The choice of land-based climate change mitigation measures influences future global biodiversity loss

Simultaneously achieving climate change mitigation and biodiversity conservation is challenging, as climate change threatens food production and increases disaster risks, while biodiversity loss degrades ecosystem services. The Paris Agreement targets limiting warming to well below 2 °C and pursuing 1.5 °C, implying stringent cumulative CO2 budgets that are difficult to meet via emissions reductions alone. Integrated assessment models (IAMs) often rely on large-scale carbon dioxide removal (CDR), especially BECCS and afforestation, although DACCS costs are falling. Large-scale land-based CDR raises concerns about habitat conversion, particularly bioenergy expansion and afforestation in naturally open ecosystems, which can harm biodiversity. It remains unclear how global-scale implementation of BECCS and afforestation interacts with regional biodiversity conservation. This study evaluates the global and regional biodiversity impacts of deploying BECCS and afforestation to meet Paris-consistent pathways, quantifying both climate-driven and land-use-driven effects and examining regional equity implications.

Prior work indicates that meeting 1.5–2 °C pathways typically requires substantial CDR, with BECCS and afforestation prominent in IAM scenarios. Concerns include biodiversity loss from conversion of natural and pasture lands to bioenergy cropland, and negative biodiversity effects of afforestation in naturally open habitats (savannas, prairies). Alternatives such as DACCS are gaining attention due to cost declines but remain uncertain at scale. Nature-based solutions (NbS), restoration, and improved forest management (including harvested wood products for carbon storage) are discussed as potentially more biodiversity-compatible options. Previous studies suggest climate stabilization can benefit biodiversity overall but may involve trade-offs from land-use changes associated with mitigation.

The study used the AIM (Asia-Pacific Integrated Model) framework to assess biodiversity impacts of climate mitigation and land-use change. Scenarios: (1) Baseline with no GHG reductions; (2) 2C-BECCS (CDR mainly via BECCS; forest area not less than baseline); (3) 2C-Aff (CDR mainly via afforestation; near-zero biofuel demand); and (4) 2C-Opt (optimal use of BECCS and afforestation with a land-sector carbon price cap of US$200/tCO2). All mitigation scenarios follow a cumulative carbon emissions budget of approximately 900–1000 GtCO2 to be compatible with Lower-2 °C pathways under SSP2 socioeconomic assumptions. Land-use modeling: AIM/Hub projects regional aggregated land demand for 17 global economic regions; AIM/PLUM downscales to 0.5° grids into 12 categories (merged into five for biodiversity: cropland, pasture, forest, other natural land, settled). Biodiversity modeling: AIM/BIO uses species distribution models (Maxent with linear and quadratic features) for 8428 species (vascular plants, amphibians, reptiles, birds, mammals), relating distributions to 19 bioclimatic variables and five land-use variables. Current climate is WorldClim v2.1 (1970–2000); future climates use CMIP6 GCMs (GFDL-ESM4, IPSL-CM6A-LR, MRI-ESM2-0), with SSP3–7.0 for the baseline and SSP1–2.6 for mitigation scenarios. Climate data at 10-arc-minute resolution were averaged to 0.5°. Species’ dispersal constraints were applied via time-step buffering. Biodiversity indicators: local species richness and the Jaccard similarity index (temporal compositional change) were computed for 2005, 2030, 2050, 2070, and 2090. To disentangle drivers, two sub-models were used: CC-model (climate change only; land use fixed at current) and LU-model (land-use change only; climate fixed at current). Sensitive grids (change in richness < −0.1 or > 0.1 by 2090) were analyzed to attribute dominant land-use transitions via Sankey diagrams and to identify sensitive ecoregions (>=15% sensitive grids). Regional equity: relationships between biodiversity loss and (a) land-use change intensity (largest categorical occupancy change per grid), (b) cumulative carbon sequestration (BECCS or afforestation) from 2005–2090, and (c) GDP per capita in 2090 were assessed across 17 AIM regions.

- Global biodiversity: Under the baseline, relative species richness declined by −0.015 (2030) and −0.069 (2090). Mitigation reduced losses: 2C-BECCS showed −0.014 (2030) and −0.033 (2090); 2C-Aff showed −0.020 (2030) and −0.047 (2090). Variance increased over time but was larger under the baseline than mitigation. - Community composition: Jaccard similarity indices declined less under mitigation. Baseline dropped to 0.692 (2070) and 0.630 (2090). 2C-BECCS: 0.762 (2070) and 0.759 (2090); 2C-Aff: 0.741 (2070) and 0.735 (2090). - Land-use change: The 2C-Aff scenario entailed larger land-use changes than 2C-BECCS, explaining its greater biodiversity impacts. - Regional climate effects (CC-model): Mitigation reduced biodiversity losses across Central/South America, Africa, Europe, SE Asia, and Oceania; however, in Arctic/sub-Arctic regions and the Tibetan Plateau, baseline warming increased local richness relative to mitigation due to poleward/upslope range expansions. - Regional land-use effects (LU-model): 2C-BECCS reduced richness versus baseline in parts of Central/South America, Sub-Saharan Africa, and Central Asia, primarily where other natural land and pasture were converted to cropland; limited areas showed gains. 2C-Aff reduced richness versus baseline across wider areas of the Americas, Sub-Saharan Africa, Europe, North/East/Southeast Asia, and Oceania, primarily where pasture/other natural land were converted to forest; a few regions (e.g., temperate grassland/desert in East Asia and Oceania) showed gains where pasture converted to forest. Overall, land-use impacts were stronger and more extensive under 2C-Aff than 2C-BECCS. - Ecoregions at risk: Grassland and savanna ecosystems in North, Central, and South America, Sub-Saharan Africa, East Asia, and Oceania are especially susceptible to biodiversity losses from conversion of pasture/other natural land to cropland (BECCS) or forest (afforestation). - Regional equity: Biodiversity loss correlated with greater land-use change intensity (baseline: R2=0.072, p=0.298; 2C-Aff: R2=0.364, p=0.010; 2C-BECCS: R2=0.316, p=0.019). Losses also tended to be greater in regions contributing more to carbon sequestration (2C-Aff: R2=0.203, p=0.069; 2C-BECCS: R2=0.194, p=0.077). Mitigation tended to reduce biodiversity losses in lower-GDP regions, albeit with large variation (baseline: R2=0.135, p=0.147; 2C-Aff: R2=0.022, p=0.567; 2C-BECCS: R2=0.080, p=0.272). - Overall, BECCS resulted in less global biodiversity loss than afforestation due to smaller land area affected by land-use change.

The study shows that meeting Paris-consistent climate targets benefits global biodiversity by reducing species loss and maintaining community composition closer to current baselines. However, high-latitude and high-altitude regions could see local richness gains under stronger warming because of range expansions of species formerly limited by cold/dry conditions, at the cost of biotic homogenization and loss of endemic compositional diversity; mitigation curtails these shifts, preserving local uniqueness. Land-based CDR measures introduce regional trade-offs: converting other natural lands and pastures to cropland (for BECCS) or to forests (afforestation) often reduces biodiversity, particularly in naturally open ecosystems (savannas, prairies). Impacts are regionally uneven, with greater biodiversity losses where land-use change and carbon sequestration contributions are higher. Strategies to reconcile mitigation with conservation include prioritizing already modified lands (existing cropland/pasture) for BECCS feedstocks, avoiding afforestation of naturally open habitats, and applying biome-appropriate zoning that favors reforestation where forests were historically present. Alternative or complementary measures—restoration of natural ecosystems, improved forest management, and increased carbon storage in harvested wood products—may deliver CDR with fewer biodiversity trade-offs if carefully designed to avoid pressures on natural forests. Equity considerations are critical to prevent disproportionate biodiversity burdens on specific regions, especially those with lower GDP, and to distribute mitigation responsibilities more fairly.

Climate change mitigation reduces global biodiversity loss relative to a no-mitigation baseline, regardless of whether CDR is achieved via BECCS or afforestation. Nonetheless, land-based CDR imposes regional biodiversity costs, with BECCS generally causing less global biodiversity loss than afforestation due to smaller land-use changes. Regions contributing more to carbon sequestration face greater biodiversity impacts, especially grassland and savanna ecoregions across the Americas, Sub-Saharan Africa, East Asia, and Oceania. To balance mitigation with conservation, prioritize already modified lands, minimize conversion of natural ecosystems, and adopt region- and biome-appropriate strategies. Nature-based solutions and improved forest management may reduce regional disparities in biodiversity impacts. Ultimately, deep emissions reductions across sectors are needed to lessen reliance on land-based CDR and its associated biodiversity trade-offs.

- Land-use representation treats bioenergy cropland as equivalent to conventional cropland due to limited empirical data on large-scale BECCS, potentially masking crop- and management-specific biodiversity effects. - Afforestation is modeled as forest expansion without distinguishing plantation types (monoculture vs. diverse native plantings), limiting assessment of forest quality effects on biodiversity. - Difficulty in directly comparing biodiversity impacts of forest use/management versus land-use conversion; potential risks that strong demand for harvested wood products could drive natural forest loss. - Uncertainty in species responses to large-scale Arctic/sub-Arctic environmental changes (e.g., permafrost thaw), and in climate projections despite using three CMIP6 GCMs. - The 2C-Opt scenario is defined but detailed biodiversity results focus mainly on baseline, 2C-BECCS, and 2C-Aff. - Assumptions about species dispersal and use of potential habitat may not capture all real-world constraints on range shifts.