Diatoms are fundamental carbon sources in a wide range of aquatic food webs and have the potential for wide application in addressing environmental change. Understanding the evolution of topics in diatom research will provide a clear and needed guide to strengthen research on diatoms. However, such an overview remains unavailable. In this study, we used Latent Dirichlet Allocation (LDA), a generative model, to identify topics and determine their trends (i.e., cold and hot topics) by analyzing the abstracts of 19,000 publications from the Web of Science that were related to diatoms during 1991–2018. A total of 116 topics were identified from a Bayesian model selection. The hot topics (diversity, environmental indicator, climate change, land use, and water quality) that were identified by LDA indicated that diatoms are increasingly used as indicators to assess water quality and identify modern climate change impacts due to intensive anthropogenic activities. In terms of cold topics (growth rate, culture growth, cell life history, copepod feeding, grazing by microzooplankton, zooplankton predation, and primary productivity) and hot topics (spatial-temporal distribution, morphology, molecular identification, gene expression, and review), we determined that basic studies on diatoms have decreased and that studies tend to be more comprehensive. This study notes that future directions in diatom research will be closely associated with the application of diatoms in environmental management and climate change to cope with environmental challenges, and more comprehensive issues related to diatoms should be considered.

Solving shallow water equations is crucial in science and engineering for understanding and predicting natural phenomena. To address the limitations of Physics-Informed Neural Network (PINN) in solving shallow water equations, we propose an improved PINN algorithm integrated with a deep learning framework. This algorithm introduces a regularization term as a penalty in the loss function, based on the PINN and Long Short-Term Memory (LSTM) models, and incorporates an attention mechanism to solve the original equation across the entire domain. Simulation experiments were conducted on one-dimensional and two-dimensional shallow water equations. The results indicate that, compared to the classical PINN algorithm, the improved algorithm shows significant advantages in handling discontinuities, such as sparse waves, in one-dimensional problems. It accurately captures sparse waves and avoids smoothing effects. In two-dimensional problems, the improved algorithm demonstrates good symmetry and effectively reduces non-physical oscillations. It also shows significant advantages in capturing details and handling complex phenomena, offering higher reliability and accuracy. The improved PINNs algorithm, which combines neural networks with physical mechanisms, can provide robust solutions and effectively avoid some of the shortcomings of classical PINNs methods. It also possesses high resolution and strong generalization capabilities, enabling accurate predictions at any given moment.

Eutrophication and salinization are serious global environmental problems in freshwater ecosystems, occasionally acting jointly to exert harmful effects on aquatic ecosystems. To elucidate the interactive effects of nutrients and salinity on phytoplankton assemblages, we conducted a four-season study during 2020–2021 of eight lakes from Yunnan Plateau (Southwest China) with a wide range of conductivities (Cond, reflecting degree of salinization), eutrophic states, and water depths and used General Additive Modeling (GAM) of the data. We found that: (1) species number (SN), density (D_Phyt), and biomass (B_Phyt) of phytoplankton showed stronger seasonal dynamics in shallow lakes than in deep lakes, all being, as expected, higher in the warm season; (2) annual and summer data revealed highly significant positive relationships between SN, D_Phyt, and B_Phyt with total nitrogen (TN) and total phosphorus (TP), which became weaker at high TP occurring when the N:P ratio was low, indicating N limitation; (3) SN, D_Phyt, and B_Phyt showed a unimodal relationship with salinity, peaking at 400–1000 μS/cm (Cond); (4) the two dominant taxa (cyanobacteria and chlorophyta) showed different patterns, with chlorophyta generally dominating at low TN and cyanobacteria at high TN and Cond, suggesting the synergistic effect of nitrogen and Cond on cyanobacterial dominance.