Exploring, harnessing and conserving marine genetic resources towards a sustainable seaweed aquaculture

1Scottish Association for Marine Science, Scottish Marine Institute, Oban, United Kingdom 2Institute of Aquaculture, College of Fisheries and Ocean Sciences, University of the Philippines Visayas, Iloilo, Philippines 3Aquaculture Department, Southeast Asian Fisheries Development Centre, Iloilo, Philippines 4Institute of Oceanography and Environmental Science, Mindanao State University, TawiTawi College of Technology and Oceanography, Boheh Sallang, SangaSanga, Bongao, TawiTawi, Philippines 5Department of Life Sciences, Natural History Museum, London, UK 6Centre for Environment, Fisheries and Aquaculture Science (Cefas), Weymouth Laboratory, Weymouth, Dorset, United Kingdom 7IRDR Centre for Gender and Disaster, University College London, London, United Kingdom 8College of Marines Life Science, Ocean University of China, Qingdao, Shandong Province, People’s Republic of China 9School of Marine Science, Sun YatSen University, Zhuhai, Guangdong Province, People’s Republic of China 10University of Dar es Salaam, Dar es Salaam, Tanzania 11Institute of Ocean and Earth Sciences, University of Malaya, Kuala Lumpur, Malaysia 12UMR 7245 Molécules de Communication et Adaptation des Microorganismes, Muséum National d'Histoire Naturelle, CNRS, Paris, France


| INTRODUC TI ON
Marine ecosystems are a widely recognised reservoir of underexploited genetic resources (Arnaud-Haond et al., 2011) the phylogenetic breadth of which exceeds that of terrestrial ecosystems (Costello & Chaudhary, 2017). A wealth of novel diversity, revealed through large-scale marine meta-barcoding projects, not only awaits to be described (de Vargas et al., 2015), but its genetic and metabolic potential also requires characterisation. The imbalance of knowledge between terrestrial and marine organisms is illustrated by the low proportion of functionally annotated genes in the few available genomes of marine organisms, for example, brown algae (Dittami et al., 2020) or corals (Cleves et al., 2020). Several authors have recently called for the urgent development of a suitable governance structure to protect the intellectual property of innovators, yet refraining unequitable appropriation of marine genetic resources (Blasiak et al., 2018;Vanagt et al., 2019), to guarantee the conservation of biodiversity (Diz, 2018), and if traded, to promote good biosecurity practice (Campbell et al., 2019).
Amongst all marine resources, algae have a long-recognised, yet still underutilised, potential for human use and biotechnological applications (Arrieta et al., 2010). Driven by increasing demand and multiple potential uses as feed, food, pharmaceuticals or nutraceuticals, aquaculture especially of marine macroalgae, that is, seaweeds has been developing at an unprecedented pace (FAO, 2020).
Between 2010 and 2018, global production of red and brown seaweeds increased by 89.5% of the total 31.3 million metric tonnes wet weight, representing a value of 12.4 billion US $ (Table 1). Despite a downturn in biomass production in 2017 and 2018, a considerable proportion of the global seaweed production is due to the expanding cultivation of the red algal genera Kappaphycus and Eucheuma (herein referred to as eucheumatoids) in Southeast Asia and the Western Indian Ocean, especially in Indonesia (FAO, 2020). Seaweed aquaculture has become an important industry that provides jobs and livelihoods to millions of families in rural coastal communities, genera Eucheuma and Kappaphycus), an activity which has been successfully initiated in many tropical countries to support their economic development. We also review the challenges currently faced by this industry and identify potential threats to the seaweed cultivation sector. Accordingly, we suggest new directions to support the continued development of an economically resilient and environmentally sustainable industry based on the utilisation of genetic resources.

K E Y W O R D S
algae, aquaculture, biosecurity, breeding, genetic diversity, marine genetic resources TA B L E 1 Global production of the main seaweed crops from 2010 to 2018, in quantity and value (FAO, 2020), and number of registered varieties in China, Korea and Japan (mostly Hwang et al., 2019, Plant Variety Database accessed June 2020, UPOV, but see Table S1) particularly in areas where few other opportunities exist, including enabling women to be economically active and gain independent economic power (Msuya & Hurtado, 2017;Rebours et al., 2014).
Seaweeds consumed as sea vegetables, especially nori (Neopyropia/ Neoporphyra) are valuable protein resources for human nutrition (Fleurence, 2004). They are also used in animal feed, for example in polyculture or integrated multi-trophic aquaculture systems to help meet the global demand for dietary protein and to contribute to global food security, aligning with several UN Sustainable Development Goals (Bjerregaard et al., 2016). As seaweed aquaculture is predicted to increase significantly, the growth of an economically and environmentally sustainable industry would create major opportunities for coastal communities and subsequently, contribute to poverty reduction (Cottier-Cook et al., 2016).
Taking the cultivation of eucheumatoid algae as a case-study, Box 1 highlights the intricate agronomical, ecological and societal challenges that need to be tackled to achieve resilient production systems, mitigate their potential impact on ecosystems and provide stable income and prospects for people working in this industry.
Many other domesticated seaweed species face similar challenges, such as a narrow genetic diversity of cultivars (e.g. Guillemin et al., 2008), an increased frequency of diseases and pests (Loureiro et al., 2015), gene flow between farms and wild populations (Grulois et al., 2011), as well as limited or inadequate governance (Campbell et al., 2019). A common theme underlying many of these issues, is the management of genetic resources and the requirement to develop robust regional, national and supranational governance towards their conservation, as well as their equitable and sustainable exploitation. New technology opens novel avenues to explore the genetic potential of seaweed resources, whether they are already cultivated or not. Here, we review the current exploitation of seaweed genetic resources and their governance in the context of a growing aquaculture sector, identify knowledge gaps and explore new directions under-pinning the continued development of a sustainable and resilient seaweed aquaculture industry that aligns with major sustainable development goals.

| CURRENT FRONTIER S IN THE E XPLOITATI ON OF ALG AL G ENE TIC RE SOURCE S
Genetic resources, breeding concepts and the resistance of seaweeds against stressors are far less studied than in major land-based crops, most of which have a long-standing research and biobanking history (Tanksley & McCouch, 1997 (Laikre et al., 2020). Efforts to characterise and identify these genetic resources of wild seaweed populations and cultivars have been accelerated lately, for example, in kelps (e.g. Guzinski et al., 2016;Zhang et al., 2017), Undaria pinnatifida (Guzinski et al., 2018;Shan et al., 2018), Sargassum (Le Cam et al., 2019) and Agarophyton (Guillemin et al., 2008(Guillemin et al., , 2014. Until recently, population genetic studies primarily targeted ecologically important algae of temperate or cold waters and aimed to reconstruct their paleogeographic histories (Hu et al., 2016). Studies aimed at underpinning breeding efforts, however, have only been initiated in recent years and knowledge is particularly limited for tropical species. In fact, cryptic diversities in many seaweed groups have only recently started to reveal their breadth, with the aid of molecular data, including groups of major economic interest, such as the eucheumatoids (Lim et al., 2017, see also Box 1) and foliose Bangiales (Yang et al., 2020).
A key challenge to enable cultivation and breeding is to control the reproduction and life history of seaweeds; indeed, the development of such knowledge is a milestone of domestication (Valero et al., 2017 and references therein). Red and brown algae exhibit diverse and complex, bi-or tri-phasic life histories, many of which have only been recently described or remain imperfectly known.
Investigations into sex determination (Shan et al., 2015;Umen & Coelho, 2019;Zhang et al., 2015), life history transition control (Cock et al., 2014) and parthenogenesis (Mignerot et al., 2019) are fundamental resources for breeding efforts (Lipinska et al., 2015). While progress has been steady for brown algae, key knowledge gaps re-  (Maggs, 1988) and to find individuals able to propagate asexually, to facilitate biobanking and germplasm amplification (Ichihara et al., 2019;Li et al., 2017;Takahashi & Mikami, 2017). Comparable to crops such as banana and potato, asexually reproducing seaweeds are far easier to propagate; however, the development of new cultivars still relies on the identification of agronomically valuable individuals in the wild, or on the ability to perform controlled crossings. Finally, and again similar to some land crops and cultured animals (e.g. oyster and salmon), the control of ploidy, whether through endo-polyploidisation, allo-polyploidisation or the generation of somatic hybrids offers hope for yield improvement and for the control of the genetic pollution of wild stocks by escapees (Goecke et al., 2020). Much can be learned from elaborate strategies deployed in animals and plants, as well as their pitfalls, particularly the dramatic erosion of genetic diversity that has plagued agriculture and animal aquaculture since the last century (Hainzelin, 2013).
Novel DNA sequencing and analytical technologies provide potential opportunities to develop efficient pipelines for a large-scale, systematic exploration of the genetic and metabolic potential of algae that can aid further biotechnological exploitation. Although it is yet to be applied at scale in seaweeds, genome breeding has demonstrably shown its potential to accelerate the introgression in crops of genetic regions encoding high performance traits or increased resistance against abiotic and biotic stressors (Hickey et al., 2017). In addition to the potential of multi-omic approaches to improve crops for long-established uses of seaweeds, the exploration of novel species for cultivation, or discovery of novel bioactive compounds is also tantalising (Kumar et al., 2016). This general trend certainly contributes to the exponential growth of the number of marine species undergoing domestication (Duarte et al., 2007), and should be further explored for seaweeds. However, with over 95% of seaweed cultivation activities taking place in low-or middle-income countries, significant investment and capacity building are needed to harness this scientific potential and ensure that seaweed farmers benefit fairly (Cottier-Cook et al., 2016).

| NOVEL CONS ERVATION CHALLENG E S B ROUG HT BY ALG AL CULTIVATI ON
The accelerated loss of marine biodiversity is a general concern (Worm et al., 2006). Key seaweed-dominated ecosystems are disappearing worldwide (Arafeh-Dalmau et al., 2020;Smale, 2020). Global warming, ocean acidification, eutrophication and other anthropogenic pressures are key drivers for rapid changes of seaweed-dominated ecosystems and their poleward shift or even retraction (Brodie, Williamson, et al., 2014;Fabricius et al., 2015;Wernberg et al., 2016).
Physiological responses of tropical seaweeds to warming and ecological responses to climate change are only starting to be understood (Kumar et al., 2020). Monitoring change and diversity loss in the marine environments continues to be a challenge, leading to the concern that vanishing tropical seaweed populations remain largely unnoticed.
In addition to these global drivers, rapid development of cultivation incurs novel risks for seaweeds in temperate as in tropical regions, which are only beginning to be identified. For most species, very little is known about the relative abundance of gametophytes as opposed to sporophytes in wild stocks, the balance between sexual versus vegetative reproduction, or ploidy (as reported above for Porphyra, for example). Some microscopic life stages, such as the gametophytes of kelps are elusive in the field, resulting in a paucity of data about their ecology, longevity and vulnerability to environmental stressors or changes (Coleman & Goold, 2019). It is clear, however, that the control exerted over the life history and ploidy by farmers might shift the balance between life stages and reproductive modes in the field, with consequences on the genetic structure of populations and their resilience to perturbations. In the most comprehensively studied species, the agar-producing alga Agarophyton chilense (formerly Gracilaria chilensis) is predominantly propagated vegetatively by farmers and farming practices favour the propagation of diploid tetrasporophytes over haploid gametophytes (Guillemin et al., 2008). Evidence suggests that over-harvesting of wild stocks, in combination with the vegetative propagation in farms, has resulted in an extreme impoverishment of the species' genetic diversity in Chile (Guillemin et al., 2014).
To date, seaweed cultivation has been widely regarded as environmentally benign, because the ability of seaweeds to absorb nutrients helps remediate eutrophication caused, for example, by fish or shellfish aquaculture (Neori et al., 2004). Seaweed cultivation also pro-  Tanzania (Halling et al., 2013;Tano et al., 2015). The morphological plasticity of many seaweeds and the difficulty to identify them in the field also contributes to unnoticed introductions for many years. For example, the first reports of introduced eucheumatoids escaping farms and of their subsequent establishment in the wild were largely made 10 to 20 years after cultivation was initiated in the area ( Figure 1). This might be due to limited environmental monitoring, but also an indication of the time-scale necessary for such an impact to become detectable (Figure 1). It should be emphasised that agriculture and non-native species are major drivers of species extinction on land (Bellard et al., 2016). Therefore, the scale at which the global seaweed cultivation progresses and its integration with other human activities arguably calls for a careful assessment of its potential long-term impacts on coastal ecosystems and their possible mitigation (Eggertsen & Halling, 2020). The specific characteristics of seaweeds, such as their complex life histories, would need to be fully considered when performing risk assessments for invasiveness (Krueger-Hadfield, 2019).
Although still poorly documented for algae (Loureiro et al., 2015, see also Box 1), the worsening of disease outbreaks due to the in-

| A G OVERNAN CE IN NEED OF ADAP TING TO R APID CHANG E S
The rapid growth of the seaweed industry is a key driver to address Providing access to genetic resources for farmers, breeders and scientists is essential to develop and sustain crop cultivation. For most seaweed species, the rapid development of the industry has yet to be matched with commensurate investment in breeding programmes; efforts to biobank germplasm and to make germplasm available to F I G U R E 1 A simplified worldwide overview of Kappaphycus and Eucheuma farming. The best documented introduction events for farming purposes are indicated by black arrows; the assumed native range depicted in blue; 'ice-ice' disease and epi-endophyte outbreaks in farming areas indicated by black and white boxes, respectively; suspected or confirmed invasions marked by grey boxes. The date of each event or first report (for continuing ones) are given in the box. Subsequent major disease outbreaks, some of which led to local industry collapse, are indicated with downward arrows, whereas horizontal arrows indicate continuing issues farmers and breeders are still in their infancy (Wade et al., 2020). Over the last few years, however, there has been a sharp rise of seaweed varieties registered with the International Union for the Protection of New Varieties of Plants (UPOV, Table 1). While the UPOV system provides protection to breeders and should thus encourage investment, it is important to note that it has been widely seen as encouraging homogeneous crops and subsequent loss of genetic diversity (Ahmadi et al., 2013). Also, only a few governments currently engage with UPOV for seaweeds (e.g. Japan and Korea). This reflects a continuing global imbalance among the countries involved in seaweed production and those engaged in research and biotechnology markets (Mazarrasa et al., 2013). Accordingly, this poses long-term challenges concerning the establishment of sustainable and equitable international partnerships. It is, therefore, important to accompany the current devel-

BOX 1 Eucheumatoid cultivation-a case-study of a globalised, vegetatively propagated crop that supports development of deprived coastal communities
Eucheumatoids are sought after for their rich content of kappa and iota carrageenans (Lim et al., 2017). Carrageen is used as an additive in manufactured food or as a stabiliser in cosmetics. Driven by high demand and over-exploitation of wild stocks, commercial cultivation of eucheumatoids was initiated in the Philippines in 1969 (Hurtado et al., 2014, Figure 1). Subsequently, cultivation was introduced in neighbouring countries in Asia, the Western Indian Ocean and the Americas, predominantly Brazil. Today, about 43 countries are engaged, or have been engaged in the cultivation of eucheumatoids (Kelly et al., 2020). Their annual production is 10.3 metric tonnes of fresh weight, for an output worth approximately 1 billion US $ (2017, FAO Fisheries statistics). Eucheumatoid farming is a major economic opportunity, specifically for low-income or middle-income regions, with proactive government policies in place, for example in Indonesia, Malaysia, the Philippines and Tanzania. Despite concerns about over-production, price stagnation (Table 1) and a controversy about carrageenan safety as a human food ingredient (Martino et al., 2017), this industry has enabled women to become economically active (Msuya, 2006), and offers livelihood opportunities to poor, often displaced populations, particularly in Asia (Nimmo, 1986;Nor et al., 2017).  (Hurtado et al., 2016;Quiaoit et al., 2016;Tan et al., 2013). Considerable plasticity in the morphology of eucheumatoids, however, impedes species identification.
Varieties, which have been gathered from wild populations, are given vernacular names by seaweed farmers. These local vernacular names are not unified between different regions and often farmers are unaware of the actual species that they are cultivating (Dumilag et al., 2016a;Ganzon-Fortes et al., 2012;Montes et al., 2008;Tan et al., 2013). This limitation in the ability to identify taxa and in the genetic characterisation of currently cultivated eucheumatoids is, therefore, a problem for the entire sector. Farmers receive lower prices for their product if they inadvertently mix species producing different types of carrageenans as Kappaphycus spp.
contain the higher valued kappa carrageenan, while Eucheuma denticulatum contains the lower priced iota carrageenan; it impedes their conscious decision in cultivar choice and also restricts supra-regional management of cultivar diversity, for example, in the establishment of biobanks.
Another serious concern is the widespread, and apparently worsening, occurrence of diseases, particularly 'ice-ice' disease and infestations by filamentous red algal endo-epiphytes (Hurtado et al., 2006;Largo, 2002;Vairappan et al., 2008). 'Ice-ice' is characterised by a whitening or loss of pigmentation of the thallus, followed by the disintegration of affected tissues and often the detachment of plants from cultivation ropes, resulting in loss of biomass. The condition appears to be the result of complex interactions between abiotic stress induced by 'unfavourable' shifts in environmental parameters, particularly a decrease in salinity or irradiance or an increase in water temperature or pH (e.g. Alibon et al., 2019; and the proliferation of 'ice-ice' promoting bacteria (often identified as Vibrio or Pseudomonas (Azizi et al., 2018; and it is thought to affect all cultivated eucheumatoid varieties (Sade et al., 2006;Tisera & Naguit, 2009 Bustamante et al., 2015;Hurtado et al., 2006;Largo, 2002;Largo et al., 2020;Tsiresy et al., 2016;Vairappan et al., 2008).
In some areas, prevailing diseases and pests have forced farmers to cease their activity. For example, in Tanzania, the production of K. alvarezii collapsed from 1,000 metric tonnes fresh weight in 2001 to ca. 13 metric tonnes fresh weight in 2010 and in Madagascar, it dropped from 1,860 metric tonnes fresh weight in 2009 to 110 metric tonnes fresh weight in 2012 . Figure 1 illustrates the generality of this trend and shows how-after an initial highly profitable period of up to ten years following the successful introduction of farming-isolated disease outbreaks are typically reported, quickly followed by regional outbreaks. Diseases are usually mitigated by switching to other cultivars, changing the location of farms seasonally, or stopping farming temporarily during the 'disease' season. Yet, production is still reduced by disease and pest outbreaks and the extent by which the mostly uncontrolled globalised movement of germplasm may, or not, have contributed to the onset and worsening of these outbreaks is unknown.
In the light of these issues, there is a strong need for breeding programmes that produce disease-resistant cultivars that can tolerate the higher temperatures as a consequence of climate change and the provision of farmers with pest-free germplasm after a disease outbreak. Completing the entire life cycle of eucheumatoids in a laboratory, however, has only been achieved at a small scale ( Sollesta, 2010). The successful application of micropropagation techniques offers new hope to establish biobanks from wild individuals (Luhan & Mateo, 2017); however, the depletion of wild eucheumatoid stocks is a matter of concern. Alarmingly, our knowledge on the conservation status of any of the cultivated eucheumatoid species is highly limited. Figure 2 provides an overview of available molecular data for the four cultivated eucheumatoid species. It shows a bias towards the molecular characterisation of farmed individuals; additionally, the overall sampling depth is insufficient to assess the genetic structure of wild populations. Throughout the assumed native ranges, there are also numerous reports of over-harvest prior to cultivation (Mshigeni, 1984;Trono, 1999), of vanishing populations, presumably as a result of global warming, and some reports of cryptic invasion of farmed genotypes into wild populations Halling et al., 2013;Tano et al., 2015). All these factors may contribute to the rarefaction of wild stocks and potentially, their genetic impoverishment.

BOX 1 (Continued)
F I G U R E 2 Haplotype networks and the geographic distribution of haplotypes of the four cultivated eucheumatoid species Eucheuma denticulatum (a), Kappaphycus alvarezii (b), K. striatus (c), K. malesianus (d), using the mitochondrial genetic sequence cox2-3 spacer as a marker. In the haplotype network, the size of the nodes relates to the number of sequences in Genbank, the colour of the inner circle relates to the geographic origin, the colour of the outer circle indicates the specimen origin (farmed, wild native, wild non-native). For the geographic distribution specimen were grouped according to their sampling location in marine ecoregions (following Spalding et al., 2007). Note that this does not necessarily reflect the indigenous diversity, as molecular information is biased towards farmed specimens and includes introduced specimens (see Figure 1 for major introduction events)