Measuring nestedness: A comparative study of the performance of different metrics

Area of Science:

• Ecology
• Network Science
• Quantitative Biology

Background:

• Nestedness is a key property of interaction networks in mutualistic communities.
• Existing nestedness metrics lack consensus, hindering cross-ecosystem comparisons.
• Understanding network structure is crucial for ecological research.

Purpose of the Study:

• To critically evaluate the behavior of six nestedness metrics.
• To compare metric performance against null models and analyze dependencies on network parameters.
• To provide a guide for measuring nestedness patterns in ecological networks.

Main Methods:

• Comparative analysis of six nestedness metrics and two variants.
• Evaluation against a maximum-entropy and maximum-likelihood null model.
• Analysis of metric dependencies on network size, fill, and eccentricity.

Main Results:

• Nestedness metrics do not universally rank networks.
• Several metrics exhibit significant dependencies on network properties.
• Dependencies explain systematic shifts observed against null models.

Conclusions:

• The study highlights the flaws and qualities of different nestedness metrics.
• It provides a critical guide for selecting and applying nestedness measures.
• It extends the application of null models to ecological networks and offers a computational repository.