Ostreopsis cf. ovata Bloom in Currais, Brazil: Phylogeny, Toxin Profile and Contamination of Mussels and Marine Plastic Litter

Ostreopsis cf. ovata is a toxic marine benthic dinoflagellate responsible for harmful blooms affecting ecosystem and human health, mostly in the Mediterranean Sea. In this study we report the occurrence of a summer O. cf. ovata bloom in Currais, a coastal archipelago located on the subtropical Brazilian coast (~25° S). This bloom was very similar to Mediterranean episodes in many aspects: (a) field-sampled and cultivated O. cf. ovata cells aligned phylogenetically (ITS and LSU regions) along with Mediterranean strains; (b) the bloom occurred at increasing temperature and irradiance, and decreasing wind speed; (c) cell densities reached up to 8.0 × 104 cell cm−2 on fiberglass screen and 5.6 × 105 cell g−1 fresh weight on seaweeds; (d) and toxin profiles were composed mostly of ovatoxin-a (58%) and ovatoxin-b (32%), up to 35.5 pg PLTX-eq. cell−1 in total. Mussels were contaminated during the bloom with unsafe toxin levels (up to 131 µg PLTX-eq. kg−1). Ostreopsis cells attached to different plastic litter, indicating an alternate route for toxin transfer to marine fauna via ingestion of biofilm-coated plastic debris.


Introduction
Benthic dinoflagellates belonging to the genus Ostreopsis are cosmopolitan, present in both tropical and temperate areas [1]. Several of the eleven Ostreopsis species currently described are reported to be toxic, although taxonomic confusion exists as some species were previously described based solely on morphological features, lacking molecular biology analyses [2]. One of the most toxic species, Ostreopsis cf. ovata, has been responsible for blooms affecting both human and animal health worldwide [3][4][5][6]. This dinoflagellate produces toxins similar to the palytoxins, i.e., isobaric palytoxin (PLTX) and ovatoxins (OVTXs) [7,8], which can intoxicate humans by inhalation or the ingestion of contaminated seafood. However, a great variability in toxin profile among species, strains, and cultured cells. We examined the vertical and temporal distribution of the bloom and the interaction of O. cf. ovata with natural and artificial substrates, including different types of plastic litter. The toxin profile was investigated in field-sampled and cultivated cells, as well as in aquatic invertebrates naturally and experimentally exposed to the toxic cells in situ.

Bloom Detection
The bloom was detected by chance during a regular SCUBA dive sampling campaign on February 16th 2017 in Currais Archipelago (Figure 1). Three days later, a yellowish biofilm was noticed covering a~50,000 m 2 area of the seafloor, extending from 0 to 8.0 m depth (Figure 2a,b). On February 19th, mucous cell aggregates were found floating abundantly at the sea surface (Figure 2c,d). Cell density over the benthic substrates increased from February 16th to the 19th, remaining similarly high until the last sampling day on February 23 rd (Figure 3). Ostreopsis reached a maximum of 5.6 × 10 5 cell g −1 of seaweeds fresh weight (fw), and 8.0 × 10 4 cell cm −2 of the artificial substrate-fiberglass screen ( Figure 3). Vertical distribution, as assessed on February 19th using artificial substrate, revealed much higher cell densities (7.4 × 10 4 cell cm −2 ) in shallower areas (1.5 m depth). Cell abundance decreased exponentially as light diminished in deeper areas, reaching 0.6 × 10 4 cell cm −2 at 6.0 m depth ( Figure 3). During the 30 days preceding the first sampling campaign, air temperature, sea surface temperature (SST) and solar radiation were increasing, while wind speed and cloud coverage were decreasing. On the first sampling day, daily-average SST had increased from 27.9 to about 28.8 • C, wind had decreased from approximately 6.6 to 4.3 m s −1 and surface radiation was~960 kJ m −2 . Another bloom recurred in February 2018 at similar environmental conditions: increasing SST, decreasing wind speed and cloud coverage. During both events, samples were obtained for cell culture establishment, however spatio-temporal distribution was only determined for the earlier event, as described above. of O. cf. ovata with natural and artificial substrates, including different types of plastic litter. The toxin profile was investigated in field-sampled and cultivated cells, as well as in aquatic invertebrates naturally and experimentally exposed to the toxic cells in situ.

Bloom Detection
The bloom was detected by chance during a regular SCUBA dive sampling campaign on February 16 th 2017 in Currais Archipelago (Figure 1). Three days later, a yellowish biofilm was noticed covering a ~50,000 m 2 area of the seafloor, extending from 0 to 8.0 m depth (Figures 2a and  2b). On February 19 th , mucous cell aggregates were found floating abundantly at the sea surface (Figures 2c and 2d). Cell density over the benthic substrates increased from February 16 th to the 19 th , remaining similarly high until the last sampling day on February 23 rd (Figure 3). Ostreopsis reached a maximum of 5.6 × 10 5 cell g −1 of seaweeds fresh weight (fw), and 8.0 × 10 4 cell cm −2 of the artificial substrate-fiberglass screen ( Figure 3). Vertical distribution, as assessed on February 19 th using artificial substrate, revealed much higher cell densities (7.4 × 10 4 cell cm −2 ) in shallower areas (1.5 m depth). Cell abundance decreased exponentially as light diminished in deeper areas, reaching 0.6 × 10 4 cell cm −2 at 6.0 m depth ( Figure 3). During the 30 days preceding the first sampling campaign, air temperature, sea surface temperature (SST) and solar radiation were increasing, while wind speed and cloud coverage were decreasing. On the first sampling day, daily-average SST had increased from 27.9 to about 28.8 °C, wind had decreased from approximately 6.6 to 4.3 m s −1 and surface radiation was ~960 kJ m −2 . Another bloom recurred in February 2018 at similar environmental conditions: increasing SST, decreasing wind speed and cloud coverage. During both events, samples were obtained for cell culture establishment, however spatio-temporal distribution was only determined for the earlier event, as described above. , showing the Ostreopsis bloom location (Currais Archipelago, detailed). In the first detailed map an arrow shows the location of Galheta Island, where mussels were firstly collected. In the second detailed map the rectangle shows the exact area affected by the bloom. , showing the Ostreopsis bloom location (Currais Archipelago, detailed). In the first detailed map an arrow shows the location of Galheta Island, where mussels were firstly collected. In the second detailed map the rectangle shows the exact area affected by the bloom.  Light percentage (L%) at a given depth (z) was calculated by the formula "L% = 100% × exp (−k × z)", were "k" is the light attenuation coefficient (1.7 divided by the Secchi disc depth) [34].

Species Identification
Cells from monoclonal cultures (4 strains) and field samples were oval and ventrally slender in apical and antapical views, as assessed by light, epifluorescence and electron microscopy ( Figure 4).  Light percentage (L%) at a given depth (z) was calculated by the formula "L% = 100% × exp (−k × z)", were "k" is the light attenuation coefficient (1.7 divided by the Secchi disc depth) [34].

Species Identification
Cells from monoclonal cultures (4 strains) and field samples were oval and ventrally slender in apical and antapical views, as assessed by light, epifluorescence and electron microscopy ( Figure 4). . Light percentage (L%) at a given depth (z) was calculated by the formula "L% = 100% × exp (−k × z)", were "k" is the light attenuation coefficient (1.7 divided by the Secchi disc depth) [34].

Species Identification
Cells from monoclonal cultures (4 strains) and field samples were oval and ventrally slender in apical and antapical views, as assessed by light, epifluorescence and electron microscopy ( Figure 4). They were 23.7-65.9 µm (mean = 43.1 µm, standard deviation (SD) = 9.1, n = 318) deep (dorso-ventral length, DV), 15.4-48.9 µm (mean = 31.4 µm, SD = 7.1, n = 270) wide (W) and 16.7-44.5 µm (mean = 26.1, SD = 5.7, n = 45) long (antero-posterior length, AP). The DV/W ratio was 1.04-1.79 (mean = 1.38, SD = 0.15, n = 326). Cultivated cells were smaller and more rounded (DV/W ratio = 1.33, SD = 0.13, n = 238) than those sampled directly from the field (DV/W ratio = 1.53, SD = 0.12, n = 88) ( Table 1). All cell dimensions were also more variable in cultures (SD of DV = 8.7, W = 7.5, AP = 6.3) than those sampled from the field (SD of DV = 6.6, W = 5.9, AP = 3.3). The thecal plate pattern was APC 3 7 5 and 2 , and the thecal surface was smooth. Mean diameter of thecal pores was 0.28 µm (SD = 0.04, n = 6), with internal structures usually splitting it into five poroids, with the presence of a few smaller pores (~0.06 µm in diameter) on the thecal surface ( Figure 4L). The first apical plate (1 ) was large and hexagonal. Suture of 1 with the third apical (3 ) plate varied from straight to curved in different specimens ( Figure 4A-H). The second apical plate (2 ) was always narrow and elongated, and located below the APC, reaching the fourth precingular plate (4"), and separating the third precingular (3") plate from the 3 plate ( Figure 4E,I). In most examined cells, 3 was pentagonal in shape and contacting 1 , 2 , 3 , 4 and the fifth precingular (5") plates ( Figure 4A,E). However, in some other cells, the suture between 1 and 5" was short or absent. In this case, 3 was more hexagonal sometimes also touching the sixth precingular (6") plate ( Figure 4B-D,F-H).     Species identification was confirmed by phylogenetic analyses based on ITS region (ITS 1, 5.8S rDNA and ITS 2) and partial LSU rDNA (D8-D10 domains). Both analyses included sequences from five monoclonal cultures, one from a cell pellet obtained from a field sample, and other sequences retrieved from GenBank. The final ITS alignment comprised 51 sequences (including one outgroup sequence) and had a length of 336 base pairs. The best-fit model was found to be GTR + G model Phylogenetic analyses were performed with two methods of reconstruction: maximum likeligood (ML) and Bayesian Inference (BI,). Considering that ML and BI analyses gave the same tree topology and relationships among clades, only the majority-rule consensus tree of the ML analysis is shown. Twelve distinct clades were found in the phylogeny inferred from ITS sequences (O. cf. ovata, O. cf.

Colonization of Plastic Litter by Microalgae
During the 2017 bloom, the abundance of attached Ostreopsis cells varied among artificial samplers made of different materials, including a fiberglass screen similar to that described in Tester et al. [32], and four types of plastic litters commonly found in the gastro-intestinal tract of green-turtles in the region [26,35]. Cell density on rigid polypropylene bottle caps (R-PP) was lower (mean 1.3 × 10 3 cell cm −2 ), with no significant differences between white (R-PPw) or red (R-PPr) colors ( Figure 7). Sections of low density polyethylene plastic bags (LDPE) accumulated up to 4.9 × 10 3 cell cm −2 , while higher densities were found on those made of flexible polypropylene plastic packing (F-PP) (mean 8.4 × 10 3 cell.cm −2 ) and fiberglass screen (mean 80 × 10 3 cell cm −2 ). Cell density of co-occurring diatoms was much higher in R-PP (mean = 9.9 × 10 2 cell cm −2 , or 76% of Ostreopsis cell abundance) compared to both flexible plastic litter-LDPE and F-PP (mean = 2.8 × 10 2 cell cm −2 , only 4% of Ostreopsis cell abundance). Proportionally, cell density of diatoms was also lower than that of Ostreopsis on fiberglass screens (26% of Ostreopsis cell abundance) after 24 h of exposure ( Figure 7).

Colonization of Plastic Litter by Microalgae
During the 2017 bloom, the abundance of attached Ostreopsis cells varied among artificial samplers made of different materials, including a fiberglass screen similar to that described in Tester et al. [32], and four types of plastic litters commonly found in the gastro-intestinal tract of greenturtles in the region [26,35]. Cell density on rigid polypropylene bottle caps (R-PP) was lower (mean 1.3 × 10 3 cell cm −2 ), with no significant differences between white (R-PPw) or red (R-PPr) colors ( Figure  7). Sections of low density polyethylene plastic bags (LDPE) accumulated up to 4.9 × 10 3 cell cm −2 , while higher densities were found on those made of flexible polypropylene plastic packing (F-PP) (mean 8.4 × 10 3 cell.cm −2 ) and fiberglass screen (mean 80 × 10 3 cell cm −2 ). Cell density of co-occurring diatoms was much higher in R-PP (mean = 9.9 × 10 2 cell cm −2 , or 76% of Ostreopsis cell abundance) compared to both flexible plastic litter -LDPE and F-PP (mean = 2.8 × 10 2 cell cm −2 , only 4% of Ostreopsis cell abundance). Proportionally, cell density of diatoms was also lower than that of Ostreopsis on fiberglass screens (26% of Ostreopsis cell abundance) after 24 h of exposure ( Figure 7).

Toxin Production and Accumulation in Marine Organisms
The presence of Ovatoxin (OVTX) -a, b, c, d and e was detected in all samples, including cell pellets from the bloom and cultured cells (Figure 8). The total toxin quota was higher in cultures (up to 35.5 pg cell −1 in the stationary phase of LM-062 culture) than in field samples (mean of 12.2 pg cell −1 ) (Figure 9). Intracellular toxin levels also varied among different strains at equivalent growth stage. Strains isolated from the 2018 bloom (LM-128 and LM-129) were slightly less toxic than those previously established (Figure 9). The toxin profile, however, was more conservative, composed mostly of OVTX-a (approx. 58% of the total toxin content in all samples) and OVTX-b (30-35% depending on the sample). Overall, OVTX-c, OVTX-d and OVTX-e contributed each to approximate

Toxin Production and Accumulation in Marine Organisms
The presence of Ovatoxin (OVTX) -a, b, c, d and e was detected in all samples, including cell pellets from the bloom and cultured cells (Figure 8). The total toxin quota was higher in cultures (up to 35.5 pg cell −1 in the stationary phase of LM-062 culture) than in field samples (mean of 12.2 pg cell −1 ) (Figure 9). Intracellular toxin levels also varied among different strains at equivalent growth stage. Strains isolated from the 2018 bloom (LM-128 and LM-129) were slightly less toxic than those previously established (Figure 9). The toxin profile, however, was more conservative, composed mostly of OVTX-a (approx. 58% of the total toxin content in all samples) and OVTX-b (30-35% depending on the sample). Overall, OVTX-c, OVTX-d and OVTX-e contributed each to approximate 3% of the total toxin content, except at exponential growth phase of culture LM-062, in which OVTX-c and OVTX-d represented up to 5% of the total toxin content each ( Figure 9). Isobaric palytoxin was not detected in any sample (Figure 8).
During the 2017 bloom (on February 22rd), we detected toxin transfer to Perna perna mussels that were collected on a nearby (~16.5 Km) location-Galheta Island-and transplanted to the bloom area in Currais Arquipelago, remaining at the bottom (1.5 m depth) for 24 h. Up to 131 µg kg −1 of total OVTX (mean 98.0 µg kg −1 , n = 5) were accumulated in mussels after the 24 h exposure period ( Table 2). In general, toxins in mussels were mostly composed of OVTX-a (68.5%) and OVTX-b (27%), with smaller amounts of OVTX-c (up to 10%) and OVTX-e (up to 4.5%) ( Table 2). Curiously, mussels sampled from Galheta Island already contained smaller OVTX amounts (up to 32.9 µg kg −1 ), even though Ostreopsis cell abundance were much lower there (up to 5.0 × 10 3 cell g −1 of seaweeds fw) compared to the bloom area in Currais (up to 560 × 10 3 cell g −1 of seaweeds fw). In addition, OVTX-a and OVTX-b (29.5 µg kg −1 in total) were also found in a single sample of coral (Palythoa sp.) naturally occurring in Currais. Conversely, toxins were undetectable (<20 ng PLTX-eq. mL −1 ) in sea urchins (Echinometra lucunter) and one sea cucumber (Holothuria grisea) sampled in the same location (Table 2). Table 2. Toxin profile in marine invertebrates (whole tissue homogenates) collected during the 2017 Ostreopsis cf. ovata bloom in Currais Archipelago, southern Brazil. Sea urchins (Echinometra lucunter), sea cucumber (Holothuria grisea) and coral (Palythoa sp.) were sampled from Currais on February 28th. Mussels (Perna perna) were collected on the same date in the nearby location of Galheta Island (where Ostreopsis cells were much less abundant), and analyzed either directly after sampling or following a 24-h transplantation ("transp.") period in Currais. Except for a pool of coral polyps, samples constituted one individual each. Average toxin amounts ("Mean total") are expressed in µg PLTX-eq. kg −1 of the animal whole tissue. LOD = limit of detection; LOQ = limit of quantitation.       [2,36,37]. Moreover, Ostreopsis spp. exhibit great intra-specific variability and inter-specific similarity in cell morphology, leading to significant problems regarding the taxonomy of the genus [2,38,39]. Recently, some clarification was obtained on oval-shaped larger-celled Ostreopsis species, with the redescription of O. lenticularis from its type locality and the indication that Ostreopsis sp. 6 may correspond to the originally described O. siamensis [2]. However, for the smaller-celled species (O. cf.   [2,36,37]. Moreover, Ostreopsis spp. exhibit great intra-specific variability and inter-specific similarity in cell morphology, leading to significant problems regarding the taxonomy of the genus [2,38,39]. Recently, some clarification was obtained on oval-shaped larger-celled Ostreopsis species, with the re-description of O. lenticularis from its type locality and the indication that Ostreopsis sp. 6 may correspond to the originally described O. siamensis [2]. However, for the smaller-celled species (O. cf. ovata, O. cf. siamensis and similar species), the confusion regarding distinctive morphological features continues.
Cell shape and size was originally noted as a distinguishing feature for Ostreopsis cf. ovata identification [40], but the cell size of Ostreopsis cf. ovata and other smaller-celled species may exhibit great variability and overlap, as shown in the Mediterranean Sea and Atlantic Iberian coast [38,39]. This variability was also observed in the present study in Brazil, in which cells from the same strain varied up to 2-fold in cell length (Table 1). Also, our cultivated cells exhibited different morphology (i.e., smaller size and more rounded) when compared to field samples, as previously reported [38]. This could be the effect of either high nutrient content in the culture media [38] or differences in cell stages [41], and emphasizes that descriptions from cultures should be interpreted cautiously. However, despite the concerns of using cell shape in Ostreopsis taxonomy, it is important to point out that O. cf. ovata cells sampled from Currais appeared wider than those originally described as O. ovata [40].
As size and shape overlap among species, plate characteristics were used as a distinguishing morphological feature in recent Ostreopsis descriptions [37,42]. However, even that feature was shown to be quite variable in the present study, and was not sufficient to separate O. cf. ovata from O. fattorussoi and O. rhodesiae: (a) for O. fattorussoi the presence of a curved suture between plates 1 and 3 , making plate 3 look hexagonal, was reported as a distinguishing characteristic [36], but in the present study ( Figure 4H) it was observed that 1 /3 curvature can be variable and should not be used as a sole characteristic; (b) for both O. rhodesiae and O. fattorussoi, the elongated second apical plate (2 ) separating the third apical plate (3 ) and the third pre-cingular plate (3") has been proposed as a distinguishing feature [36,37], however, this was also observed in our O. cf. ovata cells (Figure 4). Curiously, the 2 plate that was originally described to be short and limited to the apical pore length (not separating 3 from 3") in O. ovata (see drawings by Fukuyo [40]), have proved to be elongated in most recent pictures of cells belonging to the phylogenetic clade named as O. cf. ovata (see Figure 4A,E,I in the present study; Figure 4C,J in Penna et al. [39]; Figures 11 and 12 in Zhang et al. [43]; and Figure 55C in Hoppenrath et al. [33]).
The clade named as "O. cf. ovata" includes at least three morphologically identical but genetically distinct morphotypes. Since Sato et al. [44] found all three morphotypes in the O. ovata type locality (Ryukyu Islands, Japan [40]), it is not possible to associate either one with this taxonomic designation (cryptic diversity). Therefore, this species should be considered a species complex until further clarification. In the subclade of the O. cf. ovata species complex where the sequences from Currais aligned it is possible to find at least two strains from the Mediterranean Sea (IFR-OST01M and KC71, Figure 5) that were previously reported to be toxic [45].

Bloom Formation, Toxin Production and Contamination of Marine Organisms
In Brazil, Ostreopsis blooms have been previously reported in the oceanic archipelago of São Pedro e São Paulo (up to 9.9 × 10 4 cell g −1 of seaweeds [46]) and the coast of Rio de Janeiro State (up to 1.5 × 10 5 cell g −1 [47]), where negative effects to marine invertebrates have been documented [5]. The bloom described herein, however, is the first report for subtropical Brazilian waters. Cell abundances during the 2017 summer bloom in Currais Archipelago (5.6 × 10 5 cell g −1 of seaweeds or 8.0 × 10 4 cell cm -2 on artificial substrates) were equivalent to those reported in the Mediterranean (e.g., [48]), where extensive O. cf. ovata blooms are frequent and cell abundances of up to 7.2 × 10 6 cell g −1 of seaweeds or 6.4 × 10 4 cell cm −2 on artificial substrates can be reached [45,48,49]. Those massive blooms have often been associated with negative impacts to marine organisms and human health, due to the toxins produced by Ostreopsis [3,50,51]. Considering that cells from southern Brazil contained comparable toxin amounts, negative effects to marine fauna and human health are expected.
The 2017 O. cf. ovata bloom in Currais occurred after a period of increasing light availability. Additionally, much higher cell abundances were observed at shallower depths, suggesting that increased light intensity is an important factor triggering O. cf. ovata blooms. However, despite similar observations in previous field studies [52], laboratory experiments reported controversial results. Overall, the general environmental conditions preceding the bloom in Currais (i.e., warmer water temperatures and lower turbulence) were similar to those experienced in the Mediterranean Sea prior to O. cf. ovata blooms [10]. One year later, in February 2018 (austral summer), another O. cf. ovata bloom coincided with the period of maximum annual water temperatures (>28 • C) in Currais Archipelago; similar to what has been continuously observed in the Mediterranean (>25 • C [10]). Noteworthy, periods of high irradiance, warm temperature and low water turbulence only occur in the southern West Atlantic Ocean during short periods of the year in mid-summer and early autumn. Our results indicate that O. cf. ovata blooms should be carefully monitored over the subtropical Brazilian coast.
The southern Brazilian Ostreopsis populations sampled herein have not only the same genotype and similar environmental requirements for bloom formation, but also exhibit similar capacity to produce toxins as those causing toxic events in the Mediterranean Sea. The toxin profile in our cultures and field-sampled cells was compared to those registered in regions where harmful blooms are frequent (i.e., mainly ovatoxin-a and -b) (Table 3) (Table 3). Thus, the risks for negative impacts of O. cf. ovata blooms to marine fauna and human health should be continuously monitored in southern Brazil.
Humans and domestic animals can be intoxicated by Ostreopsis toxins upon contact with toxin-containing aerosol on the beach, as commonly documented in the Mediterranean [12,50,51,53] and suggested in the northeastern coast of Brazil [13]. Palytoxin is considered one of the most toxic naturally occurring non-peptide compounds via oral exposition, and cases of human death related to the ingestion PLTX-contaminated seafood have been reported [53,54]. In laboratory studies with marine organisms, O. cf. ovata cells exhibited acute toxicity to sea urchin gametes and larvae [55], as well as larval stages of crustaceans-Artemia salina brine shrimps, Tigriopus fulvus copepods and Amphibalanus amphitrite barnacles-and juvenile fish, Dicentrarchus labrax [3,56]. Moreover, toxins from O. cf. ovata are likely involved in massive deaths of adult sea urchins during natural blooms, as reported in New Zealand [4] and southeast Brazil [5].
There exists no current regulatory limit for PTX-like compounds in seafood [57], however, a 30 µg kg −1 safety level in seafood is recommended in Europe [58], where accumulation of these toxins has been reported in sea urchins and bivalve mollusks during an O. cf. ovata bloom [45]. In the present study, no commercial bivalve species were found in the area affected by the bloom in Currais Archipelago. We thus decided to collect commercial-sized mussels (Perna perna) from the nearby Galheta Island, a place~16.5 km distant from the bloom area and near the shore (Figure 1), where people occasionally go to collect mussels as a food source. We left some individuals in Currais for 24 h to investigate the potential accumulation of PLTX-like compounds from Ostreopsis cells, and examined others for the presence of toxins. Surprisingly, the mussels were already contaminated prior to transplantation, containing up to 32.9 µg PLTX-eq. kg −1 (average 22.3 µg kg −1 ), even though O. cf. ovata cell densities were more than 100-fold lower in Galheta Island. After 24 h of exposure to higher O. cf. ovata cell densities at the bloom area in Currais Archipelago, toxin concentrations reached up to 130 µg kg −1 in transplanted mussels. These values are within the same order of magnitude as the toxin concentration values found in mussels (up to 217 µg.kg −1 ) during an O. cf. ovata bloom on the French Mediterranean coast [45]. Considering the short exposure time of P. perna mussels in the present study and the relatively high toxin concentrations accumulated, these organisms may be considered potential intoxication vectors to humans and can be used as a sentinel for the presence of this toxin in coastal marine ecosystems. The risks for cases of human intoxication by Ostreopsis toxins in this region should be considered by local authorities engaged in seafood safety programs.  1 Inferred from graphical data. 2 No information about OVTX-b or OVTX-c. 3 Isobaric palytoxin, mascarenotoxins, ovatoxin-f and -g, -i, -j, -k.
Toxins of Ostreopsis can accumulate at lower levels in several marine organisms other than bivalves, including fishes, crustaceans, cephalopods, gastropods, echinoderms and sponges [45,63,64]. In the present study, we were not able to detect toxins in sea urchins nor in a single sea cucumber individual. Even though the animals were in close association with the Ostreopsis biofilm at the bottom of Currais Archipelago, we examined entire animals (as whole tissue homogenates), and this procedure may have diluted any toxin amount possibly present in specific tissues of these animals. Conversely, we were able to detect and quantify OVTX-a (20.0 µg kg −1 ) and -b (9.5 µg kg −1 ) in a single specimen of coral (Palythoa sp.), although it was not possible to determine whether the toxin had been assimilated by the coral or contained in Ostreopsis cells attached to the coral surface and pores. Toxin values in coral were similar to those reported in non-bivalve invertebrates during O. cf. ovata blooms in the Mediterranean [64,65].

The Plastic Litter Problem
In the ocean, plastic debris can be readily covered by a biofilm composed of bacteria and benthic microalgae, mostly diatoms [66][67][68]. Dinoflagellates, including Ostreopsis, can also attach their cells to plastic litter, but in general with less adhesion capacity [16,17]. In the present study, toxin-producing dinoflagellates were dominant over diatoms in plastic litter left in the water for 24 h during an O. cf. ovata bloom. Thus, the role of toxic cell-coated plastic debris as artificial toxin vectors for marine fauna, as well as the interactive harmful effects elicited upon their ingestion, must be thoroughly considered and examined.
The process of plastic colonization is not only dependent on the microorganisms present in the environment, but also on the plastic characteristics and the position of the plastic litter in the water column [66]. In our study, O. cf. ovata attached more abundantly to more flexible plastic materials, probably due to the movement of the plastic in the water facilitating "capture" of floating Ostreopsis cells that detach from substrate in mucous aggregates. The abundance of Ostreopsis was one order of magnitude higher on fiberglass screen, showing that its design is very efficient for sampling benthic dinoflagellates-probably due to its higher surface/volume ratio and the flexibility associated with the rough surface of the fiberglass filaments. The cell abundance of Ostreopsis was lower on rigid plastics, in which diatoms were present at equivalent numbers.
O. cf. ovata produces large quantities of mucus in static cultures, and cells aggregate into mucus strings. In the field, favored by the action of waves or currents on the sea floor, mucous Ostreopsis cell aggregates detach from the bottom and float. On their way to the surface, these sticky aggregates may come into contact with plastic litter, allowing its colonization by epibenthic Ostreopsis cells. Plastic debris (or other rafts) covered by toxic cells may thus become an alternate route for toxin transfer from benthic O. cf. ovata bloom to pelagic organisms.
Ingestion of plastic litter is common among sea turtles, seabirds, marine mammals and fish [22]. Apart from the fiberglass screen, included here for comparative purposes, the plastic materials tested in the present study (i.e., packing plastic, plastic bags and plastic bottle caps) are among the most abundant and common litter types in the stomachs of sea turtles found dead-stranded over the Brazilian coast (up to 82% of examined individuals) [26,69,70]. In the worst case, ingestion of large amounts of plastic litters can be responsible for the death of sea turtles due to suffocation or obstruction of their digestive systems [70]. Harmful effects of plastic ingestion are expected to be exacerbated in case plastic debris contain adsorbed toxic substances, such as persistent organic pollutants (reviewed in Lu et al. [31]). Likewise, marine plastic litters may be covered by moderate to large amounts of toxic micro-algal cells, as demonstrated for O. cf. ovata herein (up to 4900 cells cm −2 ) and-to a lesser extent-in another recent study in the Mediterranean Sea (up to 260 cells cm −2 [16]). The presence of abundant O cf. ovata cells covering plastic litters that drift around in sea turtle feeding grounds like Currais [71] is disturbing. Neurotoxins produced by O. cf. ovata can be highly toxic to marine animals (e.g., [14]), and plastic litter may contain high toxin doses during Ostreopsis bloom, as indicated here. For instance, a single 100-cm 2 (10 × 10 cm) low density polyethylene fragment was found to contain 8 × 10 5 cells of O cf. ovata producing 12 pg cell −1 of PLTX-like compounds, thus representing a dose of 10 ug of PLTX-like compounds. The possibility of chronic and acute intoxication of sea turtles (and other animals that ingest plastic litter) due to the ingestion of toxic microalgae-containing plastic litter should be therefore considered.

Conclusions
Ostreopsis cf. ovata is a toxic marine benthic dinoflagellate usually responsible for harmful bloom events in the Mediterrenean Sea. In the Currais Archipelago, southern Brazil, this species formed a dense bloom with similar visual characteristics: a yellowish biofilm covering an extensive area of the seafloor, and the appearance of floating mucous cell aggregates. The bloom occurred following a period dominated by similar weather conditions (increasing temperature and decreasing turbulence) to those triggering Europeans events. Parallel morphological and phylogenetic analyses indicated that O. cf. ovata cells occurring in this part of the western Atlantic Ocean belong to the same genotype as the Mediterranean bloom-forming populations. Toxin intracellular quotas and profile were also equivalent to those found in Europe, suggesting the risks for harmful effects to marine fauna and human health in this area. Moderate toxin concentrations were found in edible mussels during the bloom. Cell densities were equally high on both natural and artificial substrates. Cells of O. cf. ovata attached to different types of plastic litters that are commonly ingested by sea turtles in this area. The ingestion of biotoxin-coated plastics by these and other animals may thus cause other health damages besides suffocation or obstruction of their digestive tracts.

Sampling
Sampling was conducted from February 16th to 23rd 2017, on a site located in Currais Archipelago (25 • 44 06.75" S, 48 • 22 01.89" W), a set of three small islands on the Brazilian subtropical coast ( Figure 1). Samples of seaweeds (n = 12) and artificial substrates (n = 20; Figure 2B) were collected and processed following the procedure described in Tester et al. [32]. The artificial substrate consisted of rectangular pieces of fiberglass screens (10 × 15 cm;~2 mm mesh size; 174 cm 2 surface area; Figure 7E). Substrates were positioned in triplicate about 30 cm above the seafloor with the aid of small floats ( Figure 2B), and maintained for 24 h. Samples were vigorously shaken to detach particles from the seaweeds or artificial substrates, and the seawater containing Ostreopsis cells was divided in three aliquots of 200 mL each: (a) one was used for observation and isolation of living cells; the second one was concentrated by centrifugation to obtain cell pellets for toxin analysis; and (c) the last one was fixed with 1% lugol iodine solution for microscopic counting (1 mL aliquots in triplicate) on a Sedgewick-Rafter chamber.
On February 22nd, a field experiment was conducted with different plastic litters similar to those most commonly found in sea turtle stomachs in the region [26,35]. The plastic items (n = 16, four of each type; Figure 7) were installed in the field and processed in the same way as the artificial substrates described above, also remaining in the water~30 cm above the seafloor for 24 h ( Figure 7F). Plastic litters used included rigid polypropylene (R-PP) bottle caps of white and red colors, and sections of flexible polypropylene (F-PP) plastic packing and flexible, low-density polyethylene (LDPE) plastic bags (Figure 7).

Cultures
Cells of Ostreopsis were isolated using a capillary pipette following successive washing in sterile, local filtered seawater. After initial growth through consecutive cell divisions, the volume of culture was successively doubled by transferring the old aliquot to a larger microplate well containing an equivalent volume of sterile diluted f/2 media (f/4), without silica and~32 salinity. From 10 mL wells, cultures were transferred to 50 mL and then to 250 mL Erlenmeyer flasks, where they were maintained at 26 • C under a 12:12 h light cycle (irradiance of 70 ± 20 µmol m −2 s −1 ). For toxin analysis, cultivated cells (exponential and stationary growth phase) and field samples (100-200 mL) were harvested by centrifugation (2332× g, 5 min), the supernatant was removed, and samples were stored at −20 • C. Prior to toxin analysis, the frozen pellets were lyophilized.
Prior to electron microscopy (SEM) observations, bloom samples were preserved with neutral iodine lugol solution (1%), and cultured Ostreopsis cells with neutral and acidic lugol (1%) and glutaraldehyde solutions (5%). Small aliquots of the samples (2-5 mL) were placed on a 5-µm Millipore filter or on a 20-µm plankton net, rinsed with distilled water, and dehydrated in a series of increasing ethanol concentrations (30%, 50%, 70%, 90%, 95% and 100%), followed by critical point drying. Samples were finally mounted on a stub and sputter coated with gold palladium. Cells were observed using a JEOL ® JSM 6360-LV (Japan) microscope at 15 Kv.

DNA Amplification, Sequencing and Molecular Phylogeny
Cultivated cells and field samples (10 mL) were harvested by centrifugation (2332× g, 5 min), the supernatant was removed and replaced by ethanol to preserve the samples until the DNA analysis. Before the amplification, single cells from the ethanol-preserved samples were isolated with a glass capillary and washed six times with deionized water. Single Ostreopsis cells were placed in PCR tubes (at least two tubes for each sample) with 1-3 µL of deionized water and stored at −20 • C before the direct PCR amplifications.
Two consecutive PCR reactions (nested PCR) were performed to amplify the rDNA regions ITS1-5.8S-ITS2 (ITS) and LSU (D8-D10). For the first PCR reaction, 2.5 µL of each primer (ITSfw and OSTD10R, Table 4), 12.5 µL of PCR Master Mix 2X (Promega, Madison ® , WI, USA) containing the Taq DNA polymerase, dNTPs, MgCl 2 and reaction buffers, and 6.5 µL of nuclease free water were added to the tube. The PCR were performed in a Biometra TOne, thermocycler (Analytik Jena) as follows: one initial denaturation step at 95 • C for 2 min, followed by 35 cycles at 95 • C for 30 s, 50 • C (melting temperature, "MT") for 1 min, and 72 • C for 1 min, and a final elongation at 72 • C for 5 min. For the second PCR reaction, 1 µL of the first product were added to a new tube containing 2.5 µL of each primer (ITSfw and D3B for ITS region; D8 and OSTD10R for D8-D10; Table 4), 12.5 µL of GoTaq ® G2 Hot Start Green Master Mix (Promega ® , Madison, WI, USA) and 6.5 µL of nuclease free water. The second PCR was performed as the first, chaging the MT to 62 • C for ITS region, and 47 • C for D8-D10. DNA amplifications were controlled by electrophoresis on agarose gel. Positive samples were purified and sequenced as described in Moreira-Gonzalez et al. [73]. The alignment and phylogenetic analyses were performed as described in Chomérat et al. [2], with the following modifications: both ITS and D8-D10 rDNA region datasets were aligned using MAFFT algorithm with selection of the q-ins-i strategy; poorly aligned positions were re-moved using Gblocks algorithm; the most appropriate model of sequence evolution was selected using jModeltest2 v. 2.1.10; GTR+I+G and GTR+G were the model used for Maximum Likelihood (ML) and Bayesian Inference (BI) analysis of the D8-D10 and ITS regions, respectively; 2,000,000 generations were used in BI analysis for both alignments, with sampling every 100 generations; the posterior probabilities of each clade were calculated from the remaining 20,000 trees.

Sampling and Processing of Marine Fauna
In order to evaluate toxin uptake during the Ostreopsis bloom, sea-urchin individuals (n = 4), a pool of coral polyps and one sea cucumber individual were opportunistically sampled by snorkeling from the affected area in Currais Archipelago. Additionally, ten mussels (8-11 cm long) were collected on a nearby location in Galheta Island (distant~16.5 km from Currais and~2.5 km from the shore; 25 • 35 7.84" S, 48 • 19 17.92" W) and five of them were transplanted to the bottom of the area affected by the bloom in Currais, where they remained for 24 h before sampling. The other five individuals were immediately transported to the laboratory. All animals were promptly triturated using a tissue homogenizer (T 10 basic ULTRA-TURRAX ® , IKA, Staufen, Germany), and the homogenates were extracted in methanol (HPLC grade, Merck ® , Darmstadt, Germany) at a 9:1 (v:v) ratio, followed by sonication (130 W, CPX130, Cole Parmer ® , Vernon Hills, IL, USA) during 3 min with pulses of 3 s and intervals of 1 s, at 80% amplitude. Extracted samples were centrifuged at 2332× g for 5 min, filtered with syringe filters (PVDE, 0.22 µm, Analitica ® , São Paulo, Brazil) and kept frozen until the toxin analysis.

Toxin Analysis
Prior to toxin analysis, cell pellets (from cultures or field samples) were sonicated in bath ultrasound (Transonic TI-H-15, Elma ® , Wetzikon, Switzerland) at 45 kHz for 15 min with a methanol/water (9:1, v/v) solution. The mixture was centrifuged at 1200× g for 15 min, and the supernatant was passed through a centrifuge NanoSep filter (0.2 µm Nylon, PALL ® , Portsmouth, UK) and recovered into plastic vials with conical insert. Extracts from marine fauna were concentrated 10-fold by evaporating 1-mL aliquots with nitrogen gas at 40 • C, followed by re-suspension in 0.1 mL MeOH 90%.

Conflicts of Interest:
The authors declare no conflict of interest.