4.7 Review

Data-driven models for predicting community changes in freshwater ecosystems: A review

Journal

ECOLOGICAL INFORMATICS
Volume 77, Issue -, Pages -

Publisher

ELSEVIER
DOI: 10.1016/j.ecoinf.2023.102163

Keywords

Machine learning model; Deep learning; Artificial intelligence; Interpretability; Ecosystem management; Aquatic community

Categories

Ask authors/readers for more resources

This study reviews data-based research using models to predict the biological elements of freshwater ecosystems over the last three decades. It evaluates the ability of current models to predict changes in freshwater organisms and suggests future research directions.
Freshwater ecosystems are sensitive to disturbances related to human activities, such as climate and land-use changes. To predict and understand the potential impacts of these disturbances, models can be employed. In this study, we reviewed data-based research employing models over the last three decades to predict the biological elements of freshwater ecosystems at different scales, with a focus on phytoplankton, macroinvertebrates, and fish. Specifically, we investigated existing research trends, evaluated the ability of current models to predict changes in freshwater organisms in response to environmental changes, and suggested future research directions. Among the three aquatic organisms, phytoplankton were the focus of studies related to water quality management, whereas most studies on macroinvertebrates and fish skewed toward modeling community composition changes and habitat suitability. Considering that many studies contained more than two study objects, there was a lack of research modeling future changes, such as climate change and subsequent changes in habitat conditions. Hybrid modeling methods using both correlative and mechanistic models have recently become more important, and are likely to improve modeling performance. Advanced models have the potential to significantly enhance the conservation and management of freshwater ecosystems, while also facilitating the development of effective policies that can better address the challenges faced by these ecosystems. Model uncertainty and sensitivity analysis, as well as the interpretable techniques of machine learning, also have the potential to improve model performance. This study provides valuable insights for modeling and general scientific research based on datadriven models.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.7
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available