Assessing spirlin Alburnoides bipunctatus (Bloch, 1782) as an early indicator of climate change and anthropogenic stressors using ecological modeling and machine learning.

Combining single-species ecological modeling with advanced machine learning to investigate the long-term population dynamics of the rheophilic fish spirlin offers a powerful approach to understanding environmental changes and climate shifts in aquatic ecosystems. A new ESHIPPOClim model was developed by integrating climate change assessment into the ESHIPPO model. The model identifies spirlin as a potential early indicator of environmental changes, highlighting the interactive effects of climate change and anthropogenic stressors on fish populations and freshwater ecosystems. The ESHIPPOClim model reveals that 28.57 % of the spirlin's data indicates high resilience and ecological responsiveness, with 34.92 % showing medium-high adaptability, suggesting its substantial ability to withstand environmental stressors. With 36.51 % of the data in medium level and no data in the low category, spirlin may serve as a sentinel species, providing early warnings of environmental stressors before they severely impact other species or ecosystems. The results of uniform manifold approximation and projection (UMAP) and a decision tree show that pollution has the highest impact on the population dynamics of spirlin, followed by annual water temperature, overexploitation, and invasive species. Despite the obtained key drivers, higher abundance, dominance, and frequency values were detected in habitats with higher HIPPO stressors and climate change effects. Integrating state-of-the-art machine learning models has enhanced the predictive power of the ESHIPPOClim model, achieving approximately 90 % accuracy in identifying spirlin as an early indicator of climate change and anthropogenic stressors. The ESHIPPOClim model offers a holistic approach with broad practical applications using a simplified three-point scale, adaptable to various fish species, communities, and regions. The ecological modeling supported with advanced machine learning could serve as a foundation for rapid and cost-effective management of aquatic ecosystems, revealing the adaptability potential of fish species, which is crucial in rapidly changing environments.

Automated identification of aquatic insects: A case study using deep learning and computer vision techniques.

Deep learning techniques have recently found application in biodiversity research. Mayflies (Ephemeroptera), stoneflies (Plecoptera) and caddisflies (Trichoptera), often abbreviated as EPT, are frequently used for freshwater biomonitoring due to their large numbers and sensitivity to environmental changes. However, the morphological identification of EPT species is a challenging but fundamental task. Morphological identification of these freshwater insects is therefore not only extremely time-consuming and costly, but also often leads to misjudgments or generates datasets with low taxonomic resolution. Here, we investigated the application of deep learning to increase the efficiency and taxonomic resolution of biomonitoring programs. Our database contains 90 EPT taxa (genus or species level), with the number of images per category ranging from 21 to 300 (16,650 in total). Upon completion of training, a CNN (Convolutional Neural Network) model was created, capable of automatically classifying these taxa into their appropriate taxonomic categories with an accuracy of 98.7 %. Our model achieved a perfect classification rate of 100 % for 68 of the taxa in our dataset. We achieved noteworthy classification accuracy with morphologically closely related taxa within the training data (e.g., species of the genus Baetis, Hydropsyche, Perla). Gradient-weighted Class Activation Mapping (Grad-CAM) visualized the morphological features responsible for the classification of the treated species in the CNN models. Within Ephemeroptera, the head was the most important feature, while the thorax and abdomen were equally important for the classification of Plecoptera taxa. For the order Trichoptera, the head and thorax were almost equally important. Our database is recognized as the most extensive aquatic insect database, notably distinguished by its wealth of included categories (taxa). Our approach can help solve long-standing challenges in biodiversity research and address pressing issues in monitoring programs by saving time in sample identification.