Undaria pinnatifida: A case study to highlight challenges in marine invasion ecology and management

Abstract Marine invasion ecology and management have progressed significantly over the last 30 years although many knowledge gaps and challenges remain. The kelp Undaria pinnatifida, or “Wakame,” has a global non‐native range and is considered one of the world's “worst” invasive species. Since its first recorded introduction in 1971, numerous studies have been conducted on its ecology, invasive characteristics, and impacts, yet a general consensus on the best approach to its management has not yet been reached. Here, we synthesize current understanding of this highly invasive species and adopt Undaria as a case study to highlight challenges in wider marine invasion ecology and management. Invasive species such as Undaria are likely to continue to spread and become conspicuous, prominent components of coastal marine communities. While in many cases, marine invasive species have detectable deleterious impacts on recipient communities, in many others their influence is often limited and location specific. Although not yet conclusive, Undaria may cause some ecological impact, but it does not appear to drive ecosystem change in most invaded regions. Targeted management actions have also had minimal success. Further research is needed before well‐considered, evidence‐based management decisions can be made. However, if Undaria was to become officially unmanaged in parts of its non‐native range, the presence of a highly productive, habitat former with commercial value and a broad ecological niche, could have significant economic and even environmental benefit. How science and policy reacts to the continued invasion of Undaria may influence how similar marine invasive species are handled in the future.

it establishes is also problematic. Many species have microscopic life stages and are found in inconspicuous and often inaccessible habitats.
The incomplete taxonomy and historical records that are apparent for many marine families means that once recognized newly identified species will often be cryptogenic. It can often take considerable time for accurate identification and status of a newly identified species to be determined, requiring a wide range of genetic, ecological, and biochemical techniques, further delaying potential rapid-response management.
Identifying specific characteristics that predispose a species to being invasive is challenging. Invasive species are generally considered to have high phenotypic or genetic plasticity and a broad ecological niche in order to survive introduction, establishment, and spread in a non-native range (Kolar & Lodge, 2001;Newsome & Noble, 1986;Williamson & Fitter, 1996;Zenni, Lamy, Lamarque, & Port, 2014).
They are often described to have opportunistic life histories, including high fecundity, growth rate, and recruitment; however, there are also successful invasive species with more competitive life-history traits (Duyck, David, & Quilici, 2007;Valentine, Magierowski, & Johnson, 2007). The probability of invasion increases with the number of individuals released or reproducing, the number of introduction events, and proximity to existing populations (Kolar & Lodge, 2001;Lockwood, Cassey, & Blackburn, 2005). Resource availability, such as light, food, and physical space, is also a key factor which can influence the vulnerability of a recipient community to invasion (Levine & D'Antonio, 1999;Stachowicz, Fried, Osman, & Whitlatch, 2002).
Quantifying the ecological impacts of an invasive species is also complex. Differences in recipient communities, resource availability, environmental abiotic factors, and attributes of the invasive species itself can all create site-specific impacts. Factors such as abundance and geographical range of the invasive species may influence impacts in all cases (Parker et al., 1999), while other factors such as morphological, behavioral, or even chemical characteristics of the invasive species are more species specific (Thomsen, Olden, Wernberg, Griffin, & Silliman, 2011).
Invasive marine macroalgae (seaweeds) may function as ecosystem engineers that are able to modify the environment and alter recipient communities and, as such, have the potential to cause significant ecological and socioeconomic impacts (Dijkstra et al., 2017;Thomsen, Wernberg, Tuya, & Silliman, 2009;Williams & Smith, 2007).
Overall, there are thought to be approximately 350 different seaweed NIS accounting for around 20%-30% of all marine NIS (Schaffelke & Hewitt, 2007;Thomsen, Wernberg, South, & Schiel, 2016). The coldtemperate kelp Undaria pinnatifida ( Figure 1) is one of only two seaweeds (along with Caulerpa taxifolia) included in the Invasive Species Specialist Group list of the 100 most invasive species of the world (Lowe, Browne, Boudjekas, & De Poorter, 2000). Native to coldtemperate areas of the northwest Pacific (the coastlines of Japan, Korea, Russia, and China), the adventive kelp Undaria pinnatifida (Harvey) Suringar, 1873 (Phaecophycae, Laminariales), or "Wakame," has a worldwide non-native range ( Figure 2). First identified as an invasive species on the Mediterranean coast of France in the 1970s (Perez, Lee, & Juge, 1981), Undaria pinnatifida (hereafter referred to as Undaria) is now established on the coastlines of 13 countries across four continents (James, Kibele, & Shears, 2015). The design of efficient and effective NIS management requires a clear understanding of a species physiology, invasion dynamics, and ecological impacts.
Due to its global distribution and status as an invasive species for over 30 years, Undaria is a useful case study to highlight both successes and failures in our handling and understanding of marine NIS.

| Biology, physiology and native ecology
In its native northeast Asia, Undaria is a winter annual species that inhabits rocky substrates from the low intertidal to 18 m depth, and is widespread at depths of 1-3 m (Koh & Shin, 1990;Saito, 1975;Skriptsova, Khomenko, & Isakov, 2004). It is also a major species for seaweed mariculture in China, Japan, and Korea (Yamanaka & Akiyama, 1993), with total world yield in 2013 exceeding 2 million tonnes fresh weight (FAO FishStat). Sporophytes can grow up to 1-1.7 cm per day, reach 1.3-2 m in length, and have a maximum life span of around 6-8 months (Castric-Fey, Beaupoil, Bouchain, Pradier, & L'Hardy-Halos, 1999;Choi, Kim, Lee, & Nam, 2007;Dean & Hurd, 2007). They form large divided pinnate fronds and distinctive ruffled reproductive sporophylls ( Figure 1). As with all kelps, Undaria has a heteromorphic life cycle, with large macroscopic diploid sporophytes that produce microscopic zoospores from reproductive sporophylls. F I G U R E 2 Approximate distribution of Undaria pinnatifida. Global map: Green = native range, red = non-native range. Regional maps: Each point represents a distinct location but does not indicate precise position or entire extent. See Table S1 for more information and references The spores develop into microscopic dioecious haploid gametophytes, which, on maturation, produce motile sperm that fertilize the sessile egg and a new sporophyte will start to grow in situ of the female gametophyte (Dayton, 1985). Sporophylls develop over several months and mature sequentially from the base upwards (Saito, 1975;. Zoospores are released over approximately 20-40 days at densities of 0.13 × 10 5 -12 × 10 5 spores per cm 2 of sporophyll per hour, amounting to 1 × 10 8 -7 × 10 8 spores over the lifetime of a sporophyte (Primo, Hewitt, & Campbell, 2010;Saito, 1975;Schaffelke et al., 2005;. Once released, spores typically move at around 0.13-0.33 mm/s for 5-6 hr, but may remain motile for up to 3 days. Fixing ability starts to be reduced within a few hours, although viability can last over 10 days (Forrest, Brown, Taylor, Hurd, & Hay, 2000;Hay & Luckens, 1987;Saito, 1975;Suto, 1952). Due to the low motility and vitality of the zoospores, settlement is strongly correlated with distance from mature sporophytes, and dispersal may be limited to as little as 0.2-10 m from a spore release point Suto, 1952). Larger dispersal distances are thought to be facilitated by the drifting of entire sporophytes, which may remain viable for much longer periods. Overall, it has been estimated that maximum spore-mediated dispersal rates for populations are in the order of 10-200 m/year, while sporophyte drift may allow maximum dispersal rates of 1-10 km/year Russell, Hepburn, Hurd, & Stuart, 2008;Sliwa, Johnson, & Hewitt, 2006).
In most of its native range, Undaria sporophyte recruitment occurs in winter, becomes reproductive in spring, and goes through widespread senescence during summer, leaving only the microscopic gametophyte life stages which persist through autumn (Koh & Shin, 1990;Saito, 1975). Temperature is the key environmental factor which determines this annual population dynamic (Figure 3; Saito, 1975).
Although less defined than the influence of temperature, many abiotic factors can affect the growth and distribution of Undaria, including salinity, light, day length, nutrients, and wave exposure. Undaria is predominantly found in fully saline conditions, with mean salinities below 27 psu generally limiting its range (Floc'h, Pajot, & Wallentinus, 1991;Saito, 1975;Watanabe et al., 2014 at salinities as low as 19 psu, while gametophytes and sporophytes may survive at salinities as low as 6 psu (although below 16 psu sporophytes may start to become damaged) (Bollen et al., 2016;Peteiro & Sanchez, 2012;Saito, 1975). Undaria is viable over a wide range of light regimes; however, changes in irradiance and day length will influence the rate of recruitment, growth, and photosynthesis in both gametophyte and sporophyte stages (Baez et al., 2010;Choi et al., 2005;Morelissen, Dudley, Geange, & Phillips, 2013;Pang & Luning, 2004).

When compared to perennial or summer annual Laminarians,
Undaria has a comparatively low rate of nutrient uptake and nitrate storage, and therefore a close association between seawater and tissue nitrate (Dean & Hurd, 2007). This means that growth of sporophyte and gametophyte stages is positively related to nutrient concentration (Dean & Hurd, 2007;Gao, Endo, Taniguchi, & Agatsuma, 2013;Morelissen et al., 2013;Pang & Wu, 1996). Zoospore settlement, however, is not considered to be influenced by nutrient concentration and therefore any inhibition of recruitment by nutrient limitation would occur at the gametophyte or sporophyte stage (Morelissen et al., 2013). Increased water motion can enhance nutrient uptake in kelps (Gerard, 1982), which is highlighted by rope-based mariculture of Undaria being more efficient in moderately exposed sites with water velocities of up to 15-30 cm/s when compared to sheltered sites of 5-12 cm/s (Nanba et al., 2011;Peteiro & Freire, 2011;Peteiro, Sanchez, & Martinez, 2016 Russell et al., 2008;Saito, 1975). Due to the thin fragile nature of the sporophyte frond, Undaria is limited in highly exposed shores (Choi et al., 2007), although can still be found in low intertidal pools or lower subtidal areas, which have more shelter from wave action at exposed sites (Russell et al., 2008). Periods of low water motion are needed for high natural recruitment, with spore adhesion optimal at water velocities of 3 cm/s (Arakawa & Morinaga, 1994). Under certain conditions, spores may completely fail to adhere at flows ≥14 cm/s (Saito, 1975), however, in some cases no inhibition of adhesion rate may occur until flow rates reach over 16 cm/s, and spores may still adhere, albeit at a greatly reduced rate, at flows over 25 cm/s (Arakawa & Morinaga, 1994;Pang & Shan, 2008).
Overall, Undaria has a high growth rate, large reproductive output, high phenotypic plasticity, and a relatively wide physiological niche.
These factors are often considered characteristic of successful invasive species (Newsome & Noble, 1986;Williamson & Fitter, 1996). On the other hand, Undaria exhibits low natural dispersal ability, and its ecophysiological niche is not as broad as some other highly invasive marine macroalgae (Nyberg & Wallentinus, 2005). As such, it could be thought of as a low risk for widespread colonization; however, its invasion history demonstrates it to be a very successful invader.
Once established in these anthropogenic or modified environments, Undaria can spread into natural habitats. Due to its requirement for attachment on hard substrates, it is predominantly found invading rocky reefs; however, it can also be found more rarely to invade sea grass beds and mixed sediment communities (Farrell & Fletcher, 2006;Floc'h et al., 1996;James, Middleton, Middleton, & Shears, 2014;Russell et al., 2008). In many parts of its non-native range, Undaria Due to the low natural dispersion rates of Undaria, local spread of populations tends to occur in a step-wise manner (Fletcher & Farrell, 1999). The rate of localized natural spread is therefore far lower than human-mediated spread, with some populations having minimal range expansion for many years following their initial introduction. For example, in the UK it took over 7 years for Undaria to colonize a shoreline 200 m away from an established marina population (Farrell & Fletcher, 2006); in the USA, many marina populations remain localized following introductions over 10 years ago (Kaplanis et al., 2016); while in France, it took 10 years for Undaria to be found outside of the enclosed lagoon to which it was first introduced (Floc'h et al., 1991). In New Zealand, population expansion seems to be dependent on the area in which it is found. In Timaru Harbour, Undaria has extended less than 1 km from the harbor in over 20 years (Russell et al., 2008), in Marlborough Sound, the range of Undaria has expanded by hundreds of meters a year , and in Moeraki Harbour, expansion was around 1 km per year, while at Otago Harbour, Undaria spread around 2 km per year along adjacent exposed coastlines outside the harbor (Russell et al., 2008). Considerably faster rates of spread have also been recorded in areas of Argentina and Australia. Within the San Jose Gulf (Argentina), only 4 years after its introduction, Undaria had spread across approximately 100 km of coastline (Dellatorre et al., 2014), and in certain parts of Tasmania, local spread has been estimated to reach up to 10 km per year . Although the rate of range expansion is variable and site-specific, Undaria seems able to spread and proliferate without the direct aid of humans in all of its non-native range.
As previously discussed, temperature is the key environmental factor which determines the population dynamics of Undaria (Saito, 1975 where temperature maxima is greater than or equal to 20.6°C, an annual phenology could be expected (James et al., 2015).
Due to Undaria sporophytes living approximately 6-8 months, a recruitment period of four or more months, or multiple recruitment pulses per year could result in the year round presence of sporophytes (James et al., 2015). In Santa Barbara (California, USA) where average seasurface temperatures range from approximately 12-19°C, the presence and growth of sporophytes occur year round. There are two recruitment pulses, with a smaller autumn pulse at temperatures from 17 to 21°C, and a larger winter recruitment when temperatures are 12-17°C (Thornber et al., 2004). In this location, recruitment seems to be triggered by a fall in temperature below 15°C, with recruitment occurring around 8 weeks later (Thornber et al., 2004). A similar biannual recruitment has been recorded in New Zealand, with pulses in the autumn and spring (Hay & Villouta, 1993;

| Ecological impacts
Surveys examining the distribution of Undaria within mixed seaweed assemblages have identified that it occurs more commonly or is found in higher abundance, where there is a lower density of native canopy species (e.g., Castric-Fey et al., 1993;Cremades et al., 2006;Russell et al., 2008;Heiser et al., 2014;De Leij, Epstein, Brown, & Smale, 2017; Table 1). Due to the lack of pre-invasion data, it could be argued that Undaria may have been the cause of this reduced native canopy.
However, results indicate that Undaria is occupying substrates, depth ranges, or anthropogenically stressed habitats where native canopyforming seaweeds are limited (e.g., Castric-Fey et al., 1993;Cremades et al., 2006;Russell et al., 2008;James & Shears, 2016b; Table 1). This is supported by an investigation where data on native kelp abundance were available before the Undaria invasion. This before-after controlimpact (BACI) study showed that the introduction of Undaria led to no significant change in the abundance of native kelp species over 3 years (Forrest & Taylor, 2002).
In its native Japan and Korea, Undaria can act as a pioneer species and is part of a natural successive colonization process (Agatsuma, Matsuyama, Nakata, Kawai, & Nishikawa, 1997;Kim et al., 2016).
Where it has invaded, this pioneer-like trait is indicated by ecosystem stress or disturbance being key to Undaria's recruitment into mixed canopy assemblages (  (Curiel et al., 2001;Farrell & Fletcher, 2006

Impact on community
Heiser et al.

| Management
Management frameworks designed to control the abundance and spread of Undaria could only be found for two of the countries to which it has been introduced (  (Sinner, Forrest, & Taylor, 2000). Although not necessarily a requirement, none of these measures will reduce localized natural spread or abundance of Undaria. As previously discussed, many studies have shown that Undaria requires a level of ecosystem stress or disturbance to recruit and spread in mixed seaweed canopies. Reducing, mitigating, or preventing anthropogenic disturbance to native canopies has therefore been suggested as a management option to prevent the spread, and limit the abundance of Undaria (Valentine & Johnson, 2003). However, where Undaria has already established at high densities, or if it is acting as a "backseat driver" -suppressing native species once recruited (Bauer, 2012), maintaining native canopies alone is unlikely to be effective (Valentine & Johnson, 2003).
The management options available to directly target the local spread and abundance of Undaria are unclear. Where Undaria can be found in multiple locations and at high abundance within natural environments, it seems unlikely that eradication would be feasible. This is generally accepted by environmental managers, with widespread eradication of Undaria not currently being considered in any country to which it has been introduced ( In many regions where Undaria is now accepted (i.e., eradication is no longer being considered), commercial farming and wild harvest are being developed. Mariculture expanded across Brittany, after Undaria's initial introduction in 1981, with nine sites established into the early 1990s (Castric-Fey et al., 1993). Cultivation and mariculture have also been carried out on the Galician coast of Spain since the late 1990s and are continuing to develop along the North coast (Perez-Cirera et al., 1997;Peteiro et al., 2016). In 2010, The Ministry for Primary Industries (New Zealand) introduced a revised policy for the commercial use of Undaria which approved its wild harvest from artificial substrates or when cast ashore in selected areas. It also approved mariculture in three heavily infested areas, but prohibited harvest from natural substrates unless part of a designated control program (MAF, 2010). The rationale behind the prohibition of harvest from natural substrates was that "it could disturb or remove native canopy species leading to a proliferation of Undaria," while "harvesting when taken as part of a control program is allowed as any risks associated with harvest will be outweighed by reduced Undaria in localized areas" (MAF, 2010). It may be possible that one of the remaining options to reduce the abundance and local spread of Undaria where eradication is no longer feasible, would be through the legalization of commercial wild harvest from natural substrates. Strict biosecurity would have to be implemented to avoid its spread, and harvesting practices would need to minimize damage to native canopies-such as through a licensing system for hand harvesting only in specific areas. Timings of harvest would also have to be carefully considered, as removal or thinning of the Undaria canopy can result in a strong positive response of conspecific recruitment, and increased growth rate of the remaining stock (Gao, Endo, Taniguchi, & Agatsuma, 2014;.
However, removal before maturation could greatly reduce spore and seed-bank densities, and would perhaps limit the abundance and spread of Undaria over time.

Decisions taken by environmental managers on whether to manage
Undaria within a given jurisdiction should be made on a case-by-case basis. Where Undaria has recently arrived, or has a restricted range, it is likely that there will be a better chance of successful control or eradication. However, due to the widespread global distribution of Undaria, re-introduction is probable without the implementation of thorough biosecurity. The native community into which Undaria is introduced may also strongly influence the decisions of environmental managers.
The invasion of Undaria is likely to have greater ecological impact in areas where there are no functionally similar native species, whereas, in communities which are dominated by native canopy-forming macroalgae, Undaria may have limited impact on the community as a whole, and act as a passenger of ecosystem change. Economics and the maintenance of ecosystem services will also be factors that influence the decisions made by environmental managers. Although not covered as part of this review, Undaria can act as fouling pest to industries such as aquaculture, shipping, and recreational boating (Hay, 1990;James & Shears, 2016a;Minchin & Nunn, 2014;Zabin et al., 2009). The overall economic impacts of this interaction are poorly understood, but as has been noted above, Undaria could also have economic benefit through the development of an Undaria mariculture industry. Careful consideration and further research is needed on a site-specific basis. Clearly, the risks, costs, impacts, and benefits of all options, including potential management or eradication and possible acceptance, should be considered when developing management plans for Undaria.

| Predicting invaders and reacting to NIS
Although our understanding of marine NIS has greatly increased, Undaria is a useful case study to demonstrate that current capacity to predict the invasion dynamics of many marine NIS, and their interactions and impacts within native communities, remains limited. Once introduced, most NIS would not be expected to establish or become invasive (Lodge, 1993;Williamson & Fitter, 1996). Where invasion does occur, the time from initial introduction to when a species becomes invasive is highly variable. In some cases this ``lag time" may last decades, with little-to-no proliferation of NIS populations for a considerable time after introduction (Crooks, 2005). This is highlighted by the invasion history of Undaria, which has exhibited a wide range of expansion rates following introduction into different regions.
Undaria was considered to be an acceptable species for intentional introduction into France for mariculture purposes in 1981 (Perez et al., 1981). A better understanding of a species ecology and physiology is required before intentional introductions are conducted. However, when adventive species arrive unexpectedly, the necessity for rapidresponse management negates this consideration. A failure to react to new introductions could have major consequences. As marine invasive species can cause significant damage to the environment and economy, and due to the complex nature of species invasions, a precautionary principle should be adopted to minimize the rate of any new introductions (Bax, Williamson, Aguero, Gonzalez, & Geeves, 2003;Grosholz, 2002;Molnar et al., 2008).

| Ecological impacts
For some marine invasive species, deleterious ecological impacts can be substantial and easy to detect. Introduced voracious predators such as the northern Pacific seastar, Asterias amurensis, in Tasmania (Ross, Johnson, & Hewitt, 2003), the Lionfish, Pterois volitans, in the tropical Atlantic (Green, Akins, Maljkovi, & Ct, 2012) and the North American mud crab Rhithropanopeus harrisii in the Baltic Sea (Jormalainen, Gagnon, Sjroos, & Rothusler, 2016), prey on wide range of native species and proliferate in the absence of native predators. In these examples, clear community-wide impacts can be identified. Similarly, when invasive species greatly alter nutrient pathways, trophic interactions, or habitat structure, impacts at the community and ecosystem level are easily detectable (Crooks, 2002;Simberloff, 2011). For example, colonial ascidians of the genus Didemnum have overgrown large areas of hard substrates, particularly in the Netherlands and USA. These "mats" can greatly alter the physical habitat, cause mortality through smothering of sessile flora and fauna, and have major deleterious impact on wider ecosystem functioning with socioeconomic consequences (Bullard et al., 2007;Gittenberger, 2007). The invasion of Undaria highlights that in many other cases, ecological impacts are far harder to quantify and may vary considerably between locations and recipient communities. For these species, justifying costly eradication attempts may be challenging. However, as marine invasive species spread to new regions, decisions will have to be made on potential rapid-response management before site-specific impact studies can be carried out.
Invasive species, including Undaria, can also have facilitative impacts on the recipient community (Dijkstra et al., 2017;Irigoyen et al., 2011;Rodriguez, 2006). The invasion of bivalve molluscs onto soft sediments, such as Musculista senhousia and Crassostrea gigas, is a useful example of facilitation by a marine invasive on multiple levels. They provide complex habitats which can greatly increase infaunal and epifaunal abundance, increase organic content in sediment to the benefit of associated organisms, and can act as a trophic subsidy to predatory invertebrate and vertebrate species (Crooks & Khim, 1999;Escapa et al., 2004;Padilla, 2010). In order to understand the overall ecological impact a marine invasive species has on the recipient community, both deleterious and facilitative effects must be considered. Intrinsically, the facilitation of one species is likely to occur at the expense of others, due to changes in competition or predation. In fact for both Musculista senhousia and Crassostrea gigas, where high densities are found, a reduction in the abundance of functionally similar native species is often recorded (Creese, Hooker, De Luca, & Wharton, 1997;Crooks & Khim, 1999;Padilla, 2010). In many cases, unequivocal evidence of significant ecological impact of an invasive species on recipient communities will be difficult to attain. Prioritization of management actions will be influenced by the perceived impacts of marine invasive species, in terms of their threat to conservation and the maintenance of ecosystem services across different regions, as well as their direct socieoeconomic impacts.

| Management
Managing marine NIS is expensive and time-consuming, while eradication may be impossible once a species is established and widespread (Hulme, 2006). There are examples of successful rapid-response eradication of invasive species in the marine environment. The seaweed Caulerpa taxifolia was first identified in the USA in 2000 (Jousson et al., 2000). A rapid response only 17 days after its first discovery allowed the successful implementation of a 5-year eradication program using containment and chemical treatment, at a cost of around $7.5 million (USD) (Anderson, 2005). However, as shown by Undaria, once a marine NIS is established, proliferation and spread may be inevitable due to the natural or engineered connectivity of many water bodies. As population size increases the costs of control also increase, while attempting eradication of established populations would require significant resources and effort, and may ultimately be unsuccessful (Hobbs & Humphries, 1995). A pertinent example of a marine invasive species where targeted management was deemed to be inappropriate is the macroalgae Sargassum muticum or "Japanese wireweed" in Europe.  Viejo, 1997), however, other long-term studies recorded limited effects from the invasive species (Olabarria, Rodil, Incera, & Troncoso, 2009;Sanchez & Fernandez, 2005). Although attempts at management were made (Critchley, Farnham, & Morrell, 1986), due to its widespread distribution, uncertainties in the level of its ecological impact, as well as the costs and difficulties in its control, Sargassum now has no targeted management across most of Europe.
As with many other invasive species, Undaria has a largely opportunistic life strategy, taking advantage of resource availability in order to establish and spread (Gurevitch & Padilla, 2004). These species are sometimes considered "passengers" -promoted and maintained due to the presence of ecosystem stress or disturbance but not in themselves the cause of ecosystem change (MacDougall & Turkington, 2005). A potential management option for these species is not to directly target the species itself, but instead to manage the causes of ecosystem stress or disturbance, with the ultimate aim of restoring, maintaining or even promoting the diversity, integrity, and biotic resistance of recipient communities to invaders. Managing long-term global-scale stressors such as climate change will be challenging but crucial given the known interactions between climate and the spread of NIS (Occhipinti-Ambrogi, 2007). On a local-to-regional scale, however, managing stressors such as coastal inputs of sediments and nutrients and physical disturbances from resource extraction, fishing activities, and coastal development may allow some biotic resistance to be maintained. While designing and prioritizing targeted management options for invasive species is of significant importance, especially for those that are considered of high risk or highly damaging, it is also clear that attention should be given to preserving the integrity, diversity, and resistance of native communities through maintaining good overall environmental status. This has been shown for Undaria, as its abundance and spread is limited by the presence of diverse, native macroalgae canopies (e.g. Castric-Fey et al., 1993;De Leij et al., 2017;Russell et al., 2008;Valentine & Johnson, 2003, 2004.
As marine NIS continue to spread and extend their non-native ranges, decisions will be made on the necessity and feasibility of managing new incursions. Although a precautionary principle should be applied, it is unrealistic to assume that management and control of all species can be achieved due to the widespread establishment of many marine invasive species. Difficult choices will have to be made regarding which species should be targeted, with some potentially becoming an accepted part of the local biota. These decisions must be made on a case-by-case basis using the best information available and will depend on a variety of factors including the likely effectiveness, practicality, risk and cost of management options, as well as negative and positive ecological and socioeconomic impacts of a given species.

| Accepting NIS
Many NIS have been established in their non-native range for a considerable time and are now considered part of the natural biota in different regions across the world with major economic benefit and even cultural importance (Davis et al., 2011;Ewel et al., 1999). These species frequently occur in high abundance and over a wide distribution, and could therefore be classed as invasive. Due to the historic nature of species introductions, the widespread acceptance of certain NIS or invasive species is particularly common in the terrestrial environment.
The vast majority of the world's agricultural and horticultural species are NIS where they are grown. Many freshwater fish species have also been historically introduced for farming and sports fishing purposes and are treated essentially as part of the natural biota in many regions (Copp et al., 2005;Eustice, 2014;Gozlan, 2008 is considered as a damaging invasive, with management being developed, or enforced to reduce its spread (Guy & Roberts, 2010;NSW, 1994). However, in many parts of the USA and France, where introductions occurred in the 1920s and 1960s, respectively, they are now being seen as part of the natural biota, and are targeted by both wild capture fisheries and aquaculture using seeded bottom culture techniques (Buestel, Ropert, Prou, & Goulletquer, 2009;Cognie, Haure, & Barill, 2006;Feldman, Armstrong, Dumbauld, DeWitt, & Doty, 2000).
Although somewhat contentious, in certain cases invasive species could be considered to have benefits to nature conservation (Schlaepfer, Sax, & Olden, 2011, 2012Vitule, Freire, Vazquez, Nuez, & Simberloff, 2012). This may occur if the invasive species (i) has considerable facilitative and minimal deleterious impacts on native species; (ii) acts as a catalyst for restoration of native habitats; (iii) functionally replaces a limited or extinct native species; (iv) facilitates a species of high conservation value; or (v) acts as a biocontrol agent (Schlaepfer et al., 2011). These benefits are again more commonly identified in the terrestrial environment due to the historical and often intentional nature of introductions (e.g. Lugo, 2004;Morrison, Reekie, & Jensen, 1998). Crassostrea gigas may be another pertinent example relating to the marine environment. In many parts of Europe and America, native oysters have been over harvested and are considered endangered. It has been suggested that the spread of the invasive Pacific Oyster may have conservation benefit, functionally replacing the native species, providing habitat, a trophic subsidy and increased biofiltration, while also providing an exploitable resource, reducing further harvesting pressure on the native homolog (Paalvast, van Wesenbeeck, van der Velde, & de Vries, 2012;Shpigel & Blaylock, 1991).
As previously stated, some marine invasive species, such as voracious predators, or those with perennial life cycles and more competitive life-history traits, can have major detrimental ecological impact.
Many of these species also have minimal facilitative impacts and may lack any societal benefits. These species are unlikely to be accepted and may require prolonged management or control. Undaria, however, is a large primary producer, which may provide a trophic and habitat subsidy to native communities within some systems. Although more site-specific research is needed, in many cases, it has also been recorded as having minimal deleterious impact on native species. There is also commercial potential, with both wild harvest and rope-based mariculture conducted in parts of Undaria's non-native range (Castric-Fey et al., 1993;MAF, 2010;Perez-Cirera et al., 1997;Peteiro et al., 2016). In areas where likelihood of controlling Undaria is low due to widespread established populations, and context-specific studies show limited ecological impact, it may be that Undaria becomes one of few marine invasive species accepted as part of the local biota, with the potential for further development as a commercial resource.

| CONCLUSIONS
There are many challenges facing the future of marine invasion ecology. Total prevention of introductions of new NIS is highly unlikely, while management or eradication is extremely costly and often infeasible. Invasive species are likely to continue their spread and become conspicuous and prominent components of coastal marine communities. In many cases marine invasive species have clearly detectable deleterious impacts on recipient communities; however, in many others their influence is often limited and site-specific. Undaria has now been established for over 40 years in some of its non-native range. In these areas, rapid response or eradication is no longer an option and the need for any targeted management should be considered.
Although not yet conclusive, Undaria seems to have minimal ecological impacts in most invaded locations and does not appear to be a "driver" of ecosystem change in most contexts. If this is shown to be the case, it may be more beneficial to target management effort toward the causes of ecosystem stress that reduce native biotic resistance and allow Undaria to proliferate, rather than attempting to exclude the species itself. Further research is needed before well-considered, evidence-based management decisions can be made on a case-by-case basis. However, if Undaria was to become officially "unmanaged" in parts of its non-native range and accepted as a component of the native flora, the presence of a habitat forming, primary producer with a broad ecological niche and potential commercial value, may deliver significant economic and even environmental benefit. How science and policy reacts to the continued spread and proliferation of Undaria may influence how similar marine invasive species are handled in the future. supervision. We would also like to thank the reviewers for their insightful, constructive comments, which helped to greatly improve the quality of this manuscript.

CONFLICT OF INTEREST
None declared.

AUTHOR CONTRIBUTIONS
G.E. is the primary author and produced the majority of the content of this review. D.A.S. was involved throughout the process from first draft to final manuscript, including conception, composition, critical review, and final approval for submission.