Meeting European conservation and restoration targets under future land-use demands

The European Union is committed to achieving ambitious area-based conservation and restoration targets in the upcoming decade. Yet, there is a contention that these targets may conflict with societal needs, particularly food and timber production. Ensuring that competing demands for land are met, while the underlying objectives to mitigate climate change and reduce biodiversity loss are achieved, will require strategic planning across land uses, management measures, and jurisdictional boundaries. Here, we use an integrated spatial planning approach to identify restoration and conservation measures that maximize benefits to species as well as contributions to climate mitigation, while ensuring projected 2030 crop, pasture, and forestry production demands are met across the EU.

you can read a preprint version of the paper here

Code structure

All code used to format input data and run prioritization analysis is available in the /scripts/problem-setup/ folder:

• 1-PU.R: Sets up planning units and calculates initial land use proportions
• 2-zones.R : Sets up 25 management zones and planning unit level constraints for each zone
• 3-features-updated.R & 3b-feature-targets-disaggregated.R: formats zone contributions to feature targets and sets disaggregated feature targets (by country-biome-spp)
• 4-linear-constraints.R: sets constraints across nuts2 regions for f455 and BAU scenarios to ensure production targets are met at subnational scales ("linear constraints"). Due to some discrepancies in underlying data (and issues emerging from the decision unit scale), we adjust production targets in a small number of subnational jurisdictions to ensure feasibility to full extent possible without restoration target (see 4b-nuts2-infeasibility-*.R, for subnational solves and adjustments under different future production scenarios). Full production targets are ensured at the country scale.
• 5-problem-solve-allScenarios.R: run spatial optimization problem for all scenarios. Scripts are split by future productions scenarios (Fit455 and Reference/BAU).
• 5b-format-solutions.R: takes the outputs from the problem-solve scripts and calculates feature shortfalls, formats solutions as rasters, etc.
• 6-figures.R: makes all figures panels used in manuscript

In this code, we rely on Gurobi v 9.5 to solve optimization solutions (a proprietary LP solver, offers free academic licenses). However, it is possible to solve these problems with alternative open-source solvers (more information on supported solvers and their benchmarks here). In 5-problem-solve-allScenarios-*.R: we provide function option to run with the highs solver.

Data availability

All formatted data to create the optimization solutions presented in the paper (/scripts/problem-setup/5-*.R) is available in the /data/ folder. Underlying data for setting up input files (steps 1-4) is available upon request.