Incorporating phylogenetic conservatism and trait collinearity into machine learning frameworks can better predict macroinvertebrate traits

Abstract

In the face of rapid environmental changes, understanding and monitoring biological traits and functional diversity are crucial for effective biomonitoring. However, when it comes to freshwater macroinvertebrates, a significant dearth of biological trait data poses a major challenge. In this opinion article, we put forward a machine-learning framework that incorporates phylogenetic conservatism and trait collinearity, aiming to provide a better vision for predicting macroinvertebrate traits in freshwater ecosystems. By adopting this proposed framework, we can advance biomonitoring efforts in freshwater ecosystems. Accurate predictions of macroinvertebrate traits enable us to assess functional diversity, identify environmental stressors, and monitor ecosystem health more effectively. This information is vital for making informed decisions regarding conservation and management strategies, especially in the context of rapidly changing environments.