Variation partitioning in double-constrained multivariate analyses: linking communities, environment, space,...

Area of Science:

• Ecology
• Environmental Science
• Biodiversity Research

Background:

• Constrained multivariate analysis links ecological communities to environmental factors.
• Double-constrained correspondence analysis (dc-CA) integrates species traits as predictors.
• Existing methods incorporate phylogenetic and spatial variability.

Purpose of the Study:

• To develop a novel algorithm for decomposing community structure variation.
• To test the effects of environmental, spatial, trait, and niche predictors.
• To distinguish between traits and ecological niches as distinct predictors.

Main Methods:

• Proposed a dc-CA-based algorithm for variation decomposition.
• Integrated four predictor sets: environmental characteristics, space configuration, traits, and niche dissimilarities.
• Introduced new niche parameters and synthetic niche-based diversity.
• Developed an algorithm for testing the community-environment-niche-traits-space (CENTS) using dc-CAs.
• Created new visualization diagrams for results.

Main Results:

• Niche predictors were found to be as important as traits in explaining community structure.
• Niche predictors are not redundant and can outweigh environmental and spatial predictors.
• The study applied the method to freshwater mollusk datasets, demonstrating its utility.

Conclusions:

• The developed algorithm provides a comprehensive framework for linking life, environment, and other predictors.
• This approach offers new theoretical and practical applications, including human impact assessment.
• Ecological niches and species traits should be considered distinct and important predictors in community ecology.