Past and future of the marine bioinvasions along the Southwestern Atlantic

The first comprehensive survey of marine bioinvasions in the southern Southwest Atlantic Ocean (SWA, ca. 33°45′–ca. 54°50′S) published in 2002 reported 31 introduced and 46 cryptogenic species. In this assessment, we update this work by providing a deep historical perspective of marine biological invasions as well as a full new review of introductions that have occurred over the past nearly two decades. We reviewed a variety of sources including peer-reviewed journals and monographs, government reports, and museum databases, among others. The previous survey was re-evaluated and taxonomic experts were also consulted. For each species, the possible status as introduced or cryptogenic, likely vector, likely native region, first collection date if known, and the first authoritative reference of its SWA occurrence were determined. Species were categorized by biogeographical provinces. Our re-evaluation of species status lowered the 2002 survey to 29 introduced and 29 cryptogenic species. The result of our analyses added 100 new introduced and 43 new cryptogenic species, making a total of 129 introduced and 72 cryptogenic marine species for the SWA. Of these 100 species, 67 were found in the literature older than the 2002 survey, and 33 were found as new invasions that occurred since 2002, averaging one new invasion every 178 days. Ships are the most likely vector of invasions with an additional few species introduced with aquaculture, for ornamental purposes or for stabilizing coastal dunes. Most species are native to either the North Pacific or North Atlantic Oceans. Most introduced species (51%) occur in the warm temperate SWA marine ecoregion, while fewer (16%) occur in the cold temperate Magellanic marine ecoregion. Since ships are the main vectors in the region, we forecast that new introductions will continue until regional and international regulations to control ballast water and hull fouling begin to be effective. Our results urge the creation of long-term assessment programs focused on marine coastal biodiversity as a way to avoid inefficient environmental management based on deficient baseline information.


Introduction
The coasts of Uruguay and Argentina, spanning over 20 degrees of latitude from ca. 33°44′ to 54°50′ S in the southern Southwestern Atlantic Ocean (hereafter SWA) have played a critical role in maritime history since European exploration and colonization began in the 1500s. From the rounding of Cape Horn in 1520 by the Volcano of the Magellan-Elcano expedition to the opening of the Panama Canal in 1914, the ports and harbors of the southeast coast of South America were key components of major global shipping corridors. Not surprisingly, by the time the Beagle arrived in Buenos Aires in 1832, the European cardoon, or artichoke thistle (Cyanara cardunculus (L.) Baill.), was so common that in September 1833 Charles Darwin wrote, "I doubt whether any case is on record of an invasion on so grand a scale of one plant over the aborigines" (Darwin 1878). By 1874, the botanist Carlos Berg (1877) reported that fully half of the terrestrial plant species that he found growing wild in the Buenos Aires Province and Patagonia were European.
Despite more than 500 years of maritime traffic, and despite the early recognition of the significant presence of non-native species on lands adjacent to the sea, it was not until 2002 that Orensanz et al. provided the first comprehensive survey of exotic marine organisms in the SWA, reporting 31 introduced and 46 cryptogenic species. Here, we update this work by providing a deep historical perspective of marine biological invasions as well as a full new review of introductions that have occurred over the past nearly two decades. We report the number of introductions to exceed 120 species, more than 4 times the number previously reported.

Materials and methods
Since the publication of Orensanz et al. (2002) we have either kept records of all new reports that have come to our attention, or searched for additional records, of introduced marine viruses, protists, animals, and plants reported from Uruguay and Argentina. Particular attention was paid to taxonomic groups that were either not considered in Orensanz et al. (2002), such as the ocean-dependent plants and animals (i.e., supralittoral, marsh, and dune organisms), or were treated only briefly. For this, we pursued some of the biogeographic and taxonomic categories considered by Carlton (2009) to represent groups that often remain uninvestigated, thus potentially leading to underestimation of invasion diversity. These categories include small-bodied taxa (such as viruses, diatoms, dinoflagellates, foraminifera and others), inshore and shelf-dwelling (neritic) taxa assumed to be dispersed by ocean currents and thus often formerly (and formally) eliminated from consideration as non-native candidates (such as teredinid bivalves, or shipworms), or species with generally presumed cosmopolitan distributions (such as many seaweeds). In addition, anadromous fish were omitted in Orensanz et al. (2002), which we now consider here.
Our sources included peer-reviewed journals and monographs, government reports, conference proceedings and museum databases. We further re-evaluated the biogeographic status of the introduced and cryptogenic species earlier reported in Orensanz et al. (2002). We consulted with taxonomic experts (Supplementary material Table S1, and Acknowledgments) to confirm species identifications or to discuss species' histories and possible origins. For each species under consideration, the possible status as introduced or cryptogenic, likely vector, probable native region, first collection date if known, and the first authoritative reference of its SWA occurrence were determined. If a collection date could not be determined, the publication date was used (and so noted) as the first date of record.
For biogeographical analyses, species were categorized by marine ecoregion as Warm Temperate Southwestern Atlantic and Magellanic (Spalding et al. 2007) and by occurrence between Chuy Creek (33°44′S) to Ushuaia (ca. 54°50′S). Native areas for each introduced species were assigned according to the literature, and then largely classified according to Food and Agriculture Organization of the United Nations regions (FAO 2019). FAO regions were used instead of marine ecoregions in order to capture broader origin perspectives. For habitat analyses, species were categorized as occurring in rocky or sandy shores, mudflats, rocky or salt marshes, ports, or other (Bortolus et al. 2009, and herein). Six introduced and seven cryptogenic species of uncertain establishment (Table S1) have been conservatively retained in all of our calculations.
Probable vectors of introduction were determined by taking into account life histories, habitat, and other biological and ecological attributes following the categories of Carlton and Eldredge (2009). These vectors include ballast water, ship fouling, ship boring (organisms boring into wooden ships, or living in such burrows), solid ballast (rocks and sand), releases (intentional or accidental by private individuals or government agencies and species associated with such releases), and other (such as carriage with commercial products, in airplane cabins, resting or egg stages with luggage and clothing, etc.). Ship categories were combined when vectors could not be distinguished, thus all species were assigned to only one of the categories. Binomial names and classifications largely follow the World Register of Marine Species (WoRMS 2019), AlgaeBase (Guiry and Guiry 2019) or FishBase (Froese and Pauly 2019) and specific cases with taxonomic remarks (e.g. Bortolus et al. 2019).

Introduced and Cryptogenic Diversity and Detection
A total of 129 (64%) introduced and 72 (36%) cryptogenic species are here reported in the SWA (Table 1 and Table S1). Our re-evaluation of the status  Table S1 for first collection dates, vectors, possible origins, detailed remarks, and references. * Introduced species; unmarked taxa are cryptogenic.  of species treated in Orensanz et al. (2002), detailed below, revised the species treated therein to 29 introduced and 29 cryptogenic species (Table 2). Our present work thus adds 100 introduced and 43 cryptogenic species, for a total of 201 species now under consideration as confirmed or possible marine bioinvasions (Table 3). Overall, since Orensanz et al. (2002), 33 introduced species have been newly documented in the SWA. Of these 100 introduced species, half (51) represent numerous taxonomic groups that were either entirely omitted from, or with two exceptions, largely not treated by Orensanz et al. (2002) (Table 3). One of the exceptions is the wakame Undaria pinnatifida (Harvey) Suringar, 1873, which was included in the previous survey. We have now added 16 introduced algae species. The other exception was a salt marsh plant, Spartina anglica C.E. Hubb. which was deleted from the list because it is considered extinct in Argentina (Nicora 1978), while 11 additional introduced maritime plants were added. In turn, we note that the majority (40, or 78%) of these 51 species were present (Table 3)  The taxon that was most represented were crustaceans with 30 introduced and 16 cryptogenic species (Figure 1), mostly represented by amphipods with 12 and 6 introduced and cryptogenic species, respectively. The second most speciose group was algae, with 20 introduced and 10 cryptogenic species, followed by bryozoans with 13 introduced and 10 cryptogenics, and annelids and cnidarians with 8 introduced and 11 cryptogenics for each group. The detection of introduced and cryptogenic species increased constantly over time (Figure 2), but a pulse occurred between 1960 and 1970 when 19% of the introduced species were first observed, mainly crustaceans and tunicates ( Figure 3). Nearly 70% of introduced species were reported during the last 50 years (Figure 2). Orensanz et al. (2002) Of the 31 species treated as introduced in Orensanz et al. (2002) to the SWA, we have retained 29. The two deleted species are the salt marsh cordgrass Spartina anglica C. E. Hubb (see above), and the isopod Idotea metallica Bosc, 1802, a native pelagic, oceanic species. Of the 46 cryptogenic species, we have retained only 29. Ten species reported as cryptogenic in   1900 1901-1910 1911-1920 1921-1930 1931-1940 1941-1950 1951-1960 1961-1970 1971-1980 1981-1990 1991 or are under further study and thus were deleted. Two species of crabs previously listed as cryptogenic we now consider as native (Panopeus meridionalis Williams, 1983) or of uncertain identity regarding the original record considered (Pachygrapsus transversus (Gibbes, 1850)), thus they were also deleted.

Earliest Records
The earliest records of introduced marine species in the SWA are the red algae Melanothamnus harveyi (Bailey)

Vectors
Ships are the most likely vector (85%) for transporting introduced species, either in fouling (44.5%) or by fouling or ballast water (26.5%) (Figure 4). Aquaculture-related releases of anadromous fish and Japanese oysters are

Source Regions and Regional and Habitat Distribution
Most introduced species appear to be native to the northern hemisphere (93%, Figure 5), specifically from the Northeast and Northwest Atlantic, the Northwest Pacific, Asia or Southeast Asia, and Europe (Table S1). Regarding the biogeographical distribution, most introduced species (51%) are found only in the warm temperate SWA marine ecoregion, while fewer (16%) occur only in the cold temperate Magellanic ecoregion ( Figure 6). Most cryptogenic species are widely distributed in both marine ecoregions ( Figure 6). Mar del Plata, in the warm temperate SWA (38°S) and its surrounding areas have the highest number of both introduced and cryptogenic species (Figure 7), while this number diminishes at higher and lower latitudes. Most introduced and cryptogenic species are found in ports as well as in rocky shore habitats (Figure 8).

Invasion Diversity and Potential Underestimation
While the diversity of introduced and cryptogenic species in the Southwest Atlantic Ocean (SWA) reported here has increased the number of recognized invasions or possible invasions nearly 4.5 times and 2.5 times, respectively (based upon revised calculations of the data in Orensanz et al.  2002; Table 2 herein), we have no doubt that the actual number of bioinvasions along the southeast coast of South America is significantly higher than previously thought. In the past 17 years, 33 new invasions have been documented (based upon first new records of non-native species from 2002 and later), averaging one new invasion every 178 days (Orensanz et al. 2002;Schwindt and Bortolus 2017). This invasion rate is itself an underestimate, given the number of probable new but undetected invasions in many unstudied or understudied groups of protists, invertebrates, and algae (Carlton 2009;Bortolus et al. 2015). This said, most new records represent the results of investigations of taxa not explored in an earlier study or of re-evaluation of species previously reported as cryptogenic (Table 3). Around the world, regions that were thought to have sustained relatively few invasions were found, upon detailed study, to have significant numbers of both introduced and cryptogenic species. Examples include California (San Francisco Bay, Cohen and Carlton 1995), the Hawaiian Islands (Carlton andEldredge 2009, 2015), Australia (Hewitt et al. 2004;Sliwa et al. 2009), New Zealand (Cranfield et al. 1998Hayden et al. 2009), South Africa (Mead et al. 2011), Brazil (Lopes 2009Rocha et al. 2013), and most recently the Galapagos Islands . Despite this, in all of these regions, a very large suite of taxa continue to remain unstudied due to long-term challenges in interpreting the historical and modern biogeography of many groups (Carlton 2009;Bortolus et al. 2015) as well as a steadily increasing dearth of taxonomic expertise (Carlton and Fowler 2018).
As an example, one of many obfuscated taxonomic and biogeographic puzzles encountered in our work was the red-tide producing dinoflagellate Alexandrium. It was first recorded in the SWA as Gonyaulax excavata (Braaerud) Balech, 1971, following the first human intoxication reported in 1980 in Argentina in the Valdes Peninsula (Carreto et al. 1981). Although long considered as Alexandrium tamarense (Lebour, 1925) Balech, 1995 after global genetic and taxonomic comparisons, it has now been accepted as A. catenella (Whedon & Kofoid) Balech (John et al. 2014;Fraga et al. 2015, Table S1) and considered probably native to the North Atlantic or North Pacific oceans. Another toxic-bloom producing dinoflagellate, Dinophysis acuminata Claparede and Lachmann, 1859, was first recorded in 1941 by Balech (1944). Neither of these species were discussed in the regional phytoplankton literature as exotic invasive species, due to the assumption that they had been transported by ocean currents from some unknown overseas shores, despite the lack of evidence for this dispersal mechanism, and the concomitant abundance of data on the role of ballast water in transporting phytoplankton (including these species of dinoflagellates) globally (Bailey 2015). Further efforts should be directed to determine how many additional species of introduced phytoplankton remain undetected in the SWA. In many regions worldwide-with Argentina and Uruguay being no exception-long expanses of coastline remain unexplored. We found that most of the introduced species are concentrated in the warm temperate SWA of Uruguay and in Mar del Plata in the northern portion of the coast of Argentina. This pattern, as previously highlighted (Orensanz et al. 2002), likely reflects long-term sampling biases. Most historical and recent biofouling harbor studies were in this marine ecoregion (Bastida 1971;Brankevich et al. 1984Brankevich et al. , 1988Liuzzi 2016, 2018;Rumbold et al. 2018;Albano and Obenat 2019). Since 2002, however, there has been an increase in invasion research in colder regions, including Puerto Madryn, Comodoro Rivadavia and San Antonio Este, three important port areas of Patagonia (Rico et al. 2010(Rico et al. , 2012Schwindt et al. 2014), leading to a significant uptick in our understanding of the scale of invasions and their impact in Patagonia, with 21 introduced species now reported for this region. This is a striking shift since only two introduced species (Undaria pinnatifida (Harvey) Suringar, 1873 and Spartina anglica) were reported for Patagonia by Orensanz et al. (2002). Indeed, several new and wellknown introduced species, such as the European green crab Carcinus maenas (Linnaeus, 1758) and the club tunicate (ascidian) Styela clava Herdman, 1881 were first detected along these cold temperate shores (Hidalgo et al. 2005;Pereyra et al. 2015). Even more striking perhaps is the finding that some of the introduced species have not only colonized different environments than in their place of origin, but also showed novel unpredicted behaviors, such as a rocky shore barnacle colonizing soft bottom saltmarshes (Schwindt et al. 2008), or soft bottom benthic snails traveling fast across long distances as biofouling on sea turtles (Lezama et al. 2013).

Vectors and Donor Routes
The primary vector for marine introduced and cryptogenic species in the SWA is ships. There is no surprise there, as the same pattern has been observed worldwide (Hewitt et al. 2004;Mead et al. 2011;Carlton et al. 2019). Since the early 1500s when ships and explorers arrived to the region, many species (such as algae, barnacles, and tunicates) were transported in external hull fouling and inside the wooden hulls (such as shipworms), as well as mixed with solid ballast (such as the cordgrass Spartina alterniflora; Bortolus et al. 2015). As vessel technology evolved, water replaced rocks and sand as ballast. Nevertheless, in the past 20 years, half of the introduced species we report in this work were likely transported by hull fouling, highlighting the current importance of this vector.
The introduction of species along the SWA due to escape from aquaculture activities was not yet reported as important since this activity is not highly developed as in other countries within the region. Besides the white spot syndrome virus (WSSV) presumably associated with unregulated aquaculture activities , only one invertebrate species, the Japanese oyster Crassostrea gigas (Thunberg, 1793), was intentionally introduced in the SWA from Chile in 1982 during an aquaculture farming initiative that later failed, with the oysters then being abandoned (Orensanz et al. 2002). While the Japanese oyster is no longer economically exploited, it is spreading along SWA muddy and rocky shores making difficult to walk bare footed along some of the invaded touristic beaches (dos Santos and Fiori 2010;Schwindt et al. pers. obs.). Aquaculture farms are widespread along the Chilean coast, and they are recognized as an economically important activity. Nevertheless, none of the introduced species established in that country were reported from escapes (Castilla et al. 2005), except for salmon (Arismendi et al. 2014) where escapes are documented. For example, a massive escape of 900,000 salmon occurred recently in southern Chile after a heavy storm and with only 28% of the fish recovered (Gómez-Uchida et al. 2018). Within this context, Argentina keeps debating whether to install salmon farms in the southern part of the country despite the intense opposition of global and local communities and the lack of support from expert scientists who predict dramatic negative effects on the nearby marine protected areas and surroundings FCMPAI 2018;Summers 2019).
Most introduced species reported for the SWA appear to have originated in the northern hemisphere, mostly from the North Atlantic and specifically from Europe. Nevertheless, it would be wrong to assume that native regions are always the donor sources since many introduced species follow complex unpredicted travel routes. For instance, genetic studies on two of the most aggressive invaders worldwide, the green crab Carcinus maenas and the seaweed Undaria pinnatifida, native to the Northeast Atlantic and Northwest Pacific respectively, demonstrated that individuals found in Argentina were genetically more closely related to populations from Australia, where both species were also introduced (Voisin et al. 2005;Darling et al. 2008). The Austral cordgrass Spartina densiflora Brongn., native to the SWA, was introduced to the Pacific coast of North America where it currently has invaded most salt marshes. In this case, the donor region was Chile, and not the Atlantic coast of South America (Bortolus 2006;Fortuné et al. 2008). In a similar vein, the dune marram grass Ammophila arenaria, although native to Europe, was most likely introduced to continental Argentina from the Malvinas/Falkland Islands, where its date of arrival, amount of imported seeds and exact location were recorded in detail (Hubbard 1937).
With an increasing global connectivity during the 20 th century, the reports of introduced species in the SWA native to Southeast Asia, the Indo-Pacific and other Pacific regions became also increasingly common (Table S1). During the last 50 years, one third of the introduced species reported here arrived from the Pacific Ocean. A similar result was observed in Australia by Hewitt et al. (2004), documenting that a shift in trade led to a change of marine invasions from the North Atlantic to the North Pacific after 1950. A continuing future trend is expected for the SWA, where by 2050 a high number of introduced species are predicted to arrive from Asia (Sardain et al. 2019). A first example of this change in global trade may be the introduction of the sea slug Pleurobranchaea maculata, native to the Pacific region and recently reported in Argentina as the only country outside its native range (Farías et al. 2015(Farías et al. , 2016Battini et al. 2019).

Conclusions
We cannot make progress in our understanding of the impact of invaders if the scale of invasion diversity remains poorly assessed, and if the abundance and distribution of already established invaders remain uncertain. The most direct consequences of delays in recognizing invasions are delays in early detection of and rapid response to new invasions, and the hampering of the design of strong, effective management strategies to prevent future introductions. Our results argue strongly for regularly updated regional assessments of marine coastal biodiversity. While the United States and many European countries maintain online databases (such as NEMESIS (Fofonoff et al. 2019) and NOBANIS), for other lessknown regions such as Central and South America, Africa and much of Asia, data availability for marine introduced species remains both sorely limited and urgently needed. In this vein, the specific support to the marine biodiversity with emphasis on taxonomy has a central role (e.g. Orensanz et al. 2002;Scarabino 2006;Salazar-Vallejo et al. 2008).
Like in a rearview mirror, the notion of a pristine region associated with the SWA fades away while we progress in our analysis. However, Argentina and Uruguay have indeed started to work in the right direction with the National Strategy of Invasive Exotic Species (https://www.argentina.gob.ar/ ambiente/biodiversidad/especiesinvasoras and https://bit.ly/2NgLum9) and the online database provided by the Inter-American Biodiversity Information Network (IABIN) through the Invasives Information Network (I3N, http:// www.inbiar.uns.edu.ar/), as well as the production of new identification field guides focused on marine-coastal introduced species . However, since these initiatives tend to be threatened by disrupted or unreliable long-term governmental support, marine biodiversity is still often either unrepresented or under-represented in many cases. Since ships are the main vectors, we forecast that new introductions will continue to occur until regional and international regulations directed to control ballast water and hull fouling are effectively and globally integrated across oceans, seas, river basins and political borders.