This repository contains the code for the simulation and visualization of an ODE- and network-based model of tree-to-tree nutrient transfer through ectomycorrhizal networks (EMNs). The model's dynamics are based on the source-sink hypothesis, where older/bigger trees act as sources of carbon for smaller/younger trees (saplings). The model is largely based on the data and results from "Architecture of the wood-wide web: Rhizopogon spp. genets link multiple Douglas-fir cohorts" (https://doi.org/10.1111/j.1469-8137.2009.03069.x). The model dynamics are greatly simplified and do not fully capture the complex and diverse dynamics in real-world EMNs. However, this simple model displays directed carbon (C) transfer through common EMNs favoring saplings that have higher respiration demands compared to older trees. This in turn leads to competition between saplings, where some saplings outcompete others due to being connected to better or more trees. These dynamics emerge from simple, local node-to-node rules.
growth_and_carbon.mp4
data/: Contains network data from paper used in experiments and saved data from computationally demanding experiments.figures/: Plots and network visualizations used in presentation.movie_frames/: Contains animation(s) of network.src/: Python source files used in experiments/simulations.
Notebooks containing visualization and analysis code for various experiments are found in the root of the repository. The notebooks contain the following experiments:
Forest Data Experiments.ipynb: Main results and experiments on the forest data from the Beiler paper.Forest Attack and Failure Experiment.ipynb: Attack and failure experiments on forest data.Average Degree vs. Average Growth Experiment.ipynb: Experiment on the importance of connectedness on growth.Characteristic Time Experiment.ipynb: Experiment on the influence of connectedness on the characteristic time.Network Dynamics Animation.ipynb: Notebook for creating the animation of carbon spread and growth throughout the forest network.Barabasi-Albert Network Experiments.ipynb: Experiments repeated on generated Barbasi-Albert scale-free network.Erdos-Renyi Network Experiments.ipynb: Experiments repeated on generated Erdos-Renyi random network.