GRINanalysis (Grafted Root Interaction networks) - A VWStifftung funded project -

This repository contains data and analyses of the manuscript entitled "Cooperative root graft networks benefit mangrove trees under stress", currently under Review at Communications Biology 2021.

Co-Authored by: AG Vovides, MC Wimmler, F Schrewe, T Balke, M Zwanzig, C Piou, E Delay, J López-Portillo and U Berger

The following lines briefly describe the files included in this repository.

The setup.R file installs all required packages at once.

1) Data

This folder contains the processed data (rda-files) to run the analysis and create the figures.

Processed data

• LMtrees.rda: contains data (e.g. unique ID, coordinates, species, DBH, graft-status, …) for each tree in the study area, resulting from merging raw data bases of tree attributes (publicly available at https://doi.org/10.5525/gla.researchdata.657) and field mapping of root connections through root grafts (Networks.rda)
• LMavis.rda: is a subset of LMtrees.rda containing only Avicennia germinans trees, but also additional information as the salinity group of the study plot and the cumulated grafting probability of each tree.
• LMlinks.rda: contains all grafted pairs of trees with additional information like the distance between them or the group they belong to.
• LMgroups.rda: contains all groups of grafted trees with additional information as the number of trees and the number of links within each group.
• AlldistLines.rda and Plot_data.rda: contains node degree destribution
• data_gam.rda: contains gam model results

2) Analysis

This folder contains

• The multilevel logistic regression model described in the extended data section (glmer_grafting.R)
• The computations to extract the indices of competition (Hegyi) and neighborhood asymmetry (neighborhood_gamm_slenderness.R). The latter used on the Generalized additive mixed effects model (GAMM). This script also contains the necessary code to reproduce extended data Figures 3 and 4, as well as Supplementary information Figures 1 and 2.
• The comparison between power-law distribution and other random-process distribution functions (power_law_distributions.R).
• Descriptive statistics and simple correlations (numbers.R)

3) Figures

This folder contains the scripts needed to reproduce figures present in the manuscript (including those in Supplementary Information Files).