllospora are described in Chapter 4.

Using genomics to restore and future-proof underwater seaweed forests

Figure 5.1: Map of sites where the seaweed Phyllospora comosa was sampled for a study on host genetics, phenotype and surface- associated microbial communities: PM: Port Macquarie; FO: Forster; AB: Anna Bay: CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

Using genomics to restore and future-proof underwater seaweed forests

5.3.2 Processing and bioinformatics DNA was extracted in random order from each swab sample using the Powersoil DNA Isolation Kit (Mo Bio Laboratories #12888-100) following the manufacturers guidelines. DNA extracts were stored in a -20∘C freezer until amplification with Polymerase Chain Reaction (PCR). We used the primers 341F (5’ - CCTACGGGNGGCWGCAG- 3’) and 805R (5’ -GACTACHVGGGTATCTAATCC- 3’), which target V3-V4 regions of the 16S rRNA gene (Klindworth et al., 2013). Agarose gel electrophoresis, Nanodrop 1000 and the Qubit 2.0 Fluorometer (Thermo Fisher Scientific) were used to check the quality and quantity of the amplicons before being sent to the Ramaciotti Centre for Genomics (UNSW) for sequencing via the Illumina MiSeq 2000 Platform. Gene sequence reads were quality filtered using Trimmomatic (Bolger et al., 2014) with a sliding window trim of 4:15 base pairs (bp) and removal of sequences with <36 bp. Paired-end reads were merged with a minimum length of 400bp and maximum of 500bp using USEARCH (Edgar, 2010). UNOISE was then used to remove chimeras and produce amplicon sequence variants (ASVs), i.e. operational taxonomic units at a unique sequence level (0% distance) (Edgar, 2016). USEARCH was used to map the original reads back to ASVs, generating a table of 3170 ASVs. ASV sequences were searched with BlastN against the SILVA SSU Ref NR99 database for taxonomic classification. ASVs assigned to chloroplasts and rare ASVs (less than 0.1 % total abundance) were removed. We standardised individual counts by their respective sample sequencing depth (total number of ASV counts) to obtain relative abundances and this dataset was used for compositional and diversity statistical analyses. For analyses of individual ASVs, we used raw counts normalised using DESeq2 (McMurdie and Holmes, 2014). We analysed the SNP data with the same genotyping and bioinformatic process as Chapter 4 but including only the individuals from the 8 sites where microbial samples had been taken in 2018. Following filtering steps to correct for genotyping errors, Linkage Disequilibrium and Hardy Weinberg deviations, a dataset of 114 loci across 156 samples were retained.

Using genomics to restore and future-proof underwater seaweed forests

5.3.3 Statistical analyses

All analyses were conducted using the statistical platform R (version 3.6; R Core Team, 2019). We first calculated descriptive statistics including (i) how much genetic and phenotypic variation there was amongst sites and individuals, (ii) correlations between geography, phenotype and genetics, (iii) how many ASVs were identified across all individuals in total and (iv) how many ASVs were shared by all individuals (putative “core” taxa). Differences in genetic and phenotypic variation among sites were tested using permutational multivariate analysis of variance (PERMANOVAs; Anderson, 2001) using adonis in vegan (999 permutations; Oksanen et al., 2018). Variance components were used to determine the amount of variation due to site. We also tested for differences in individual phenotypic traits between sites using a series of linear and (where data conformed to a binomial distribution) generalized linear models. Correlations between traits and latitude were tested using the Spearman correlation coefficient. Dissimilarity matrices for genetic allele frequencies and normalised phenotypic data were calculated based on Euclidean distances between samples. Analyses used 999 permutations, and data were visualised using Principal Component Analysis (PCA) ordinations. Post-hoc pairwise tests were performed with 999 permutation and a false discovery rate (FDR; Benjamini and Hochberg, 1995) applied to correct for multiple comparisons. We also plotted mean variances per site and used permutational multivariate dispersion analyses (betadisper) to test for homogeneity of multivariate dispersion among groups (Anderson, 2006), followed where necessary by post-hoc tests using the emmeans package (Length, 2018). Correlations between host trait distances and geographic distance were tested via Mantel tests in vegan. Data for pairwise geographic distances between sites and site centroid pairwise Euclidean genetic and phenotypic distances among site centroids were calculated using the vegdist function. We examined the relationship between microbial communities and geography by comparing differences in overall microbial communities between sites with PERMANOVA. Similarity matrices were calculated based on Bray–Curtis distances on square-root transformed relative abundance data (‘community

Using genomics to restore and future-proof underwater seaweed forests

structure’) and on Jaccard distances (presence/absence; ‘community composition’). Analyses used 999 permutations and data were visualised using non-metric multidimensional scaling (nMDS) ordinations. Post-hoc pairwise tests were performed with 999 permutations and an FDR applied to correct for multiple comparisons. We also plotted mean variances and used permutational multivariate dispersion analysis (betadisper) to test for homogeneity of multivariate dispersion within groups (Anderson, 2006), followed where necessary by post-hoc tests. We further tested for correlations between overall microbial community dissimilarity (based on the centroid of the Bray-Curtis Dissimilarity matrix) and geographic distance between sites with Mantel tests in vegan.

To examine the relationship between microbial communities and host- related factors, we calculated host genetic diversity (expected heterozygosity, HE) per site with DartR (Gruber et al., 2018) and estimated overall and core microbial community alpha diversity using species richness and Simpson’s diversity indices calculated with vegan. We then tested for relationships between population-level genetic diversity (HE) and average community alpha-diversity at each site via linear regression.

To determine if there was a relationship between phenotypic or genetic distance and microbial community distance (beta diversity) between individual samples, we calculated genetic and phenotypic (Euclidean) distances between hosts and Bray–Curtis distance matrices on square-root transformed relative abundances for all microbiota samples with the vegdist function. We then tested for correlations between (i) genetic diversity and overall microbial community beta-diversity and (ii) phenotypic diversity and microbial community beta- diversity across all sites using Mantel tests in vegan. There was a significant correlation between geographic distance and host genetic distance between sites (see Results), which we were unable to disentangle due to low numbers of strongly genetically related individuals amongst sites. We therefore ran an additional series of Mantel tests that tested for correlations between the factors above using only comparisons between hosts within each site. Overall, these analyses allowed assessing the relationship between host genetic/phenotypic differentiation and 98

Using genomics to restore and future-proof underwater seaweed forests

microbial community dissimilarity at both continental (1300 km) and local (500 m2) scales.

To identify specific host traits that may be associated with microbial community structure, we performed distance-based linear models and Redundancy Analyses (dbRDA; McArdle and Anderson, 2001) on the Bray-Curtis dissimilarity matrix of square-root transformed ASV relative abundances in vegan, including the factor Site first in the model to account for geographic effects. Variables that were strongly correlated (r2 > 0.7) were removed, leaving one in the set to represent those removed. To determine the variance explained by geographic, genetic and phenotypic aspects separately, we first fit the dbRDA model on each component. We then identified which of these aspects explained microbial communities overall, identifying the most parsimonious model including all possible variables using model selection with a stepwise procedure (direction = both) based on p-values. Missing phenotype data (0.9%) were imputed with the average of each metric at each site. Allele frequencies were imputed by using the most common allele frequency observed within each genetic cluster using sNMF (Frichot et al., 2014) in the LEA (Frichot and François, 2015) package.

To determine which ASVs were associated with the host variables selected by the dbRDA model above, we fitted multivariate generalised linear models in DEseq2 assuming a negative binomial distribution for each ASV. There were seven variables identified as important using model selection; each variable was fitted as a single predictor in separate models, which also included the factor Site first in the model to account for geographic effects. Likelihood ratio tests were used to test for significance of each variable, using adjusted p-values to account for multiple testing. Prior to analyses, raw ASV count data was normalised with size factor dispersions calculated for each treatment combination to account for differences in sequencing depth (Love et al., 2014). Dozens of taxa turned out to significantly differ among each of the variables, so we only focused on the five most abundant ASVs that were very significantly affected (adjusted p < 0.01) by each variable. Wald pairwise tests were used to compare relative abundances of taxa among treatments with adjusted p-values to account for multiple testing. 99

Using genomics to restore and future-proof underwater seaweed forests

5.4 Results Host genetics varied significantly among sites, with 50% of the genetic variation explained by site-to-site variation (F7,148 = 21.02, p < 0.01; Fig. 5.2a and Table S5.1). There were differences in the dispersion of genetic data between sites (permdist

F7,148 = 16.95, p < 0.01, Fig. S5.2 and Table S5.2).

Overall host phenotype varied significantly between sites, with 31% of the total phenotypic variation explained by site-to-site variation (F7,148 = 9.51, p < 0.01; Fig. 5.2b). Only two pairs of sites were not phenotypically different from each other (Eden and Bateman’s Bay, p = 0.24, Forster and Port Macquarie, p = 0.12, Table

S5.3). There were differences in the dispersion of phenotypic data between sites

(F7,148 =3.96, p < 0.01, Fig. S5.3), with dispersion in one site (Cronulla) significantly lower than at the other sites (Table S5.4). Individually, all phenotypic traits except total number of branches varied significantly between sites (Fig. S5.4, Table S5.5). Several morphological traits were significantly correlated with each other and with latitude (Table S5.6). Broadly, the total length of individuals was much greater at the Tasmanian (cooler range-edge) populations. Maximum photosynthetic quantum yield was highest at Bicheno, Southport and Anna Bay. Evidence of bleaching and thallus fouling was found across all sites and was relatively high on average (~20%) at Anna Bay. Evidence of herbivory/grazing was highest at Bicheno. Stipe rot disease was also present at all sites except for Port Macquarie and was also highest at Anna Bay.

There were 3169 unique ASVs found across all 156 samples. These were reduced to 2058 ASVs following filtering. The “core” microbial community was represented by 23 ASVs (1.1% of all ASVs), which were found in every sample and represented an average of 45.2% (SE 1.3) of the relative abundance within the community. Most of these (60%) belonged to the class of Proteobacteria; the classes Verrucomicrobia, Planctomycetes and Cyanobacteria were also present (Table S5.7).

There was a significant effect of site on microbial community structure and composition (F7,148 = 11.64, p<0.01 and F7,148 = 7.85, p < 0.01; Fig. 5.2c). There were

100

Using genomics to restore and future-proof underwater seaweed forests

differences between all sites (Table S5.8) and there were also differences in the dispersion of microbial community structure and composition between some sites

(F7,148 = 3.73, p < 0.01, F7,148 = 3.97, p<0.01; Fig. S5.4), with dispersion in one site (Cronulla) again being generally lower than other sites (Table S5.9).

Host genetic distance among sites and overall microbial community dissimilarity were significantly correlated with geographic distance (p < 0.01, r =

0.78; p < 0.05 , r = 0.51, respectively; Fig 5.3), while phenotypic distance was not (p

= 0.06 , r = 0.43; Fig. 5.3).

There was no significant relationship between the genetic diversity of host populations and average microbial ASV species richness or Simpson diversity at each site F1,6 = 0.57, p = 0.48, Fig 5.4a and F1,6 = 1.00, p = 0.36, Fig 5.4b, respectively).

Pairwise host genetic distance and microbial community dissimilarity were positively correlated on a latitudinal scale (p < 0.01, r = 0.42, Fig. 5.5a). Within sites, however, only one location (Anna Bay) exhibited a significant correlation between genetic distance and microbial community dissimilarity (Fig. 5.5b and Table 5.1). Phenotypic distance and microbial community dissimilarity were also positively correlated on a latitudinal scale (p < 0.01, r = 0.15, Fig. 5.6a). Within sites, however, only one location (Bicheno) maintained a significant correlation between phenotypic distance and microbial community dissimilarity (Fig. 5.6b and Table 5.1).

The individual dbRDA models revealed that Site explained 32.57% of variation in microbial communities, while Site + all SNPs explained 53.63% and Site + all phenotypic traits 34.9% (Table 5.2). After model selection, the final dbRDA model of all ASVs including the covariates site, maximum quantum yield, herbivory, stipe base length and three SNP loci (28125_un_3937436, 40713_un_5699768 and 52118_un_7296457) explained a significant amount (35%) of variation in overall microbial community structure. Of these covariates, site specific differences and allele frequencies at locus 28125_un_3937436 had the greatest influence on microbial communities (Fig. 5.7).

101

Using genomics to restore and future-proof underwater seaweed forests

Geography (differences between sites) influenced a large number of ASV’s (1199), the most abundant of which were assigned to the classes Planctomycetacia, Oxyphotobacteria and Alphaproteobacteria (Table S5.10 and Fig. S5.6a). A comparatively lower number of ASVs were associated with individual SNPs than with geography (between 13-50 ASVs associated with each SNP locus, versus > 1000 ASVs related to differences between sites). The most abundant ASVs associated with specific loci were assigned to the classes Planctomycetacia, Bacteroidia, Proteobacteria, Oxyphotobacteria, Verrucomicrobia, Gammaproteobacteria and Alphaproteobacteria and the family Caldilineaceae (Table S5.10 and Fig. S5.6b).

Overall, phenotypic traits were associated with the least number of ASVs relative to other model covariates (13-31). Of those ASVs that were associated with phenotypic traits, the most abundant were assigned to the classes Bacteroidia, Proteobacteria, Verrucomicrobia, Gammaproteobacteria and Alphaproteobacteria (Table S5.10 and Fig. S5.6c).

102

Using genomics to restore and future-proof underwater seaweed forests

(a) (b)

Figure 5.2: (a) Principal Component Analysis (PCA) of the seaweed Phyllospora comosa’s genetic structure, based on allele frequencies at 114 SNP loci; (b) PCA of phenotype, based on 11 traits describing morphology and condition; (c) nMDS analysis of Phyllospora- associated microbial communities, based on (i) Bray-Curtis and (iii) Jaccard dissimilarity matrices for square-root transformed relative abundances; Sites North to South: PM: Port Macquarie; FO: Forster; AB: Anna Bay: CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

103

Using genomics to restore and future-proof underwater seaweed forests

Figure 5.3: Relationship between geographic distance (km) and pairwise distance/dissimilarity between centroids of the eight Phyllospora comosa populations for genetics (Euclidean distances), overall surface-associated microbial communities (Bray-Curtis dissimilarity of square-root transformed relative abundances) and phenotype (Euclidean distances). Data with significant associations detected using Mantel tests fitted with linear regression. 95% Confidence Intervals shaded in grey.

104

Using genomics to restore and future-proof underwater seaweed forests

Simpson diversity Species richness

Host genetic diversity (HE) Host genetic diversity (HE)

Figure 5.4: Relationship between site-level host genetic diversity (HE) and a) species richness and b) Simpson diversity of microbial communities associated with the seaweed Phyllospora comosa’s surface. Data collected at eight sites.

105

Using genomics to restore and future-proof underwater seaweed forests

(a)

(b)

Figure 5.5: Relationship between microbial community dissimilarity (Bray-Curtis on square-root data) based on all ASVs and genetic distance (Euclidean) between all Phyllospora comosa individuals (a) across and within all sites (b) within each site. Sites are ordered north to south: PM: Port Macquarie; FO: Forster; AB: Anna Bay: CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport. Data with significant associations detected using Mantel tests fitted with linear regression. 95% Confidence Intervals shaded in grey. 106

Using genomics to restore and future-proof underwater seaweed forests

(a)

(b)

Figure 5.6: Relationship between microbial community dissimilarity (Bray-Curtis on square root relative abundances) based on all ASVs and phenotypic distance (Euclidean) between all Phyllospora comosa individuals (a) across and within all sites (b) within each site. Sites are ordered north to south: PM: Port Macquarie; FO: Forster; AB: Anna Bay: CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport. Data with significant associations detected using Mantel tests fitted with linear regression. 95% Confidence Intervals shaded in grey. 107

Using genomics to restore and future-proof underwater seaweed forests

Table 5.1: Mantel tests results describing relationship between (i) Phyllospora comosa genetic distance (Euclidean) and (ii) phenotypic distance (Euclidean) on surface-associated microbial community dissimilarities calculated using Bray-Curtis on square-root transformed relative abundances.*

Site a Genetics Phenotype Overall microbiome Overall microbiome Mantel r p Mantel r p PM -0.166 0.775 -0.1767 0.87 FO 0.089 0.194 0.233 0.056 AB 0.468 0.002 0.023 0.4 CR -0.004 0.545 0.094 0.283 MB 0.014 0.498 -0.061 0.664 ED 0.125 0.089 0.1721 0.078 BI -0.055 0.703 0.221 0.045 SO -0.2044 0.979 -0.0156 0.536

108

Using genomics to restore and future-proof underwater seaweed forests

Table 5.2: ANOVA output table for (i) initial and (ii) final dbRDA models, showing geographic, host phenotypic and host genetic associations with Phyllospora comosa’s associated microbial community structure. The final model was selected via model selection based on p-values, using a stepwise selection procedure.

Df SS F p Adjusted R2 (i) Site 0.326 Phenotype 0.115 SNPs 0.536 Site + phenotype 0.349 Site + SNPs 0.536

(ii) Overall ASVs 16 10.36 6.302 0.001 0.350 ~ Site + PAM + 139 14.28 28125_un_3937436 + 40713_un_5699768 + Herbivory+ Stipe base length + 52118_un_7296457

109

Using genomics to restore and future-proof underwater seaweed forests

Figure 5.7: Associations between geography, genetics and phenotype of the Phyllospora comosa host with 156 microbial community samples isolated from the surface of host fronds; as inferred by distance-based Redundancy Analysis (dbRDA). Vectors indicate the direction and strength (length) of each significant variable in explaining variation between microbial community samples. Vector labels starting with “site” indicate specific site association; PAM = maximum quantum yield; logstipebaselength = stipe base length (data log transformed), Herbivory = amount of grazing on thallus; all vectors starting with X indicate SNP loci. Numbers at the end of SNP names indicate allele frequencies of the major allele at that locus. PM: Port Macquarie; FO: Forster; AB: Anna Bay: CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

110

Using genomics to restore and future-proof underwater seaweed forests

5.5 Discussion

The importance of microbial communities to the development, health and survival of eukaryotic hosts is a key theme emerging in contemporary ecology. Understanding the mechanisms underpinning these interactions is critical, particularly in the context of environmental change, species decline and the development of active restoration and future-proofing strategies (Wood et al., 2019). This study investigated the combined association of geography and host genetic and phenotypic traits on the microbiome of a forest-forming seaweed, Phyllospora. Using these multifaceted data, we were able to explain over half of the microbial community variation even at a high taxonomic resolution (ASV). In most cases, Phyllospora’s microbiome was most strongly associated with local environmental conditions (i.e., was site-specific). Nevertheless, similar host genotypes and phenotypes generally had more similar microbial communities, although the effect of these traits on community dissimilarity were most apparent at regional rather than site specific scales. While there were strong correlations between Phyllospora genetic structure and geography, our modelling approach allowed us to gain an idea of the respective influences of these two factors on overall microbial communities. For example, Site plus host genetics (SNPs) explained 21% more variation than just Site alone, suggesting that at least 21% of the variance in microbial communities is explained by host genetics. We were further able to identify key loci and phenotypic traits that could be used to predict the relative abundance of specific ASVs found on Phyllospora’s surface, including taxa with known associations to seaweed disease and cell degradation. These results indicate that Phyllospora’s surface-associated microbial communities are jointly shaped by local environmental conditions and host-specific differences. The strength of each of these factors is, however, context- specific - and likely dependant on trade-offs between each of these traits as well as additional ecological or environmental influences.

111

Using genomics to restore and future-proof underwater seaweed forests

5.5.1 Composition of Phyllospora’s microbiome

Although Phyllospora’s microbial communities varied widely, we found that, in terms of abundance, almost half of the microbiome (~45%) was made up of just 23 ASVs that were found on every individual. While this level of homogeneity is at least one order of magnitude greater than has been found for other holobiont taxa such as corals (Hernandez-Agreda et al., 2018) and humans (Rothchild et al., 2018), similarly high levels of microbiome homogeneity have been found for sponges (e.g Marino et al., 2017 - although taxa were grouped with lower resolution; but see Griffiths et al., 2019) and other seaweeds (e.g Roth-Schulze et al., 2018). It is possible that these taxa in particular are involved in co-adapted relationships with their seaweed hosts, and further research focusing on this or separating out the influences on variable taxa could be critical for understanding Phyllospora’s biology.

5.5.2 Influence of geography

We found a strong association between geography and the structure and composition of Phyllospora’s microbial communities. Differences in the overall microbiome were linked to both large-scale patterns that could be attributed to geographic distance between sites, but also site-to-site differences. Teasing out the specific environmental drivers of these patterns was beyond the scope of this study, but previous work has highlighted that temperature, light (Brown et al., 2012; Gilbert et al., 2012; Ghiglione et al., 2012; Rusch et al., 2007), salinity (Weigel and Pfister 2019), wave motion and phosphate concentrations (Marzinelli et al., 2015) can be influential over small and medium spatial scales. It is important to note that some driving factors may be more important than others; many of the differences between sites were attributed to changes in abundance of ASVs that have been previously linked to seaweed defence (Hyphomonadaceae), heat stress (Blastopirellula) and cell wall degradation (Cyanobacteria; Hollants 2013). This suggests that local‐scale differences in temperature and grazing may be particularly important drivers of microbial communities in this system.

112

Using genomics to restore and future-proof underwater seaweed forests

5.5.3 Genetic associations

There was also a strong correlation (42%) between genetic distance and microbial community dissimilarity at latitudinal scales, although whether this was causal could not be discerned. At the site level, there was little relationship between host genetic distance and microbial community dissimilarity, which suggests that the associations observed at latitudinal scales were either simply due to geography or that they break down at local scales. A lack of sufficient host genetic differentiation within sites is unlikely to be the reason for this, given that the northern-most site (Port Macquarie) had several very genetically distinct individuals that still provided no evidence for linked patterns in microbial community differentiation. Nevertheless, host genetics were significantly correlated with microbial community dissimilarity at one site (Anna Bay), which indicates that host genetic factors are associated with the microbiome at local scales under certain conditions. Previous work suggests that healthy specimens of seaweeds maintain microbial communities more specific to their environment (McGeoch et al., 2019), but when seaweeds become stressed or bleach this specificity breaks down and communities may become more homogeneous (Marzinelli et al., 2015). As hosts at this site (Anna Bay) had the highest levels of fouling, bleaching and disease relative to all other sampling locations, genetics may be playing a role in determining resistance or susceptibility to these fouled/disease phenotypes, with effects only becoming observable when ”healthy” interactions based on non-genetic (e.g. environmental) factors break down. This could also be because other factors which normally drive community patterns (e.g. local environment, phenotype) may not be as important at this site, or that host genetic effects on the microbiome are stronger here. Indeed, Anna Bay had very low variation in many morphological traits (Fig. S5.5), so an additional explanation of the patterns observed here could be that when phenotype is less variable, genetic differences become important or are easier to discern. The idea that genetics may have context-specific effects is further supported by the lack of a detectable link between overall genetic diversity and microbial diversity or species richness at each site. This contrasts with some contemporary

113

Using genomics to restore and future-proof underwater seaweed forests

ecological theories, which predict that increasing genetic diversity translates to higher phenotypic diversity and available niche space for different species (Evans et al., 2016; Whitham, 2006). In this study, a few specific host loci were associated with changes in microbial community structure. Of these, the greatest association was seen at one specific locus in which being heterozygous or homozygous for the minor allele was associated with higher abundances of several taxa (e.g. the families Saprospiracae and Hyphomonadaceae) that have been linked to warm environmental conditions and increased herbivore associations (Castro et al., in prep.) but lower levels of bleaching disease (Marzinelli et al., 2015) in other species of brown seaweeds. We propose that a few alleles across the genome may influence community structure via exclusion or affinity for particular microbial taxa, which is in line with recent work suggesting that specific environmental or physicochemical characteristics of the host, rather than the overall genetics or host phylogeny explain microbial community composition (Hacquard et al., 2015). While at this stage we cannot infer causality, current estimates of overall allelic control over microbial diversity in human systems are also similarly small (0.65– 0.97%, Wang et al., 2016). Although we had no specific hypotheses about which ASVs might be affected by differences in alleles, our data provides a powerful platform upon which to explore these ideas.

5.5.4 Phenotypic associations

Microbial communities from phenotypically distant hosts were generally more dissimilar at a latitudinal scale, indicating that some of the phenotypic traits measured were associated with microbial community structure. This association was generally quite weak (explained 15% of variance), however this is to be expected given that our phenotypic measure represented the combination of morphology, disease, photosynthetic efficiency, sex and reproductive capacity – traits which likely had associations of varying direction and magnitude on different microbial taxa. Further, as was the case for genetically dissimilar hosts, this association broke down almost completely at local (500 m2 within-site) scales. Phenotypic variation was much lower within sites than between them, potentially

114

Using genomics to restore and future-proof underwater seaweed forests

presenting insufficient variation to create observable effects. The Bicheno site was the only exception, where phenotypic distance had a significant correlation (22%) with overall microbial community dissimilarity. Phenotypic variation was high at this site, particularly for morphological characteristics and levels of grazing (see Fig. S5.5). Phenotype-microbial associations may thus still exist at local scales and may only be observable if (i) other drivers of microbial community structure are weaker, and/or (ii) one or more of the phenotypic associations are strengthened. Future surveys could target these associations by sampling more broadly across phenotypes, as we randomly targeted reproductive individuals, or conducting manipulative experiments with the ASVs identified as associated with herbivory and other phenotypic traits here. While we can only speculate about the specific factors contributing to this pattern, it is interesting to note that the Bicheno site had the highest levels of herbivory observed across all locations. Herbivory can be strongly related with microbial community structure on other species of brown seaweeds (Langley et al., in prep.). In this study, herbivory was one of the key phenotypic variables associated with microbial community composition. Herbivory may thus have stronger associations to the microbiome than other phenotypic/functional traits measured here. It is important to note that although we found no significant association between geographic distance and phenotypic distance, seaweed phenotype is generally shaped by localised environmental conditions (Flukes et al., 2015; Fowler- Walker et al., 2006). Thus, microbial communities that are associated with particular phenotypic traits (e.g. herbivory, stipe length or photosynthetic efficiency) may be confounded by underlying factors that may also shape morphology or the presence of disease phenotypes at local scales (e.g. number of herbivores, temperature, nutrients or wave exposure). This idea is supported by the fact that Site plus phenotype only explained a small amount (~ 2.3%) more variation in microbial community structure than just Site alone; further experimental work is thus necessary to untangle these relationships. Microbial communities can also mediate seaweed traits such as morphology, physiological health and susceptibility to disease or grazing. Indeed, many of the most abundant ASVs that were associated with phenotypic traits in 115

Using genomics to restore and future-proof underwater seaweed forests

this study belonged to genera involved in these functions. It is important to note that mechanistic host phenotype-microbiome associations may operate in either direction; i.e. associated phenotypes may shape microbial community structure, or microbes may be involved in shaping the phenotype of the host, or both (see Egan et al., 2013 and references within). The direction and function of any mechanisms underpinning these associations should therefore be explored with more detailed research.

5.5.5 Study limitations and future directions

The present study provides a powerful dataset collected from a wide variety of environments and individuals, but the nature of sampling wild habitat-formers means that disentangling geographic from host effects from purely observational data remains a challenge. This is particularly so for species which exhibit relatively low genetic diversity and strong isolation by distance. Future manipulative experiments, for example experimental transplants of hosts (Campbell et al., 2015), are thus needed to tease out the effect of host genetics versus geographic and environmental influences. Another option could be to investigate these interactions in a clonal species, so that replicate genotypes could be compared to others that are genetically dissimilar. This study focused on the important drivers of microbial community assemblages, but the ecological consequences of these remain largely unknown. Our work provides a strong basis to design future experiments with more targeted hypotheses in mind. For example, as bleaching, associated heat stress and herbivory represent one of the greatest threats to seaweed forests worldwide, further investigations which tease apart geographic effects on allele frequencies to determine whether susceptibility with the major allele, or increased defences due to the minor allele is conferred at locus 28125_un_3937436 (or other linked loci) are of critical importance. This may enable a better understanding of the biological effects of ASVs on hosts, and whether disruptions of host-specific taxa have biological effects. Incorporating taxa grouped by functional traits may also find different patterns of association.

116

Using genomics to restore and future-proof underwater seaweed forests

5.6 Conclusion

The incorporation of genome-wide genetic data to microbiome studies is a novel approach (Awany et al., 2018) that is only beginning to be applied to wild systems. Our results are consistent with recent work focusing on humans that show that the local environment, not host traits, contribute most of the compositional variation in the microbiome (Rothschild et al., 2018). Interactions between host genetics, phenotype, geography and significantly associated ASVs identified in this study provide a platform to support and guide future avenues of research. This includes experimental work to determine if the presence of specific microbial taxa has an underlying functional importance to the host (e.g. Burke et al., 2011 a,b). Although the development of techniques to harness the beneficial effects of microbial interactions are still in their infancy in seaweed systems, management or manipulation of microbial functions and communities have become well- established in bioremediation of terrestrial and aquatic ecosystems (e.g Tyagi et al., 2011). This work thus provides an important baseline to work towards similar restoration-enhancing interventions, as well as to the development of rapid, cost- effective assessment and monitoring tools under changing environmental conditions.

117

Using genomics to restore and future-proof underwater seaweed forests

Chapter 6:

General Discussion

118

Using genomics to restore and future-proof underwater seaweed forests

Human activities are changing our world on spatial and temporal scales that are unprecedented in human history (Zalasiewicz et al., 2008). It is becoming increasingly evident that the only clear way to protect and maintain key ecosystem functions and services provided by our natural systems now involves collective actions to intervene at many levels of human and natural organization (Steffen et al., 2018). This includes large-scale efforts to tackle global stressors that cause widespread degradation (e.g. stopping and reversing carbon emissions), as well as comparatively small-scale efforts, such as the active management of local systems that are impacted by other, interactive stressors operating at smaller scales. Ecological understanding is critical to help guide these actions towards a more stable and sustainable future (Schmitz, 2017). The overarching aim of this thesis was to investigate and develop approaches to improve restoration and inform future-proofing of underwater forests, by enhancing our understanding of the ecology of natural seaweed populations. In the Introduction and in Chapter 2, I discussed the global decline of underwater seaweed forests and established the need to implement restoration and future-proofing strategies for foundation species generally. Since restoring to account for current and future conditions is becoming critical to address local and global stressors, developing guides for restoration and future-proofing strategies is a critical priority for research. In Chapter 2, I reviewed the literature and derived a framework that described several critical research themes that need attention to progress marine macrophyte restoration and future-proofing strategies. I found that a range of promising approaches to restoration and future-proofing already exist and may be appropriated from other systems. Genomics tools emerged as being highly versatile, with the potential to tackle multiple issues within the restoration and future-proofing framework. I then used this framework to guide a series of empirical studies (Chapter 3-5) that demonstrate how genomic data, analysed through the lens of population genetics, landscape genomics and microbial ecology, can be used to develop and guide underwater forest restoration and future-proofing solutions, using restoration of the fucoid Phyllospora comosa (hereafter, Phyllospora) as an example.

119

Using genomics to restore and future-proof underwater seaweed forests

In this final General Discussion, I build on the restoration and future- proofing trajectory framework outlined in Chapter 2 and provide a synthesis of my major findings from subsequent chapters and their limitations. These can be classified into three broad themes: (1) informing donor selection for restoration and future-proofing strategies, (2) monitoring and assessing the success of such strategies and (3) understanding and harnessing host-microbiome interactions. Under these themes, I discuss how genomic information may be used to enhance restoration and future-proofing success in underwater forests and postulate future avenues of research. Finally, I consider how this work fits within the broader context of ecological restoration and future-proofing of underwater forests.

6.1. Incorporating genomic data into restoration and future-proofing strategies for underwater forests

6.1.1. Donor selection Ecological restoration requires the use of donor material from wild or lab-cultured sources. Where wild material is used, the sites for donor collection should be carefully chosen to avoid causing further impacts to extant populations (as discussed in Chapter 2) and if possible, to also harness the underlying ecology of the species to establish desired characteristics in the restored population. Two key aspects to consider are donor provenance and genetic diversity. These two aspects can profoundly influence the restoration success of foundation species, for example, by directly affecting establishment rates and fitness, by enhancing environmental resilience and/or by increasing the adaptive capacity of macrophyte populations (Hughes and Stachowicz, 2004; Forsman and Wennersten, 2016; Wernberg et al., 2018; Vander Mijnsbrugge; see Chapters 2 and 3). Decisions regarding donor selection will vary depending upon project goals, e.g. whether it is to replicate genetic diversity and structure of extant populations, or explicitly introduce novel characteristics in order to future-proof restored populations. Regardless of which of these goals is chosen, an understanding of genetic patterns is needed and restoration guidelines therefore typically recommend incorporating

120

Using genomics to restore and future-proof underwater seaweed forests

genetic understanding where possible (e.g. Society for Ecological Restoration, 1993; International Union for Conservation of Nature, 2013; see Chapter 2). Existing seaweed restoration programs have generally collected donor plants from one or more sites nearby the restoration site, most often in the absence of empirical genetic information. Often, this is because such programs are still in early, experimental stages and this is logistically the easiest option. In addition, this could also be due to a generally accepted paradigm within traditional restoration that “local is best” (Evans et al., 2018; Vander Mijnsbrugge et al., 2010). In rare cases, decisions have been made based on existing knowledge of genetic population connectivity e.g. data derived from neutral genetic markers such as microsatellites (e.g. Coleman and Kelaher, 2009). Using nearby populations as donors in the absence of genetic or genomic information, however, might not achieve desired restoration outcomes because these populations may not be genetically diverse enough, or may not possess functional characteristics that will make them resistant or resilient to future conditions (e.g. tolerance to rising ocean temperatures; see Chapter 4). In Chapters 3 and 4, I explicitly collected and applied genomic data to delineate donor source zones and characterised putatively adaptive genomic regions that may be desirable for restoration and future-proofing strategies. Both studies revealed that Phyllospora’s genetic diversity is relatively low overall with high connectivity, mirroring previous studies that characterized Phyllospora’s genetic patterns using microsatellites (neutral genetic markers; Coleman and Kelaher, 2009; Coleman et al., 2011a). Although at small scales (10’s of kms and particularly around Sydney), genetic diversity and structure was relatively similar amongst populations, at larger (species-range-wide) scales these metrics varied greatly. For instance, in comparison to central populations, genetic diversity was significantly reduced and there was significant genetic structuring at the range edges. These patterns suggest that sourcing donors from central, but not range- edge population is likely to achieve high genetic diversity that may optimise restoration success and enhance future adaptive capacity to a range of stressors in Sydney or other locations over time. These data can therefore be used to (i) target specific donor populations based on genetic structure and desired genetic diversity 121

Using genomics to restore and future-proof underwater seaweed forests

levels, (ii) allow for the spreading out of impacts of donor material collection across genetically appropriate local populations and (iii) identify areas at risk of localized collapse without intervention (based on within population genetic diversity; see Chapter 4) - although consideration of traits/factors other than host genetics are also likely to be necessary (see Chapter 5). Many seaweed species, particularly brown seaweeds, have been characterised for population genetic or phylogeographic studies using putatively neutral genetic markers such as microsatellites, the choice of population genetic marker over the past decades (see Hu et al., 2016 for a review). The advent of cheaper sequencing and SNPs (Single Nucleotide Polymorphisms) now allows both neutral and functional/adaptive portions of the genome to be simultaneously analysed. In this thesis, using SNPs to delineate genetic boundaries at relatively small scales (<100 km radius of the restoration zone) yielded similar estimates of genetic variation (via AMOVA) to previous genetic studies of Phyllospora that used neutral microsatellite markers (Coleman and Kelaher, 2009, Coleman et al., 2011a). The similarity between the current and previous data also demonstrated that genetic diversity and structure has likely not changed in the 10 years since the last study of Phyllospora’s genetics and connectivity (Coleman and Kelaher, 2009). These points are encouraging as they suggest that in the absence of modern genomic data, other restoration programs may be able to draw from previous neutral data, where available, for their design. Both Chapters 3 and 4 also demonstrated, however, how modern genomic markers, such as SNPs, can provide powerful advantages over these traditional genetic approaches. Even with a relatively small number of SNPs, the range-wide SNP data presented in Chapter 4 produced information at a much higher resolution (i.e. higher FST estimates) compared to previous estimates of genetic structure based on microsatellite markers. It also allowed me to explicitly test for natural selection processes. While I did not find any evidence of early selection occurring in the first (F1) generation of restored populations using nearby donors (Chapter 3), data from populations along the entire distribution of the species did reveal that some Phyllospora populations may be locally adapted, particularly at the range edges. While this study was not designed to explore the causes or 122

Using genomics to restore and future-proof underwater seaweed forests

functional effects of adaptation, by applying gene-environment association analyses I established that these patterns may be linked to sea surface temperature. By considering levels of genetic diversity derived from both neutral and adaptive regions of the genome, or focusing on genomic regions that are potentially under selection, genomics can thus be used not only to delineate donor source areas for restoration (as in Chapter 3), but also to provide information for future-proofing strategies such as assisted evolution (Chapter 4). For example, incorporation of desirable genetic/ functional characteristics identified here, and further validated experimentally, could lead to restored populations that are stressor-tolerant or highly resilient. Considering that climate change impacts are already producing large but cryptic effects on the genetic diversity and structure of seaweed forests (Coleman et al., in review, Gurgel et al., in press; Wernberg et al., 2010), the use of modern genomic tools to inform and enable development of these strategies will become increasingly important.

6.1.2 Evaluating the success of restoration and future-proofing strategies

Ensuring the success of any active interventions into ecological systems requires some level of monitoring and evaluation in order to determine whether the defined goal was achieved (see Chapman and Underwood, 2010); however, genomic success is rarely measured (Mijangos et al., 2015). There is little point in investing resources into planning and executing strategic restoration and future-proofing techniques if the subsequent genetic patterns are not assessed to ensure they have been achieved. Thus, restoration programs that aim to achieve desired levels of genetic diversity and/or structure, or to incorporate specific genotypes, must use genetic tools to ensure such aims are met and inform the future direction of restoration. In Chapter 3, I used genetic assignment, selection tests and assessment of genetic diversity and structure to demonstrate that the F1 generation of restored Phyllospora populations had similar genetic structure and diversity to extant populations, with no evidence of natural selection. This is encouraging for Phyllospora’s restoration program (“Operation Crayweed”) as it demonstrates that current restoration techniques for this species are successful in introducing diverse

123

Using genomics to restore and future-proof underwater seaweed forests

local genotypes to degraded sites. Further, it shows that we can effectively “design” populations (to some degree) using careful selection of donor plants. This last point is particularly encouraging in the context of environmental change as it suggests that we can manipulate the genetic makeup of restored populations to future-proof these underwater forests. The relevance of these techniques and findings for achieving broader restoration objectives remain, however, unknown. For instance, it is not clear how this translates to restoring associated biodiversity and ecosystem functions, and experimental tests which consider different levels of genetic diversity or structure on the broader ecosystem are needed. Incorporating non-local genotypes may also present challenges associated with maladaptation or genetic swamping if mixed with genotypes from local populations (Weeks et al., 2011). Further, extending these methods to other seaweed species may also yield different results if reproduction is not as rapid and synchronous as it is for Phyllospora, or if genotypic variation results in differential fitness and reproduction. Nevertheless, these findings are encouraging in the broader context of underwater forest restoration and potential future-proofing strategies.

6.1.3. Harnessing host-microbiome interactions

Recent research has revealed that microbial interactions are critical to the development, function, survival and evolution of eukaryotes. Thinking of eukaryotic hosts as “holobionts”, that is, the host plus its associated microbiota, and understanding host-microbial interactions is beginning to provide new innovative solutions to enhance restoration programs in many systems (see Bashan et al., 2014; Holguin et al., 2001; Hong and Lee, 2014 and Chapters 2 and 5). In particular, work in coral reefs is paving the way in marine coastal systems. For example, experimental inoculations of bacterial cocktails can reduce negative impacts of a simulated oil spill (dos Santos et al., 2015), mitigate coral bleaching and alleviate pathogenic infection (Rosado et al., 2018). Recent work has also shown that coral-associated bacterial communities are linked to intraspecific variation in coral heat tolerance (Ziegler et al., 2017) and this may provide a platform for inferring/testing resistance to this and other stressors.

124

Using genomics to restore and future-proof underwater seaweed forests

In marine systems, macroorganisms are embedded in a “soup” of microbes, which have the potential to interact with and influence the structure of host- associated microbiota. A fundamental question is, therefore: What are the main drivers of the structure of host-associated microbiomes? Is it the environment, e.g. where the hosts are, is it characteristics of the host (such as genetics or phenotype), or a combination of these? What is their relative importance? This is not only relevant for management interventions in marine systems, but arguably for any intervention, including manipulation of the human microbiome to improve human health outcomes (e.g. see Relling and Evans, 2015). For the former, this may have implications on how we design restoration and/or future-proofing programs. For instance, if microbial communities or important taxa are mainly influenced by the environment, attempts to harness microbial interactions to improve restoration or future-proofing outcomes may fail as microbial communities are likely to be swamped by local microbial taxa. This is especially pertinent if strategies involve introducing microbes that are not already found in similar abundances in the target environment. Alternately, if microbial communities are influenced by host-specific traits, harnessing microbial interactions may be as simple as including genotypes or phenotypes with beneficial communities; but it could also mean that strategies attempting to manipulate the microbiome may need to be tailored to specific host types, as is often done in human medicine (Relling and Evans, 2015). In Chapter 5, I investigated these ideas with the seaweed holobiont and found that geography, genetics and seaweed phenotype all play a significant role in shaping Phyllospora’s microbiome. However, independently, they explained ~32.5 %, 21% and 2.5% of variation in surface-associated microbial communities, respectively, suggesting that genetics, together with the environment, are both strong drivers of microbiomes and therefore both should be considered in a restoration context. Often, as was the case here, environmental and genetic differences are correlated. Experimental work is therefore needed to decouple environmental versus host-genetics influences on the associated microbiota. For example, experimental transplants among populations (see e.g. Campbell et al., 2015) can help disentangle the effects of the host versus the environment in 125

Using genomics to restore and future-proof underwater seaweed forests

influencing host-microbiomes and thus determine when/where each of these matters most. Importantly, several key genetic loci and phenotypic traits were strongly related to taxa with known associations to seaweed defence, disease and tissue degradation (Hollants, 2013; Marzinelli., et al., 2015; see Chapter 5). While the incorporation of microbial associations into underwater forest restoration and future-proofing schemes still remains hampered by a lack of understanding of which taxa or functions are truly important for seaweeds, the findings in Chapter 5 suggest that restoration and future-proofing strategies for managing Phyllospora may benefit from accommodating host genetics in their design. Considering the potential associations between microbes, host genetics and increased stress/disease and grazed phenotypes that were observed, potential strategies may include manipulating genotypes or determining if genotypes that do not naturally have beneficial microbiomes may be manipulated early on in life. Although marine systems present significant challenges to manipulation of microbial communities, recent work in corals has demonstrated that coral-associated microbiomes can be influenced to develop in distinct directions following microbial dosing at early larval stages in experimental conditions (Damjanovic et al., 2017). Focusing on early life stages can thus enhance the feasibility of using such solutions in seaweed systems.

6.2 The value of incorporating ecological knowledge

By matching genomic data with ecological knowledge, we can enhance our understanding and practice of restoration and future-proofing strategies at ecologically relevant scales. This was demonstrated in several ways throughout this thesis, for example: (i) in Chapter 3, choices to source donors ultimately depended upon the state (based on informal quantification of population density and condition measures) of donor sites as well as genetic source-delineation so as to avoid potential overharvesting issues (as Discussed in Chapter 2). (ii) Genetic measures also failed to reflect that initial recruitment was lower than is likely to be necessary for a self-sustaining population in all restoration sites in this study

126

Using genomics to restore and future-proof underwater seaweed forests

(Chapter 3). This indicated that repeated restoration efforts whereby restored plots are periodically topped up with new donor individuals is likely to yield faster, more sustainable results than single-event restoration (this is now being implemented at all restoration sites described in this study). While we may have deduced this from genetic analysis during later stages of the restoration, combining genomic with ecological knowledge early on is clearly necessary to enhance restoration efforts and reduce resource costs (such as those associated with genomic sequencing), particularly during early developmental stages. (iii) The unexpected finding that sex ratios were biased towards males in the northern-most, range-edge site (Port Macquarie) also indicated interesting potential underlying mechanisms of selection and further investigations of this may lead to new solutions and future- proofing innovations. It is also important to highlight that although this thesis focused on restoration success in terms of seaweed biology and processes, restoration of foundation species is generally driven by the overarching goal of improving associated ecosystem functions and biodiversity. Assessing the restoration via these ecosystem-wide components was beyond the scope of this thesis, however I did conduct ecological surveys at all restoration sites, as well as multiple reference (extant Phyllospora forests surrounding Sydney) and control (unrestored in Sydney) sites during my candidature and this monitoring program is ongoing.

6.3 Limitations and future directions

6.3.1 Genomic markers and data resolution Quantitative genetics studies take years (even decades) to complete, so until recently it was largely impractical to base restoration or future-proofing strategies on genomic data that informs adaptive processes. This thesis demonstrates the usefulness of landscape genomic approaches to rapidly inform restoration and future-proofing strategies, using contemporary genomic techniques. However, here I was only able to use a relatively low number of SNPs compared to most landscape genomic studies conducted in terrestrial systems (which use 10’s-100’s of thousands of SNPs; e.g. Feng et al., 2015; Selechnik et al., 2019) because of major

127

Using genomics to restore and future-proof underwater seaweed forests

challenges involved in obtaining DNA of sufficient quality and quantity and limits on resources to extract multiple samples (see Chapter 3). Brown algal (phaeophycean) tissue is known to create inherent challenges to obtaining good quality DNA material as it can contain polysaccharides that interfere with PCR and DNA digestion (Coleman et al., 2018; Wilson et al., 2016). Even though higher numbers of SNPs are beginning to be used on macroalgae (Fraser et al., 2016; Coleman et al., in prep.), missing data ratios are still high (e.g. more than 95% missing data with 75,000 SNPs were reported in Fraser et al., 2016). This represents a challenge to accurately identifying outlier loci and applying landscape genomics techniques, which are well known to have high false-positive rates (Narum and Hess, 2011; Whitlock and Lotterhos, 2015). This possibly occurred in Chapter 4 as patterns of adaptive diversity and structure differed with each method used to identify outlier loci. Another limitation with using a low number of SNPs is that it may lead to underestimation of phenotypic or functional variance, especially if phenotypic traits are controlled by many genes of small effect (which is the reigning paradigm in quantitative genetics). Further, SNPs have consistently been found to underestimate phenotypic variation when compared to traditional heritability estimates derived from traditional quantitative genetic studies (Wood et al., 2014; Sandoval-Motta et al., 2017) and this must be considered during any investigations that search for heritable functional variation. While acknowledging these limitations, however, in Chapter 4 I was still able to characterize adaptive diversity and loci under putative selection for this species. This is an important first step and provides avenues of prioritization for future research efforts into genomic areas of interest. These types of studies would likely be improved by using larger quantities of SNPs and functional studies based on RNAseq or whole-genome data. Reducing sample size to increase the number of SNPs genotyped may provide a possible compromise between resource limitation and data resolution. Further, although I focused on the use of genomic tools to design strategies based on natural/standing genetic variation, it is important to note that an understanding of adaptive genomic patterns and functional implications will be increasingly useful in the context of the development of more transformative gene editing and synthetic 128

Using genomics to restore and future-proof underwater seaweed forests

biological solutions (Coleman and Gould, 2019). Ultimately, working towards higher genomic resolution will provide the best information for restoration and future-proofing research.

6.3.2 Time constraints and the need for research to match restoration timescales Chapter 3 represents a crucial step in early restoration assessment. It is important to emphasize, however, that restoration (and potential future-proofing) programs are long-term endeavours and genomics tools should accordingly be used to assess success at multiple stages throughout their lifetimes. While the timeframe necessary to do this was not an appropriate timescale for my thesis, this would take the form of repeated genetic surveys across life stages and generations, with repeated or larger transplant efforts. Restoration projects with more established populations (i.e. that are at least in the stage of having an F3 generation) can use genetics to estimate effective population sizes. Projects in their early stages will also clearly benefit from frequent field surveys, because they will provide information on recruitment density which will ultimately play a large role in restoration success or failure.

6.3.3 Understanding how genomics can enhance restoration and future-proofing – ongoing challenges Environmental management interventions have inherent risks and should only be considered after planning for success and failures (IUCN, 2013). Interventionist strategies to enhance Phyllospora’s survival under future climates will also be enhanced if they are based on a wide range of scientific evidence and testing to enhance management strategies. Key themes and questions for the future include:

• Determining the optimum genetic conservation unit Many questions remain about how best to use genomic data for conservation and management (Waples and Lindley, 2018). In particular, this includes prioritizing the relative importance of conserving key adaptive alleles or genetic diversity (Moritz, 2002). While some work suggests that genetic heterogeneity plays a role

129

Using genomics to restore and future-proof underwater seaweed forests

in determining the tolerance of chimeric seaweeds to abiotic stressors (Medina et al., 2015) and that genetic diversity may confer adaptive capacity through climate stress (Wernberg et al., 2018), the influence of genetic traits and diversity of seaweed populations in a restoration/management context needs to be explicitly explored. Testing this will involve embedding further experiments that compare low and high genetic diversity (for example, sourced from the central versus range edges) to discern best practices around future-proofing the population, or by creating a common garden including individuals found across the species range. The incorporation of transcriptomics and proteomics to study variation in gene activity as a response to environmental variables would enable more specific identification of beneficial genetic traits.

• Determining key microbial taxa/functions that shape health, survival and adaptation of seaweeds

Our understanding of which microbial taxa are important (i.e. having beneficial or negative impacts) to the seaweed holobiont is still developing and may be improved by (i) exploring all components of the associated microbiota, i.e. not only bacteria and archaea, but also fungi and viruses, (ii) understanding the relationship between microbial taxa and function, i.e. if specific taxa are necessary, or if this does not matter as long as they can provide the same function, and (iii) utilizing techniques commonly used in medical/virology contexts such as host-genome- association (Awany et al., 2018). Chapter 5 drew on (iii) above and demonstrated how analysis of individual seaweed traits and genetic loci can reveal associations with microbial taxa of interest. In this study I focused on the most abundant microbial taxa, however, ultimately the same dataset can also be used to investigate more specific hypotheses about taxa that are emerging as likely drivers of disease, such as Nautella or Vibrio (Case et al., 2011; Fernandes et al., 2011; Wang et al., 2008), or those that are likely to be fundamental to the host or enhance resistance/resilience to stressors. While association studies are useful tools to determine taxa or functions of interest, experimental studies are necessary to determine causation. 130

Using genomics to restore and future-proof underwater seaweed forests

For example, inoculation experiments using multiple taxa strongly associated with seaweed diseases demonstrated that while some indeed caused disease, many others did not (Kumar et al., 2016; Qiu, 2017; Qiu et al., in prep.). Isolation of microbial taxa that are correlated with host traits of interest and further inoculation experiments (or selective removal of particular taxa, e.g. using combinations of antibiotics) to determine causation is needed to progress this avenue of research (Kumar et al., 2016). Such work would yield further understanding of potential microbial drivers of selection in the holobiont or could be used to identify microbes that may facilitate adaptation to change, e.g. via protective biofilms which may defend against heat and/or cell-degradation- associated pathogens. Addressing (i) and (ii) is likely to be critical to developing successful management solutions. For example, (i) exploring all components of the associated microbiota is probably important given, for example, that an endophytic fungus is already known to cause disease (“stipe rot”) and likely contributed to historical population declines of Phyllospora (Peters 2015; Campbell et al., in review). Importantly, components of the microbiome may interact and these potential associations could be harnessed, for example by identifying bacteria that inhibit pathogenic fungal growth. (ii) Understanding if specific taxa are necessary would simplify strategies to harness microbial solutions, especially if these involve manipulating microbiota across variable environments. To address this, further studies should look at taxa-function relationships e.g. using metagenomics, to determine whether this is important from a solution/restoration point of view (e.g Burke et al., 2011b).

• Determining the usefulness of interventionist strategies To avoid wasting resources and potential risks of interventionist strategies, it is essential to consider the likelihood of desirable conservation units (be they adapted phenotypes, genetically diverse populations, etc.) being spread naturally to determine if interventionist strategies such as genetic rescue are truly necessary. For example, rapid selection may occur naturally in response to large and acute environmental stressors and this can be assessed by monitoring population genetic 131

Using genomics to restore and future-proof underwater seaweed forests

patterns on a regular basis (e.g. Gurgel et al., in press, Coleman et al., in review). Modelling predicted patterns of dispersal from genetically desirable populations to vulnerable populations under ecologically relevant timescales can also help to assess the natural spread of genotypes to target areas (e.g. Quigley et al., 2019).

• Refining restoration and future-proofing techniques Future work is still needed to determine best practice methods for restoration. For example, when considering genomic patterns, quantification of how many donors/ replants, etc. are necessary to achieve sustainable target levels of genetic diversity or to maintain target adaptive loci in a population. When considering the development of strategies that harness host-microbial interactions, working with seaweeds is likely to need much refinement owing to challenges associated with working in the marine environment. Additionally, the short life cycle of many seaweed species (Schiel and Foster, 2006) means that solutions that have an effect on wild populations would either need to be applied quite frequently in the field or need to be rapidly self-propagating and such techniques will need to be explored in future research. While using genomics to harness genetic resources and microbial interactions may aid in providing an amount of mitigation for some issues, such as increasing temperatures, disease and possibly even grazing, they will also need to be combined with other strategies to limit individual and synergistic stressors, such as carbon emission reduction and herbivore exclusion.

• Expanding restoration and future-proofing approaches to other foundation species In Australia and elsewhere, many other species of habitat-forming seaweeds are likely to continue to decline, particularly due to ocean warming impacts predicted over the next century (Martinez et al., 2018). Given the fact that Phyllospora populations are genetically structured, likely harbor adaptive variation (Chapter 4) and are known to have higher dispersal and connectivity than other canopy forming seaweeds in the region (Coleman et al., 2011a; Durrant et al., 2015) other

132

Using genomics to restore and future-proof underwater seaweed forests

species are also likely to exhibit genetic differentiation that may influence their future survival and deserve similar research and management attention. However, techniques to overcome specific stressors, incorporate life histories, species- specific microbiomes, etc., will have to be considered on a case-by-case basis. For example, although we successfully achieved the desired mix of genetic diversity and structure in the restoration of Phyllospora populations (Chapter 3), methods to “design” populations of other species with a complex haplodiplontic life histories (e.g. laminarians, or “true” kelps) may involve more complex techniques than transplanting roughly similar proportions of individuals with desired genetic characteristics. For species other than Phyllospora, studies similar to those described here should also be embedded into ongoing restoration projects in the future.

6.4 Conclusion

As seaweed restoration and future-proofing programs grow worldwide, the topic of how to appropriately utilise our understanding of genetic and microbial resources is increasingly receiving more attention. In this thesis I have demonstrated that the combination of ecological knowledge and genomics can provide clear directives for such programs. For any management strategies to be useful however, there is an urgent need to mitigate the human effects that caused the loss in the first instance. First and foremost, this means preventing/reducing carbon emissions and other anthropogenic stressors (e.g. water pollution) that may hamper success of any of the solutions discussed here. Given the current state and trajectory of most ecosystems on Earth this needs to be coupled with management interventions to restore, enhance and/or future-proof ecosystems. This thesis stands as a small demonstration of how the integration of multiple disciplines such as ecology, genetics, genomics and microbiology can be harnessed to develop innovative solutions. By enabling us to better understand how to guide and assess the efficacy of restorative solutions, we are taking a small but solid step towards the regeneration of our planet.

133

Using genomics to restore and future-proof underwater seaweed forests

Literature cited

Abelson, A., Halpern, B. S., Reed, D. C., Orth, R. J., Kendrick, G. A., Beck, M. W., Belmaker, J., Krause, G., Edgar, G. J., Airoldi, L., Brokovich, E., France, R., Shashar, N., de Blaeij, A., Stambler, N., Salameh, P., Shechter, M., and Nelson, P. A. (2016). Upgrading marine ecosystem restoration using ecological‐social concepts. Bioscience, 66, 156–163.

Airoldi, L., and Beck, M. W. (2007). Loss, status and trends for coastal habitats of Europe. Oceanography and Marine Biology, 45, 345–405.

Aitken, S.N., and Whitlock, M.C. (2013). Assisted Gene Flow to Facilitate Local Adaptation to Climate Change. Annual Review of Ecology, Evolution, and Systematics 44(1), 367-388.

Anderson, M. J. (2001). A new method for non-parametric multivariate analysis of variance. Austral Ecology, 26, 32-46.

Anderson, M. J. (2006). Distance‐based tests for homogeneity of multivariate dispersions. Biometrics, 62, 245-253.

Anthony, K., Bay, L. K., Costanza, R., Firn, J., Gunn, J., Harrison, P., Heyward, A., Lundgren, P., Mead, D., Moore, T., Mumby, P. J., van Oppen, M. J. H., Robertson, J., Runge, M. C., Suggett, D. J., Schaffelke, B., Wachenfeld, D., and Walshe, T. (2017). New interventions are needed to save coral reefs. Nature Ecology and Evolution 1, 1420–1422.

Armstrong, D. P., Hayward, M. W., Moro, D., and Seddon, P. J. (2015). ‘Advances in Reintroduction Biology of Australian and New Zealand Fauna.’ (CSIRO Publishing: Melbourne, Vic., Australia.)

Arrigo, K. R. (2005). Marine microorganisms and global nutrient cycles. Nature, 437, 349–355.

Arumugam, M., Raes, J., Pelletier, E., Le Paslier, D., Yamada, T., Mende, D. R., and Meta, H. I. T. C. (2011). Enterotypes of the human gut microbiome. Nature, 473(7346), 174-180.

Assis, J., Berecibar, E., Claro, B., Alberto, F., Reed, F., Raimondi, P., and Serrão, E. A. (2017). Major shifts at the range edge of marine forests: the combined effects of climate changes and limited dispersal. Scientific Reports, 7, 44348. 134

Using genomics to restore and future-proof underwater seaweed forests

Assis, J., Coelho, N.C., Lamy, T., Valero, M., Alberto, F. and Serrão, E.Á. (2016), Deep reefs are climatic refugia for genetic diversity of marine forests. Journal of Biogeography, 43, 833-844.

Assis, J., N. Castilho Coelho, F. Alberto, M. Valero, P. Raimondi, D. Reed, and Serrão E. A. (2013). High and distinct range-edge genetic diversity despite local bottlenecks. PLoS ONE, 8: e68646.

Awany, D., Allali, I., Dalvie, S., Hemmings, S., Mwaikono, K.S., Thomford, N.E., Gomez, A., Mulder, N., and Chimusa, E.R. (2019). Host and Microbiome Genome- Wide Association Studies: Current State and Challenges. Frontiers in Genetics, 9, 637.

Azam, F., and Malfatti, F. (2007). Microbial structuring of marine ecosystems. Nature Reviews. Microbiology, 5, 782–791.

Babcock, R. C., Shears, N. T., Alcala, A. C., Barrett, N. S., Edgar, G. J., Lafferty, K. D., McClanahan, T. R., and Russ, G. R. (2010). Decadal trends in marine reserves reveal differential rates of change in direct and indirect effects. Proceedings of the National Academy of Sciences of the United States of America, 107, 18256–18261.

Baine, M. S. P. (2001). Artificial reefs: a review of their design, application, management and performance. Ocean and Coastal Management, 44, 241–259.

Bashan, Y., de-Bashan, L. E., Prabhu, S. R., and Hernandez, J. (2014). Advances in plant growth-promoting bacterial inoculant technology: formulations and practical perspectives (1998–2013). Plant and Soil, 378, 1–33.

Bauer, M.A., Kainz, K., Carmona-Gutierrez, D. and Madeo, F. (2018). Microbial wars: Competition in ecological niches and within the microbiome. Microbial Cell, 5(5): 215-219.

Bayraktarov, E., Saunders, M. I., Abdullah, S., Mills, M., Beher, J., Possingham, H. P., Mumby, P. J., and Lovelock, C. E. (2016). The cost and feasibility of marine coastal restoration. Ecological Applications, 26, 1055–1074.

Beacham, T. D., Withler, R. E., Murray, C. B. and Barner, L. W. (1988). Variation in Body Size, Morphology, Egg Size, and Biochemical Genetics of Pink Salmon in British Columbia. Transactions of the American Fisheries Society, 117, 109-126.

135

Using genomics to restore and future-proof underwater seaweed forests

Beardmore, J. A., and Porter, J. S. (2003). Genetically modified organisms and aquaculture. FAO Fisheries Circular number 989, FIRI/C989(En), Food and Agriculture Organization of the United Nations, Rome, Italy.

Bellgrove, A., Mckenzie, P. F., Cameron, H., and Pocklington, J. B. (2017). Restoring rocky intertidal communities: lessons from a benthic macroalgal ecosystem engineer. Marine Pollution Bulletin, 117, 17–27.

Benestan, L., Quinn, B. K., Maaroufi, H., Laporte, M., Clark, F. K., Greenwood, S. J., and Bernatchez, L. (2016). Seascape genomics provides evidence for thermal adaptation and current‐mediated population structure in American lobster (Homarus americanus). Molecular Ecology, 24, 5073–5092.

Bengtsson, M.M., Sjotun, K., and Ovreas, L. (2010). Seasonal dynamics of bacterial biofilms on the kelp Laminaria hyperborea. Aquatic Microbial Ecology, 60: 71–83.

Benjamini, Y. and Hochberg, Y. (1995) Controlling the false discovery rate—a practical and powerful approach to multiple testing. J. R. Stat. Soc., B Met., 57 (1), 289-300

Bennett, A.F., and Lenski, R.E. (2007). An experimental test of evolutionary trade- offs during temperature adaptation. Proceedings of the National Academy of Sciences USA, 104 (suppl 1), 8649-8654.

Bennett, S., Wernberg, T., Arackal Joy, B., de Bettignies, T., and Campbell, A. H. (2015). Central and rear-edge populations can be equally vulnerable to warming. Nature Communications 6, 10280.

Bennett, S., Wernberg, T., Connell, S. D., Hobday, A. J., Johnson, C. R., and Poloczanska, E. S. (2016). The ‘Great Southern Reef’: social, ecological and economic value of Australia’s neglected kelp forests. Marine and Freshwater Research, 67, 47–56.

Benson, A. K., Kelly, S. A., Legge, R., Ma, F., Low, S. J., Kim, J., Zhang, M., Oh, P.L., Nehrenberg, D., Kunjie, H., Kachman, S. D., Moriyama, E. N., Walter, J., Peterson, D.A., and Pomp, D. (2010). Individuality in gut microbiota composition is a complex polygenic trait shaped by multiple environmental and host genetic factors. Proceedings of the National Academy of Sciences U.S.A., 107, 18933– 18938.

136

Using genomics to restore and future-proof underwater seaweed forests

Bischoff, A., Steinger, T. and Müller-Schärer, H. (2010). The Importance of Plant Provenance and Genotypic Diversity of Seed Material Used for Ecological Restoration. Restoration Ecology, 18, 338-348.

Bishop, M. J., Coleman, M. A. and Kelaher, B. P. (2010). Cross-habitat impacts of species decline: response of estuarine sediment communities to changing detrital resources. Oecologia,, 163, 517-25.

Blekhman, R., Goodrich, J. K., Huang, K., Sun, Q., Bukowski, R., Bell, J. T., Spector, T. D., Keinan, A. L., Ruth, E., and Gevers, D.(2015). Host genetic variation impacts microbiome composition across human body sites. Genome Biology, 16 1–12.

Bolger, A. M., Lohse, M., and Usadel, B. (2014). Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics, 30(15), 2114-2120.

Bouma, T.J., De Vries, M.B., Low, E., Peralta, G., Tánczos, I.V., van de Koppel, J. and Herman, P.M.J., 2005. Trade‐offs related to ecosystem engineering: A case study on stiffness of emerging macrophytes. Ecology, 86(8), 2187-2199.

Bourne, D.C., M. Garren, T.M. Work, E. Rosenberg, G.W. Smith, C.D. (2009). Microbial disease and the coral holobiont. Trends in Microbiology, 17(12), 554- 562.

Bourque, A. S., Kenworthy, W. J., and Fourqurean, J. W. (2015). Impacts of physical disturbance on ecosystem structure in subtropical seagrass meadows. Marine Ecology Progress Series, 540, 27–41.

Brancalion, P. H. S., and Van Melis, J. (2017). On the need for innovation in ecological restoration. Annals of the Missouri Botanical Garden, 102, 227–236.

Brancalion, P. H. S., Viani, R. A. G., Strassburg, B. B. N., and Rodrigues, R. R. (2012). Finding the money for tropical forest restoration. Unasylva, 63, 41–49.

Breed, M. F., Harrison, P. A., Bischoff, A., Durruty, P., Gellie, N. J., Gonzales, E. K., Havens, K., Karmann, M., Kilkenny, F. F., Krauss, S. L. and Lowe, A. J., (2018). Priority actions to improve provenance decision-making. BioScience, 68(7), 510- 516.

Breed, M.F., Harrison, P. A., Blyth, C., Byrne, M., Gaget, V., Gellie, N.J., Groom, S.V., Hodgson, R., Mills, J.G., Prowse, T.A. and Steane, D.A., (2019). The potential

137

Using genomics to restore and future-proof underwater seaweed forests

of genomics for restoring ecosystems and biodiversity. Nature Reviews Genetics, 20(10), 615-628.

Broadhurst, L. M., Lowe, A., Coates, D. J., Cunningham, S. A., McDonald, M., Vesk, P. A. and Yates, C. (2008). Seed supply for broadscale restoration: maximizing evolutionary potential. Evolutionary Applications, 1, 587-597.

Brown, B., Fadillah, R., Nurdin, Y., Soulsby, I., and Ahmad, R. (2014). Case study: community based ecological mangrove rehabilitation (CBEMR) in Indonesia from small (12–33 ha) to medium scales (400 ha) with pathways for adoption at larger scales (>5000 ha). Sapiens 7, 1–12.

Brown, M.V., Lauro, F.M., DeMaere, M.Z., Muir, L., Wilkins, D., Thomas, T., Riddle, M.J., Fuhrman, J.A., Andrews-Pfannkoch, C., Hoffman, J.M., McQuiad, J.B., Allan, A., Rintoul, S.R. and Cavicchiolo, R. (2012). Global biogeography of SAR11 marine bacteria. Mol Syst Biol 8: 595.

Brudvig, L. A. (2011). The restoration of biodiversity: where has research been and where does it need to go? American Journal of Botany 98, 549–558.

Bucharova, A., Bossdorf, O., Hölzel, N., Kollmann, J., Prasse, R. and Durka, W. (2019). Mix and match: regional admixture provenancing strikes a balance among different seed-sourcing strategies for ecological restoration. Conservation Genetics, 20, 7-17.

Bulgarelli, D., Schlaeppi, K., Spaepen, S., van Themaat, E. V. L., and Schulze- Lefert, P. (2013). Structure and functions of the bacterial microbiota of plants. Annual Review of Plant Biology. 64, 807–838.

Bulleri, F., and Chapman, M. G. (2010). The introduction of coastal infrastructure as a driver of change in marine environments. Journal of Applied Ecology, 47, 26– 35.

Bullock, J. M., Aronson, J., Newton, A. C., Pywell, R. F., and Rey-Benayas, J. M. (2011). Restoration of ecosystem services and biodiversity: conflicts and opportunities. Trends in Ecology and Evolution, 26, 541–549.

Burke, C., Thomas, T., Lewis, M., Steinberg, P. and Kjelleberg, S. (2011a) Composition, uniqueness and variability of the epiphytic bacterial community of the green alga Ulva australis. ISME, J 5: 590–600.

138

Using genomics to restore and future-proof underwater seaweed forests

Burke, C., Steinberg, P., Rusch, D., Kjelleberg, S., and Thomas, T. (2011b) Bacterial community assembly based on functional genes rather than species. Proceedings of the National Academy of Sciences USA, 108: 14288–14293. Burkholder, J., Libra, B., Weyer, P., Heathcote, S., Kolpin, D., Thorne, P. S., and Wichman, M. (2007). Impacts of waste from concentrated animal feeding operations on water quality. Environmental Health Perspectives 115, 308–312.

Burridge, T. R. (1990). Reproduction, development and growth in Phyllospora comosa, Seiroccus axillaris and Scytothalia dorycarpa (Seirococcaceae, Phaeophyta). PhD Thesis, Monash University, Melbourne.

Burrows, M., Schoeman, D., Buckley, L., Moore, P., Poloczanska, E., Brander, K., Bruno, J. F., Duarte, C. M., Halpern, B. S., Holding, J., Kappel C. V., Kiessling, W., O'Connor, M. I., Pandolfi, J. M., Parmesan, C., Schwing, F. B., Sydeman, W. J. and Richardson, A. (2011). The Pace of Shifting Climate in Marine and Terrestrial Ecosystems. Science, 334(6056), new series, 652-655.

Calumpong, H. P., and Fonseca, M. S. (2001). Seagrass transplantation and other seagrass restoration methods. In ‘Global Seagrass Research Methods’. (Eds F. T. Short and R. G. Coles.) pp. 425–443. (Elsevier: Amsterdam, Netherlands.)

Camill, P., McKone, M. J., Sturges, S. T., Severud, W. J., Ellis, E., Limmer, J., Martin, C. B., Navratil, R. T., Purdie, A. J., Sandel, B. S. and Talukder, S., (2004). Community‐and ecosystem‐level changes in a species‐rich tallgrass prairie restoration. Ecological applications, 14(6), pp.1680-1694.

Campbell, A. H., Marzinelli, E. M., Peters, T., Ferrari, J., Hill, R., Lachnit, T., Kjelleberg, S., Goncalves, P., Coleman, M. A., Egan, S., Pereira, R. C., Thomas, T. and Steinberg, P. D. (in review) Widespread fungal disease of a dominant seaweed and declining underwater forests.

Campbell, A. H., Marzinelli, E. M., Gelber, J., and Steinberg, P. D. (2015). Spatial variability of microbial assemblages associated with a dominant habitat-forming seaweed. Frontiers in Microbiology 6, 230.

Campbell, A. H., Marzinelli, E. M., Vergés, A., Coleman, M. A., and Steinberg, P. D. (2014a). Towards restoration of missing underwater forests. PLoS One 9, e84106.

Campbell, A. H., Vergés, A., and Steinberg, P. D. (2014b). Demographic consequences of disease in a habitat‐forming seaweed and impacts on interactions between natural enemies. Ecology 95, 142–152. 139

Using genomics to restore and future-proof underwater seaweed forests

Campbell, A.H., Harder, T., Nielsen, S., Kjelleberg, S. and Steinberg, P.D. (2011), Climate change and disease: bleaching of a chemically defended seaweed. Global Change Biology, 17: 2958-2970.

Cang, F. A., Wilson, A. A., and Wiens, J. J. (2016). Climate change is projected to outpace rates of niche change in grasses. Biology letters, 12(9), 20160368.

Carney, L. T., Waaland, J. R., Klinger, T., and Ewing, K. (2005). Restoration of the bull kelp Nereocystis luetkeana in nearshore rocky habitats. Marine Ecology Progress Series 302, 49–61.

Case, R. J., Longford, S. R., Campbell, A. H., Low, A., Tujula, N., Steinberg, P. D. and Kjelleberg, S. (2011). Temperature induced bacterial virulence and bleaching disease in a chemically defended marine macroalga. Environmental Microbiology 13: 529–537.

Casey, J. M., Connolly, S. R., and Ainsworth, T. D. (2015). Coral transplantation triggers shift in microbiome and promotion of coral disease associated potential pathogens. Scientific Reports 5, 11903.

Castro, L., Vergés, A., Campbell, A., Wernberg, T., Steinberg, P., Thomas, T., Straub, S., Cetina-Heredia, P., Roughan, M. and Marzinelli, E. M. (in prep.). Ocean warming and heatwaves alter the kelp microbiome and increase grazing by a tropical herbivore.

Caye, K., Jumentier, B., Lepeule, J., François, O. (2019). LFMM 2: Fast and Accurate Inference of Gene-Environment Associations in Genome-Wide Studies, Molecular Biology and Evolution, Volume 36, Issue 4, 852–860.

Cetina-Heredia, P., Roughan, M., Sebille, E., Feng, M., and Coleman, M. (2015). Strengthened currents override the effect of warming on lobster larval dispersal and survival. Global Change Biology 21, 4377–4386.

Cetina‐Heredia, P., Roughan, M., Van Sebille, E., and Coleman, M. A. (2014). Long‐term trends in the East Australian Current separation latitude and eddy driven transport. Journal of Geophysical Research. Oceans 119, 4351–4366.

Chapman, M.G. and Underwood, A.J. (2010). The need for a practical scientific protocol to measure successful restoration. Wetlands Australia, 19(1), pp.28–49.

140

Using genomics to restore and future-proof underwater seaweed forests

Chaves, R. B., Durigan, G., Brancalion, P. H. S., and Aronson, J. (2015). On the need of legal frameworks for assessing restoration projects success: new perspectives from São Paulo state (Brazil). Restoration Ecology 23, 754–759.

Chefaoui, R. M., Duarte, C. M., and Serrao, E. A. (2018). Dramatic loss of seagrass habitat under projected climate change in the Mediterranean Sea. Global Change Biology 24, 4919–4928.

Chen, K.-Y., Marschall, E. A., Sovic, M. G., Fries, A. C., Gibbs, H. L. and Ludsin, S. A. (2018). assignPOP: An r package for population assignment using genetic, non- genetic, or integrated data in a machine-learning framework. Methods in Ecology and Evolution, 9, 439-446.

Choi, Y. D. (2004). Theories for ecological restoration in changing environment: toward ‘futuristic’ restoration. Ecological Research 19, 75–81.

Choi, Y. D. (2007). Restoration ecology to the future: a call for new paradigm. Restoration Ecology 15, 351–353.

Cinner, J. E., Huchery, C., Macneil, M. A., Graham, N. A., Mcclanahan, T. R., Maina, J., Maire, E., Kittinger, J. N., Hicks, C. C., Mora, C., Allison, E. H., D’agata, S., Hoey, A., Feary, D. A., Crowder, L., Williams, I. D., Kulbicki, M., Vigliola, L., Wantiez, L., Edgar, G., Stuart-Smith, R. D., Sandin, S. A., Green, A. L., Hardt, M. J., Beger, M., Friedlander, A., Campbell, S. J., Holmes, K. E., Wilson, S. K., Brokovich, E., Brooks, A. J., Cruz-Motta, J. J., Booth, D. J., Chabanet, P., Gough, C., Tupper, M., Ferse, S. C., Sumaila, U. R., and Mouillot, D. (2016). Bright spots among the world’s coral reefs. Nature 535, 416–419.

Clark, J. S., Poore, A. G., Ralph, P. J., and Doblin, M. A. (2013). Potential for adaptation in response to thermal stress in an intertidal macroalga. Journal of Phycology 49, 630–639.

Cole, L. J. (2017). Cut adrift: the distribution and abundance of rafting algae and its associated fauna now and in a future ocean, Doctor of Philosophy thesis, School of Biological Sciences, University of Wollongong, https://ro.uow.edu.au/theses1/121

Coleman, M. A. and Wernberg, T. (2017). Forgotten underwater forests: The key role of fucoids on Australian temperate reefs. Ecology and evolution, 7, 8406- 8418.

141

Using genomics to restore and future-proof underwater seaweed forests

Coleman, M. A., and Kelaher, B. P. (2009). Connectivity among fragmented populations of a habitat-forming alga, Phyllospora comosa (Phaeophyceae, Fucales) on an urbanised coast. Marine Ecology Progress Series 381, 63–70.

Coleman, M. A., Bates, A. E., Stuart-Smith, R. D., Malcolm, H. A., Harasti, D., Jordan, A., Knott, A., Edgar, G. J., and Kelaher, B. P. (2015). Functional traits reveal early responses in marine reserves following protection from fishing. Diversity and Distributions 21, 876–887.

Coleman, M. A., Cetina-Heredia, P., Roughan, M., Feng, M., Van Sebille, E., and Kelaher, B. P. (2017). Anticipating changes to future connectivity within a network of marine protected areas. Global Change Biology 23, 3533–3542.

Coleman, M. A., Chambers, J., Knott, N. A., Malcolm, H. A., Harasti, D., Jordan, A., and Kelaher, B. P. (2011a). Connectivity within and among a network of temperate marine reserves. PLoS One 6, e20168.

Coleman, M. A., Kelaher, B. P., Steinberg, P. D., and Millar, A. J. K. (2008). Absence of a large brown macroalga on urbanized rocky reefs around Sydney, Australia, and evidence for historical decline. Journal of Phycology 44, 897–901.

Coleman, M. A., Palmer-Brodie, A., and Kelaher, B. P. (2013). Conservation benefits of a network of marine reserves and partially protected areas. Biological Conservation 167, 257–264.

Coleman, M. A., Roughan, M., Macdonald, H. S., Connell, S. D., Gillanders, B. M., Kelaher, B. P., and Steinberg, P. D. (2011b). Variation in the strength of continental boundary currents determines continent-wide connectivity in kelp. Journal of Ecology 99, 1026–1032.

Coleman, M., Minne, A.J.P., Vranken, S. and Wernberg, T. (In review b). Genetic tropicalisation of kelp forests.

Coleman, M., Wood, G., Filbee-Dexter, K., Minne, A. J. P., Goold, h.d., Vergés, A., Marzinelli, E. M., Steinberg, P. D. and Wernberg, T. (In review a). Restore or redefine: future trajectories for restoration.

Coleman, M. A. and Brawley S. H. (2005) Spatial and temporal variability in dispersal and population genetic structure of a rockpool alga. Mar. Ecol. Prog. Ser. 300: 63-77

142

Using genomics to restore and future-proof underwater seaweed forests

Coleman, M. A., and Goold, H. (2019). Harnessing synthetic biology for kelp forest conservation. Journal of Phycology 0(ja). doi: 10.1111/jpy.12888.

Coleman, M. A., Weigner, K. E., Kelaher, B. P. (2017). Optimising DNA extraction from a critically endangered marine alga. Cons. Gen. Res. 10(3), 309-311. DOI 10.1007/s12686-017-0810-5

Côté, I. M., Darling, E. S., and Brown, C. J. (2016). Interactions among ecosystem stressors and their importance in conservation. Proceedings of the Royal Society of London – B. Biological Sciences 283, 20152592.

Coyte, K. Z., Schluter, J., Foster, K. R. (2015). The ecology of the microbiome: Networks, competition, and stability. Science, 350(6261):663–666.

Crain, C. M., Kroeker, K., and Halpern, B. S. (2008). Interactive and cumulative effects of multiple human stressors in marine systems. Ecology Letters 11, 1304– 1315.

Crouzeilles, R., Curran, M., Ferreira, M. S., Lindenmayer, D. B., Grelle, C. E. V., and Rey Benayas, J. M. (2016). A global meta-analysis on the ecological drivers of forest restoration success. Nature Communications 7, 11666.

Cumming, E. E., Matthews, T. G., Sanderson, C. J., Ingram, B. A. and Bellgrove, A. (2019). Optimal spawning conditions of Phyllospora comosa (Phaeophyceae, Fucales) for mariculture. Journal of Applied Phycology, 1-10.

Cumming, Erin E., Matthews, Ty G., Sanderson, John C., Ingram, Brett A. and Bellgrove, Alecia 2020, Growth and survivorship of Phyllospora comosa (Phaeophyceae, Fucales) on different mariculture seeding twines in a hatchery setting, Aquaculture, vol. 523, pp. 1-8, doi: 10.1016/j.aquaculture.2020.735216.

Cunningham, S. (2002). ‘The Restoration Economy: The Greatest New Growth Frontier.’ (Berrett-Koehler Publishers: San Francisco, CA, USA.)

Warren, D. R., Kraft, C. E., Josephson, D. C., Driscoll, C. T. (2017) Acid rain recovery may help to mitigate the impacts of climate change on thermally sensitive fish in lakes across eastern North America. Glob. Change Biol., 23, pp. 2149-2153

Dafforn, K. A., Glasby, T. M., Airoldi, L., Rivero, N. K., Mayer-Pinto, M., and Johnston, E. L. (2015). Marine urbanization: an ecological framework for

143

Using genomics to restore and future-proof underwater seaweed forests

designing multifunctional artificial structures. Frontiers in Ecology and the Environment 13, 82–90.

Damjanovic, K., Blackall, L. L., Webster, N. S., and van Oppen, M. J. H. (2017). The contribution of microbial biotechnology to mitigating coral reef degradation. Microbial Biotechnology 10, 1236–1243. doi: 10.1111/1751-7915.12769

Datta, D., Chattopadhyay, R., and Guha, P. (2012). Community based mangrove management: a review on status and sustainability. Journal of Environmental Management 107, 84–95.

Dave, R., Saint-Laurent, C., Murray, L., Antunes Daldegan, G., Brouwer, R., de Mattos Scaramuzza, C. A., Raes, L., Simonit, S., Catapan, M., García Contreras, G., Ndoli, A., Karangwa, C., Perera, N., Hingorani, S. and Pearson, T. (2019). Second Bonn Challenge progress report. Application of the Barometer in 2018. Gland, Switzerland: IUCN. xii + 80pp.

Davis, K. L., Coleman, M. A., Connell, S. D., Russell, B. D., Gillanders, B. M., and Kelaher, B. P. (2017). Ecological performance of construction materials subject to ocean warming and acidification. Marine Environmental Research 131, 177–182. de Vries, F. T., Manning, P., Tallowin, J. R. B., Mortimer, S. R., Pilgrim, E. S., Harrison, K. A., Hobbs, P. J., Quirk, H., Shipley, B., Cornelissen, J. H. C., Kattge, J., Bardgett, R. D. (2012). Abiotic drivers and plant traits explain landscape-scale patterns in soil microbial communities. Ecology Letters, 15: 1230–1239.

DeFries, R. S., Ellis, E. C., Chapin, I. I. I. F. S., Matson, P. A., Turner, I. B. L., Agrawal, A., Crutzen, P. J., Field, C., Gleick, P., Kareiva, P. M., Lambin, E., Liverman, D., Ostrom, E., Sanchez, P. A., and Syvitski, J. (2012). Planetary opportunities: a social contract for global change science to contribute to a sustainable future. Bioscience 62, 603–606. dela Cruz, D. W., Villanueva, R. D., and Baria, M. V. B. (2014). Community-based, low-tech method of restoring a lost thicket of Acropora corals. ICES Journal of Marine Science 71, 1866–1875.

Delgado, C. L., Wada, N., Rosegrant, M. W., Meijer, S., and Ahmed, M. (2003). Fish to 2020: supply and demand in changing global markets. Worldfish Center Technical Report 62, Penang. International Food Policy Research Institute, Washington, DC, USA.

144

Using genomics to restore and future-proof underwater seaweed forests

Deutsch, C., Ferrel, A., Seibel, B., Pörtner, H., and Huey, R. (2015). Climate change tightens a metabolic constraint on marine habitats. Science 348, 1132-1135.

Diekmann, O. E., and E. A. Serrao. (2012). Range-edge genetic diversity: locally poor extant southern patches maintain a regionally diverse hotspot in the seagrass Zostera marina. Molecular Ecology 21: 1647– 1657.

Dobson, A. P., Bradshaw, A. D., and Baker, A. J. M. (1997). Hopes for the future: restoration ecology and conservation biology. Science 277, 515–522.

Dornelas, M., Gotelli, N. J., Mcgill, B., Shimadzu, H., Moyes, F., Sievers, C., and Magurran, A. E. (2014). Assemblage time series reveal biodiversity change but not systematic loss. Science 344, 296–299. dos Santos, H. F., Duarte, G. A., Rachid, C. T., Chaloub, R. M., Calderon, E. N., Marangoni, L. F., et al., (2015). Impact of oil spills on coral reefs can be reduced by bioremediation using probiotic microbiota. Sci. Rep. 5:18268.

Dray, S. E. and Dufour, A.-B. E. (2007). The ade4 Package: Implementing the Duality Diagram for Ecologists. Journal of Statistical Software, 22, 1-20.

Duarte, B., Martins, I., Rosa, R., Matos, A. R., Roleda, M. Y., Reausch, T. B. H., Engelen, A. H., Serrao, E. A., Pearson, G. A., Marques, J. C., Cacador, I., Duarte, C. M., and Jueterbock, A. (2018). Climate change impacts on seagrass meadows and macroalgal forests: an integrative perspective on acclimation and adaptive potential. Frontiers in Marine Science 5, 190–202.

Duarte, C. M., and Chiscano, C. L. (1999). Seagrass biomass and production: a reassessment. Aquatic Botany 65, 159–174.

Duarte, C. M., Losada, I. J., Hendriks, I. E., Mazarrasa, I., and Marbà, N. (2013). The role of coastal plant communities for climate change mitigation and adaptation. Nature Climate Change 3, 961–968.

Duarte, C., Dennison, W., Orth, R., and Carruthers, T. (2008). The charisma of coastal ecosystems: addressing the imbalance. Estuaries and Coasts 31, 233–238.

Dubilier, N., Bergin, C., and Loot, C. (2008). Symbiotic diversity in marine animals: the art of harnessing chemosynthesis. Nature Reviews. Microbiology 6, 725–740.

145

Using genomics to restore and future-proof underwater seaweed forests

Duffy, J. E., and Hay, M. E. (1990). Seaweed adaptation to herbivory. Bioscience 40, 368–375.

Duggins, D., Eckman, J. E., Siddon, C. E., and Klinger, T. (2001). Interactive roles of mesograzers and current flow in survival of kelps. Marine Ecology Progress Series 223, 143–155.

Durrant, H. M. S., Barrett, N. S., Edgar, G. J., Coleman, M. A., and Burridge, C. P. (2015). Shallow phylogeographic histories of key species in a biodiversity hotspot. Phycologia 54(6), 556-565.

Edgar, R. C. (2010). Search and clustering orders of magnitude faster than BLAST. Bioinformatics, 26(19), 2460-2461. doi:10.1093/bioinformatics/btq461

Edgar, R. C. (2016). UNOISE2: improved error-correction for Illumina 16S and ITS amplicon sequencing. bioRxviv. doi:10.1101/081257

Egan, S., Harder, T., Burke, C., Steinburg, P. D., Kjelleberg, S., and Thomas, T. (2013). The seaweed holobiont: understanding seaweed–bacteria interactions. FEMS Microbiology Reviews 37, 462–476.

Egan, S., Thomas, T., and Kjelleberg, S. (2008) Unlocking the diversity and biotechnological potential of marine surface associated microbial communities. Current Opinions in Microbiology 11: 219–225.

Eger, A. M., Marzinelli, E., Steinberg, P. D., and Vergés, A. (2019). Worldwide Synthesis of Kelp Forest Reforestation. [Dataset]

Engel, S., Jensen, P. R., and Fenical, W. (2002). Chemical ecology of marine microbial defense. Chemistry and Ecology 28, 1971–1985.

Evanno, G., Regnaut, S. and Goudet, J. (2005) Detecting the number of clusters of individuals using the software structure: a simulation study. Molecular Ecology, 14, 2611-2620.

Evans, S. M., E. A. Sinclair, A. G. Poore, K. F. Bain, and A. Vergés. (2018). Assessing the effect of genetic diversity on the early establishment of the threatened seagrass Posidonia australis using a reciprocal‐transplant experiment. Restoration Ecology 26:570-580.

146

Using genomics to restore and future-proof underwater seaweed forests

Evans, S. M., Sinclair, E. A., Poore, A. G., Bain, K. F., and Vergés, A. (2016). Genotypic richness predicts phenotypic variation in an endangered clonal plant. PeerJ, 4, e1633. doi:10.7717/peerj.1633.

Fahimipour, A. K., Kardish, M. R., Lang, J. M., Green, J. L., Eisen, J. A., Stachowicz, J. J., (2017). Global-scale structure of the eelgrass microbiome. Applied Environmental Microbiology, 83:e03391-16.

Falace, A., Kaleb, S., De la Fuente, G., Asnaghi, V., and Chiantore, M. (2018). Ex situ cultivation protocol for Cystoseira amentacea var. stricta (Fucales, Phaeophyceae) from a restoration perspective. PLoS One 13, e0193011.

Falkenberg, L. J., Connell, S. D., and Russell, B. D. (2013). Disrupting the effects of synergies between stressors: improved water quality dampens the effects of future CO2 on a marine habitat. Journal of Applied Ecology 50, 51–58.

FAO (2019). New UN Decade on Ecosystem Restoration offers unparalleled opportunity for job creation, food security and addressing climate change. News Article: http://www.fao.org/news/story/en/item/1182090/icode/.

Feng, X., Jiang, G. and Fan, Z. (2015). Identification of outliers in a genomic scan for selection along environmental gradients in the bamboo locust, Ceracris kiangsu. Scientific Reports, 5: 13758.

Fernandes, N., Case, R. J., Longford, S. R., Seyedsayamdost, M. R., Steinberg, P. D., Kjelleberg, S. and Thomas, T. (2011) Genomes and virulence factors of novel bacterial pathogens causing bleaching disease in the marine red alga Delisea pulchra. PLoS ONE 6: e27387.

Ferrario, F., Iveša, L., Jaklin, A., Perkol‐Finkel, S., Airoldi, L., and Siqueira, T. (2016). The overlooked role of biotic factors in controlling the ecological performance of artificial marine habitats. Journal of Applied Ecology 53, 16–24.

Filbee-Dexter, K., and Scheibling, R. (2012). Hurricane-mediated defoliation of kelp beds and pulsed delivery of kelp detritus to offshore sedimentary habitats. Marine Ecology Progress Series 455, 51–64.

Filbee-Dexter, K., and Smajdor, A. (2019). Ethics of Assisted Evolution in Marine Conservation. Frontiers in Marine Science 6(20).

147

Using genomics to restore and future-proof underwater seaweed forests

Firth, L.B., Knights, A.M., Bridger, D., Evans, A., Mieskowska, N., Moore, P.J., O'Connor, N.E., Sheehan, E.V., Thompson, R.C. and Hawkins, S.J., 2016. Ocean sprawl: challenges and opportunities for biodiversity management in a changing world. Oceanography and Marine Biology: an Annual Revue. 54, 189-262

Firth, L. B., Thompson, R. C., Bohn, K., Abbiati, M., Airoldi, L., Bouma, T. J., Bozzeda, F., Ceccherelli, V. U., Colangelo, M. A., Evans, A., Ferrario, F., Hanley, M. E., Hinz, H., Hoggart, S. P. G., Jackson, J. E., Moore, P., Morgan, E. H., Perkol- Finkel, S., Skov, M. W., Strain, E. M., van Belzen, J., and Hawkins, S. J. (2014). Between a rock and a hard place: environmental and engineering considerations when designing coastal defence structures. Coastal Engineering 87, 122–135.

Fischer, M. C., Heckel, G. and Excoffier, L. (2011). Enhanced AFLP genome scans detect local adaptation in high-altitude populations of a small rodent (Microtus arvalis). Molecular Ecology, 20, 1450-1462.

Flukes, E.B., Wright, J.T. and Johnson, C.R. (2015), Phenotypic plasticity and biogeographic variation in physiology of habitat‐forming seaweed: response to temperature and nitrate. Journal of Phycology, 51: 896-909.

Foll, M. and Gaggiotti, O. E. (2008). A genome scan method to identify selected loci appropriate for both dominant and codominant markers: A Bayesian perspective. Genetics, 180, 977-993.

Foll, M., Fisher, M. C., Heckel, G. and Excoffier, L. (2010). Estimating population structure from AFLP amplification intensity. Molecular Ecology, 19, 4638-4647.

Fonseca, M. S., Kenworthy, W. J., and Thayer, G. W. (1998). Guidelines for the conservation and restoration of seagrasses in the United States and adjacent waters. NOAA Coastal Ocean Program Decision Analysis Series number 12, NOAA Coastal Ocean Office, Silver Spring, MD, USA.

Fonseca, M. S., Kenworthy, W. J., Courtney, E. X., and Hall, M. O. (1994). Seagrass planting in the southeastern United States: methods for accelerating habitat development. Restoration Ecology 2, 198–212.

Ford, T., and Meux, B. (2010). Giant kelp community restoration in Santa Monica Bay. Urban Coast 2, 43–46.

Forester, B. R., Lasky, J. R., Wagner, H. H., Urban, D. L. (2018). Comparing methods for detecting multilocus adaptation with multivariate genotype– environment associations. Molecular Ecology 2018; 27: 2215– 2233.

148

Using genomics to restore and future-proof underwater seaweed forests

Forsman, A., and Wennersten, L. (2016). Inter-individual variation promotes ecological success of populations and species: evidence from experimental and comparative studies. Ecography 39, 630–648.

Foster, M.S. and Schiel, D.R., 2010. Loss of predators and the collapse of southern California kelp forests: alternatives, explanations and generalizations. Journal of Experimental Marine Biology and Ecology, 393(1-2), 59-70

Fowler-Walker, M. J., Wernberg, T. and Connell, S. D. (2006). Differences in kelp morphology between wave sheltered and exposed localities: morphologically plastic or fixed traits? Marine Biology 148, 755–767.

Frankham, R. (2015). Genetic rescue of small inbred populations: meta‐analysis reveals large and consistent benefits of gene flow. Molecular Ecology 24, 2610– 2618.

Frankham, R. (2016). Genetic rescue benefits persist to at least the F3 generation, based on a meta-analysis. Biological Conservation 195, 33–36.

Frankham, R., Ballou, J. D., Eldridge, M. D. B., Lacy, R. C., Ralls, K., Dudash, M. R., and Fenster, C. B. (2011). Predicting the probability of outbreeding depression. Conservation Biology, 25(3), 465-475.

Frankham, R., Ballou, J. D., Ralls, K., Eldridge, M., Dudash, M. R., Fenster, C. B., Lacy, R. C., and Sunnucks, P. (2017). ‘Genetic Management of Fragmented Animal and Plant Populations.’ (Oxford University Press: Oxford, UK.)

Fraser, C.I., McGaughran, A., Chuah, A. and Waters, J.M. (2016), The importance of replicating genomic analyses to verify phylogenetic signal for recently evolved lineages. Mol Ecol, 25: 3683-3695.

Frichot, E. and François, O. (2015), LEA: An R package for landscape and ecological association studies. Methods Ecol Evol, 6: 925-929.

Frichot, E., Mathieu, F., Trouillon, T., Bouchard, G., and François, O. (2014). Fast and efficient estimation of individual ancestry coefficients. GeneticS, 196 (4): 973- 983.

Frichot, E., Schoville, S. D., Bouchard, G., Francois, O. (2013). Testing for associations between loci and environmental gradients using latent factor mixed models. Molecular Biology and Evolution, 30 1687-1699.

149

Using genomics to restore and future-proof underwater seaweed forests

Gallagher, R. V., Hancock, N., Makinson, R. O., and Hogbin, T. (2014). Assisted colonisation as a climate change adaptation tool. Report to the Biodiversity Hub of the NSW Office of Environment and Heritage. Appendix to the NSW Draft Translocation Policy (2007). (NSW Biodiversity Research Hub: NSW, Australia.) Available at: https://www.mq.edu.au/about/about-the-university/faculties-and- departments/faculty-of-science-and-engineering/departments-and- centres/department-of-biological-sciences/our-research/biodiversity-node- archived/Assisted-Colonisation-as-a-Climate-Change-Adaptation-Tool.- Gallagher-et-al-2014.pdf [Verified 26 February 2019].

Gellie, N. J., Mills, J. G., Breed, M. F., and Lowe, A. J. (2017). Revegetation rewilds the soil bacterial microbiome of an old field. Molecular Ecology 26, 2895–2904. Gerland, P., Raftery, A. E., Ševčíková, H., Li, N., Gu, D., Spoorenberg, T., Alkema, L., Fosdick, B. K., Chunn, J., Lalic, N., Bay, G., Buettner, T., Heilig, G. K., and Wilmoth, J. (2014). World population stabilization unlikely this century. Science 346, 234–237.

Ghiglione, J. F., Galand, P. E., Pommier, T., Pedros‐Alio, C., Maas, E. W., Bakker, K., Bertilson, S., Kirchman, D. L., Lovejoy, C., Yager, P. L., and Murray, A. E. (2012). Pole‐to‐pole biogeography of surface and deep marine bacterial communities. Proc Natl Acad Sci USA 109: 17633–17638.

Gianni, F., Bartolini, F., Airoldi, L., Ballesteros, E., Francour, P., Guidetti, P., Meinesz, A., Thibaut, T., and Mangialajo, L. (2013). Conservation and restoration of marine forests in the Mediterranean Sea and the potential role of marine protected areas. Advances in Oceanography and Limnology 4, 83–101.

Gibbs, H. K. and Salmon, J. M. Mapping the world’s degraded lands. Applied Geography 57, 12–21 (2015).

Gilbert, J. A., Steele, J. A., Caporaso, J. G., Steinbrueck, L., Reeder, J., Temperton, B., Huse, S., McHardy, A. C., Knight, R., Joint, I., Somerfield, P., Fuhrman, J. A., and Field, D. (2012). Defining seasonal marine microbial community dynamics. Isme Journal 6: 298–308.

Gillies, C., Fitzsimons, J., Branigan, S., Hale, L., Hancock, B., Creighton, C., Alleway, H., Bishop, M., Brown, S., Chamberlain, D., Cleveland, B., Crawford, C., Crawford, M., Diggles, D. B. K., Ford, J., Hamer, P., Hart, A., Johnston, E., Mcdonald, T., and Winstanley, R. (2015). Scaling-up marine restoration efforts in Australia. Ecological Management and Restoration 16, 84–85.

150

Using genomics to restore and future-proof underwater seaweed forests

Gilman, S. E., Urban, M. C., Tewksbury, J., Gilchrist, G. W., and Holt, R. D. (2010). A framework for community interactions under climate change. Trends in Ecology and Evolution 25, 325–331.

Goecke, F., Labes, A., Wiese, J. and Imhoff, J. F. (2010) Chemical interactions between marine macroalgae and bacteria. Marine Ecology Progress Series 409: 267–299.

Goodsell, P. J., and Chapman, M. G. (2009). Rehabilitation of habitat and the value of artificial reefs. In ‘Marine Hard Bottom Communities: Patterns, Dynamics, Diversity, and Change’. (Ed. M. Wahl.) pp. 333–344. (Springer-Verlag.) Goudet, J. (2005). hierfstat, a package for r to compute and test hierarchical F- statistics. Molecular Ecology Notes, 5, 184-186.

Granado, R., Neta, L. C. P., Nunes-Freitas, A. F., Voloch, C. M. and Lira, C. F. (2018). Assessing genetic diversity after mangrove restoration in Brazil: why is it so important? Diversity, 10, 27.

Grant, J., Wilson, K., Grover, A., and Togstad, H. (1982). Early development of Pendleton artificial reef. Marine Fisheries Review 44, 53–60.

Griffiths, S. M., Antwis,, R. E., Lenzi, L., Lucaci, A., Behringer, D. C., Butler IV, M. J. and Preziosi, R.F. (2019). Host genetics and geography influence microbiome composition in the sponge Ircinia campana. Journal of Animal Ecology 2019; 88: 1684–1695.

Groffman, P. M., Stylinski, C., Nisbet, M. C., Duarte, C. M., Jordan, R., Burgin, A., Previtali, M. A., and Coloso, J. (2010). Restarting the conversation: challenges at the interface between ecology and society. Frontiers in Ecology and the Environment 8, 284–291.

Gruber, B., P. J. Unmack, O. F. Berry, and A. Georges. (2018). dartr: an R package to facilitate analysis of SNP data generated from reduced representation genome sequencing. Molecular Ecology Resources 18:691–699.

Gurgel, C. F. D., Camacho, O., Minne, A. J. P., Wernberg, T. and Coleman, M. (In press). Marine heatwave drives cryptic loss of genetic diversity in underwater forests. Current Biology.

Gusareva, E. S, Acerbi, E., Lau, K., Luhung, I., Premkrishnan, B. N. V., Kolundžija, S., Purbojati, R. W., Wong, A., Houghton, J. N. I., Miller, D., Gaultier, N. E.,

151

Using genomics to restore and future-proof underwater seaweed forests

Heinle, C. E., Clare, M. E., Vettath, V. K., Kee, C., Lim, S. B. Y., Chénard, C., Phung, W. J., Kushwaha, K. K., Nee, A. P., Putra, A., Panicker, D., Yanqing, K., Hwee, Y. Z., Lohar, S. R., Kuwata, M., Kim, H. L., Yang, L., Uchida, A., Drautz- Moses, D. I., Junqueira, A. C. M. and Schuster, S. C. (2019). Microbial communities in the tropical air ecosystem follow a precise diel cycle. Proceedings of the National Academy of Sciences, 116 (46) 23299-23308.

Hacker, J., Dobrindt, U., Steinert, M., Merkert, H., and Hentschel, U. ( 2005 ) Co- evolution of bacteria and their hosts: a marriage made in heaven or hell? The Influence of Cooperative Bacteria on Animal Host Biology ( McFall-Ngai MJ Henderson B Ruby E-G , eds), pp. 57 – 72 . Cambridge University Press , New York.

Hacquard, S., Garrido-Oter, R., Gonzalex, A., Spaepen, S., Ackermann, G., Lebeis, S., McHardy, A. C., Dangl, J. L., Knight, R., Ley, R. and Schulze-Lefert, P. (2015). Microbiota and Host Nutrition across Plant and Animal Kingdoms. Cell Host and Microbe, 17(5), 603-616.

Halekoh, U. and Højsgaard, S. (2014). A Kenward-Roger Approximation and Parametric Bootstrap Methods for Tests in Linear Mixed Models - The R Package pbkrtest. Journal of Statistical Software. 59, 1-30.

Halpern, B. S., Silliman, B. R., Olden, J. D., Bruno, J. P., and Bertness, M. D. (2007). Incorporating positive interactions in aquatic restoration and conservation. Frontiers in Ecology and the Environment 5, 153–160.

Halpern, B. S., Walbridge, S., Selkoe, K. A., Kappel, C. V., Micheli, F., D'Agrosa, C., Bruno, J. F., Casey, K. S., Ebert, C., Fox, H. E., Fujita, R., Heinemann, D. Lenihan, H. S., Madin, E. M. P., Perry, M. T., Selig, E. R., Spalding, M., Steneck, R., Watson, R. (2008). A global map of human impact on marine ecosystems. Science 319, 948–952.

Handley, L. L. (2015). How will the ‘molecular revolution’ contribute to biological recording? Biological Journal of the Linnean Society, 115: 3, 750–766.

Harley, C., Anderson, K., Demes, K., Jorve, J., Kordas, R. A., Coyle, T., and Graham, M. (2012). Effects of climate change on global seaweed communities. Journal of Phycology 48, 1064–1078.

Heck, K. L., and Orth, R. J. (1980). Seagrass habitats: the roles of habitat complexity, competition and predation in structuring associated fish and motile macroinvertebrate assemblages. Estuarine Perspectives 1980, 449–464. 152

Using genomics to restore and future-proof underwater seaweed forests

Hellweger, F. L., van Sebille, E., and Fredrick, N. D. (2014) Biogeographic patterns in ocean microbes emerge in a neutral agent‐based model. Science 345: 1346– 1349.

Hernandez-Agreda, A., Leggat, W., Bongaerts, P., Herrera, C., Ainsworth, T. D. (2018). Rethinking the coral microbiome: simplicity exists within a diverse microbial biosphere. mBio 9:e00812-18.

Hernandez-Agreda, A., Leggat, W., Bongaerts, P. and Ainsworth, T.D. (2016). The microbial signature provides insight into the mechanistic basis of coral success across reef habitats. mBio 7(4):e00560-16. doi:10.1128/mBio.00560-16.

Higgs, E., Falk, D. A., Guerrini, A., Hall, M., Harris, J., Hobbs, R. J., Jackson, S. T., Rhemtulla, J. M., and Throop, W. (2014). The changing role of history in restoration ecology. Frontiers in Ecology and the Environment 12, 499–506.

Hijmans, Robert J. 2017. Raster: Geographic Data Analysis and Modeling. https://CRAN.R-project.org/package=raster.

Hobbs, R. J. (2018) Restoration Ecology’s silver jubilee: innovation, debate, and creating a future for restoration ecology. Restoration Ecology 26, 801–805.

Hobbs, R. J., and Harris, J. A. (2001). Restoration ecology: repairing the earth’s ecosystems in the new millennium. Restoration Ecology 9, 239–246.

Hobday, A. J. and Pecl, G. T. (2014). Identification of global marine hotspots: sentinels for change and vanguards for adaptation action. Reviews in Fish Biology and Fisheries, 24: 415.

Hoffmann, A., Griffin, P., Dillon, S., Catullo, R., Rane, R., Byrne, M., Jordan, R., Oakeshott, J., Weeks, A., Joseph, L., Lockhart, P., Borevitz, J., and Sgro, C. (2015). A framework for incorporating evolutionary genomics into biodiversity conservation and management. Climate Change Responses 2, 1.

Holguin, G., Vazquez, P., and Bashan, Y. (2001). The role of sediment microorganisms in the productivity, conservation, and rehabilitation of mangrove ecosystems: an overview. Biology and Fertility of Soils 33, 265–278.

Holl, K. D., and Howarth, R. B. (2000). Paying for restoration. Restoration Ecology 8, 260–267.

153

Using genomics to restore and future-proof underwater seaweed forests

Hollants, J., Leliaert, F., De Clerck, O. and Willems, A. (2013), What we can learn from sushi: a review on seaweed–bacterial associations. FEMS Microbiol Ecol, 83: 1-16.

Holling, C.S. (1978). Adaptive Environmental Assessment and Management. John Wiley and Sons, Chichester, UK. ISBN 9781932846072

Hong, S. H., and Lee, E. Y. (2014). Vegetation restoration and prevention of coastal sand dunes erosion using ion exchange resins and the plant growth- promoting rhizobacteria Bacillus sp. SH1RP8 isolated from indigenous plants. International Biodeterioration and Biodegradation 95, 262–269.

Houde, A. L., Garner, S. R., and Neff, B. D. (2015). Restoring species through reintroductions: strategies for source population selection. Restoration Ecology 23, 746–753.

Hu, Z.M., Duan. D.L. and Lopez-Bautista J. (2016). Seaweed Phylogeography from 1994 to 2014: An Overview. In: Hu ZM., Fraser C. (eds) Seaweed Phylogeography. Springer, Dordrecht

Huenneke, L.F. (1991). Ecological implications of genetic variation in plant populations. In Genetics and Conservation of Rare Plants, Falk DA, Holsinger, KE (eds). Oxford, UK: Oxford University Press, pp 31- 44.

Hughes, A., Inouye, B. T. J., Johnson, M., Underwood, N., and Vellend, M. (2008). Ecological consequences of genetic diversity. Ecology Letters 11, 609–623.

Hughes, R. A., and Stachowicz, J. J. (2004). Genetic diversity enhances the resistance of a seagrass ecosystem to disturbance. Proceedings of the National Academy of Sciences of the United States of America 101, 8998–9002.

Hultine, K. R., Grady, K. C., Wood, T. C., Shuster, S. M., Stella, J. C., Whitham, T. G. (2016), Climate change perils for dioecious plant species. Nature Plants, 2 (8), 1-8

Hyndes, G. A., Heck, K. L., Vergés, A., Harvey, E. S., Kendrick, G. A., Lavery, P. S., McMahon, K., Orth, R. J., Pearce, A., and Vanderklift, M. (2016). Accelerating tropicalization and the transformation of temperate seagrass meadows. Bioscience 66, 938–948.

154

Using genomics to restore and future-proof underwater seaweed forests

International Union for Conservation of Nature (IUCN), (2013). ‘Guidelines for Reintroductions and Other Conservation Translocations. Version 1.0. IUCN Species Survival Commission, Gland, Switzerland.

IPBES (2018). The IPBES assessment report on land degradation and restoration. In: Montanarella, L., Scholes, R., and Brainich, A. (Eds.). Secretariat of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services, Bonn, Germany. IPCC (2019). "IPCC, 2019: Summary for Policymakers.," in IPCC Special Report on the Ocean and Cryosphere in a Changing Climate.).

Jackson, J. B. C., Kirby, M. X., Berger, W. H., Bjorndal, K. A., Botsford, L. W., Bourque, B. J., Bradbury, R. H., Cooke, R., Erlandson, J., Estes, J. A., Hughes, T. P., Kidwell, S., Lange, C. B., Lenihan, H. S., Pandolfi, J. M., Peterson, C. H., Steneck, R. S., Tegner, M. J., and Warner, R. R. (2001). Historical overfishing and the recent collapse of coastal ecosystems. Science 293, 629–637.

Johnson, C. R., Chabot, R. H., Marzloff, M. P., and Wotherspoon, S. (2017). Knowing when (not) to attempt ecological restoration. Restoration Ecology 25, 140–147.

Johnson, C., Ling, S., Ross, J., Shepherd, S., and Miller, K. (2005). Establishment of the long‐spined sea urchin (Centrostephanus rodgersii) in Tasmania: first assessment of potential threats to fisheries. FRDC Final Report, 2001/044, School of Zoology and Tasmanian Aquaculture and Fisheries Institute, Univesity of Tasmania, Hobart, Tas., Australia.

Jombart, T. (2008). adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics, 24, 1403-5.

Kamvar, Z., JF, T. and NJ., G. (2014). Poppr: an R package for genetic analysis of populations with clonal, partially clonal, and/or sexual reproduction. Front. Genet, 6:208.

Keenan, K., McGinnity, P., Cross, T. F., Crozier, W. W. and Prodöhl, P. A., (2013). diveRsity: An R package for the estimation and exploration of population genetics parameters and their associated errors. Methods in Ecology and Evolution, 4(8), pp.782-788.

Keenan, R., Lamb, D., Woldring, O., Irvine, A., and Jensen, R. (1997). Restoration of plant biodiversity beneath tropical tree plantations in Northern Australia. Forest Ecology and Management 99, 117–131. 155

Using genomics to restore and future-proof underwater seaweed forests

Kendrick, G., and Statton, J. (2019). 2. Seagrass meadows. In ‘The Role of Restoration for Conserving Matters of National Environmental Significance’. (Eds I. M. McLeod, L. Boström-Einarsson, C. Johnson, G. Kendrick, C. Layton, A. A. Rogers, and J. Statton.) Report to the National Environmental Science Programme, pp. 17–41. Marine Biodiversity Hub, Hobart, Tas., Australia.

Kenworthy, W. J., Hall, M. O., Hammerstrom, K. K., Merello, M., and Schwartzschild, A. (2018). Restoration of tropical seagrass beds using wild bird fertilization and sediment regrading. Ecological Engineering 112, 72–81.

Kettenring, K. M., Mercer, K. L., Adams, C. R., and Hines, J. (2014). Application of genetic diversity- ecosystem function research to ecological restoration. Journal of Applied Ecology 51, 339–348.

Kleynhans, E. J., Otto, S. P., Reich, P. B., and Vellend, M. (2016). Adaptation to elevated CO2 in different biodiversity contexts. Nature Communications 7, 12358.

Klindworth, A., Pruesse, E., Schweer, T., Peplies, J., Quast, C., Horn, M., and Glockner, F. O. (2013). Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Res, 41(1).

Koch, M., Bowes, G., Ross, C., and Zhang, X. (2013). Climate change and ocean acidification effects on seagrasses and marine macroalgae. Global Change Biology 19, 103–132.

Kopelman, N. M., Mayzel, J., Jakobsson, M., Rosenberg, N. A. and Mayrose, I. (2015). Clumpak: a program for identifying clustering modes and packaging population structure inferences across K. Molecular ecology resources, 15, 1179- 1191.

Kovács-Hostyánszki, A., Espíndola, A., Vanbergen, A. J., Settele, J., Kremen, C. and Dicks, L. V. (2017). Ecological intensification to mitigate impacts of conventional intensive land use on pollinators and pollination. Ecology Letters, 20, 673-689.

Krause-Jensen, D., and Duarte, C. M. (2016). Substantial role of macroalgae in marine carbon sequestration. Nature Geoscience 9, 737–742.

156

Using genomics to restore and future-proof underwater seaweed forests

Kronenberger, J. A., Funk, W. C., Smith, J. W., Fitzpatrick, S. W., Angeloni, L. M., Broder, E. D., and Ruell, E. W. (2017). Testing the demographic effects of divergent immigrants on small populations of Trinidadian guppies. Animal Conservation 20, 3–11.

Krumhansl, K. A. et al., (2016). Global patterns of kelp forest change over the past half-century. Proceedings of the National Academy of Sciences, 113, 13785-13790.

Krumhansl, K. A., Okamoto, D. K., Rassweiler, A., Novak, M., Bolton, J. J., Cavanaugh, K. C., Connell, S. D., Johnson, C. R., Konar, B., Ling, S. D., Micheli, F., Norderhaug, K. M., Pérez-Matus, A., Sousa-Pinto, I., Reed, D. C., Salomon, A. K., Shears, N. T., Wernberg, T., Anderson, R. J., Barrett, N. S., Buschmann, A. H., Carr, M. H., Caselle, J. E., Derrien-Courtel, S., Edgar, G. J., Edwards, M., Estes, J. A., Goodwin, C., Kenner, M. C., Kushner, D. J., Moy, F. E., Nunn, J., Steneck, R. S., Vásquez, J., Watson, J., Witman, J. D., and Byrnes, J. E. K. (2016). Global patterns of kelp forest change over the past half-century. Proceedings of the National Academy of Sciences of the United States of America 113, 13785–13790.

Kumar, V., Zozaya-Valdes, E., Kjelleberg, S., Thomas, T. and Egan, S. (2016). Multiple opportunistic pathogens can cause a bleaching disease in the red seaweed Delisea pulchra. Environmental Microbiology, 18(11), 3962-3975.

Lachnit, T., Blümel, M., Imhoff, J.F. and Wahl, M. (2009). Specific epibacterial communities on macroalgae: phylogeny matters more than habitat. Aquatic Biology, 5: 181–186.

Lachnit, T., Meske, D., Wahl, M., Harder, T., and Schmitz, R. (2011) Epibacterial community patterns on marine macroalgae are host‐specific but temporally variable. Environmental Microbiology 13: 655–665.

Lacoursière‐Roussel, A., Côté, G., Leclerc, V. and Bernatchez, L. (2016), Quantifying relative fish abundance with eDNA: a promising tool for fisheries management. Journal of Applied Ecology, 53: 1148-1157.

Laegdsgaard, P. (2006). Ecology, disturbance and restoration of coastal saltmarsh in Australia: a review. Wetlands Ecology and Management 14, 379–399. Lamb, D. (2014). ‘Large-Scale Forest Restoration.’ (Routledge: Oxford, UK.)10.4324/9780203071649

Langley, M., Vergés, A. and Marzinelli, E. (in prep.). Kelp microbiome variation associated with fish herbivory and latitude.

157

Using genomics to restore and future-proof underwater seaweed forests

Layton, C., Coleman, M., Marzinelli, E. M., Steinberg, P., Swearer, S. E., Vergés, A., Wernberg, T. and Johnson, C. R. (2018). Restoring Kelp Habitat in Australia. In: McLeod I. M., Boström-Einarsson L., Johnson C. R., Kendrick G., Layton C., Rogers A. A., Statton J. (2018). The role of restoration in conserving matters of national environmental significance. Report to the National Environmental Science Programme, Marine Biodiversity Hub, 42-54.

Lefcheck, J. S., Orth, R. J., Dennison, W. C., Wilcox, D. J., Murphy, R. R., Keisman, J., Gurbisz, C., Hannam, M., Landry, J. B., Moore, K. A., Patrick, C. J., Testa, J., Weller, D. E., and Batiuk, R. A. (2018). Long-term nutrient reductions lead to the unprecedented recovery of a temperate coastal region. Proceedings of the National Academy of Sciences of the United States of America 115, 3658–3662.

Lefcheck, J. S., Orth, R. J., Dennison, W. C., Wilcox, D. J., Murphy, R. R., Keisman, J., Gurbisz, C., Hannam, M., Landry, J.B., Moore, K. A. and Patrick, C. J., (2018). Long-term nutrient reductions lead to the unprecedented recovery of a temperate coastal region. Proceedings of the National Academy of Sciences, 115(14), pp.3658-3662.

Lenth, R. (2018). emmeans: Estimated Marginal Means, aka Least-Squares Means. Retrieved from https://CRAN.R-project.org/package=emmeans

Lesen, A. E., Rogan, A., and Blum, M. J. (2016). Science communication though art: objectives, challenges, and outcomes. Trends in Ecology and Evolution 31, 657–660.

Lesica, P., and Allendorf, F. W. (1999). Ecological genetics and the restoration of plant communities: mix or match? Restoration Ecology 7, 42–50.

Levin, R. A., Voolstra, C. R., Agrawal, S., Steinberg, P. D., Suggett, D. J., and van Oppen, M. J. H. (2017). Engineering strategies to decode and enhance the genomes of coral symbionts. Frontiers in Microbiology 8, 1220.

Lin, H., and Qin, S. (2014). Tipping points in seaweed genetic engineering: scaling up opportunities in the next decade. Marine Drugs 12, 3025–3045.

Lindenmayer, D. B., and Likens, G. E. (2009). Adaptive monitoring: a new paradigm for long-term research and monitoring. Trends in Ecology and Evolution 24, 482–486.

158

Using genomics to restore and future-proof underwater seaweed forests

Ling, S. D., Johnson, C. R., Frusher, S. D., and Ridgway, K. R. (2009). Overfishing reduces resilience of kelp beds to climate-driven catastrophic phase shift. Proceedings of the National Academy of Sciences of the United States of America 106, 22341–22345.

Lotze, H. K., Coll, M., Magera, A. M., Warde-Paige, C., and Airoldi, L. (2011). Recovery of marine animal populations and ecosystems. Trends in Ecology and Evolution 26, 595–605.

Lotze, H. K., Lenihan, H. S., Bourque, B. J., Bradbury, R. H., Cooke, R. G., Kay, M. C., Kidwell, S. M., Kirby, M. X., Peterson, C. H., and Jackson, J. B. (2006). Depletion, degradation, and recovery potential of estuaries and coastal seas. Science 312, 1806–1809.

Lourenço, C.R., Zardi, G.I., McQuaid, C.D., Serrão, E.A., Pearson, G.A., Jacinto, R. and Nicastro, K.R. (2016), Upwelling areas as climate change refugia for the distribution and genetic diversity of a marine macroalga. Journal of Biogeography, 43: 1595-1607.

Love, M., Anders, S., and Huber, W. (2014). DESeq2: Differential gene expression analysis based on the negative binomial distribution. R package version 1.20.0.

Lovelock, C. E., Atwood, T., Baldock, J., Duarte, C. M., Hickey, S., Lavery, P. S., Masque, P., Macreadie, P. I., Ricart, A. M., Serrano, O., and Steven, A. (2017). Assessing the risk of carbon dioxide emissions from blue carbon ecosystems. Frontiers in Ecology and the Environment 15, 257–265.

Lu, T. T., and Williams, S. L. (1994). Genetic diversity and genetic structure in the brown alga Halidrys dioica (Fucales: Cystoseiraceae) in Southern California. Marine Biology 121, 363–371.

Luu, K., Bazin, E. and Blum, MG. (2017). pcadapt: an R package to perform genome scans for selection based on principal component analysis. Molecular Ecology Resources. 7(1):67-77.

Lyerly, C. M., Hernández Cordero, A. L., Foreman, K. L., Phillips, S. W., and Dennison, W. C. (Eds) (2014). New insights: science-based evidence of water quality improvements, challenges, and opportunities in the Chesapeake. Available at http://ian.umces.edu/press/reports/publication/438/new_insights_science_bas

159

Using genomics to restore and future-proof underwater seaweed forests

ed_evidence_of_water_quality_improvements_challenges_and_opportunities_in_ the_chesapeake_2014-02-24/ [Verified 10 June 2018].

Macreadie, P. I., Nielsen, D. A., Kelleway, J. J., Atwood, T. B., Seymour, J. R., Petrou, K., Connolly, R. M., Thomson, A. C. G., Trevathan-Tackett, S. M., and Ralph, P. J. (2017). Can we manage coastal ecosystems to sequester more blue carbon? Frontiers in Ecology and the Environment 15, 206–213.

Mao, X., Augyte, S., Huang, M., Hare, M., Bailey, D., Umanzor, S., Marty-Rivera, M., Robbins, K., Yarish, C., Lindell, S. and Jannink, J.L., 2020. Population genetics of sugar kelp in the Northwest Atlantic region using genome-wide markers. bioRxiv.

Mamo, L. T., Kelaher, B. P., Coleman, M. A., and Dwyer, P. G. (2018). Protecting threatened species from coastal infrastructure upgrades: the importance of evidence-based conservation. Ocean and Coastal Management 165, 161–166.

Mann, K. H. (1973). Seaweeds: their productivity and strategy for growth. The role of large marine algae in coastal productivity is far more important than has been suspected. Science 182, 975–981.

Marino, C., Pawlik, J., López‐Legentil, S., and Erwin, P. (2017). Latitudinal variation in the microbiome of the sponge Ircinia campana correlates with host haplotype but not anti‐predatory chemical defense. Marine Ecology Progress Series, 565, 53–66.

Marion, S. R., and Orth, R. J. (2010). Innovative techniques for large‐scale seagrass restoration using Zostera marina (eelgrass) seeds. Restoration Ecology 18, 514– 526.

Marshall, K., Joint, I., Callow, M., and Callow, J. (2006). Effect of marine bacterial isolates on the growth and morphology of axenic plantlets of the green alga Ulva linza. Microbial Ecology 52, 302–310.

Martin, D. M. (2017). Ecological restoration should be redefined for the twenty‐ first century. Restoration Ecology 25, 668–673.

Martínez, B., Radford, B., Thomsen, M. S., Connell, S. D., Carreño, F., Bradshaw, C. J. A., Fordham, D. A., Russell, B. D., Gurgel, C. F. D., and Wernberg, T. (2018). Distribution models predict large contractions of habitat‐forming seaweeds in response to ocean warming. Diversity and Distributions 24, 1350–1366.

160

Using genomics to restore and future-proof underwater seaweed forests

Martínez, B., Radford, B., Thomsen, M. S., Connell, S. D., Carreno, F., Bradshaw, C. J. A., Fordham, D. A., Russell, B. D., Gurgel, C. F. D. and Wernberg, T. (2018). Distribution models predict large contractions in habitat‐forming seaweeds in response to ocean warming. Diversity and Distributions., 24: 1350– 1366.

Marzinelli E. M., Campbell A. H., Vergés A., Coleman M. A., Kelaher B. P., Steinberg P.D. (2014) Restoring seaweeds: does the declining fucoid Phyllospora comosa support different biodiversity than other habitats? Journal of Applied Phycology 26: 1089– 1096

Marzinelli, E. M., Leong, M. R., Campbell, A. H., Steinberg, P. D., and Vergés, A. (2016). Does restoration of a habitat-forming seaweed restore associated faunal diversity? Restoration Ecology 24, 81–90.

Marzinelli, E. M., Qiu, Z., Dafforn, K. A., Johnston, E. L., Steinberg, P. D., and Mayer-Pinto, M. (2018). Coastal urbanisation affects microbial communities on a dominant marine holobiont. Biofilms and Microbiomes 4, 1.

Marzinelli, E. M., Campbell, A. H., Zozaya Valdes, E., Vergés, A., Nielsen, S., Wernberg, T., de Bettignies, T., Bennett, S., Caporaso, J. G, Thomas, T. and Steinberg, P. D. (2015). Continental-scale variation in seaweed host-associated bacterial communities is a function of host condition, not geography. Environmental microbiology, vol. 17, pp. 4078 – 4088.

Matheson, F. E., Reed, J., Dos Santos, V. M., MacKay, G., and Cummings, V. J. (2017). Seagrass rehabilitation: successful transplants and evaluation of methods at different spatial scales. New Zealand Journal of Marine and Freshwater Research 51, 96–109.

Matthieu Foll. (2012, January 21). BayeScan Official Page. Retrieved 16:47, August 8, 2018 from http://cmpg.unibe.ch/software/BayeScan/.

McArdle, B. H. and Anderson, M. J. (2001). Fitting Multivariate Models to Community Data: A Comment on Distance-Based Redundancy Analysis. Ecology, 82: 290-297.

McCoy, S. J., Widdicombe, S. (2019). Thermal plasticity is independent of environmental history in an intertidal seaweed. Ecology and Evolution, 9: 13402– 13412.

161

Using genomics to restore and future-proof underwater seaweed forests

McFall-Ngai, M., Hadfield, M. G., Bosch, T. C. G., Carey, H. V.; Domazet-Lošo, T., Douglas, A. E.; Dubilier, N.; Eberl, G., Fukami, T.; Gilbert, S. F.; Hentschel, U., King, N., Kjelleberg, S., Knoll, A. H., Kremer, N., Mazmanian, S. K., Metcalf, J. L.; Nealson, K., Pierce, N. E., Rawls, J. F., Reid, A., Ruby, E. G., Rumpho, M., Sanders, J. G.; Tautz, D., Wernegreen, J. J. (2013). Animals in a bacterial world, a new imperative for the life sciences. Proceedings of the National Academy of Sciences 110, 3229–3236.

McGeoch, M. A., Latombe, G., Andrew, N. R., Nakagawa, S., Nipperess, D. A., Roige, M., …Hui, C. (2019). Measuring continuous compositional change using decline and decay in zeta diversity. Ecology. McHugh, D. J. (2003). A guide to the seaweed industry. FAO Fisheries Technical Paper number 441, Food and Agriculture Organization of the United Nations, Rome, Italy.

McKay, J. K., Christian, C. E., Harrison, S., and Rice, K. J. (2005). ‘How local is local?’ A review of practical and conceptual issues in the genetics of restoration. Restoration Ecology 13, 432–440.

McKenzie, P. F., and Bellgrove, A. (2006). No outbreeding depression at a regional scale for a habitat-forming intertidal alga with limited dispersal. Marine and Freshwater Research 57, 655–663.

McMurdie, P.J. and Holmes, S. (2014) Waste Not, Want Not: Why Rarefying Microbiome Data Is Inadmissible. PLoS Computational Biology 10(4): e1003531.

Medina, F. J., Flores, V., Gonzalez, A. V. and Santelices, B. (2015). Coalescence increases abiotic stress tolerance in sporelings of Mazzaella laminarioides (Gigartinales, Rhodophyta). Journal of Applied Phycology, 27, 1593-1598.

Mendez, F. J. & Losada, I. J. (2004). Transformation of random and non-random breaking waves over vegetation fields. Coast. Eng. 51, 103–118.

Menz, M. H. M., Dixon, K. W., and Hobbs, R. J. (2013). Hurdles and opportunities for landscape-scale restoration. Science 339, 526–527.

Merritt, D. J., and Dixon, K. W. (2011). Restoration seed banks – a matter of scale. Science 332, 424–425.

162

Using genomics to restore and future-proof underwater seaweed forests

Michel G., Nyval‐Collen P., Barbeyron T., Czjzek M. and Helbert W. (2006). Bioconversion of red seaweed galactans: a focus on bacterial agarases and carrageenases. Applied Microbial Biotechnology 71: 23–33.

Mijangos, J. L., Carlo, P., S., S. P. B. and D., C. M. (2015). Contribution of genetics to ecological restoration. Molecular Ecology, 24, 22-37.

Millennium Ecosystem Assessment (2005). Summary for decision makers. In ‘Ecosystems and Human Well-being: Synthesis’. pp. 1–24. (Island Press: Washington DC, USA.)

Miller, A., Coleman, M., Clark, J., Cook, R., Naga, Z., Doblin, M., et al., (In press). Local thermal adaptation and limited gene flow constrain future climate responses of a marine ecosystem engineer. Evolutionary Applications.

Miller, B. P., Sinclair, E. A., Menz, M. H., Elliott, C. P., Bunn, E., Commander, L. E., Dalziell, E., David, E., Davis, B., Erickson, T. E. and Golos, P. J., (2017). A framework for the practical science necessary to restore sustainable, resilient, and biodiverse ecosystems. Restoration Ecology, 25(4), pp.605-617.

Morandi, B., Piegay, H., Lamouroux, N. and Vaudor, L. (2014). How is success or failure in river restoration projects evaluated? Feedback from French restoration projects. Journal of Environmental Management, 137; 178-188.

Morin, P. A., Luikart, G., and Wayne, R. K. (2004). SNPs in ecology, evolution and conservation. Trends in Ecology and Evolution 19, 208–216.

Morin, P.A., Martien, K.K. and Taylor, B.L., (2009). Assessing statistical power of SNPs for population structure and conservation studies. Molecular Ecology Resources, 9(1), 66-73.

Moritz C. (2002). Strategies to protect biological diversity and the evolutionary processes that sustain it. Systematic Biology, 51, 238–254.

Narum, S. R. and Hess, J. E. (2011), Comparison of FST outlier tests for SNP loci under selection. Molecular Ecology Resources, 11: 184-194

Nazareno, A.G., Bemmels, J.B., Dick, C.W. and Lohmann, L.G., (2017). Minimum sample sizes for population genomics: an empirical study from an Amazonian plant species. Molecular Ecology Resources, 17(6), 1136-1147.

163

Using genomics to restore and future-proof underwater seaweed forests

Neefjes, J., Jongsma, M. L., Paul, P., Bakke, O. (2011) Towards a systems understanding of MHC class I and MHC class II antigen presentation. Nat Rev Immunol 11:823–836.

Nicastro, K. R., Zardi, G. I., Teixeira, S., Neiva, J., Serrao, E. A., and Pearson, G. A. (2013). Shift happens: trailing edge contraction associated with recent warming trends threatens a distinct genetic lineage in the marine macroalga Fucus vesiculosus. BMC Biology 11: 6.

O’Brien, K. R., Waycott, M., Maxwell, P., Kendrick, G. A., Udy, J. W., Fergusun, A. J. P., Kilminster, K., Scanes, P., McKenzie, L. J., McMahon, K., Adams, M. P., Samper-Villarreal, J., Collier, C., Lyons, M., Mumby, P. J., Radke, L., Christianen, M. J. A., and Dennison, W. C. (2018). Seagrass ecosystem trajectory depends on the relative timescales of resistance, recovery and disturbance. Marine Pollution Bulletin 134, 166–176.

O’Brien, P. A., Webster, N. S., Miller, D. J. and Bourne, D. G. (2019). Host-microbe coevolution: applying evidence from model systems to complex marine invertebrate holobionts. mBio 10:e02241-18.

O’Leary, J. K., Micheli, F., Airoldi, L., and Wong, J. (2017). The resilience of marine ecosystems to climatic disturbances. Bioscience 67, 208.

Oksanen, J., Blanchet, F. G., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., . . . Wagner, H. (2018). vegan: Community Ecology Package (Version R Package version 2.5-2).

On-prom, S. (2014). Community-based mangrove forest management in Thailand: key lesson learned for environmental risk management. In ‘Sustainable Living with Environmental Risks’. (Eds N. Kaneko, S. Yoshiura, and M. Kobayashi.) pp. 87–96. (Springer: Tokyo, Japan.)

Pearson, G. A. and Serrao, E. A. (2006). Revisiting synchronous gamete release by fucoid algae in the intertidal zone: fertilization success and beyond? Integrated Computational Biology, 46, 587-97.

Pecl, G. T., Araujo, M. B., Bell, J. D., Blanchard, J., Bonebrake, T. C., Chen, I. C., Clark, T. D., Colwell, R. K., Danielsen, F., Evengard, B., Falconi, L., Ferrier, S., Frusher, S., Garcia, R. A., Griffis, R. B., Hobday, A. J., Janion-Scheepers, C., Jarzyna, M. A., Jennings, S., Lenoir, J., Linnetved, H. I., Martin, V. Y., Mccormack, P. C., Mcdonald, J., Mitchell, N. J., Mustonen, T., Pandolfi, J. M., Pettorelli, N., 164

Using genomics to restore and future-proof underwater seaweed forests

Popova, E., Robinson, S. A., Scheffers, B. R., Shaw, J. D., Sorte, C. J. B., Strugnell, J. M., Sunday, J. M., Tuanmu, M.-N., Vergés, A., Villanueva, C., Wernberg, T., Wapstra, E., and Williams, S. E. (2017). Biodiversity redistribution under climate change. Science 355, eaai9214.

Perkol‐Finkel, S., Ferrario, F., Nicotera, V., and Airoldi, L. (2012). Conservation challenges in urban seascapes: promoting the growth of threatened species on coastal infrastructures. Journal of Applied Ecology 49, 1457–1466.

Perring, M. P., Standish, R. J., Price, J. N., Craig, M. D., Erickson, T. E., Ruthrof, K. X., Whiteley, A. S., Valentine, L. E., and Hobbs, R. J. (2015). Advances in restoration ecology: rising to the challenges of the coming decades. Ecosphere 6, art131.

Peters, M. A., Hamilton, D., and Eames, C. (2015). Action on the ground: a review of community environmental groups’ restoration objectives, activities and partnerships in New Zealand. New Zealand Journal of Ecology 39, 179–189.

Peters, T. A. (2015). Patterns, mechanisms and consequences of disease in a habitat-forming macroalga. PhD, University of New South Wales.

Pfeiffer, L., Holden, H., and Jackson, E. (2017). Seagrass and aluminium are strange bedfellows: science–art collaboration via the power of steam. In ‘2017 Science, Technology, Engineering Arts and Mathematics (STEAM) Education Proceedings – Hawaii University International Conferences 2017’, 8–10 June 2017, Honolulu, HI, USA. (Hawaii University International Conferences, Science Technology and Engineering, Arts Mathematics and Education.) Available at https://huichawaii.org/wp-content/uploads/2017/09/Pfeiffer-Linda-2017- STEAM-HUIC.pdf [Verified 24 February 2019].

Phillips, J.A. (2001). Marine macroalgal biodiversity hotspots: why is there high species richness and endemism in southern Australian marine benthic flora? Biodiversity and Conservation 10(9), 1555-1577.

Pritchard, J. K., Stephens, M. and Donnelly, P. (2000). Inference of population structure using multilocus genotype data. Genetics, 155, 945-59.

Provan, J., and Maggs, C. A. (2011). Unique genetic variation at a species' rear edge is under threat from global climate change. Proceedings of the Royal Society B 279: 39– 47.

165

Using genomics to restore and future-proof underwater seaweed forests

Provost, E. J., Kelaher, B. P., Dworjanyn, S. A., Russel, B. D., Connell, S. D., Ghedini, G., Gillanders, B. M., Figueira, W., and Coleman, M. A. (2017). Climate‐ driven disparities among ecological interactions threaten kelp forest persistence. Global Change Biology 23, 353–361.

Qin, S., Lin, H., and Jiange, P. (2012). Advances in genetic engineering of marine algae. Biotechnology Advances 30, 1602–1613.

Qiu, Z. (2017) Effect of environmental changes on the microbiome and condition of the kelp Ecklonia radiata. PhD Thesis UNSW

Qiu, Z., Coleman, M., Provost, E., Campbell, A., Kelaher, B., Dalton, S., Thomas, T., Steinberg, P., Marzinelli, E. (2019). Future climate change is predicted to affect the microbiome and condition of habitat-forming kelp. Proceedings of the Royal Society B: Biological Sciences, 286(1896), 1-10.

Qiu, Z., Marzinelli, E. M., Campbell, A. H., Kjelleberg, S., Thomas, T., Steinberg, P. D.(in prep.) Microbiome characterization and inoculation experiments reveal bacteria affecting the condition of a habitat-forming host.

Quigley, K. M., Bay, L. K., van Oppen, M. J. H. (2019). The active spread of adaptive variation for reef resilience. Ecology and Evolution. 9: 11122– 11135.

R Core Team. (2019). R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from https://www.R-project.org/

Radchuk, V., Reed, T., Teplitsky, C. et al., Adaptive responses of animals to climate change are most likely insufficient. Nature Communications 10, 3109 (2019). https://doi.org/10.1038/s41467-019-10924-4

Ralls, K., Ballou, J. D., Dudash, M. R., Eldridge, M. D., Fenster, C. B., Lacy, R. C., Sunnucks, P., and Frankham, R. (2018). Call for a paradigm shift in the genetic management of fragmented populations. Conservation Letters 11, e12412.

Rawls J. F., Mahowald M. A., Ley R. E., Gordon J. I. (2006). Reciprocal gut microbiota transplants from zebrafish and mice to germ-free recipients reveal host habitat selection. Cell 127 423–433.

166

Using genomics to restore and future-proof underwater seaweed forests

Reigersman, C. J. A., Houben, G. F. H., and Havinga, B. (1939). Rapport omtrent den invloed van de wierziekte op den achteruitgang van de wierbedrijven, met Bijlagen. Provinciale Waterstaat in Noord-Holland, Haarlem, Netherlands.

Relling M. V., Evans W. E. (2015). Pharmacogenomics in the clinic. Nature 526(7573):343-50.

Rellstab, C., Gugerli, F., Eckert, A.J., Hancock, A.M. and Holderegger, R. (2015), A practical guide to environmental association analysis in landscape genomics. Molecular Ecology, 24: 4348-4370.

Reusch, T. B. H., Ehlers, A., Hammerli, A., and Worm, B. (2005). Ecosystem recovery after climatic extremes enhanced by genotypic diversity. Proceedings of the National Academy of Sciences of the United States of America 102, 2826–2831.

Rey Benayas, J. M., Newton, A. C., Diaz, A., and Bullock, J. M. (2009). Enhancement of biodiversity and ecosystem services by ecological restoration: a meta-analysis. Science 325, 1121–1124.

Reynolds, L. K., Mcglathery, K. J., and Waycott, M. (2012). Genetic diversity enhances restoration success by augmenting ecosystem services. PLoS One 7, e38397.

Rhode, C., Bester-van der Merwe, A.E. and Roodt-Wilding, R., (2017). An assessment of spatio-temporal genetic variation in the South African abalone (Haliotis midae), using SNPs: implications for conservation management. Conservation genetics, 18(1), 17-31.

Richardson, B. J., and Lefroy, T. (2016). Restoration dialogues: improving the governance of ecological restoration. Restoration Ecology 24, 668–673.

Ridgway, K. R. ( 2007), Long‐term trend and decadal variability of the southward penetration of the East Australian Current, Geophysical Research Letters, 34, L13613.

Rinkevich, B. (2014). Rebuilding coral reefs: does active reef restoration lead to sustainable reefs? Current Opinion in Environmental Sustainability 7, 28–36.

Rodriguez-Verdugo, A., Carrillo-Cisneros, D., Gonzalez-Gonzalez, A., Gaut, B. S. and Bennett, A. F. (2014). Different tradeoffs result from alternate genetic adaptations to a common environment. PNAS August 19, 2014 111 (33) 12121-12126.

167

Using genomics to restore and future-proof underwater seaweed forests

Rohwer F., Seguritan V., Azam F., Knowlton N. (2002). Diversity and distribution of coral-associated bacteria. Marine Ecology Progress Series, 243: 1–10.

Rosado, P. M., Leite, D. C. A., Duarte, G. A. S., Chaloub, R. M., Jospin, G., Nunes da Rocha, U., et al., (2018). Marine probiotics: increasing coral resistance to bleaching through microbiome manipulation. ISME J. 13, 921–936.

Rosenberg, M. and Falkovitz, L. (2004). The Vibrio shiloi/Oculina patagonica Model System of Coral Bleaching. Annual Review of Microbiology 2004 58:1, 143-159.

Rosenberg, M. and Zilber-Rosenberg (2018). The hologenome concept of evolution after 10 years. Microbiome, 6:78.

Rosenzweig, M. (2003). ‘Win–Win Ecology, How the Earth’s Species can Survive in the Midst of Human Enterprise.’ (Oxford University Press: Oxford, UK.)

Rossetto, M., Bragg, J., Kilian, A., Mcpherson, H., Van der Merwe, M. and Wilson, P. D. (2019). Restore and Renew: a genomics-era framework for species provenance delimitation. Restoration Ecology, 27, 538-548.

Rothschild, D., Weissbrod, O., Barkan, E., Kurilshikov, A., Korem, T., Zeevi, D., Costea, P. I., Godneva, A., Kalka, I. N., Bar, N. and Shilo, S. (2018). Environment dominates over host genetics in shaping human gut microbiota. Nature, 555(7695), pp.210-215.

Roth-Schulze, A. J., Zozaya-Valdés, E., Steinberg, P. D., and Thomas, T. (2016). Partitioning of functional and taxonomic diversity in surface-associated microbial communities. Environmental Microbiology 18, 4391–4402.

Roth-Schulze, A.J., Pintado, J., Zozaya-Valdés, E., Cremades, J., Ruiz, P., Kjelleberg. S., Thomas, T. (2018). 'Functional biogeography and host specificity of bacterial communities associated with the Marine Green Alga Ulva spp.', Molecular Ecology, 27, 1952 – 1965.

Rousset, F. (2008). genepop’007: a complete re-implementation of the genepop software for Windows and Linux. Molecular Ecology Resources, 8, 103-106.

168

Using genomics to restore and future-proof underwater seaweed forests

Ruby, E. G., and Nealson, K. H. (1976). Symbiotic association of Photobacterium fischeri with the marine luminous fish Monocentris japonica: a model of symbiosis based on bacterial studies. The Biological Bulletin 151, 574–586.

Rusch, D. B., Halpern, A. L., Sutton, G., Heidelberg, K. B., Williamson, S., Yooseph, S., et al. (2007) The sorcerer II global ocean sampling expedition: northwest atlantic through eastern tropical pacific. PLoS Biology, 5: 398–431.

Sala, E., Boudouresque, C. F., and Harmelin-Vivien, M. (1998). Fishing, trophic cascades, and the structure of algal assemblages: evaluation of an old but untested paradigm. Oikos 82, 425–439.

Sanderson, J., Ling, S., Dominguez, J., and Johnson, C. (2016). Limited effectiveness of divers to mitigate ‘barrens’ formation by culling sea urchins while fishing for abalone. Marine and Freshwater Research 67, 84–95.

Sandoval-Motta, S., Aldana, M., Martinez-Romero, E. and Frank, A. (2017). The Human Microbiome and the Missing Heritability Problem. Frontiers in Genetics. https://doi.org/10.3389/fgene.2017.00080

Scanes, P. R. and Philip, N. (1995). Environmental impact of deepwater discharge of sewage off Sydney, NSW, Australia. Marine Pollution Bulletin, 31, 343-346.

Schiel, D., and Foster, M. (2006). The Population Biology of Large Brown Seaweeds: Ecological Consequences of Multiphase Life Histories in Dynamic Coastal Environments. Annual Review of Ecology, Evolution, and Systematics, 37, 343-372.

Schmitz, O. J. (2017). The New Ecology. Rethinking a Science for the Anthropocene. Princeton University Press, Princeton, New Jersey. 237pp.

Seaman, W. (2007). Artificial habitats and the restoration of degraded marine ecosystems and fisheries. Hydrobiologia 580, 143–155.

Selechnik, D., Richardson, M. F., Shine, R., DeVore, J. L., Ducatez, S. and Rollins, L. A. (2019). Increased adaptive variation despite reduced overall genetic diversity in a rapidly adapting invader. Frontiers in Genetics, https://doi.org/10.3389/fgene.2019.01221

Selkoe, K. A., d’Aloia, C. C., Crandall, E. D., Iacchei, M., Liggins, L., Puritz, J. B., von der Heyden, S., and Toonen, R. (2016). A decade of seascape genetics:

169

Using genomics to restore and future-proof underwater seaweed forests

contributions to basic and applied marine connectivity. Marine Ecology Progress Series 554, 1–19.

Sgrò, C. M., Lowe, A. J. and Hoffmann, A. A. (2011). Building evolutionary resilience for conserving biodiversity under climate change. Evolutionary Applications, 4, 326-337.

Sharp, K. H., Eam, B., Faulkner, D. J., and Haygood, M. G. (2007). Vertical transmission of diverse microbes in the tropical sponge Corticium sp. Applied and Environmental Microbiology 73, 622–629.

Shears, N. T., and Babcock, R. C. (2003). Continuing trophic cascade effects after 25 years of no-take marine reserve protection. Marine Ecology Progress Series 246, 1–16.

Simonson, E. J., Metaxas, A. and Scheibling, R. E. (2015). Kelp in hot water: II. Effects of warming seawater temperature on kelp quality as a food source and settlement substrate. Marine Ecology Progress Series, 537, 105-119.

Singh, R. P. and Reddy, C. R. K., (2016). Unraveling the functions of the macroalgal microbiome. Frontiers in Microbiology, 6: 1488.

Smale, D. A., and Wernberg, T. (2013). Extreme climatic event drives range contraction of a habitat-forming species. Proceedings. Biological sciences, 280(1754), 20122829.

Smale, D. A. (2020), Impacts of ocean warming on kelp forest ecosystems. New Phytology, 225: 1447-1454. Society for Ecological Restoration (1993). Environmental policies of the society for ecological restoration. Restoration Ecology 1, 206–207.

Spyksma, A. J. P., Shears, N. T., and Taylor, R. B. (2017). Predators indirectly induce stronger prey through a trophic cascade. Proceedings of the Royal Society of London. Series B, Biological Sciences 284, 1866–1874.

Srinivas G., Möller S., Wang J., Künzel S., Zillikens D., Baines J. F., et al., (2013). Genome wide mapping of gene–microbiota interactions in susceptibility to autoimmune skin blistering. Nature communications 4 2462. 10.1038/ncomms3462

170

Using genomics to restore and future-proof underwater seaweed forests

Staufenberger, T., Thiel, V., Wiese, J. and Imhoff, J. F. (2008). Phylogenetic analysis of bacteria associated with Laminaria saccharina. FEMS Microbiol Ecol 64: 65–77.

Steffen, W., Broadgate, W., Deutsch, L., Gaffney, O., and Ludwig, C. (2015). The trajectory of the Anthropocene: The Great Acceleration. The Anthropocene Review, 2(1), 81–98.

Steffen, W., Rockström, J., Richardson, K., Lenton, T. M., Folke, C., Liverman, D., Summerhayes, C. P., Barnosky, A. D., Cornell, S. E., Crucifix, M., Donges, J. F., Fetzer, I., Lade, S. J., Scheffer, M., Winkellmann, R. and Joachim Schnellnhuber, H. (2018). Trajectories of the Earth System in the Anthropocene. PNAS, 115 (33) 8252-8259.

Steneck, R. S. and Johnson, C. R. 2013. Kelp forests: dynamic patterns, processes, and feedbacks, Marine Community Ecology and Conservation, Sinauer Associates, Inc., MD Bertness, JF Bruno, BR Silliman, JJ Stachowicz (ed), Massachusetts, USA, pp. 315-336. ISBN 9781605352282 (2014).

Steneck, R. S., Graham, M. H., Bourque, B. J., Corbett, D., Erlandson, J. M., Estes, J. A., and Tegner, M. J. (2002). Kelp forest ecosystems: biodiversity, stability, resilience and future. Environmental Conservation 29:436-459.

Storey, J. D. (2002). A direct approach to false discovery rates. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 64, 479-498.

Strain, E. M. A., Olabarria, C., Mayer‐Pinto, M., Cumbo, V., Morris, R. L., Bugnot, A. B., Dafforn, K. A., Heery, E., Firth, L. B., Brooks, P. R., and Bishop, M. J. (2018). Eco‐engineering urban infrastructure for marine and coastal biodiversity: which interventions have the greatest ecological benefit? Journal of Applied Ecology 55, 426–441.

Straub, S. C., Wernberg, T., Thomsen, M. S., Moore, P. J., Burrows, M. T., Harvey, B. P., Smale, D. A. (2019). Resistance, extinction, and everything in between – the diverse responses of seaweeds to marine heatwaves. Frontiers in Marine Science: Global Change and the Future Ocean, accepted 25/11/19 Manuscript ID: 473578.

Sutherland, J.P (1990). Perturbations, resistance, and alternative views of the existence of multiple stable points in nature. American Naturalist., 136, 270-275

171

Using genomics to restore and future-proof underwater seaweed forests

Teagle, H., Hawkins, S. J., Moore, P. J., and Smale, D. A. (2017). The role of kelp species as biogenic habitat formers in coastal marine ecosystems. Journal of Experimental Marine Biology and Ecology 492, 81–98.

Thompson, L. R., Sanders, J. G., Mcdonald, D., Amir, A., Ladau, J., Locey, K. J., Prill, R. J., Tripathi, A., Gibbons, S. M., Ackermann, G., Navas-Molina, J. A., Janssen, S., Kopylova, E., Vázquez-Baeza, Y., González, A., Morton, J. T., Mirarab, S., Zech Xu, Z., Jiang, L, ...Shade, A., Pollard, K. S., Goodwin, K. D., Jansson, J. K., Gilbert, J. A., Knight, R., Earth Microbiome Project Consortium (2017). A communal catalogue reveals earth’s multiscale microbial diversity. Nature 551, 457–463.

Thompson, L., Sanders, J., McDonald, D., Amir, A., Ladau, J., Locey, K. J., Prill, R. J., Tripathi, A., Gibbons, S. M., Ackermann, G., Navas-Molina, J. A., Janssen, S., Kopylova, E., Vasquez-Baeza, Y., Gonzalez, A., Morton, J. T., Mirarab, S., Zech Xu, Z., Jiang, L., Haroon, M. F., Kanbar, J., Zhu, Q., Jin Song, S., Kosciolek, T., Bokulich, N. A., Lefler, J.,… Goodwin, K. D., Jansson, J. K., Gilbert, J. A., Knight, R. and The Earth Microbiome Project Consortium. (2017). A communal catalogue reveals Earth’s multiscale microbial diversity. Nature, 551, 457–463.

Timpane-Padgham, B. L., Beechie, T., and Klinger, T. (2017). A systematic review of ecological attributes that confer resilience to climate change in environmental restoration. PLoS One 12, e0173812. Tol, S. J., Jarvis, J. C., York, P. H., Grech, A., Congdon, B. C., and Coles, R. G. (2017). Long distance biotic dispersal of tropical seagrass seeds by marine mega- herbivores. Scientific Reports 7, 4458.

Trevathan-Tackett, S. M., Kelleway, J., Macreadie, P. I., Beardall, J., Ralph, P., and Bellgrove, A. (2015). Comparison of marine macrophytes for their contributions to blue carbon sequestration. Ecology 96, 3043–3057.

Tujula, N. A., Crocetti, G. R., Burke, C., Thomas, T., Holmstrom, C. and Kjellebergm, S. (2010) Variability and abundance of the epiphytic bacterial community associated with a green marine Ulvacean alga. ISME J 4: 301–311.

Tuya, F., Vila, F., Bergasa, O., Zarranz, M., Espino, F., and Robaina, R. R. (2017). Artificial seagrass leaves shield transplanted seagrass seedlings and increase their survivorship. Aquatic Botany 136, 31–34.

172

Using genomics to restore and future-proof underwater seaweed forests

Tyagi, M., da Fonseca, M. M. R. and de Carvalho, C. C. C. R. (2011). Bioaugmentation and biostimulation strategies to improve the effectiveness of bioremediation processes. Biodegradation 22, 231–241.

Tylianakis, J. M., Didham, R. K., Bascompte, J., and Wardle, D. A. (2008). Global change and species interactions in terrestrial ecosystems. Ecology Letters 11, 1351– 1363. UN 2019. New UN Decade on Ecosystem Restoration offers unparalleled opportunity for job creation, food security and addressing climate change. In: Eisele, F. and Hwang, B. S. (eds.). FAO Liaison Office New York: The Food and Agriculture Organisation, UN Environment.

Underwood, A. J., Kingsford, M. J. and Andrew, N. L. (1991). Patterns in shallow subtidal marine assemblages along the coast of New South Wales. Australian Journal of Ecology, 16, 231-249. van Katwijk, M. M., Bos, A. R., de Jonge, V. N., Hanssen, L. S. A. M., Hermus, D. C. R., and de Jong, D. J. (2009). Guidelines for seagrass restoration: importance of habitat selection and donor population, spreading of risks, and ecosystem engineering effects. Marine Pollution Bulletin 58, 179–188. van Katwijk, M., Thorhaug, A., Marba, N., Orth, R., Duarte, C., Kendrick, G., Althuizen, I. H. J., Balestri, E., Bernard, G., Cambridge, M., Cunha, A., Durance, C., Giesen, W., Han, Q., Hosokawa, S., Kiswara, W., Komatsu, T., Lardicci, C., Lee, K.-S., and Verduin, J. (2016). Global analysis of seagrass restoration: the importance of large-scale planting. Journal of Applied Ecology 53, 567–578. van Oppen, M. J. H., Oliver, J. K., Putnam, H. M., and Gates, R. D. (2015). Building coral reef resilience through assisted evolution. Proceedings of the National Academy of Sciences of the United States of America 112, 2307–2313.

Vander Mijnsbrugge, K., Bischoff, A., and Smith, B. (2010). A question of origin: where and how to collect seed for ecological restoration. Basic and Applied Ecology 11, 300–311.

Vasquez, J. A., and McPeak, R. H. (1998). A new tool for kelp restoration. California Fish and Game 84, 149–158.

Venter, J. C., Remington, K., Heidelber, J. F. Halpern, A. L., Rusch, D., Eisen, J. A., Wu, D., Paulsen, I., Nelson, K. E., Nelson, W., Fouts, D. E., Levy, S., Knap, A. H., Lomas, M. W., Nealson, K., Nelson, White, O., Peterson, J., Hoffman, J., Parsons, R., Baden-Tillson, H., Pfannkoch, C., Rogers, Y. and Smith, H.O. (2004).

173

Using genomics to restore and future-proof underwater seaweed forests

Environmental Genome Shotgun Sequencing of the Sargasso Sea. Science, 304(5667), 66-74.

Verduin, J. J., Paling, E. I., van Keulen, M., and Rivers, L. E. (2012). Recovery of donor meadows of Posidonia sinuosa and Posidonia australis contributes to sustainable seagrass transplantation. International Journal of Ecology 2012, 837317.

Verdura, J., Sales, M., Ballesteros, E., Cefalì, M. A., and Cebrian, E. (2018). Restoration of a canopy-forming alga based on recruitment enhancement: methods and long-term success assessment. Frontiers of Plant Science 9, 1832.

Vergés, A., Campbell, A.H., Wood, G., Kajlich, L., Eger, A.M., Cruz, D., Langley, M., Bolton, D., Coleman, M.A., Turpin, J. and Crawford, M., 2020. Operation Crayweed: Ecological and sociocultural aspects of restoring Sydney’s underwater forests. Ecological Management & Restoration, 21(2), pp.74-85.

Vergés, A., Doropoulos, C., Malcolm, H. A., Skye, M., Garcia-Pizá, M., Marzinelli, E. M., Campbell, A. H., Ballesteros, E., Hoey, A. S., Vila-Concejo, A., Bozec, Y. M., and Steinberg, P. D. (2016). Long-term empirical evidence of ocean warming leading to tropicalization of fish communities, increased herbivory, and loss of kelp. Proceedings of the National Academy of Sciences of the United States of America 113, 13791–13796.

Vergés, A., Steinberg, P. D., Hay, M. E., Poore, A. G. B., Campbell, A. H., Ballesteros, E., Heck, K. L., Booth, D. J., Coleman, M. A., Feary, D. A., Figueira, W., Langlois, T., Marzinelli, E. Z., Mizerek, T., Mumby, P. J., Nakamura, Y., Roughan, M., van Sebille, E., Sen Gupta, A., Smale, D. A., Tomas, F., Wernberg, T., and Wilson, S. K. (2014). The tropicalisation of temperate marine ecosystems: climate-mediated changes in herbivory and community phase shifts. Proceedings of the Royal Society of London – B. Biological Sciences 281, 20140846.

Wahl, M., Goecke, F., Labes, A., Dobretsov, S., and Weinberger, F. (2012) The second skin: ecological role of epibiotic biofilms on marine organisms. Front Microbiol 3: 1–21.

Wang, G., Shuai, L., Li, Y., Lin, W., Zhao, X. and Duan, D. (2008) Phylogenetic analysis of epiphytic marine bacteria on Hole‐Rotten diseased sporophytes of Laminaria japonica. J Appl Phycol 20: 403–409. 174

Using genomics to restore and future-proof underwater seaweed forests

Wang, J., Thingholm L. B., Skiecevièienë J., Rausch P., Kummen M., Hov J. R., et al., (2016). Genome-wide association analysis identifies variation in vitamin d receptor and other host factors influencing the gut microbiota. Nat. Genet. 48 1396.

Waples, R. S., and Lindley, S. T. (2018). Genomics and conservation units: The genetic basis of adult migration timing in Pacific salmonids. Evolutionary applications, 11(9), 1518–1526. doi:10.1111/eva.12687

Watanuki, A., Aota, T., Otsuka, E., Kawai, T., Iwahashi, Y., Kuwahara, H., and Fujita, D. (2010). Restoration of kelp beds on an urchin barren: removal of sea urchins by citizen divers in southwestern Hokkaido. Bulletin of Fisheries Research 32, 83–87.

Water, S. (2017). Sydney Water annual report 2016–17. (Sydney Water.) Available at https://www.sydneywater.com.au/web/groups/publicwebcontent/documents/ document/zgrf/mty4/~edisp/dd_168714.pdf [Verified 24 February 2019].

Waters, C. N., Zalasiewicz, J., Summerhayes, C., Barnosky, A. D., Poirier, C., Gałuszka, A., Cearreta, A., Edgeworth, M., Ellis, E. C., Ellis, M., Jeandel, C., Leinfelder, R., Mcneill, J. R., Richter, D. D., Steffen, W., Syvitski, J., Vidas, D., Wagreich, M., Williams, M., Zhisheng, A., Grinevald, J., Odada, E., Oreskes, N., and Wolfe, A. P. (2016). The Anthropocene is functionally and stratigraphically distinct from the Holocene. Science 351, aad2622.

Waters, J.M. (2008), Marine biogeographical disjunction in temperate Australia: historical landbridge, contemporary currents, or both? Diversity and Distributions, 14: 692-700.

Waycott, M., Duarte, C. M., Carruthers, T. J. B., Orth, R. J., Dennison, W. C., Olyarnik, S., Calladine, A., Fourqurean, J. W., Heck, K. L., Hughes, A. R., Kendrick, G. A., Kenworthy, W. J., Short, F. T., and Williams, S. L. (2009). Accelerating loss of seagrasses across the globe threatens coastal ecosystems. Proceedings of the National Academy of Sciences of the United States of America 106, 12377–12381.

Webster, M. S., Colton, M. A., Darling, E. S., Armstrong, J., Pinsky, M. L., Knowlton, N., and Schindler, D. E. (2017). Who should pick the winners of climate change? Trends in Ecology and Evolution 32, 167–173.

Weeks, A. R., Sgro, C. M., Young, A. G., Frankham, R., Mitchell, N. J., Miller, K. A., Byrne, M., Coates, D. J., Eldridge, M. D., Sunnucks, P., Breed, M. F., James, E. A., and Hoffmann, A. A. (2011). Assessing the benefits and risks of translocations in changing environments: a genetic perspective. Evolutionary Applications 4, 709–725.

175

Using genomics to restore and future-proof underwater seaweed forests

Weigner, K. (2016). Facilitating modern genetic analysis of the habitat-forming macroalga, P. comosa. Honours, Southern Cross University.

Wernberg, T., Thomsen, M. S., Connell, S. D., Russell, B. D., Waters, J. M., Zuccarello, G. C., Kraft, G. T., Sanderson, C., West, J. A. and Gurgel, C. F., (2013). The footprint of continental-scale ocean currents on the biogeography of seaweeds. PloS one, 8(11).

Wernberg, T., Bennett, S., Babcock, R. C., De Bettignies, T., Cure, K., Depczynski, M., Dufois, F., Fromont, J., Fulton, C. J., Hovey, R. K. and Harvey, E. S., (2016). Climate-driven regime shift of a temperate marine ecosystem. Science, 353(6295), pp.169-172.

Wernberg, T., Coleman, M., Bennett, S., Thomsen, M. S., Tuya, F., and Kelaher, B. P. (2018). Genetic diversity and kelp forest vulnerability to climatic stress. Scientific Reports 8, 1851.

Wernberg, T., Krumhansl, K., Filbee-Dexter, K., and Pedersen, M. F. (2019a). Chapter 3 - Status and Trends for the World’s Kelp Forests. Pages 57-78 in C. Sheppard, editor. World Seas: an Environmental Evaluation (Second Edition). Academic Press.

Wernberg, T., Russell, B. D., Thomsen, M. S., Gurgel, C. F. D., Bradshaw, C. J., Poloczanska, E. S. and Connell, S. D., (2011). Seaweed communities in retreat from ocean warming. Current biology, 21(21), pp.1828-1832.

Wernberg, T., Thomsen, M. S., Tuya, F., Kendrick, G. A., Staehr, P. A., and Toohey, B. D. (2010). Decreasing resilience of kelp beds along a latitudinal temperature gradient: potential implications for a warmer future. Ecology Letters 13, 685–694.

Wernberg, T., Coleman, M. A., Babcock, R. C., Bell, S. Y., Bolton, J. J., Connell, S. D., Hurd, C. L., Johnson, C. R., Marzinelli, E. M., Shears, N. and Steinberg, P., (2018). Biology and ecology of the globally significant kelp Ecklonia radiata. Oceanography and Marine Biology: An Annual Review, 57. Westermeier, R., Patiño, D., Murúa, P., Muñoz, L., Ruiz, A., and Atero, C. (2013). Uso de algas pardas de cultivo para la biorremediación del ambiente costero en la Bahía de Chañaral. Informe final FIC 2011 33-01-211. (Universidad Austral de Chile: Copiapó, Chile.) Available at https://goreatacama.gob.cl/wp-content/uploads/08- 10-2013_17-36-24_10488986.pdf [Verified 18 February 2019].

Whiteley, A. R., Fitzpatrick, S. W., Funk, W. C. and Tallmon, D. A. (2015). Genetic rescue to the rescue. Trends in Ecology and Evolution, 30(1), 42-49.

176

Using genomics to restore and future-proof underwater seaweed forests

Whitham, T. G., Bailey, J. K., Schweitzer, J. A., Shuster, S. M., Bangert, R. K., … Wooley, S. C. (2006). A framework for community and ecosystem genetics: from genes to ecosystems. Nature Reviews – Genetics 7, 510–523.

Whitlock, M., Lotterhos, K., and Editor: Judith L. Bronstein. (2015). Reliable Detection of Loci Responsible for Local Adaptation: Inference of a Null Model through Trimming the Distribution of FST. The American Naturalist, 186(S1), S24-S36.

Wiens, J. A., and Hobbs, R. J. (2015). Integrating conservation and restoration in a changing world. Bioscience 65, 302–312.

Wilkins, L. G., Leray, M., O’Dea, A., Yuen, B., Peixoto, R. S., Pereira, T. J., Bik, H. M., Coil, D. A., Duffy, J. E., Herre, E. A. and Lessios, H. A., (2019). Host-associated microbiomes drive structure and function of marine ecosystems. PLoS biology, 17(11).

Williams, S. E., Shoo, L. O., Isaac, J. L., Hoffmann, A. A., and Langham, G. (2008). Towards an integrated framework for assessing the vulnerability of species to climate change. PLoS Biology 6, e325.

Williams, S. L. (2001). Reduced genetic diversity in eelgrass transplantations affects both population growth and individual fitness. Ecological Applications, 11, 1472-1488.

Williams, S. L., Ambo-Rappe, R., Sur, C., Abbott, J. M., and Limbong, S. R. (2017). Diversity enhances restoration. Proceedings of the National Academy of Sciences of the United States of America 114, 11986–11991.

Wilsonm, L. J., Weber, X. A., King, T. M. and Fraser, C. I. (2016). DNA extraction techniques for genomic analyses of macroalgae. In: Seaweed Phylogeography: Adaptation and Evolution of Seaweeds under Environmental Change. (Ed.s Zi- Min, H and Fraser, C). Springer: Beijing. pp 363-386.

Wilson, A. M. W., and Forsyth, C. (2018). Restoring near-shore marine ecosystems to enhance climate security for island ocean states: aligning international processes and local practices. Marine Policy 93, 284–294.

Wilson, K. A., Lulow, M., Burger, J., Fang, Y. C., Andersen, C., Olson, D., O’Connell, M., and McBride, M. F. (2011). Optimal restoration: accounting for space, time and uncertainty. Journal of Applied Ecology 48, 715–725.

177

Using genomics to restore and future-proof underwater seaweed forests

Wood, A. R., Esko, T., Yang, J., Vedantam, S., Pers, T. H., Gustafsson, S., Chu, A. Y., Estrada, K., Kutalik, Z., Amin, N. and Buchkovich, M. L., (2014). Defining the role of common variation in the genomic and biological architecture of adult human height. Nature genetics, 46(11), p.1173.

Wood, G., Marzinelli, E. M., Coleman, M. A., Campbell, A. H., Santini, N. S., Kajlich, L., Verdura, J., Wodak, J., Steinberg, P. D. and Vergés, A. (2019). Restoring subtidal marine macrophytes in the Anthropocene: trajectories and future-proofing. Marine and Freshwater Research, 70, 936-951.

Wormersley, H. B. S. (1987). The marine benthic flora of Southern Australia. Part 2. Southern Australian Government Printing Division, Adelaide, Australia.

Wylie, L., Sutton-Grier, A. E., and Moore, A. (2016). Keys to successful carbon projects: lessons learned from global case studies. Marine Policy 65, 76–84.

Yang, S. L., Shi, B. W., Bouma, T. J., Ysebaert, T. & Luo, X. X. 2012. Wave attenuation at a salt-marsh margin: A case study of an exposed coast on the Yangtze estuary. Estuar. Coasts 35, 169–182.

Yoon, J. T., Sun, S. M., and Chung, G. (2014). Sargassum bed restoration by transplantation of germlings grown under protective mesh cage. Journal of Applied Phycology 26, 505–509.

Zalasiewicz, J., Williams, M., Smith, A., Barry, T. L., Coe, A. L., Bown, P. R., Brenchley, P., Cantrill, D., Gale, A., Gibbard, P…M., Rawson, P., and Stone, P. (2008). Are we now living in the Anthropocene? GSA Today 18, 4–8.

Zerebecki, R. A., Crutsinger, G. M., Hughes, A. R., and Dam, N. (2017). Spartina alterniflora genotypic identity affects plant and consumer responses in an experimental marsh community. Journal of Ecology 105, 661–673.

Zhang, Y., Cioffi, W., Cope, R., Daleo, P., Heywood, E., Hoyt, C., Smith, C., and Silliman, B. (2018). A Global synthesis reveals gaps in coastal habitat restoration research. Sustainability 10, 1040.

Zheng X, Levine D, Shen J, Gogarten S, Laurie C, Weir B (2012). “A High- performance Computing Toolset for Relatedness and Principal Component Analysis of SNP Data.” Bioinformatics, 28(24), 3326-3328.

Ziegler, M., Seneca, F. O., Yum, L. K., Palumbi, S. R., and Voolstra, C. R. (2017). Bacterial community dynamics are linked to patterns of coral heat tolerance. Nat. Commun. 8:14213.

Zilber-Rosenberg, I. and Rosenberg, E. (2008). Role of microorganisms in the evolution of animals and plants: the hologenome theory of evolution. FEMS Microbiology Reviews, 32: 5, 723–735. 178

Using genomics to restore and future-proof underwater seaweed forests

Appendix A: Supplementary material to accompany Chapter 3

179

Using genomics to restore and future-proof underwater seaweed forests

Modified DNA extraction protocol DNA was extracted using the Qiagen DNeasy Plant Mini kit with the following modifications to increase yield and quality: 450 ul of lysis buffer and 150 ul of ispropanol were added during the cell lysis step and incubated for 12-24 hours before 195ul of the neutralisation buffer was added. An extra wash step with 95% etOH was included and DNA was then incubated for 30 minutes at room temperature before elution. DNA was extracted over 2-4 reactions and pooled for each algal sample, then cleaned using a Qiagen PowerClean Pro Cleanup kit with an added wash step with 95% etOH. DNA was then incubated for 30 minutes at room temperature before elution.

180

Using genomics to restore and future-proof underwater seaweed forests

(a) 1600 1400 1200 1000 800

600

400 Wet weight biomass (g) biomass weight Wet 200 0 BB TE PB CR SP SH

(b)

12 )

2 10 8

25mm (per 6 4 2

Density of reproductive conceptacles ofreproductive Density 0 BB TE PB CR SP SH

Donor Site

Fig. S3.1: Average (a) wet weight biomass and (b) density of reproductive conceptacles of individuals sampled at six extant sites surrounding Sydney. BB: Bateau Bay, TE: Terrigal, PB: Palm Beach, CR: Cronulla, SP: Shark Park, SH: Shell Harbour. Coloured bars represent donor sites. In panel (b), dark columns indicate males and light colours indicate females.

181

Using genomics to restore and future-proof underwater seaweed forests

Table S3.1: Population structure (pairwise FST) among extant Phyllospora populations.*

Site a BB TE PB CR SP SH

BB - TE 0.022 - PB 0.025 0.048 - CR 0.029 0.035 0.022 - SP 0.044 0.058 0.037 0.042 - SH 0.085 0.100 0.061 0.072 0.072 -

* All values significant. a BB: Bateau Bay, TE: Terrigal, PB: Palm Beach, CR: Cronulla, SP: Shark Park, SH: Shell Harbour. b Pairwise estimates between sites from north versus the south of the distributional gap shown in grey.

182

Using genomics to restore and future-proof underwater seaweed forests

Table S3.2: AMOVA between extant Phyllospora sites

Source of variation df SS Variance % variation* component* Among clusters 2 421.5 1.19 3.15 Among sites within 3 300.7 1.13 2.99 clusters Among samples 171 5988.72 - 0.36 - 0.95 within sites Within samples 177 6325.73 35.73 94.81

*Significant values shown in bold.

183

Using genomics to restore and future-proof underwater seaweed forests

Appendix B:

Supplementary material to accompany Chapter 4

184

Using genomics to restore and future-proof underwater seaweed forests

Table S4.1: Proportions of males and female Phyllsopora comosa adults sampled at 13 sites.

Site Males (%) Females (%) Port Macquarie 80 20 Forster 60 40 Anna Bay 50 50 Bateau Bay 52 48 Terrigal 54 46 Palm Beach 50 50 Cronulla 38 62 Shark Park 57 43 Shellharbour 30 70 Malua Bay 55 45 Eden 30 70 Bicheno 63 37 Southport 45 55

185

Using genomics to restore and future-proof underwater seaweed forests

Table S4.2: Private alleles in unfiltered Phyllsopora comosa data from adults sampled at 13 sites.

Site Total Locus 125545_un 12662_un 138584_un 163100_un 22701_un 31844_un 35503_un 37225_un 40745_un 43676_un 70284_un 91778_un PM 2 0 1 0 0 0 0 0 2 0 0 0 0 FO 0 0 0 0 0 0 0 0 0 0 0 0 0 AB 2 0 0 0 0 0 1 0 0 0 0 1 0 BB 0 0 0 0 0 0 0 0 0 0 0 0 0 TE 1 0 0 0 0 0 0 1 0 0 0 0 0 PB 0 0 0 0 0 0 0 0 0 0 0 0 0 CR 1 0 0 0 0 0 0 0 0 2 0 0 0 SP 4 2 0 0 1 2 0 0 0 0 0 0 4 SH 1 0 0 0 0 0 0 0 0 0 3 0 0 MB 1 0 0 1 0 0 0 0 0 0 0 0 0 ED 0 0 0 0 0 0 0 0 0 0 0 0 0 BI 0 0 0 0 0 0 0 0 0 0 0 0 0 SO 0 0 0 0 0 0 0 0 0 0 0 0 0 * Number of individuals with expressed private allele in bold.

aPM: Port Macquarie; FO: Forster; AB: Anna Bay: BB: Bateau Bay: TE: Terrigal: PB: Palm Beach; CR: Cronulla; SP: Shark Park; SH: Shellharbour; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

186

Using genomics to restore and future-proof underwater seaweed forests

Table S4.3: Comparison of observed heterozygosity (HO) estimates and sample size for Phyllsopora comosa adults sampled at 13 sites.

Site Rank n HO Southport 13 19 0.175 Port Macquarie 12 20 0.175 Forster 11 20 0.201 Bicheno 10 20 0.233 Anna Bay 9 20 0.248 Shark Park 8 29 0.281 Palm Beach 7 29 0.295 Eden 6 20 0.308 Shell Harbour 5 30 0.309 Bateau bay 4 26 0.327 Terrigal 3 29 0.328 Cronulla 2 49 0.33 Malua Bay 1 20 0.338

187

Using genomics to restore and future-proof underwater seaweed forests

Fig. S4.1: Principal Component Analysis (PCA) of the seaweed Phyllospora comosa’s genetic structure, based on allele frequencies at 109 SNP loci. Sites North to South: PM: Port Macquarie; FO: Forster; AB: Anna Bay: BB: Bateau Bay; TE: Terrigal; PB: Palm Beach; CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

188

Using genomics to restore and future-proof underwater seaweed forests

Table S4.4: Phyllospora comosa’s putatively adaptive loci identified with different methods.

Bayescan Pcadapt Overlapping Identified by multiple

SST-associated loci methods 20189_un_2826394 17842_un_2497795 102358_un_14330055 20189_un_2826394 8588_un_1202259 19803_un_2772369 10554_un_1477479 8588_un_1202259 29052_un_4067206 50062_un_7008617 29052_un_4067206 109470_un_15325751 69129_un_9677972 17842_un_2497795 12378_un_1732835 90233_un_12632545 50062_un_7008617 13493_un_1888940 92542_un_12955815 92542_un_12955815 150914_un_21127894 92850_un_12998925 15622_un_2187006 9777_un_1368712 17842_un_2497795

18198_un_2547658

20122_un_2817001

20189_un_2826394

21052_un_2947210

2396_un_335353

28359_un_3970173

28834_un_4036686

29052_un_4067206

3682_un_515415

40009_un_5601211

40713_un_5699768

42507_un_5950921

43432_un_6080431

44027_un_6163695

45748_un_6404632

4755_un_665635

47703_un_6678358

48949_un_6852791

50062_un_7008617

50239_un_7033372

53085_un_7431836

54483_un_7627558

68250_un_9554917

74449_un_10422773

8588_un_1202259

92542_un_12955815

97284_un_13619681 9866_un_1381175

189

Using genomics to restore and future-proof underwater seaweed forests

Table S4.5.1: Population structure (pairwise Fst) among extant Phyllospora populations based on outlier loci identified using Bayescan (3) *

Site a PM FO AB BB TE PB CR SP SH MB ED BI SO PM FO 0.500 AB -0.010 0.419 BB 0.335 0.027 0.265 TE 0.150 0.140 0.087 0.051 PB 0.314 0.035 0.236 0.006 0.025 CR 0.329 0.027 0.257 0.003 0.040 -0.010 SP 0.354 0.018 0.277 0.005 0.047 -0.007 -0.007 SH 0.147 0.157 0.086 0.057 -0.010 0.031 0.048 0.066 MB 0.016 0.529 0.030 0.382 0.192 0.365 0.381 0.390 0.213 ED 0.612 0.795 0.584 0.696 0.575 0.692 0.676 0.681 0.628 0.479 BI 0.612 0.787 0.587 0.684 0.572 0.686 0.670 0.673 0.624 0.487 0.020 SO 0.658 0.837 0.625 0.740 0.611 0.727 0.707 0.719 0.664 0.524 0.085 0.232

* Significant values in bold.

aPM: Port Macquarie; FO: Forster; AB: Anna Bay: BB: Bateau Bay: TE: Terrigal: PB: Palm Beach; CR: Cronulla; SP: Shark Park; SH: Shellharbour; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

190

Using genomics to restore and future-proof underwater seaweed forests

Table S4.5.2: Population structure (pairwise Fst) among extant Phyllospora populations based on outlier loci (8) identified using PCadapt *

Site a PM FO AB BB TE PB CR SP SH MB ED BI SO PM FO 0.375 AB 0.197 0.385 BB 0.318 0.151 0.375 TE 0.249 0.150 0.365 0.046 PB 0.298 0.171 0.306 0.064 0.071 CR 0.254 0.243 0.364 0.059 0.058 0.056 SP 0.274 0.200 0.200 0.100 0.135 0.036 0.141 SH 0.301 0.081 0.287 0.045 0.069 0.022 0.116 0.028 MB 0.256 0.127 0.172 0.082 0.130 0.055 0.154 -0.003 0.014 ED 0.206 0.409 0.344 0.474 0.409 0.494 0.490 0.446 0.427 0.388 BI 0.416 0.453 0.301 0.403 0.350 0.324 0.420 0.195 0.276 0.223 0.533 SO 0.471 0.119 0.426 0.355 0.303 0.307 0.421 0.279 0.190 0.220 0.458 0.425

* Significant values in bold.

aPM: Port Macquarie; FO: Forster; AB: Anna Bay: BB: Bateau Bay: TE: Terrigal: PB: Palm Beach; CR: Cronulla; SP: Shark Park; SH: Shellharbour; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

191

Using genomics to restore and future-proof underwater seaweed forests

Table S4.5.3: Population structure (pairwise Fst) among extant Phyllospora populations based on SST-associated loci (36) identified using lfmm *

Site a PM FO AB BB TE PB CR SP SH MB ED BI SO PM FO 0.110 AB 0.031 0.092 BB 0.069 0.054 0.033 TE 0.072 0.066 0.039 0.005 PB 0.089 0.082 0.039 0.011 0.018 CR 0.072 0.061 0.033 0.002 0.011 0.003 SP 0.101 0.069 0.050 0.009 0.009 0.009 0.003 SH 0.126 0.118 0.082 0.033 0.017 0.031 0.030 0.021 MB 0.103 0.127 0.082 0.060 0.046 0.061 0.056 0.066 0.039 ED 0.202 0.183 0.162 0.153 0.129 0.168 0.155 0.159 0.150 0.092 BI 0.201 0.186 0.167 0.159 0.118 0.166 0.156 0.167 0.144 0.107 0.090 SO 0.203 0.217 0.194 0.204 0.158 0.215 0.195 0.203 0.200 0.166 0.108 0.075

* Significant values in bold.

aPM: Port Macquarie; FO: Forster; AB: Anna Bay: BB: Bateau Bay: TE: Terrigal: PB: Palm Beach; CR: Cronulla; SP: Shark Park; SH: Shellharbour; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

192

Using genomics to restore and future-proof underwater seaweed forests

Table S4.6: Test statistics for pairwise comparisons of observed heterozygosity estimates between sites.

Site a PM FO AB BB TE PB CR SP SH MB ED BI SO PM FO -0.013 AB -0.089 -0.076 BB -0.146 -0.134 -0.057 TE -0.156 -0.143 -0.067 -0.009 PB -0.133 -0.12 -0.044 0.014 0.023 CR -0.143 -0.13 -0.054 0.003 0.013 -0.01 SP -0.119 -0.107 -0.031 0.027 0.037 0.013 0.024 SH -0.119 -0.107 -0.031 0.027 0.036 0.012 0.024 0 MB -0.152 -0.139 -0.063 -0.006 0.004 -0.019 -0.009 -0.033 -0.032 ED -0.112 -0.099 -0.024 0.034 0.044 0.042 0.03 0.007 0.007 0.0399 BI -0.047 -0.034 0.042 0.099 0.109 0.086 0.096 0.072 0.072 0.105 0.065 SO 0.012 0.025 0.1 0.158 0.168 0.145 0.155 0.131 0.131 0.164 0.124 0.059

* Significant values following FDR correction in bold.

aPM: Port Macquarie; FO: Forster; AB: Anna Bay: BB: Bateau Bay: TE: Terrigal: PB: Palm Beach; CR: Cronulla; SP: Shark Park; SH: Shellharbour; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

193

Using genomics to restore and future-proof underwater seaweed forests

(a) (b)

(c)

Figure S4.2: Relationship between geographic distance (km) and genetic distance

(pairwise Fst) between Phyllospora individuals for selective loci datasets; a) outlier loci identified using Bayescan; b) outlier loci identified using PCAdapt and c) sea surface temperature-associated loci. Data collected at thirteen sites and plots fitted with linear regression (fitted values: blue line; 95% Confidence Intervals: grey shade).

194

Using genomics to restore and future-proof underwater seaweed forests

Table S4.7.1: Genetic diversity of thirteen Phyllospora populations across its latitudinal range, based on outlier loci (3) identified using Bayescan.

a b c n Global FST HO HE AR Port Macquarie 20 0.267 0.259 1.667 Forster 20 0.200 0.172 1.667 Anna Bay 20 0.367 0.317 1.983 Bateau Bay 26 0.295 0.280 1.977 Terrigal 29 0.373 0.378 1.999 Palm Beach 29 0.276 0.284 1.886 Cronulla 49 0.410 0.367 0.293 1.955 Shark Park 29 0.351 0.299 1.999 Shellharbour 30 0.348 0.328 1.667 Malua Bay 20 0.267 0.312 2.000 Eden 20 0.167 0.160 1.333 Bicheno 20 0.167 0.165 1.333 Southport 19 0.123 0.100 1.333

a Observed heterozygosity (HO) b Expected heterozygosity (HE) cRarefied allelic richness (AR)

195

Using genomics to restore and future-proof underwater seaweed forests

Table S4.7.2: Genetic diversity of thirteen Phyllospora populations across its latitudinal range, based outlier loci identified using PCAdapt (8).

a b c n Global FST HO HE AR Port Macquarie 20 0.478 0.428 2.000 Forster 20 0.354 0.306 1.987 Anna Bay 20 0.232 0.290 1.862 Bateau Bay 26 0.319 0.295 1.864 Terrigal 29 0.399 0.363 1.828 Palm Beach 29 0.276 0.304 1.976 Cronulla 49 0.253 0.339 0.319 1.987 Shark Park 29 0.308 0.310 1.991 Shellharbour 30 0.345 0.333 1.992 Malua Bay 20 0.332 0.347 1.999 Eden 20 0.258 0.338 1.875 Bicheno 20 0.181 0.149 1.600 Southport 19 0.197 0.208 1.618

a Observed heterozygosity (HO) b Expected heterozygosity (HE) cRarefied allelic richness (AR)

196

Using genomics to restore and future-proof underwater seaweed forests

Table S4.7.3: Genetic diversity of thirteen Phyllospora populations across its latitudinal range, based on SST-associated loci (36).

a b c n Global FST HO HE AR Port Macquarie 20 0.340 0.197 1.513 Forster 20 0.376 0.217 1.597 Anna Bay 20 0.321 0.202 1.654 Bateau Bay 26 0.375 0.247 1.761 Terrigal 29 0.339 0.238 1.811 Palm Beach 29 0.360 0.236 1.732 Cronulla 49 0.058 0.367 0.246 1.777 Shark Park 29 0.345 0.230 1.685 Shellharbour 30 0.342 0.238 1.797 Malua Bay 20 0.364 0.258 1.907 Eden 20 0.331 0.231 1.822 Bicheno 20 0.326 0.228 1.767 Southport 19 0.285 0.180 1.629

aObserved heterozygosity (HO) b Expected heterozygosity (HE) cRarefied allelic richness (AR)

197

Using genomics to restore and future-proof underwater seaweed forests

Appendix C:

Supplementary material to accompany Chapter 5

198

Using genomics to restore and future-proof underwater seaweed forests

Figure S5.1: Figure depicting traits measured or characterised for each of 156 Phyllospora comosa individuals. (A) Host genetics; (B) phenotypic traits, including (i) maximum photosynthetic quantum yield, (ii) thallus circumference, (iii) frond width, (iv) primary length, (v) total length, (vi) wet weight, (vii) number of vesicles, (viii) stipe base length, (ix) stipe base width, (x) number of branches, (xi) density of reproductive conceptacles, (xii) presence of stipe rot disease, (xiii) percentage of thallus bleached, (xiv) percentage of thallus fouled by epibionts, (xv) scaled presence of grazing, (xvi) sex and (C) microbial communities.

199

Using genomics to restore and future-proof underwater seaweed forests

Table S5.1: Adjusted p-values for pairwise comparisons of host genetic data between 8 sites.*

Site a AB CR ED FO MB BI PM CR 0.001 ------ED 0.001 0.001 - - - - - FO 0.001 0.001 0.001 - - - - MB 0.001 0.001 0.001 0.001 - - - BI 0.001 0.001 0.001 0.001 0.001 - - PM 0.001 0.001 0.001 0.001 0.001 0.001 - SO 0.001 0.001 0.001 0.001 0.001 0.001 0.001

* Significant values in bold. aPM: Port Macquarie; FO: Forster; AB: Anna Bay; CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

200

Using genomics to restore and future-proof underwater seaweed forests

AB CR ED FO MB BI PM SO

Figure S5.1: Dispersion of Phyllospora comosa genetic data for 8 sites. PM: Port Macquarie; FO: Forster; AB: Anna Bay; CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

201

Using genomics to restore and future-proof underwater seaweed forests

Table S5.2: Results of pairwise comparisons of dispersion in host genetic data between 8 sites. *

Site a AB CR ED FO MB BI PM CR 0.999 ------ED 1.000 0.987 - - - - - FO <0.001 <.001 0.002 - - - - MB 0.962 0.999 0.843 <.001 - - - BI 0.278 0.083 0.469 0.149 0.018 - - PM <.001 <.001 <.0001 0.999 <.001 0.038 - SO <.001 <.001 <.0001 0.957 <.001 0.007 0.999

* Significant values in bold. aPM: Port Macquarie; FO: Forster; AB: Anna Bay; CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

202

Using genomics to restore and future-proof underwater seaweed forests

Table S5.3: Adjusted p-values for pairwise comparisons of host phenotype data between 8 sites.*

Site a AB CR ED FO MB BI PM CR 0.001 ------ED 0.001 0.001 - - - - - FO 0.001 0.001 0.001 - - - - MB 0.001 0.001 0.249 0.013 - - - BI 0.001 0.001 0.001 0.001 0.001 - - PM 0.001 0.001 0.001 0.124 0.010 0.001 - SO 0.001 0.001 0.001 0.001 0.001 0.001 0.001

* Significant values in bold. aPM: Port Macquarie; FO: Forster; AB: Anna Bay; CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

203

Using genomics to restore and future-proof underwater seaweed forests

AB CR ED FO MB BI PM SO

Figure S5.2: Dispersion of Phyllospora comosa phenotype data for 8 sites. PM: Port Macquarie; FO: Forster; AB: Anna Bay; CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

204

Using genomics to restore and future-proof underwater seaweed forests

Table S5.4: Results of pairwise comparisons of dispersion in host phenotype data between 8 sites. *

Site a AB CR ED FO MB BI PM CR 0.1587 ------ED 0.9474 0.8002 - - - - - FO 1 0.3129 0.9941 - - - - MB 0.5496 0.0003 0.0511 0.3004 - - - BI 0.8556 0.0021 0.1748 0.6243 0.9995 - - PM 1 0.1804 0.964 1 0.4713 0.7999 - SO 1 0.1866 0.9634 1 0.4992 0.8189 1

* Significant values in bold. aPM: Port Macquarie; FO: Forster; AB: Anna Bay; CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

205

Using genomics to restore and future-proof underwater seaweed forests

Figure S5.3: Plots of all phenotypic traits measured, ordered by site. Y-axes are labelled with the description of traits (not replicated here, for simplicity). Dark red plots indicate traits that were significantly associated with microbial community structure. Sites ordered north to south; PM: Port Macquarie; FO: Forster; AB: Anna Bay: CR: Cronulla; MB: Malua Bay; ED: Eden; NT: Bicheno; ST: Southport

206

Using genomics to restore and future-proof underwater seaweed forests

Table S5.5: Results of tests for differences between individual host phenotype data between 8 sites. *

Test Trait d.f Stat p PAM 7,142 F = 10.62 1E-10 * Thallus circumference 7,147 F = 8.13 2E-08 * Frond width 7,148 F = 11.56 1E-11 * 3.25E- F = 8.99 Log10 Primary length 7,148 09 * < F = 54.73 2.2e- Log10 Total length 7,148 16 * Wet weight 7,148 F = 5.909 5E-06 * lm Vesicles per 100g 7,145 F = 11.03 4E-11 * Stipe base length 7,148 F = 13.52 2E-13 * Stipe base width 7,148 F = 6.508 1E-06 * Thallus bleaching 7,147 F = 6.609 9E-07 * Thallus fouling 7,146 F = 8.655 7E-09 * Herbivory 7,147 F = 10.04 3E-10 * < Reproductive F = 51.93 2.2e- conceptacles 7,146 16 * Number of branches 7,148 F = 1.958 6E-02

Trait d.f glm Proportion with stipe rot 7,143 Dev =24.933 8E-04 * Proportion of males 7,145 Dev = 20.017 6E-03 *

* Significant values in bold.

207

Using genomics to restore and future-proof underwater seaweed forests

Table S5.6: Spearman correlation coefficient for all phenotypic traits measured.

a b c d e f g h i j k l m n o p Latitudea PAMb 0.210 Circumferencec -0.371 -0.271 Frond widthd -0.253 -0.237 0.115 Primary lengthe 0.430 0.103 -0.034 -0.170 Total lengthf 0.391 0.198 0.040 -0.243 0.688 Wet weightg -0.044 0.011 0.627 -0.093 0.376 0.557 Vesicles per 0.356 -0.110 -0.039 -0.146 0.383 0.520 0.147 100gh Stipe base 0.481 0.251 -0.243 -0.256 0.509 0.572 0.200 0.359 lengthi Stipe base 0.044 -0.011 0.081 -0.061 0.210 0.199 0.187 -0.134 0.022 widthj # branchesk -0.046 -0.006 0.242 -0.054 0.417 0.222 0.299 0.246 -0.014 0.071 Conceptaclesl 0.806 0.182 -0.312 -0.122 0.414 0.432 0.002 0.370 0.524 0.084 -0.066 Stipe rotm 0.044 0.093 -0.050 -0.135 -0.111 -0.133 -0.133 -0.096 -0.003 0.080 -0.002 0.016 Bleachingn -0.301 -0.041 0.008 0.142 -0.247 -0.447 -0.168 -0.330 -0.365 -0.053 -0.058 -0.368 0.152 Foulingo 0.066 0.039 -0.072 -0.125 -0.047 -0.174 -0.036 -0.301 -0.102 0.101 0.050 0.000 0.159 0.365 Herbivoryp 0.111 0.323 -0.366 0.000 -0.227 -0.243 -0.498 0.067 -0.115 0.036 -0.024 -0.002 0.320 0.125 0.031 Sex -0.168 0.060 0.173 0.040 0.256 0.215 0.163 0.016 -0.074 0.166 0.240 -0.192 0.027 -0.031 -0.094 0.020

* Significant values in bold.

208

Using genomics to restore and future-proof underwater seaweed forests

Table S5.7: List of “core” microbial taxa. Each amplicon sequence variant (ASV) was present in all 156 individuals of the seaweed Phyllospora comosa sampled during a study spanning their entire latitudinal distribution.

ASV ID Taxonomic details ASV1 Proteobacteria Alphaproteobacteria Caulobacterales Hyphomonadaceae uncultured ASV10 Proteobacteria Gammaproteobacteria Arenicellales Arenicellaceae Arenicella ASV13 Verrucomicrobia Verrucomicrobiae Verrucomicrobiales Rubritaleaceae Rubritalea ASV141 Planctomycetes Planctomycetacia Pirellulales Pirellulaceae Blastopirellula ASV15 Proteobacteria Gammaproteobacteria Arenicellales Arenicellaceae Arenicella ASV155 Proteobacteria Alphaproteobacteria Caulobacterales Hyphomonadaceae Hellea ASV182 Planctomycetes Planctomycetacia Pirellulales Pirellulaceae Blastopirellula ASV2 Cyanobacteria Oxyphotobacteria ASV23 Proteobacteria Alphaproteobacteria Rhodobacterales Rhodobacteraceae uncultured ASV3 Planctomycetes Planctomycetacia Pirellulales Pirellulaceae Blastopirellula ASV31 Proteobacteria Gammaproteobacteria Arenicellales Arenicellaceae Arenicella ASV351 Proteobacteria Alphaproteobacteria Caulobacterales Hyphomonadaceae Hellea ASV4 Planctomycetes Planctomycetacia Pirellulales Pirellulaceae Blastopirellula ASV43 Planctomycetes Planctomycetacia Pirellulales Pirellulaceae Blastopirellula ASV44 Verrucomicrobia Verrucomicrobiae Verrucomicrobiales Rubritaleaceae Rubritalea ASV5 Proteobacteria Alphaproteobacteria Caulobacterales Hyphomonadaceae Hellea ASV529 Proteobacteria Alphaproteobacteria Caulobacterales Hyphomonadaceae Litorimonas ASV57 Proteobacteria Alphaproteobacteria Caulobacterales Hyphomonadaceae Hellea ASV6 Proteobacteria Alphaproteobacteria Caulobacterales Hyphomonadaceae Litorimonas ASV68 Proteobacteria Alphaproteobacteria Caulobacterales Hyphomonadaceae uncultured ASV7 Proteobacteria Gammaproteobacteria Thiohalorhabdales Thiohalorhabdaceae Granulosicoccus ASV8 Planctomycetes Planctomycetacia Pirellulales Pirellulaceae Blastopirellula ASV9 Proteobacteria Gammaproteobacteria Thiohalorhabdales Thiohalorhabdaceae Granulosicoccus

209

Using genomics to restore and future-proof underwater seaweed forests

Table S5.8: Adjusted p-values for pairwise comparisons of microbial community data between 8 sites, based on (a) Bray-Curtis dissimilarity matrix based on overall microbial data and (b) Jaccard distances based on overall microbial data. *

(a) Site a AB CR ED FO MB BI PM CR 0.001 ------ED 0.001 0.001 - - - - - FO 0.001 0.001 0.001 - - - - MB 0.001 0.001 0.001 0.001 - - - BI 0.001 0.001 0.001 0.001 0.001 - - PM 0.001 0.001 0.001 0.001 0.001 0.001 - SO 0.001 0.001 0.001 0.001 0.001 0.001 0.001

(b) Site a AB CR ED FO MB BI PM CR 0.001 ------ED 0.001 0.001 - - - - - FO 0.001 0.001 0.001 - - - - MB 0.001 0.001 0.001 0.001 - - - BI 0.001 0.001 0.001 0.001 0.001 - - PM 0.001 0.001 0.001 0.001 0.001 0.001 - SO 0.001 0.001 0.001 0.001 0.001 0.001 0.001

* Significant values in bold. aPM: Port Macquarie; FO: Forster; AB: Anna Bay; CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

210

Using genomics to restore and future-proof underwater seaweed forests

(a) (b)

Figure S5.4: Dispersion of Phyllospora comosa microbial community data for 8 sites, based on (a) Bray-Curtis dissimilarity matrix based on overall microbial data and (b) Jaccard distances based on overall microbial data. PM: Port Macquarie; FO: Forster; AB: Anna Bay: CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

211

Using genomics to restore and future-proof underwater seaweed forests

Table S5.9: Results of pairwise comparisons of dispersion in microbial community data between 8 sites, based on (a) Bray-Curtis dissimilarity matrix based on overall microbial data and (b) Jaccard distances based on overall microbial data. *

(a) Site a AB CR ED FO MB BI PM CR 0.2374 ------ED 0.9046 0.9379 - - - - - FO 0.9988 0.0521 0.5464 - - - - MB 0.9204 0.0082 0.1909 0.9981 - - - BI 0.9395 0.0093 0.2115 0.9991 1.0000 - - PM 1.0000 0.1583 0.8266 0.9998 0.9583 0.9707 - SO 0.7138 0.9945 0.9999 0.3094 0.0815 0.0912 0.5934

(b) Site a AB CR ED FO MB BI PM CR 0.1825 ------ED 0.8333 0.9476 - - - - - FO 0.9986 0.0347 0.424 - - - - MB 0.9231 0.0055 0.1343 0.9985 - - - BI 0.9780 0.0117 0.2262 1.0000 1.0000 - - PM 1.0000 0.1059 0.7045 0.9999 0.9679 0.9941 - SO 0.6147 0.9955 0.9999 0.2251 0.0558 0.1024 0.4613

* Significant values in bold. aPM: Port Macquarie; FO: Forster; AB: Anna Bay; CR: Cronulla; MB: Malua Bay; ED: Eden; BI: Bicheno; SO: Southport.

212

Using genomics to restore and future-proof underwater seaweed forests

Table S5.10: Microbial taxa significantly related to geography, host phenotypic and genetic traits shown to influence overall microbial community, as identified using DEseq2.

Variable Numb Abundant Taxa Pairwise test er of associated result associ ASVs (SNPs only) ated ASVs Bacteria; Proteobacteria; Alphaproteobacteria; Caulobacterales; Site 1199 ASV1 Hyphomonadaceae

Bacteria; Cyanobacteria; ASV2 Oxyphotobacteria; Bacteria; Planctomycetes; Planctomycetacia; Pirellulales; Pirellulaceae; Blastopirellula; ASV3 uncultured Bacteria; Planctomycetes; Planctomycetacia; Pirellulales; Pirellulaceae; Blastopirellula; ASV4 uncultured Bacteria; Proteobacteria; Alphaproteobacteria; Caulobacterales; Hyphomonadaceae; Hellea; ASV5 uncultured

Bacteria; Proteobacteria; Alphaproteobacteria; Maximum Caulobacterales; quantum Hyphomonadaceae; Litorimonas; yield 17 ASV6 uncultured Bacteria; Bacteroidetes; Bacteroidia; Chitinophagales; Saprospiraceae; Rubidimonas; ASV21 uncultured Bacteria; Proteobacteria; Gammaproteobacteria; Thiohalorhabdales; Thiohalorhabdaceae; ASV50 Granulosicoccus; uncultured Bacteria; Proteobacteria; Alphaproteobacteria; Caulobacterales; Hyphomonadaceae; Fretibacter; ASV61 Hyphomonadaceae Bacteria; Bacteroidetes; Bacteroidia; Chitinophagales; Saprospiraceae; Rubidimonas; ASV89 uncultured 213

Using genomics to restore and future-proof underwater seaweed forests

Bacteria; Proteobacteria; Alphaproteobacteria; Caulobacterales; Hyphomonadaceae; Hellea; Herbivory 13 ASV5 uncultured Bacteria; Proteobacteria; Alphaproteobacteria; Caulobacterales; Hyphomonadaceae; Hellea; ASV57 uncultured

Bacteria; Cyanobacteria; ASV54 Oxyphotobacteria Bacteria; Verrucomicrobia; Verrucomicrobiae; Verrucomicrobiales; Rubritaleaceae; Rubritalea; ASV62 uncultured Bacteria; Bacteroidetes; Bacteroidia; Flavobacteriales; Flavobacteriaceae; ASV2162 Tenacibaculum; uncultured

Bacteria; Proteobacteria; Alphaproteobacteria; Rhizobiales; Stipe base Rhizobiaceae; uncultured; length 31 ASV11 uncultured Bacteria; Planctomycetes; Planctomycetacia; Pirellulales; Pirellulaceae; Blastopirellula; ASV18 Planctomycetaceae Bacteria; Proteobacteria; Gammaproteobacteria; Thiohalorhabdales; Thiohalorhabdaceae; ASV28 Granulosicoccus; uncultured Bacteria; Planctomycetes; Planctomycetacia; Pirellulales; Pirellulaceae; Blastopirellula; ASV35 Planctomycetaceae Bacteria; Proteobacteria; Alphaproteobacteria; Caulobacterales; Hyphomonadaceae; Fretibacter; ASV61 Hyphomonadaceae

Locus Bacteria; Proteobacteria; 28125_un_3 Alphaproteobacteria; 937436 36 ASV5 Caulobacterales; 0≠1; 0≠2; 2=1 214

Using genomics to restore and future-proof underwater seaweed forests

Hyphomonadaceae; Hellea; uncultured

Bacteria; Chloroflexi; Anaerolineae; Caldilineales; Caldilineaceae; uncultured; ASV12 uncultured 0≠1; 0≠2; 2=1 Bacteria; Proteobacteria; Gammaproteobacteria; Thiohalorhabdales; Thiohalorhabdaceae; ASV16 Granulosicoccus; uncultured 0≠1; 0=2; 2=1 Bacteria; Proteobacteria; Gammaproteobacteria; Thiohalorhabdales; Thiohalorhabdaceae; ASV19 Granulosicoccus; uncultured 0≠1; 0≠2; 2=1

Bacteria; Bacteroidetes; Bacteroidia; Chitinophagales; Saprospiraceae; Rubidimonas; ASV21 uncultured 0≠1; 0≠2; 2=1 Locus Bacteria; Chloroflexi; 40713_un_5 Anaerolineae; Caldilineales; 699768 13 ASV12 Caldilineaceae; uncultured 0≠1; 0=2; 2≠1 Bacteria; Proteobacteria; Alphaproteobacteria; Rhizobiales; ASV86 Rhizobiaceae; uncultured 0≠1; 0=2; 2≠1 Bacteria; Chloroflexi; Anaerolineae; Caldilineales; ASV142 Caldilineaceae; uncultured 0≠1; 0=2; 2≠1 Bacteria; Proteobacteria; Gammaproteobacteria; Thiohalorhabdales; Thiohalorhabdaceae; ASV197 Granulosicoccus; uncultured 0≠1; 0=2; 2≠1 Bacteria; Verrucomicrobia; Verrucomicrobiae; Verrucomicrobiales; Rubritaleaceae; Roseibacillus; ASV275 uncultured 0≠1; 0=2; 2≠1

Bacteria; Chloroflexi; Locus Anaerolineae; Caldilineales; 52118_un_7 Caldilineaceae; uncultured; 296457 50 ASV12 uncultured 0=1; 0≠2; 2≠1 Bacteria; Planctomycetes; Planctomycetacia; Pirellulales; Pirellulaceae; Blastopirellula; ASV14 uncultured 0=1; 0≠2; 2≠1

215

Using genomics to restore and future-proof underwater seaweed forests

Bacteria; Proteobacteria; Gammaproteobacteria; ASV19 Thiohalorhabdales; Thiohalorhabdaceae; Granulosicoccus; uncultured 0=1; 0≠2; 2≠1 Bacteria; Bacteroidetes; Bacteroidia; Chitinophagales; ASV21 Saprospiraceae; Rubidimonas; uncultured 0=1; 0≠2; 2≠1 Bacteria; Proteobacteria; Gammaproteobacteria; ASV28 Thiohalorhabdales; Thiohalorhabdaceae; Granulosicoccus; uncultured 0=1; 0≠2; 2≠1

216

Using genomics to restore and future-proof underwater seaweed forests

(S5.5a)

217

Using genomics to restore and future-proof underwater seaweed forests

(S5.5b)

218

Using genomics to restore and future-proof underwater seaweed forests

(S5.5c)

219

Using genomics to restore and future-proof underwater seaweed forests

Figure S5.5: Relationship between (a) geography, (b) host genetics and (c) phenotype on Phyllospora comosa’s overall microbial communities. Taxa shown are the five most abundant ASVs that were significantly with each respective variable of interest. Sites ordered north to south; PM: Port Macquarie; FO: Forster; AB: Anna Bay: CR: Cronulla; MB: Malua Bay; ED: Eden; NT: Bicheno; ST: Southport.

220