Smart Aquatic Knowledgebase and Artificial Intelligence for Yielding Advanced Navigation (SakAi-yan): Predicting Marine Biodiversity and Fishing Impact

This study explores the intersection of environmental factors and biodiversity, utilizing IoT and geospatial analytics to predict biodiversity across various regions. Advanced machine learning techniques were applied to develop predictive models by integrating a biodiversity index into an already robust dataset of environmental variables. The most crucial model developed is described as the stacking model, showing a very high level of predictive accuracy, with a remarkable correlation between predicted and actual biodiversity value. Our findings imperatively stress the controlling roles of temperature, atmospheric pressure, and fishing intensity on biodiversity. The results emphasize targeted conservation attention in those areas where both high biodiversity and overlap of fishing activities are found. Such model developments can not only deepen our understanding of biodiversity dynamics but also pave the way to improve the predictive models established here through future studies to ensure the sustainable management of marine ecosystems.