WeatherXBiodiversity: Soroye et al. (2020) Replication for Iberian Bombus

Python port and Iberian replication of Soroye et al. 2020 — bumble bee climate exposure → local extinction

Tests whether the thermal-exposure mechanism reported by Soroye, Newbold and Kerr (2020, Science) — an increasing frequency of temperatures exceeding species-specific historical tolerances predicts local bumble bee extinction — replicates on the Iberian Peninsula using an independent Python re-implementation of the authors' R pipeline, GBIF occurrence data, and CRU TS 3.24.01 climate.

Headline result. The claim replicates: the Iberian coefficient (sc_TEI_delta = +0.48, 95 % CI [0.27, 0.69]) is positive and significant, and larger in magnitude than the continental mean Soroye reports (+0.15 in our Python-port validation on his own data).

Reference paper: Soroye P., Newbold T., Kerr J. (2020). Climate change contributes to widespread declines among bumble bees across continents. Science 367(6478):685–688. DOI: 10.1126/science.aax8591

What's in this release

• + "soroye_port/" + — five-script Python port of Soroye's R pipeline: occurrence cleaning, 100 km equal-area presence–absence grids, sampling-effort rasters, TEI/PEI climate indices, and the mixed-effects logistic GLMM.
• Two entry points: + "01_clean_data.py" + (Soroye's NA + Europe data, Phase 2 port validation) and + "01_clean_data_iberia.py" + (GBIF Iberia, Phase 3 regional replication). Scripts 02–05 are shared.
• Dockerfile, Snakefile, continuous-integration smoke test, and a Jupyter Book that narrates the FORRT reproduction + replication story.

What changed from v0.1.0 Replaces the earlier simplified monthly-means analysis with a line-by-line port of Soroye's five R scripts. The preliminary v0.1.0 coefficient sign was a methodology artefact and is superseded by the values above.

Citation If you use this software, please cite both this repository (via the Zenodo concept DOI, which will be linked in the README once minted) and Soroye et al. (2020) above.