Adaptive Crown of Thorn Starfish control: ReefMod-GBR outputs, model, and code

Full description

ReefMod-GBR is a spatially explicit ecological model that simulates the dynamics of coral colonies on 3806 reefs of the Great Barrier Reef over time (Bozec et al., 2022). The model represents coral demographic processes, including recruitment, growth and survival, ecological interactions with algae and Crown-of-Thorns Starfish (CoTS), and disturbances such as heat stress, cyclones and sedimentation. ReefMod-GBR was used to assess the ecological benefits and logistical costs of multiple CoTS control strategies over time. Simulations were run from 2027 to 2050 at six-monthly timesteps using three global climate models (GCM) (CNRM-ESM2-1, MRI-ESM2-0 and MIROC-ES2L) under Shared Socioeconomic Pathway 2-4.5. The model outputs include coral cover, coral composition and CoTS control activities, which were processed and analysed in R to compare the performance of different control strategies. This dataset includes: 1. ReefMod scenario outputs (MAT and CSV) The outputs are provided in both MATLAB MAT and CSV formats. Two sets (1HS and 0HS) of outputs are included. 1HS: Simulations account for the effects of heat stress on CoTS detection and culling, with greater heat stress reducing the effectiveness of CoTS control operations. 0HS: Simulations assume that heat stress does not affect CoTS detection or culling. Heat stress continues to affect coral dynamics in the model. MAT files contain the raw outputs generated by ReefMod-GBR. File names follow the format [s/l][CoTS control strategy number]_[number of simulation timsteps]_SSP245_[GCM], where s / l: Short format (annual outputs) / long format (six-monthly outputs). CSV files contain processed and reformatted model outputs converted from MAT files for further analysis in R. They include reef-level outputs (annual results for individual reefs, such as coral area and outbreak area), GBR-level and scenario outputs (results across the 3,806 reefs, such as total coral area and lists of reefs visited or culled in each year), and persistence outputs (data used to analyse how long the effects of CoTS suppression persist following culling). Note that for the naming of the CSV files, CoTS control strategies have been recoded to match the scenario IDs used in the project summary results report. These IDs therefore differ from the original strategy numbers used in ReefMod-GBR. The project summary results report and supplementary materials can be found via this link: https://www.dropbox.com/scl/fo/zkgjvq5ggd1uwr4ubhvk1/ANSPkm7Rblp4yntCvwij-A8?rlkey=tqjz1fjz4idhgqs9dkea5ag9l&st=0ltyrvc0&dl=0 2.ReefMod-GBR model code This includes a modified version of ReefMod-GBR based on ReefMod-GBR 7.3, available at: https://github.com/ymbozec/REEFMOD.7.3_GBR.git. The model code is written in MATLAB and uses MAT and CSV data files for model inputs and outputs. Compared with the original ReefMod 7.3, the model has substantially expanded CoTS control capabilities. Key modifications include multiple adaptive control strategies with reef-level and regional-level bleaching triggers, as well as reef relocation rules with different scoring criteria. CoTS detection and culling can be affected by heat stress, with new CoTS-per-tow/density conversions and survey (manta tow & RHIS), culling, and travel effort calculations. CoTS strategies can be adjusted at different timesteps. The model inputs and hindcast boundary were also updated substantially. This includes new 2007 CoTS initial conditions based on CoTS control data (instead of CoCoNet), coral and CoTS observations through 2026, 2026 coral and reef geometry data, updated DHW and cyclone datasets, revised workable reef dataset, and updated CoTS risk/reference tables. The code was also reorganised for repeated, simultaneous HPC runs of different alternative scenarios and customised output extraction. 3. Data processing and analysis scripts. This includes MATLAB and R scripts used to process and analyse the model outputs. scenario_matlab2R_clean: MATLAB scripts for converting raw MAT outputs into CSV files and preparing the data for analysis in R. scenario_analysis: R scripts for generating summary datasets and plots for visualising and comparing scenario results. 4. Summary of initial discussions that formed the basis of modelling the adaptive scenarios. Early consultations with reef managers and vessel operators through online meetings took place between November 2025 and January 2026. The team identified three broad adaptive responses to bleaching: improving existing control methods, relocating effort to other reefs, and switching from culling to surveillance. Following these initial discussions, ongoing engagement with reef managers through email correspondence, as well as online and in-person meetings helped refine the design of the modelling scenarios, identify relevant datasets, and ensure that operational assumptions reflected current management practices. The findings from the initial discussions are summarised into this document: CCIP-BR-07_Attachment 1_Document_Summary of initial discussions_final.docx