Generalism drives abundance: A computational causal discovery approach

Area of Science:

• Ecology
• Causal Inference
• Community Ecology

Background:

• Abundant species often exhibit generalist traits, interacting with more species or occupying diverse habitats.
• A key ecological question is whether generalism leads to abundance (selection) or abundance leads to generalism (drift).
• Previous studies using formal logic suggested abundance drives generalism, but this is debated.

Purpose of the Study:

• To resolve the causal relationship between species generalism and abundance.
• To investigate the roles of selection versus drift in structuring ecological communities.
• To apply and refine computational causal discovery methods in ecology.

Main Methods:

• Refined a formal logic-based causal discovery method, correcting for skewed distributions.
• Employed two independent causal discovery methods: nonparametric regression and information theory.
• Applied these methods to empirical datasets from plant-hummingbird and reef fish communities.

Main Results:

• All three independent causal discovery methods consistently indicated that generalism drives abundance.
• Contrary to prior research, evidence strongly supports a selection process where generalism enhances abundance.
• Selection processes were found to be more influential than drift processes in variable environments.

Conclusions:

• Species generalism is a primary driver of species abundance in ecological communities.
• Computational causal discovery offers powerful tools for resolving long-standing ecological questions.
• Understanding the interplay of selection and drift is crucial for predicting community dynamics in changing environments.