Playing hide and seek: Distribution with depth of potentially harmful epibenthic dinoflagellates of Southern El Hierro Island, Canary Islands (NE Atlantic)

eDNA metabarcoding reveals biodiversity and depth stratification patterns of dinoflagellate assemblages within the epipelagic zone of the western Coral Sea

BACKGROUND

Dinoflagellates play critical roles in the functioning of marine ecosystems but also may pose a hazard to human and ecosystem health by causing harmful algal blooms (HABs). The Coral Sea is a biodiversity hotspot, but its dinoflagellate assemblages in pelagic waters have not been studied by modern sequencing methods. We used metabarcoding of the 18 S rRNA V4 amplicon to assess the diversity and structure of dinoflagellate assemblages throughout the water column to a depth of 150 m at three stations in the Western Coral Sea. Additionally, at one station we compared metabarcoding with morphological methods to optimise identification and detection of dinoflagellates.

RESULTS

Stratification of dinoflagellate assemblages was evident in depth-specific relative abundances of taxonomic groups; the greatest difference was between the 5-30 m assemblages and the 130-150 m assemblages. The relative abundance of Dinophyceae (photosynthetic and heterotrophic) decreased with increasing depth, whereas that of Syndiniales (parasitic) increased with increasing depth. The composition of major taxonomic groups was similar among stations. Taxonomic richness and diversity of amplicon sequence variants (ASVs) were similar among depths and stations; however, the abundance of dominant taxa was highest within 0-30 m, and the abundance of rare taxa was highest within 130-150 m, indicating adaptations to specific depth strata. The number of unclassified ASVs at the family and species levels was very high, particularly for Syndinian representatives.

CONCLUSIONS

Dinoflagellate assemblages in open water of the Coral Sea are highly diverse and taxonomically stratified by depth; patterns of relative abundance along the depth gradient reflect environmental factors and ecological processes. Metabarcoding detects more species richness than does traditional microscopical methods of sample analysis, yet the methods are complementary, with morphological analysis revealing additional richness. The large number of unclassified dinoflagellate-ASVs indicates a need for improved taxonomic reference databases and suggests presence of dinoflagellate-crypto and-morphospecies.

Molecular Phylogeny, Morphology, Growth and Toxicity of Three Benthic Dinoflagellates Ostreopsis sp. 9, Prorocentrum lima and Coolia monotis Developing in Strait of Gibraltar, Southwestern Mediterranean

Few works have been carried out on benthic harmful algal blooms (BHAB) species in the southern Mediterranean and no data are available for the highly dynamic Strait of Gibraltar (western Mediterranean waters). For the first time, Ostreopsis sp. 9, Prorocentrum lima and Coolia monotis were isolated in this key region in terms of exchanges between the Atlantic Ocean and the Mediterranean and subject to intense maritime traffic. Ribotyping confirmed the morphological identification of these three dinoflagellates species. Monoclonal cultures were established and the maximum growth rate and cell yield were measured at a temperature of 24 °C and an irradiance of 90 µmol photons m-2 s-1, for each species: 0.26 ± 0.02 d-1 (8.75 × 103 cell mL-1 after 28 days) for Ostreopsis sp. 9, 0.21 ± 0.01 d-1 (49 × 103 cell mL-1 after 145 days) for P. lima and 0.21 ± 0.01 d-1 (10.02 × 103 cell mL-1 after 28 days) for C. monotis. Only P. lima was toxic with concentrations of okadaic acid and dinophysistoxin-1 measured in optimal growth conditions ranging from 6.4 pg cell-1 to 26.97 pg cell-1 and from 5.19 to 25.27 pg cell-1, respectively. The toxin content of this species varied in function of the growth phase. Temperature influenced the growth and toxin content of P. lima. Results suggest that future warming of Mediterranean coastal waters may lead to higher growth rates and to increases in cellular toxin levels in P. lima. Nitrate and ammonia affected the toxin content of P. lima but no clear trend was noted. In further studies, we have to isolate other BHAB species and strains from Strait of Gibraltar waters to obtain more insight into their diversity and toxicity.

Diversity and distribution of benthic dinoflagellates in Tonga include the potentially harmful genera Gambierdiscus and Fukuyoa

Benthic dinoflagellates that can cause illness, such as ciguatera poisoning (CP), are prevalent around the Pacific but are poorly described in many locations. This study represents the first ecological assessment of benthic harmful algae species in the Kingdom of Tonga, a country where CP occurs regularly. Surveys were conducted in June 2016 in the Tongatapu island group, and in June 2017 across three island groups: Ha'apai, Vava'u, and Tongatapu. Shallow subtidal coastal habitats were investigated by measuring water quality parameters and conducting quadrat surveys. Microalgae samples were collected using either macrophyte collection or the artificial substrate method. Benthic dinoflagellates (Gambierdiscus and/or Fukuyoa, Ostreopsis, and Prorocentrum) were counted using light microscopy, followed by molecular analyses (real-time PCR in 2016 and high throughput sequencing (metabarcoding) in 2017) to identify Gambierdiscus and Fukuyoa to species level. Six species were detected from the Tongatapu island group in 2016 (G. australes, G. carpenteri, G. honu, G. pacificus, F. paulensis, and F. ruetzleri) using real-time PCR. Using the metabarcoding approach in 2017, a total of eight species (G. australes, G. carpenteri, G. honu, G. pacificus, G. cheloniae, G. lewisii, G. polynesiensis, and F. yasumotoi) were detected. Species were detected in mixed assemblages of up to six species, with G. pacificus and G. carpenteri being the most frequently observed. Ha'apai had the highest diversity with eight species detected, which identifies this area as a Gambierdiscus diversity 'hotspot'. Vava'u and Tongatapu had three and six species found respectively. Gambierdiscus polynesiensis, a described ciguatoxin producer and proposed causative agent of CP was found only in Ha'apai and Vava'u in 2017, but not in Tongatapu in either year. Ostreopsis spp. and Prorocentrum spp. were also frequently observed, with Prorocentrum most abundant at the majority of sites. In 2016, the highest number of Gambierdiscus and/or Fukuyoa cells were observed on seagrass (Halodule uninervis) from Sopu, Tongatapu. In 2017, the highest numbers of Gambierdiscus and/or Fukuyoa from artificial substrate samples were recorded in the Halimeda dominant habitat at Neiafu Tahi, Vava'u, a low energy site. This raised the question of the effect of wave motion or currents on abundance measurements from artificial substrates. Differences in detection were noticed between macrophytes and artificial substrates, with higher numbers of species found on artificial substrates. This study provides a baseline of benthic dinoflagellate distributions and diversity for Tonga that may be used for future studies and the development of monitoring programmes.

Taxonomy and abundance of epibenthic Prorocentrum (Dinophyceae) species from the tropical and subtropical Southwest Atlantic Ocean including a review of their global diversity and distribution

In the tropical and subtropical South Atlantic Ocean, studies on the taxonomy and abundance of benthic harmful algae are scarce and the region has been largely under investigated. In this study, morphological descriptions, molecular (LSU rDNA and ITS region) and abundance data of benthic Prorocentrum species from the tropical and subtropical Southwest Atlantic and three oceanic islands are presented. Moreover, a review of global benthic Prorocentrum species richness and distribution is presented. Eleven benthic Prorocentrum species were found in Brazil. Morphological and molecular data on P. borbonicum, P. hoffmannianum, P. lima species complex and P. rhathymum were provided. Prorocentrum panamense, P. cf. caipirignum, P. cf. concavum, P. cf. norrisianum, P. emarginatum/fukuyoi/sculptile complex and two not identified species were observed using scanning electron and/or light microscopy, and morphological descriptions are presented. Prorocentrum lima species complex was found at all investigated sites, in abundances up to 2 × 104 cells g-1 FW at the Northeast Brazil, while maximum abundance of all the remaining species did not exceed 1 × 103 cells g-1 FW. The Fernando de Noronha archipelago can be considered a hotspot of benthic Prorocentrum species diversity, with ten species registered. Data compiled in the literature review shows a clear latitudinal gradient with higher species richness in tropical and subtropical regions relative to temperate areas. It is also evident that there is a bias caused by taxonomic impediment and an uneven sampling effort, with many regions still to be investigated using a combined morphological and molecular effort. Therefore, the current knowledge on the global distribution of benthic Prorocentrum species is likely underestimated.

Model of the Origin of a Ciguatoxic Grouper (Plectropomus leopardus)

Published data were used to model the transfer of ciguatoxins (CTX) across three trophic levels of a marine food chain on the Great Barrier Reef (GBR), Australia, to produce a mildly toxic common coral trout (Plectropomus leopardus), one of the most targeted food fishes on the GBR. Our model generated a 1.6 kg grouper with a flesh concentration of 0.1 µg/kg of Pacific-ciguatoxin-1 (P-CTX-1 = CTX1B) from 1.1 to 4.3 µg of P-CTX-1 equivalents (eq.) entering the food chain from 0.7 to 2.7 million benthic dinoflagellates (Gambierdiscus sp.) producing 1.6 pg/cell of the P-CTX-1 precursor, P-CTX-4B (CTX4B). We simulated the food chain transfer of ciguatoxins via surgeonfishes by modelling Ctenochaetus striatus feeding on turf algae. A C. striatus feeding on ≥1000 Gambierdiscus/cm² of turf algae accumulates sufficient toxin in <2 days that when preyed on, produces a 1.6 kg common coral trout with a flesh concentration of 0.1 µg/kg P-CTX-1. Our model shows that even transient blooms of highly ciguatoxic Gambierdiscus can generate ciguateric fishes. In contrast, sparse cell densities of ≤10 Gambierdiscus/cm² are unlikely to pose a significant risk, at least in areas where the P-CTX-1 family of ciguatoxins predominate. The ciguatera risk from intermediate Gambierdiscus densities (~100 cells/cm²) is more difficult to assess, as it requires feeding times for surgeonfish (~4-14 days) that overlap with turnover rates of turf algae that are grazed by herbivorous fishes, at least in regions such as the GBR, where stocks of herbivorous fishes are not impacted by fishing. We use our model to explore how the duration of ciguatoxic Gambierdiscus blooms, the type of ciguatoxins they produce, and fish feeding behaviours can produce differences in relative toxicities between trophic levels. Our simple model indicates thresholds for the design of risk and mitigation strategies for ciguatera and the variables that can be manipulated to explore alternate scenarios for the accumulation and transfer of P-CTX-1 analogues through marine food chains and, potentially, for other ciguatoxins in other regions, as more data become available.

Using artificial substrates to quantify Gambierdiscus and other toxic benthic dinoflagellates for monitoring purposes

Collecting methods generally used to determine cell abundances of toxic benthic dinoflagellates (BHAB) use cells dislodged from either macrophytes or artificial substrates. This article compares the advantages of the macrophyte and artificial substrate methods and discusses which method is more appropriate for use in monitoring programs that focus on toxic BHAB species identification and quantification. The concept of benthic dinoflagellate "preference" for specific macrophytes was also reviewed. Examination of data from 75 field studies showed macrophytes with higher surface area per unit biomass harbored higher concentrations of Gambierdiscus cells. There was no definitive evidence that cells were actively selecting one macrophyte over another. This observation supports the use of artificial substrates (AS) as a means of assessing cell abundances in complex habitats because cell counts are normalized to a standardized surface area, not macrophyte biomass. The artificial substrate method represents the most robust approach, currently available, for collecting toxic, benthic dinoflagellates for a cell-based early warning system.