Inferring Birth Versus Death Dynamics for Ecological Interactions in Stochastic Heterogeneous Populations

Area of Science:

• Ecology
• Mathematical Biology
• Population Dynamics

Background:

• Ecological interactions and population dynamics are crucial but often implicitly represented in population-level data.
• Distinguishing between birth and death dynamics, and the nature of interspecies interactions (competitive, antagonistic, mutualistic), is challenging.
• Stochasticity plays a significant role in heterogeneous populations, influencing observed dynamics.

Purpose of the Study:

• To develop and validate an inference method for disentangling interaction types and birth-death processes in stochastic heterogeneous populations.
• To analyze population size time series data to infer underlying ecological mechanisms.
• To explore how different birth and death rate combinations affect population time series statistics.

Main Methods:

• Utilized general birth-death processes to model stochastic heterogeneous populations.
• Proposed an inference method analyzing population size time series data.
• Validated the method using a stochastic Lotka-Volterra interaction dynamics model.

Main Results:

• Demonstrated that distinct birth and death rate pairs, despite having the same net growth rate, yield different time series statistics.
• Successfully validated the inference method on a two-type Lotka-Volterra model.
• Showed that stochastic fluctuations allow estimation of parameters not identifiable in deterministic models.

Conclusions:

• The proposed inference method effectively disambiguates ecological interaction types and birth-death processes from population data.
• Stochastic fluctuations are essential for a comprehensive understanding and parameter estimation in ecological models.
• This approach offers a powerful tool for analyzing complex ecological systems and uncovering hidden interaction mechanisms.