Reply2_F26

Reproducibility code and outputs for:

Martínez-Lanfranco JA (preprint). Habitat amount control is necessary but not sufficient to resolve the fragmentation debate. EcoEvoRxiv. [DOI — to be updated upon posting]

This repository contains the reproducibility script, computed output tables, and figures for Supplementary Material S1 of the above preprint. The preprint is a comment on:

Fahrig L, et al. (2026) Why controlling for habitat amount is critical for resolving the fragmentation debate. Conservation Biology e70245.

---

Contents

Code/
  Reply2_F26_SupplS1_reproducibility.Rmd   # R Markdown reproducibility script
  supplement_S1_numbers.csv                # all numbers produced by the script
  appendix_table_A1.csv                    # per-study predictor data (37 study pairs)
  session_info.txt                         # R session and package versions

Figures/
  figure_S1.png    # Figure S1 — realised predictor space by landscape type
  figure_S2.png    # Figure S2 — bidirectional linear vs GAM fits

---

Data

This repository does not redistribute the original dataset. All analyses use the publicly archived dataset of Gonçalves-Souza et al. (2025):

Gonçalves-Souza T, Vancine M (2025). Dataset and code for "Species turnover does not rescue biodiversity in fragmented landscapes." Zenodo. https://doi.org/10.5281/zenodo.14885581

To reproduce the analyses:

1. Download Data/diversity_of_2.csv from the companion GitHub repository at https://github.com/thiago-goncalves-souza/ms-biodiversity-loss-fragmented-landscapes
2. Place it in a Data/ subfolder relative to the Rmd script
3. Knit Reply2_F26_SupplS1_reproducibility.Rmd

---

Requirements

R 4.4.0 or later. The following packages are required (see Code/session_info.txt for exact versions):

tidyverse, lme4, glmm.hp, performance, car, mgcv, patchwork, knitr, kableExtra

Install all at once:

install.packages(c("tidyverse", "lme4", "glmm.hp", "performance",
                   "car", "mgcv", "patchwork", "knitr", "kableExtra"))

---

What the script produces

Knitting the Rmd generates a Reply2_F26_outputs/ folder containing:

File	Contents
supplement_S1_numbers.csv	All numbers in Tables S1–S5
appendix_table_A1.csv	Per-study predictor data (37 study pairs)
figure_S1.png	Figure S1
figure_S2.png	Figure S2
session_info.txt	R session information

No values are entered manually. Every number in Supplementary Material S1 is produced by this script.

---

Scope

This repository documents design-level predictor structure only. The script does not analyse biodiversity responses and does not reproduce any biodiversity models from Gonçalves-Souza et al. (2025) or Fahrig et al. (2026).

---

Contact

Juan A. Martínez-Lanfranco — jacoloml@gmail.com