Abstract

Agricultural biologicals represent a rapidly growing field providing potential solutions to food security and sustainable agriculture. Microbial products are one of the key components of agricultural biologicals, where beneficial microbes enhance crop production and resistance through plant-microbe interactions. However, the effectiveness depends on beneficial microorganisms successfully navigating through complex soil environments to reach plant targets such as roots and rhizosphere. Microbial movement poses significant challenges to monitor and study in real-world environments.
This study developed a flexible Agent-Based Model (ABM) that simulates the dynamics of microbial communities via a large number of microbial agents, each with their own attributes, including age, location, and functional class. The model incorporates interactions between inoculated microbes, native background microbial communities, and environmental factors. The microbial dynamics are governed by two mechanisms: population dynamics and directional movement preferences. The model simulated population dynamics via birth-death processes. Directional preference was driven by two factors: resource attraction describes chemotactic responses to plant signals such as plant root exudates, modeled via attraction strength parameter, and density-dependent repulsion force from neighboring microbes, modeled with repulsion intensity parameter.
The model was validated through systematic parameter analysis and laboratory experiments confirming that growth patterns matched simulated results. Comprehensive simulations replicating real world agricultural scenarios demonstrated that the model successfully reproduced theoretical population growth patterns and captured realistic chemotactic responses. Results showed that the attraction strength parameter effectively modulates microbial movement preferences toward resources, with small parameter values resulting in tight clustering. Large repulsion intensity values show realistic avoidance behaviors between microbes. Laboratory validation using fungal strains confirmed preferential growth toward nutrient sources, with optimal attraction strength value between 2.5 and 3.5 to accurately reproduce observed experimental growth patterns across multiple timepoints.
This ABM provides a flexible computational framework for understanding microbial behavior in agricultural systems and offers a computationally tractable approach for large scale population studies. The framework can incorporate additional mechanisms for studying microbial behaviors and environmental interactions, allowing researchers to tailor it for different research applications. The ABM enables predicting microbial behaviors in complex soil environments and contributes to the development of more effective agricultural biologicals for sustainable farming practices. The model is available at: https://github.com/ComputationalAgronomy/simulation_bacteria.