Alongot Intarachart

Accurate monitoring of chlorophyll-a (Chl-a) is essential for managing coastal aquaculture, as Chl-a indicates phytoplankton biomass and water quality. This study developed a hybrid deep learning model integrating convolutional neural networks (CNN), bidirectional long short-term memory (BiLSTM), and an attention mechanism (Attention) to classify Chl-a using hourly, water quality datasets collected from the GOT001 station in Si Racha Bay, Eastern Gulf of Thailand (2020–2024). A random forest (RF) identified sea surface temperature (SEATEMP), dew point temperature (DEWPOINT), and turbidity (TURB) as the most influential variables, accounting for over 90% of the accuracy. Chl-a concentrations were categorized into ecological groups (low, medium, and high) using quantile-based binning and K-means clustering to support operational classification. Model performance comparison showed that the CNN–BiLSTM model achieved the highest classification accuracy (81.3%), outperforming the CNN–LSTM model (59.7%). However, the addition of the Attention did not enhance predictive performance, likely due to the limited number of key predictive variables and their already high explanatory power. This study highlights the potential of CNN–BiLSTM as a near-real-time classification tool for Chl-a levels in highly variable coastal ecosystems, supporting aquaculture management, early warning of algal blooms or red tides, and water quality risk assessment in the Gulf of Thailand and comparable coastal regions.

Sea cucumbers play a vital role in marine and coastal ecosystems, with some species holding significant economic value. Accurate growth assessment, particularly weight estimation, is crucial for their management and conservation. However, direct measurement poses challenges, as sea cucumbers expel internal fluids when handled, altering their body size and weight. This study evaluates the effectiveness of image analysis combined with ArUco markers to estimate the weight of three economically and ecologically important sea cucumber species found in Thailand: black sea cucumber (Holothuria leucospilota), pink warty sea cucumber (Cercodemas anceps), and sandfish (Holothuria scabra). The proposed method demonstrated high accuracy, with R² values of 0.9699, 0.9774, and 0.9882, respectively. Furthermore, no significant differences (p > 0.05) were observed between traditional hand measurements and image-based assessments, with relative errors of 7.71 ± 4.30% for black sea cucumber, 5.06 ± 3.37% for pink warty sea cucumber, and 4.50 ± 3.23% for sandfish. Unlike deep learning, which requires large datasets and high computation, this simple, cost-effective, and adaptable method highlights the potential of image analysis combined with ArUco markers as a non-invasive and accurate tool for estimating sea cucumber weight. This approach minimizes stress on the animals and can be extended to other aquatic species. However, challenges such as shadows, water turbidity, and the presence of similarly shaped objects near sea cucumbers should be considered when applying this technique in field conditions.

The nitrogen budget and the effects of varying densities of sea grape (Caulerpa lentillifera) on water quality and the growth performance of Asian seabass (Lates calcarifer) in a polyculture system were evaluated. Four treatments were tested, each stocked with 20 fish (average weight: ~20 g; density: 2.66 kg/m³). Treatment 1 (control) contained no sea grapes, while treatments 2, 3, and 4 included 100, 200, and 400 g of sea grapes, corresponding to 0, 667 g/m³, 1333 g/m³, and 2666 g/m³, respectively. Significant (p < 0.05) reductions in total ammonia nitrogen (TAN), total suspended solids (TSS), and turbidity were observed with increasing sea grape density; however, no significant differences (p > 0.05) were found among the three sea grape treatments. Higher sea grape densities increased nitrite (NO₂–N) and nitrate (NO₃–N) concentrations due to enhanced nitrification. Approximately 65% of the nitrogen input originated from feed, but only 47.78–48.96% was assimilated into fish biomass. Nitrogen losses included 1.17–1.46% via water exchange and final drainage, while 45.27–50.76% was likely retained in sediments, volatilized, or lost through denitrification. Sea grapes effectively absorbed total nitrogen (TN), demonstrating their potential as biofilters for improving water quality without compromising fish growth performance. The optimal density was 100 g of sea grapes per 2.66 kg/m³ of seabass biomass, offering a sustainable strategy to enhance productivity while mitigating environmental impacts.