Climate resilience of European wine regions

Geographical indications (GIs) underpin the identity of many European wines by legally linking products to specific places, practices, and grape varieties (terroir). Climate change is altering viticultural conditions (phenology, composition, suitability), potentially decoupling traditional grape–place–practice combinations and challenging GI resilience. Legal rigidity in PDO regulations—restricted varieties, fixed techniques, and strict provenance—can impair adaptation. Prior work has emphasized bioclimatic pressures (temperature, precipitation) and varietal diversity, but comprehensive, multi-dimensional vulnerability assessments for wine GIs remain sparse and mostly region-specific. This study addresses this gap by assessing the climate change vulnerability of 1085 European PDO wine regions using an IPCC-based framework (exposure, sensitivity, adaptive capacity), integrating biophysical, socioeconomic, and regulatory data to inform climate-resilient pathways for the GI system.

Existing studies document climate change impacts on European viticulture, including shifts in temperature and precipitation affecting yields, phenology, berry composition, and suitability. Research has linked bioclimatic indices (e.g., Huglin Index, Cool Night Index, Dryness Index) to viticultural outcomes and explored how varietal diversity buffers regions under future climates. While vulnerability assessments are common in other sectors and crops, applications to wine GIs are limited and typically focus on single regions. Moreover, previous adaptation studies have emphasized climatic exposure while often neglecting legal and socioeconomic dimensions that constrain or enable adaptation. The relationship between adaptive capacity and GI resilience remains underexplored, underscoring the need for integrated assessments at scale.

Design: A continental-scale, index-based vulnerability assessment for 1085 European PDO wine regions following the IPCC framework. Vulnerability integrates three components: exposure (projected climate change), sensitivity (degree to which PDOs are affected based on traditional varietal climate ranges), and adaptive capacity (socioeconomic and biophysical resources to adapt). All indicators are standardized to 0–1 for comparative analysis.

Study regions: PDO wine regions across the EU (Italy 35%, France 31%, Spain 8%, Bulgaria 4%, Romania 3%, Hungary 3%, Portugal 3%). Spatial boundaries from Candiago et al.

Exposure: Quantified as projected change between present (1981–2010) and future (2071–2100) under SSP3-7.0 using three viticulture-specific bioclimatic indices: Huglin Index (HI), Cool Night Index (CNI), and Dryness Index (DI). Future climate derived from CHELSA downscaled data (1 km) using the ensemble mean of 5 CMIP6 GCMs (GFDL-ESM4, IPSL-CM6A-LR, MPI-ESM1-2-HR, MRI-ESM2-0, UKESM1-0-LL), largely representing ISIMIP primary models. Index changes computed per grid cell, averaged within each PDO, min–max normalized (0 = smallest change, 1 = largest), and averaged across HI, CNI, DI to yield exposure. Model spread for temperature and precipitation assessed to characterize uncertainty.

Sensitivity: Represents how PDOs are affected given the historic bioregional climate ranges of their primary grape varieties. Steps:

• Primary varieties and cultivation areas extracted from PDO product specifications; areas linked using Anderson & Nelgen dataset, harmonizing synonyms via multiple sources (coverage ~70% of listed varieties). When areas were given at macro-region level, allocated to PDOs by authorized status weighted by vineyard area (CORINE Land Cover and OpenStreetMap).
• Bioregional climate range derivation: PDOs categorized into 17 bioclimatic groups based on HI, CNI, DI (1981–2010) per Fraga et al. Indices computed from CHELSA (1 km) and averaged per PDO. For each variety and group, a weighted average and standard deviation were computed for each index, weighting by cultivation area, to approximate historic growing conditions. Ranges validated against literature for selected varieties.
• Sensitivity calculation: For each PDO, the difference between current climate and the upper limits of the bioregional climate ranges of its primary varieties was computed. Varieties near the upper limit imply higher sensitivity. PDO sensitivity is the cultivation-area–weighted average across primary varieties, min–max normalized (0 = lowest, 1 = highest sensitivity). A supplementary analysis assessed potential positive shifts toward varietal reference conditions under SSP1-2.6, SSP3-7.0, SSP5-8.5 for 2041–2070 and 2071–2100.

Adaptive capacity: Fifteen indicators across five dimensions, scaled using 5th–95th percentile thresholds to reduce outliers and averaged with equal weights, then min–max normalized (0–1):

• Social: Aging index; Dependency ratio.
• Physical: Population density (per agricultural area); Road length; Mechanization Index (€/ha); Naturalness (% natural/semi-natural in winegrowing areas).
• Natural: Shift in space (km² of cooler suitable areas); Water availability (mm excess precipitation); Available climatic niches (intra-regional thermal variability, °C).
• Human: Labor force (% regular farm labor); Education level (farm managers); Research accessibility (km to nearest wine/vine research center).
• Financial: Debt ratio (% liabilities/total assets); Return on assets (%); Subsidy dependence (% net income from subsidies). An AHP expert survey across five PDOs (Maremma Toscana, Douro, Südsteiermark, Moselle Luxembourgeoise, Târnave) showed highly heterogeneous indicator weights and weak cross-region correlation; hence equal weighting was adopted for pan-European neutrality.

Integrated vulnerability index: Each component (exposure, sensitivity, adaptive capacity) categorized as low, moderate, high using terciles. Rules: Very high = high exposure and sensitivity with low adaptive capacity; High = any two unfavorable categories (or low adaptive capacity); Low = at least two favorable categories (or high adaptive capacity); Moderate = remaining cases. Regions additionally grouped into six clusters with similar component profiles to discuss pathways. Outputs provide comparative rankings and country shares by vulnerability classes.

• Exposure: Highest projected climate change exposure (mean >0.7 on 0–1 scale) in Romania, Croatia, Bulgaria, Italy, and Hungary—often in mountainous terrains (Apennines, Alps, Carpathians). Lower exposure (mean <0.4) in ocean-influenced areas such as Portugal, the Canary Islands, and higher-latitude countries like Belgium and the Netherlands. General trends: higher temperatures (HI, CNI increase) and drier conditions (DI decrease). Model spread increases for temperature toward century’s end, especially in Central/Eastern Europe; precipitation uncertainty highest in mountainous regions.
• Sensitivity: Many Southern European PDOs show higher sensitivity due to limited varietal spectra and/or already warm climates near the upper bounds of varietal climate ranges. Some Southern PDOs (e.g., Do Tejo, PT) have low sensitivity, while some higher-latitude PDOs (e.g., Champagne, FR) are relatively sensitive. Where sensitivity is low, some authorized varieties could initially benefit from warming, but potential benefits decline under stronger climate change scenarios and are constrained by limited authorized varietal ranges.
• Adaptive capacity: Highest levels concentrated within/near the Alps and along the Apennines (e.g., Conegliano Valdobbiadene Prosecco, Alto Adige). Slovenia and Italy have the largest shares of PDOs in the upper quartile of adaptive capacity (65% and 14%, respectively); France has less than 10%. Central Spain and Eastern Europe (e.g., Slovakia, Greece, Romania, Bulgaria, Hungary) show low adaptive capacity (<0.3 on average). Spanish PDOs often have stronger financial capacity but weaker physical and natural dimensions, lowering overall adaptive capacity. Higher-latitude regions (parts of FR, DE, DK, BE, NL) tend to be moderate (~0.5).
• Integrated vulnerability: 5% of PDOs exhibit very high vulnerability (Group 6) with high exposure and sensitivity and low adaptive capacity, including examples in Bulgaria (Southern Black Sea), Romania (Oltina), Hungary (Hajós-Baja), parts of Italy (Trebbiano d'Abruzzo, Lambrusco Mantovano) and Spain (Sierra de Salamanca). Groups 3–5 are highly vulnerable but heterogeneous: Group 3 has high exposure and sensitivity but comparatively higher adaptive capacity (e.g., parts of SE France such as Côtes de Provence; Conegliano Valdobbiadene Prosecco; East Slovak; Alentejo; Rioja). Group 4 combines lower sensitivity with limited adaptive resources; Group 5 shows lower exposure but high sensitivity and low adaptive capacity. Groups 1–2 (low to moderate vulnerability) include many higher-latitude and Alpine regions (e.g., Rheinhessen, Crémant de Wallonie, Côtes d'Auvergne, Alsace, Alto Adige) with either higher adaptive capacity or lower sensitivity, offering better prospects to maintain identity. Even in cooler regions, long-term benefits are not guaranteed without regulatory flexibility due to declining varietal benefits under higher emissions.

The study addresses the central question of which European PDO wine regions are most vulnerable to climate change and why, by jointly evaluating exposure, sensitivity (rooted in traditional varietal–climate relationships), and adaptive capacity (resources enabling adaptation). Findings reveal that vulnerability is highest where warming and drying intensify (SE and parts of E Europe), where varietal portfolios are narrow and near upper climatic thresholds, and where adaptive resources are scarce. Conversely, regions with higher adaptive capacity or lower sensitivity are better positioned to sustain product identity. These insights inform adaptation pathways aligned with each region’s component profile: high-exposure regions may mitigate via viticultural process adjustments (canopy management, irrigation where feasible, vineyard structure, rootstock choice, cover crops) to buffer microclimate and water stress; high-sensitivity regions can adjust blends/compositions and experiment with introducing new varieties; where both exposure and sensitivity are high, more transformative measures (elevation shifts, microclimate niches, site relocation) may be necessary. However, legal rigidity of PDO regulations constrains such options. Thus, enhancing resilience will require regulatory flexibility—expanding authorized varieties, permitting innovative practices, and potentially redefining geographical or product specifications—while safeguarding terroir values. Regions with strong adaptive capacity across natural, physical, financial, human, and social dimensions can deploy broader, costlier strategies; low-capacity regions will need prioritized, timely support and targeted measures. Overall, the integrated, comparative assessment highlights where policy, research, and investment should focus to sustain the GI system under accelerating climate pressures.

This work delivers the first pan-European, integrated vulnerability assessment for 1085 PDO wine regions, combining exposure, sensitivity anchored in bioregional varietal climate ranges, and a multi-dimensional adaptive capacity index. It identifies geographic hotspots of vulnerability in Southern and Eastern Europe and underscores the pivotal role of GI system rigidity in amplifying sensitivity. It also maps where adaptive resources are relatively strong (Alpine and Apennine regions) versus weak (central Spain, parts of Eastern Europe), guiding tailored adaptation pathways. Future research should close key data gaps: variety-specific phenological and distribution data to refine sensitivity; local-scale socioeconomic and management data to parameterize adaptive capacity and derive region-specific indicator weights; and deeper evaluation of management practices (rootstocks, training systems, inter-row vegetation) and their efficacy under diverse topoclimatic contexts. Policymaking should consider evolving PDO regulations to enable innovation—broader varietal authorization, adaptive viticultural/enological practices, and flexible geographic specifications—preserving the terroir–consumer link while enhancing climate resilience.

• Variety-specific data limitations: Incomplete, uneven information on spatial distribution and phenology for many authorized varieties (only ~70% coverage linked to cultivation area), limiting precision of sensitivity estimates.
• Adaptive capacity data gaps: Lack of local-scale, viticulture-specific socioeconomic and management data across Europe prevents region-specific weighting; equal weights may obscure local priorities.
• Aggregation choices: Vulnerability outcomes are sensitive to indicator standardization, weighting, and aggregation (additive vs multiplicative), potentially affecting rankings.
• Climate model uncertainty: Increasing temperature projection spread toward century’s end in parts of Central/Eastern Europe; precipitation uncertainty higher in mountains; results reflect ensemble means.
• Regulatory dynamics: Static treatment of PDO regulations may not capture future legal adaptations that alter sensitivity/adaptive capacity.
• Comparative scope: Indicators scaled relatively (0–1) for European comparison; results are high-level and not a substitute for region-specific, detailed assessments.