Fitting stochastic predator-prey models using both population density and kill rate data.

Area of Science:

• Ecology
• Mathematical Biology
• Population Dynamics

Background:

• Mechanistic predator-prey modeling typically uses process rate data or inverse modeling.
• Stochasticity in population growth and predation processes presents unique modeling challenges.

Purpose of the Study:

• To explore the benefits of combining population density and kill rate data for stochastic predator-prey models.
• To assess the identifiability and precision of model parameters with and without kill rate data.

Main Methods:

• Fitting a discrete-time, stochastic predator-prey model (Leslie type) to simulated time series data.
• Incorporating environmental stochasticity and interaction stochasticity (stochastic functional response).
• Employing both Bayesian and frequentist estimation techniques.

Main Results:

• Kill rate data significantly enhances parameter estimation quality for models with fixed-point attractors.
• Noisy limit cycle attractors can be identified from population data alone, but kill rate data improves precision.
• Identifiability is possible with population data, but kill rate data is crucial for practical parameter estimation.

Conclusions:

• Interaction data, such as kill rates, is vital for obtaining identifiable dynamical models under process stochasticity.
• These findings can extend to other biotic interactions beyond predation.
• Combining diverse datasets strengthens ecological modeling capabilities.