Red Tide Detection Method for HY−1D Coastal Zone Imager Based on U−Net Convolutional Neural Network

Existing red tide detection methods have mainly been developed for ocean color satellite data with low spatial resolution and high spectral resolution. Higher spatial resolution satellite images are required for red tides with fine scale and scattered distribution. However, red tide detection methods for ocean color satellite data cannot be directly applied to medium–high spatial resolution satellite data owing to the shortage of red tide responsive bands. Therefore, a new red tide detection method for medium–high spatial resolution satellite data is required. This study proposes the red tide detection U−Net (RDU−Net) model by considering the HY−1D Coastal Zone Imager (HY−1D CZI) as an example. RDU−Net employs the channel attention model to derive the inter−channel relationship of red tide information in order to reduce the influence of the marine environment on red tide detection. Moreover, the boundary and binary cross entropy (BBCE) loss function, which incorporates the boundary loss, is used to obtain clear and accurate red tide boundaries. In addition, a multi−feature dataset including the HY−1D CZI radiance and Normalized Difference Vegetation Index (NDVI) is employed to enhance the spectral difference between red tides and seawater and thus improve the accuracy of red tide detection. Experimental results show that RDU−Net can detect red tides accurately without a precedent threshold. Precision and Recall of 87.47% and 86.62%, respectively, are achieved, while the F1−score and Kappa are 0.87. Compared with the existing method, the F1−score is improved by 0.07–0.21. Furthermore, the proposed method can detect red tides accurately even under interference from clouds and fog, and it shows good performance in the case of red tide edges and scattered distribution areas. Moreover, it shows good applicability and can be successfully applied to other satellite data with high spatial resolution and large bandwidth, such as GF−1 Wide Field of View 2 (WFV2) images.

Use of intravenous lipid emulsion therapy as a novel treatment for brevetoxicosis in sea turtles

The southwest coast of Florida experiences annual red tides, a type of harmful algal bloom that results from high concentrations of Karenia brevis. These dinoflagellates release lipophilic neurotoxins, known as brevetoxins, that bind to sodium channels and inhibit their inactivation, resulting in a variety of symptoms that can lead to mass sea turtle strandings. Traditional therapies for brevetoxicosis include standard and supportive care (SSC) and/or dehydration therapy; however, these treatments are slow-acting and often ineffective. Because red tide events occur annually in Florida, our objective was to test intravenous lipid emulsion (ILE) as a rapid treatment for brevetoxicosis in sea turtles and examine potential impacts on toxin clearance rates, symptom reduction, rehabilitation time, and survival rates. Sea turtles exhibiting neurological symptoms related to brevetoxicosis were brought to rehabilitation from 2018–2019. Upon admission, blood samples were collected, followed by immediate administration of 25 mg ILE/kg body mass (Intralipid® 20%) at 1 mL/min using infusion pumps. Blood samples were collected at numerous intervals post-ILE delivery and analyzed for brevetoxins using enzyme-linked immunosorbent assays. In total, nine (four subadults, one adult female, four adult males) loggerheads (Caretta caretta), five (four juvenile, one adult female) Kemp’s ridleys (Lepidochelys kempii), and four juvenile green turtles (Chelonia mydas) were included in this study. We found that plasma brevetoxins declined faster compared to turtles that received only SSC. Additionally, survival rate of these patients was 94% (17/18), which is significantly higher than previous studies that used SSC and/or dehydration therapy (47%; 46/99). Nearly all symptoms were eliminated within 24–48 h, whereas using SSC, symptom elimination could take up to seven days or more. The dosage given here (25 mg/kg) was sufficient for turtles in this study, but the use of a higher dosage (50–100 mg/kg) for those animals experiencing severe symptoms may be considered. These types of fast-acting treatment plans are necessary for rehabilitation facilities that are already resource-limited. Intravenous lipid emulsion therapy has the potential to reduce rehabilitation time, save resources, and increase survival of sea turtles and other marine animals experiencing brevetoxicosis.

Red tide of the Lingulodinium polyedrum (Dinophyceae) in Odessa Bay (Black Sea)

One of the most massive red tides at the Odessa Bay was observed in September October 2020. It was caused by a toxic dinoflagellate Lingulodinium polyedrum (Stein) Dodge. The maximum abundance (56.1 x 106 cells L-1) of L. polyedrum was registered at the Odessa port area on October 6 when a water temperature and a salinity were 19.7°C and 14.3 ‰ respectively. The red tide was so huge and dense that the water glowed at night due to the bioluminescence characteristic of this species. The article briefly describes the history of the study of L. polyedrum in this area and provides a detailed morphological description with original photographs of this species sampled from bloom. We associate the appearance of the red tide with an increased temperature of sea water and air, a high content of nutrients, the presence of viable L. polyedrum cysts, and a slight decrease in salinity in the bay during the period of a mass development of the species in autumn of 2020. The red tide was accompanied by Protoperidinium steini, P. divergens, Prorocentrum cordatum, P. minimum, P. micans, Gonyaulas scrippsae, Diplopsalis lenticula, Azadinium spinosum, Dinophysis rotundata, D. acuminata, Oblea rotunda, Scrippsiella trochoidea, Ceratium furca.

Automatic Identification of Harmful Algae Based On Multiple Convolutional Neural Networks and Transfer Learning

The monitoring of harmful algae is very important for the maintenance of the aquatic ecological environment. Traditional algae monitoring methods require professionals with substantial experience in algae species, which are time-consuming, expensive and limited in practice. The automatic classification of algae cell images and the identification of harmful algae images were realized by the combination of multiple Convolutional Neural Networks (CNNs) and deep learning techniques based on transfer learning in this work. 11 common harmful and 31 harmless algae genera were collected as input samples, the five CNNs classification models of AlexNet, VGG16, GoogLeNet, ResNet50, and MobileNetV2 were fine-tuned to automatically classify algae images, and the average accuracy was improved 11.9% when compared to models without fine-tuning. In order to monitor harmful algae which can cause red tides or produce toxins severely polluting drinking water, a new identification method of harmful algae which combines the recognition results of five CNN models was proposed, and the recall rate reached 98.0%. The experimental results validate that the recognition performance of harmful algae could be significantly improved by transfer learning, and the proposed identification method is effective in the preliminary screening of harmful algae and greatly reduces the workload of professional personnel.

An overview of harmful algal blooms and eutrophication in Jakarta Bay, Indonesia

Algal blooms have been occurring in Jakarta Bay for twenty years. However, recently the occurrence of algal blooms, their harmful effects, and their duration have been intensified. Algal blooms have devastated the marine environment, caused fish mortality, and been detrimental to local tourism, local fishing, and other industries along the coast. It comes to speculation that the increase of anthropogenic activity from surrounding areas is taking a toll on the environment. So, this research aimed to study the recent rise of algal blooms in Jakarta Bay and the possible anthropogenic links, mainly through cultural eutrophication, to the increasing occurrence of red tides and their impact. Observation has been conducted to study the dynamic of algal blooms concerning eutrophication and the existing seasons. Collecting samples were performed using a canonical plankton net from 2008 until 2015. The results showed that the abundance of phytoplankton ranged from 40.90 x 106 up to 1699.10 x 106 cells.m−3. The highest quantity of cells was observed in May 2010 between rainy to dry seasons. There is evidence that the reported increase in frequency and magnitude of algal bloom events in Jakarta Bay is linked to cultural eutrophication. The recent exponential growth of the city may be a contributing factor in the increasing intensity of algal blooms. The cultural eutrophication of coastal waters increased, leading to the intensity and frequency of algal bloom.

Superoxide Production by the Red Tide-Producing Chattonella marina Complex (Raphidophyceae) Correlates with Toxicity to Aquacultured Fishes

The marine raphidophyte Chattonella marina complex forms red tides, causing heavy mortalities of aquacultured fishes in temperate coastal waters worldwide. The mechanism for Chattonella fish mortality remains unresolved. Although several toxic chemicals have been proposed as responsible for fish mortality, the cause is still unclear. In this study, we performed toxicity bioassays with red sea bream and yellowtail. We also measured biological parameters potentially related to ichthyotoxicity, such as cell size, superoxide (O2•−) production, and compositions of fatty acids and sugars, in up to eight Chattonella strains to investigate possible correlations with toxicity. There were significant differences in moribundity rates of fish and in all biological parameters among strains. One strain displayed no ichthyotoxicity even at high cell densities. Strains were categorized into three groups based on cell length, but this classification did not significantly correlate with ichthyotoxicity. O2•− production differed by a factor of more than 13 between strains at the late exponential growth phase. O2•− production was significantly correlated with ichthyotoxicity. Differences in fatty acid and sugar contents were not related to ichthyotoxicity. Our study supports the hypothesis that superoxide can directly or indirectly play an important role in the Chattonella-related mortality of aquacultured fishes.

Red Tide Events and Seasonal Variations in the Partial Pressure of CO2 and Related Parameters in Shellfish-Farming Bays, Southeastern Coast of Korea

Mixed results have been reported on the evaluation of the coastal carbon cycle and its contribution to the global carbon cycle, mainly due to the shortage of observational data and the considerable spatiotemporal variability arising from complex biogeochemical factors. In this study, the partial pressure of carbon dioxide (pCO2) and related environmental factors were measured in the Jinhae–Geoje–Tongyeong bay region of the southeastern Korean Peninsula in February 2014, August 2014, April 2015, and October 2015. The mean pCO2 of surface seawater ranged from 215 to 471 μatm and exhibited a high correlation with the surface seawater temperature when data for August were excluded (R2 = 0.69), indicating that the seasonal variation in CO2 could be largely attributed to the variation in seawater temperature. However, a severe red tide event occurred in August 2014, when the lowest pCO2 value was observed despite a relatively high seawater temperature. It is considered that the active biological production of phytoplankton related to red tides counteracted the summer increase in pCO2. Based on the correlation between pCO2 and temperature, the estimated decrease in pCO2 caused by non-thermal factors was approximately 200 μatm. During the entire study period, the air–sea CO2 flux ranged from −14.2 to 3.7 mmol m–2 d–1, indicating that the study area served as an overall sink for atmospheric CO2, and only functioned as a weak source during October. The mean annual CO2 flux estimated from the correlation with temperature was −5.1 mmol m–2 d–1. However, because this estimate did not include reductions caused by sporadic events of biological production, such as red tides and phytoplankton blooms, the actual uptake flux is considered to be higher. The mean saturation state (ΩAr) value of carbonate aragonite was 2.61 for surface water and 2.04 for bottom water. However, the mean ΩAr of bottom water was <2 in August and October, and the ΩAr values measured at some of the bottom water stations in August were <1. Considering that the period from August to October corresponds to the reproduction and growth stages of shellfish, such low ΩAr values could be very damaging to shellfish production and the aquaculture industry.

A Remote Sensing and Machine Learning-Based Approach to Forecast the Onset of Harmful Algal Bloom

In the last few decades, harmful algal blooms (HABs, also known as “red tides”) have become one of the most detrimental natural phenomena in Florida’s coastal areas. Karenia brevis produces toxins that have harmful effects on humans, fisheries, and ecosystems. In this study, we developed and compared the efficiency of state-of-the-art machine learning models (e.g., XGBoost, Random Forest, and Support Vector Machine) in predicting the occurrence of HABs. In the proposed models the K. brevis abundance is used as the target, and 10 level-02 ocean color products extracted from daily archival MODIS satellite data are used as controlling factors. The adopted approach addresses two main shortcomings of earlier models: (1) the paucity of satellite data due to cloudy scenes and (2) the lag time between the period at which a variable reaches its highest correlation with the target and the time the bloom occurs. Eleven spatio-temporal models were generated, each from 3 consecutive day satellite datasets, with a forecasting span from 1 to 11 days. The 3-day models addressed the potential variations in lag time for some of the temporal variables. One or more of the generated 11 models could be used to predict HAB occurrences depending on availability of the cloud-free consecutive days. Findings indicate that XGBoost outperformed the other methods, and the forecasting models of 5–9 days achieved the best results. The most reliable model can forecast eight days ahead of time with balanced overall accuracy, Kappa coefficient, F-Score, and AUC of 96%, 0.93, 0.97, and 0.98 respectively. The euphotic depth, sea surface temperature, and chlorophyll-a are always among the most significant controlling factors. The proposed models could potentially be used to develop an “early warning system” for HABs in southwest Florida.

Effects of Multiple Karenia brevis Red Tide Blooms on a Common Bottlenose Dolphin (Tursiops truncatus) Prey Fish Assemblage: Patterns of Resistance and Resilience in Sarasota Bay, Florida

Red tide blooms caused by the toxic dinoflagellate Karenia brevis are natural disturbance events that occur regularly along Florida’s west coast, often resulting in massive fish kills and marine mammal, seabird, and sea turtle mortalities. Limited prior work on the ecological effects of red tides suggests they play an important role in structuring ecosystem dynamics and regulating communities, however specific effects on prey populations and potential alterations to predator-prey interactions are unknown. We surveyed the prey fish assemblage of a top marine predator, the common bottlenose dolphin (Tursiops truncatus), in shallow seagrass habitat in Sarasota Bay, Florida, during 2004–2019, collecting data on prey density, species composition, K. brevis cell densities, and environmental variables. Across eight distinct red tide bloom events, resistance, resilience, and the ecological effects on the prey assemblage varied depending on bloom intensity, season, and frequency. Prey assemblage structure showed significant and distinct short-term shifts during blooms independent of the normal seasonal shifts in prey structure seen during non-bloom conditions. Canonical correspondence analysis indicated a strong influence of K. brevis density on assemblage structure. Blooms occurring primarily in the summer were associated with less initial prey resistance and higher than average annual catch per unit effort (CPUE) 1–3 years following bloom cessation, with bloom frequency prolonging the time needed to reach higher than average annual CPUE. Regardless of season, recovery to pre-bloom prey abundances occurred within 1 year. Sample-based rarefaction and extrapolation indicated significant differences in prey diversity among summer bloom events. This study is a first step in identifying differences in resistance, resilience, and the ecological effects of multiple red tide bloom events of various temporal scales and intensity on a dolphin prey assemblage. Improved understanding of the influence of red tides on estuarine structural dynamics and function can better inform management, and potentially guide mitigation efforts post-bloom.