D1.12 IA 1.5 brief: Report on innovative post-fire restoration and adaptation strategies in the actual context of increasing environmental uncertainty

This document was produced under the terms and conditions of Grant Agreement No. 101037419 of the European Commission. It does not necessarily reflect the view of the European Union and in no way anticipates the Commission’s future policy in this area. www.fire-res.eu [email protected] : FIRE-RES : Innovative technologies and socio-ecological-economic solutions for fire resilient territories in Europe : H2020-LC-GD-1-1-2020 (Preventing and fighting extreme wildfires with the integration and demonstration of innovative means) : 1 : 1.4.2 : Forest Science and Technology Centre of Catalonia (CTFC) : Cabildo Insular Gran Canaria (CIGC) : 30/04/2025 : Pedro Tardós, Eva Samblàs Vives, Lluís Coll, Lena Vilà-Vilardell, Eduardo Balguerías Quintero, Luis Fernando Arencibia Aguilar, Federico Grillo Delgado, Míriam Piqué. : This deliverable is the second of two focusing on post-fire strategies and adaptation to increasing environmental uncertainties. As large fires and extreme wildfire events (EWE) become more frequent, understanding the factors that influence forest resistance and recovery is crucial. Both pre-fire and post-fire interventions play a key role in enhancing forest resilience and adaptation to EWE. This deliverable presents three innovative management strategies to help forests adapt to the increasing occurrence of EWE, namely (i) silvicultural treatments for EWE prevention and pasture improvement, (ii) prescribed burning and grazing introduction for EWE prevention, and (iii) restoration of vegetation cover after an EWE. : fire adaptation, grazing, pasture improvement, prescribed burning, restoration, silvicultural treatments : Tardós, P., Samblàs, E., Coll, L., Vilà-Vilardell, L., Balguerías, E., Arencibia, L.F., Grillo, F., Piqué, M. (2025). Report on innovative post-fire restoration and adaptation strategies in the actual context of increasing environmental uncertainty. D1.12 IA brief 1.5 FIRE-RES project. https://doi.org/10.5281/zenodo.15348087 : 10.5281/zenodo.15348087 [X] PUPublic: must be available in the website [ ] COConfidential: Only for members of the Consortium and the Commission Services [ ] CI – Classified: As referred in to Commission Decision 2001/844/EC 07/03/2025 Draft Pedro Tardós (CTFC), Eva Samblàs (CIGC), Lluís Coll (CTFC), Lena Vilà-Vilardell (CTFC), Eduardo Balguerías (CIGC), Luis Fernando Arencibia (CIGC), Federico Grillo (CIGC), Míriam Piqué (CTFC) 08/04/2025 Revision Marius Hauglin (NIBIO), Martí Rosell (CFRS) 30/04/2025 Final Version Pedro Tardós (CTFC), Eva Samblàs (CIGC), Lluís Coll (CTFC), Lena Vilà-Vilardell (CTFC), Eduardo Balguerías (CIGC), Luis Fernando Arencibia (CIGC), Federico Grillo (CIGC), Míriam Piqué (CTFC) Copyright © All rights reserved. This document or any part thereof may not be made public or disclosed, copied or otherwise reproduced or used in any form or by any means, without prior permission in writing from the FIRE-RES Consortium. Neither the FIRE-RES Consortium nor any of its members, their officers, employees or agents shall be liable or responsible, in negligence or otherwise, for any loss, damage or expense whatever sustained by any person as a result of the use, in any manner or form, of any knowledge, information or data contained in this document, or due to any inaccuracy, omission or error therein contained. All Intellectual Property Rights, know-how and information provided by and/or arising from this document, such as designs, documentation, as well as preparatory material in that regard, is and shall remain the exclusive property of the FIRE-RES Consortium and any of its members or its licensors. Nothing contained in this document shall give, or shall be construed as giving, any right, title, ownership, interest, license or any other right in or to any IP, know-how and information. The information and views set out in this publication does not necessarily reflect the official opinion of the European Commission. Neither the European Union institutions and bodies nor any person acting on their behalf, may be held responsible for the use which may be made of the information contained therein. 2.3.1. Study site ..................................................................................................................................... 3 2.3.2. Treatments .................................................................................................................................. 4 2.3.3. Sampling methods ..................................................................................................................... 6 2.3.4. Custom fuel models ................................................................................................................... 6 2.3.5. Meteorological scenarios and fire behaviour simulation ..................................................... 7 2.4.1. Forest and understory structure before and after the treatments ...................................... 8 2.4.2. Fuel models ............................................................................................................................... 10 2.4.3. Fire behaviour ........................................................................................................................... 12 2.5.1. Forest structure changes ......................................................................................................... 14 2.5.2. Fire prevention expected effectivity ....................................................................................... 14 2.5.3. Practical implications for management ................................................................................ 15 3.3.1. Study site: Gran Canaria ......................................................................................................... 23 3.3.2. Strategic Management Points (SMP) ...................................................................................... 24 3.3.3. Treatments implemented ........................................................................................................ 25 3.4.1. Case study: Cortijo de Huertas 2023 wildfire ....................................................................... 33 4.3.1. Study site ................................................................................................................................... 40 4.3.2. Treatments design .................................................................................................................... 41 4.3.3. Plantation description ............................................................................................................. 44 Figure 1. Location of the study site. ............................................................................................ 4 Figure 2. Zoning of the different treatments and Control. ..................................................... 5 Figure 3. Aerial view of an example of the treatments. At the bottom-left, very high intensity thinning distributed by groups (CC ≈ 30%); at the top right, homogeneously distributed high intensity thinning (CC ≈ 50%); at the bottom-right, untreated stand (initial conditions). ..................................................................................................................................... 9 Figure 4. Relationship between heat per unit area and rate of spread across preand posttreatment scenarios. The lower graph zooms in on the lower-left section of the upper graph, focusing exclusively on surface fire scenarios. AC denotes scenarios where the fire type is active crown fire. The remaining scenarios represent surface fires. ...................... 13 Figure 5. Location of the study sites managed with prescribed burns (red polygons). .... 27 Figure 6. Prescribed burns carried out in the selected sites. Source: Cabildo de Gran Canaria – UOFF. ........................................................................................................................... 28 Figure 7. Prescribed grazing in a ravine (right) and in a main ridge axis (left). Source: Cabildo de Gran Canaria. ........................................................................................................... 29 Figure 8. Parameters conditioning the inclusion of a site in “Gran Canaria Pastorea” and preventive measures to be taken. Source: Cabildo de Gran Canaria.................................. 30 Figure 9. Prescribed grazing treatments in a ravine dominated by the invasive alien species Arundo donax. ............................................................................................................... 32 Figure 10. Prescribed grazing treatments in a grassland located within the perimeter of Cortijo de Huertas wildfire occurred in 2023. ......................................................................... 32 Figure 11. Perimeter of the Cortijo de Huertas wildfire and sites with prescribed burns in the previous years. ...................................................................................................................... 33 Figure 12. Severity of the Cortijo de Huertas wildfire and level of damage on prescribed burning sites. Source: Balguerías (2024). ................................................................................. 34 Figure 13. Wildfire severity transition in the prescribed burn sites located in Pico de la Gorra. Source: Balguerías (2024). ............................................................................................. 35 Figure 14. Limit of the prescribed burned sites in Altos del Pozo. The figure corresponds to the treated area with shrubs. Note the difference between the untreated (left) and treated (right) areas. Source: Balguerías (2024). .................................................................... 35 Figure 15. Limit of the prescribed burned sites in Pico de la Gorra. Note the difference between the untreated (right) and treated (left) areas. Source: Balguerías (2024). .......... 36 Figure 16. Aerial view of Cortijo de Huertas Wildfire. Areas treated with prescribed burns in yellow. Source: UOFF - Cabildo de Gran Canaria. .............................................................. 36 Figure 17. Location of the study site (green polygon) within the wildfire perimeter (red line). The inset shows the location of the Santa Coloma de Queralt wildfire in Catalonia, NE Spain. ....................................................................................................................................... 41 Figure 18. Provenance regions selected: two from local climate (Interior Catalonia, Coastal Catalonia), two from moderate emissions scenario (Bárdenas-Ribagorça, Interior Levante), two from severe emissions scenario (Monegros-Ebro valley, Meridional Betica). ........................................................................................................................................................ 42 Figure 19. Experimental design with random distribution of the blocks within the study site. ................................................................................................................................................ 43 Figure 20. Experimental design with random distribution of the six Aleppo pine provenances and the seed orchard seedlings. Each point represents a seedling (total of 2,500 seedlings were planted). .................................................................................................. 43 Figure 21. Example of the planted seedlings within a block. The tape defines the areas where different provenances are planted. .............................................................................. 44 Figure 22. Aleppo pine seedling planted. ................................................................................ 44 Figure 23. Survival rate of provenances in June 2024. ........................................................... 45 Figure 24. Survival rate of provenances in October 2024. .................................................... 46 Table 1. Fuel category and particle diameter relationship...................................................... 6 Table 2. Summary of stand structure variables for the pre-treatment scenario (Pre) and post-treatment scenarios (SOH, SOVH, WTH, WTVH). N, number of trees per hectare; CC, canopy cover; BA, basal area; Dg, quadratic mean diameter; Hm, mean height; CBH, crown base height; VOB, volume over bark; VOBe, extracted volume over bark. Some values are presented as "mean value ± standard deviation". ................................................ 8 Table 3. Summary of cover (%) and height (H) of the main understory components for the pre-treatment scenario (Pre) and post-treatment scenarios involving stem only harvesting (SOH and SOVH). ........................................................................................................ 9 Table 4. Summary of the surface fuel model across different scenarios. 1h, 10h, 100h, dead fuel load for the corresponding category; Lh, live herbaceous fuel load; Lw, live woody fuel load; FBD, fuel bed depth. Red and green indicate unfavourable or favourable variations, respectively, in percentage, relative to pre-treatment values. Some values are presented as "mean ± standard deviation". ............................................................................ 10 Table 5. Canopy fuel variables across different scenarios. CC, canopy cover; Hm, mean canopy height; CBH, canopy base height; ACFL, available canopy fuel load; CBD, canopy bulk density. Red and green indicate unfavourable or favourable variations, respectively, in percentage, relative to pre-treatment values. .................................................................... 11 Table 6. Fuel moisture values across different meteorological scenarios (P50, P75 and P99). Fuel abbreviations as in Table 4. WS, wind speed; RH, relative humidity; T, air temperature. Red, yellow, and green indicate unfavourable, neutral, or favourable variations, respectively, in percentage, relative to pre-treatment values. ......................... 11 Table 7. Simulated fire behaviour parameters for each treatment (PRE, SOH, SOVH, WTH, WTVH) and meteorological scenario (P50, P75, P99). FT, fire type; ROS, rate of spread; HPUA, heat per unit area; FLI, fireline intensity; FL, flame length; SH, scorch height. S, surface fire; CC, conditional crown fire. Red, yellow, and green indicate unfavourable, neutral, or favourable variations relative to pre-treatment values. .................................... 12 Table 8. Simulated fire behaviour parameters for the PRE scenario. Values correspond to active crown fire cases under conditional crown fire conditions in meteorological scenarios P75 and P99. Abbreviations: see Table 7. .............................................................. 13 Table 9. Details of the prescribed burn sites. ......................................................................... 27 Table 10. Severity of the Cortijo de Huertas wildfire over the prescribed burn area and the percentage relative to the total treated area affected. Source: Balguerías (2024). .... 33 Table 11. Regions of origin of Aleppo pine provenances planted ....................................... 42 1 The increased frequency of large fires and extreme wildfire events (EWE) demands a good understanding of the factors that drive the capacity of forests to respond and recover after fires. The impact of a large fire or an EWE on forest and landscape dynamics is driven by several factors associated to the vegetation present before the fire, the structure of the landscape, the fire event itself, soil characteristics, topography, and preand post-fire climatic conditions. Deliverable 1.12 part I “Innovative post-fire strategies and adaptation to the current context of increasing environmental uncertainties” identified the key factors, metrics, and associated thresholds used to assess areas at higher risk of erosion and runoff after fire. This was achieved through a systematic literature review and a questionnaire distributed to forest fire experts. The thresholds were determined based on the post-fire regeneration strategy of the ecosystem’s dominant species, that can be classified into four types: species without post-fire related traits, post-fire colonizers, seeders, and resprouters. Building on the identified factors and thresholds, three innovative management interventions were designed and implemented in fire-prone and new fire areas predicted to be vulnerable to EWE in two different living labs: Catalonia and Canarias. Management interventions that enhance forests resistance and resilience to EWE can be implemented either before or after the fire. Pre-fire interventions focus on modifying vegetation composition and structure to reduce the risk of high-intensity fires and EWE and mitigate their negative impacts. Post-fire interventions aim to promote vegetation recovery in areas at risk of soil erosion. This deliverable presents management actions implemented within the Innovation Action 1.5 “Demonstration of innovative post-fire restoration and adaptation strategies”, addressing both preventive and post-fire restoration strategies with the aim to provide a scientifically-based framework to guide managers in the definition of adaptation strategies to EWEs. Each action addresses one or more of the identified factors driving EWE impacts and influencing landscape resistance and resilience to EWE: • Silvicultural treatments for EWE prevention and pasture improvement in a new fire area (Section 2) • Prescribed burning treatments and grazing introduction for EWE prevention in a fire-prone area (Section 3) • Restoration of vegetation cover after an EWE in a fire-prone area (Section 4) 8 2.4.1. Forest and understory structure before and after the treatments Both treatment intensities (high and very high) significantly reduce stand density and stock variables (i.e., N, CC, BA, VOB). The removed basal area is approximately 44% of the initial value in the high-intensity thinning and 67% in the very-high-intensity thinning. Mean height (Hm) and crown base height (CBH) tend to increase under both treatment intensities. The remaining standing volume over bark (VOB) is greater after the highintensity treatment than after the very-high-intensity treatment, whereas the extracted volume follows the opposite trend. Table 2 compares stand structure variables before and after the treatments. Figure 3 provides an aerial view of the applied treatments. Since the thinning treatments are identical for both harvesting techniques, post-treatment variable values are also the same. Table 2. Summary of stand structure variables for the pre-treatment scenario (Pre) and posttreatment scenarios (SOH, SOVH, WTH, WTVH). N, number of trees per hectare; CC, canopy cover; BA, basal area; Dg, quadratic mean diameter; Hm, mean height; CBH, crown base height; VOB, volume over bark; VOBe, extracted volume over bark. Some values are presented as "mean value ± standard deviation". Trees·ha-1 756 ±163.1 366 213 % 71 ±2.5 50 30 m2·ha-1 54.3 ±7.7 30.4 18.1 cm 30.5 ±2.3 32.5 32.9 m 13.1 ±1.8 14.1 ±1.1 13.5 ±2.2 m 6.9 ±1.3 7.9 ±1.5 7.4 ±1.5 m3·ha-1 333.8 ±42.1 189.2 115.0 m3·ha-1 - 149.0 233.6 Additionally, in the case of stem-only harvesting treatments (SOH and SOVH), both significantly reduce the cover of herbaceous layer and woody understory species while increasing the cover of woody debris (Table 3). The mean height of shrub species is negatively affected by both treatments, whereas the height of logging slash increases. The high variability in woody understory species cover in the pre-treatment scenario reflects the presence of Pinus sylvestris regeneration groups in canopy gaps. Whole-tree harvesting treatments would have a similar effect but with a much smaller increase in woody debris cover. 9 Table 3. Summary of cover (%) and height (H) of the main understory components for the pre-treatment scenario (Pre) and post-treatment scenarios involving stem only harvesting (SOH and SOVH). 76.8 ±11.4 0.049 ±0.013 42.3 0.05 28.3 0.05 11.9 ±13.3 0.93 ±0.81 2.5 0.32 1.8 0.37 8.3 ±2.9 0.16 ±0.14 36.7 0.23 50.9 0.29 Figure 3. Aerial view of an example of the treatments. At the bottom-left, very high intensity thinning distributed by groups (CC ≈ 30%); at the top right, homogeneously distributed high intensity thinning (CC ≈ 50%); at the bottom-right, untreated stand (initial conditions). 10 2.4.2. Fuel models Table 4 and Table 5 show the surface and canopy fuel variables of the custom fuel models. All treatments increase dead fuel loads (1h, 10h, 100h), with the 10h fuel being particularly affected. This increase is significantly greater in stem-only harvesting compared to wholetree harvesting. All treatments reduce live fuel loads and fuel bed depth (except for SOVH, which does not reduce FBD). Fuel bed depth is very low across all models, both before and after the treatments. This is due to the consistently high herbaceous layer cover across all treatments, which has a relatively low height, and the low percentage of shrub cover. In post-treatment scenarios, the reduced height of logging residues further contributes to this outcome. Both treatments (highand very-high-intensity thinning) reduced canopy cover while slightly increasing mean canopy height and mean crown base height. They also significantly reduced available canopy fuel load and, consequently, canopy bulk density. Table 4. Summary of the surface fuel model across different scenarios. 1h, 10h, 100h, dead fuel load for the corresponding category; Lh, live herbaceous fuel load; Lw, live woody fuel load; FBD, fuel bed depth. Red and green indicate unfavourable or favourable variations, respectively, in percentage, relative to pre-treatment values. Some values are presented as "mean ± standard deviation". 2.19 ±0.51 5.61 157% 8.31 280% 2.53 16% 2.80 28% 0.63 ±0.50 4.21 571% 11.04 1661% 0.99 57% 1.67 166% 2.11 ±1.83 10.30 388% 21.35 911% 2.93 39% 4.04 91% 0.53 ±0.15 0.30 -44% 0.20 -63% 0.30 -44% 0.20 -63% 0.69 ±0.69 0.18 -73% 0.10 -85% 0.18 -73% 0.10 -85% 0.15 ±0.16 0.11 -22% 0.17 15% 0.06 -62% 0.06 -60% 11 Table 5. Canopy fuel variables across different scenarios. CC, canopy cover; Hm, mean canopy height; CBH, canopy base height; ACFL, available canopy fuel load; CBD, canopy bulk density. Red and green indicate unfavourable or favourable variations, respectively, in percentage, relative to pre-treatment values. 70 50 -29% 30 -57% 13.1 14.1 7% 13.5 2% 6.9 7.9 15% 7.4 8% ·-2 1.173 0.638 -46% 0.378 -68% ·-3 0.187 0.104 -45% 0.063 -67% Table 6 presents fuel moisture values under different meteorological scenarios. Dead fuel moisture (1h, 10h, and 100h) is influenced by stand variables, meteorological conditions, and exposure (canopy cover). Live fuel moisture is influenced only by exposure and species composition. Table 6. Fuel moisture values across different meteorological scenarios (P50, P75 and P99). Fuel abbreviations as in Table 4. WS, wind speed; RH, relative humidity; T, air temperature. Red, yellow, and green indicate unfavourable, neutral, or favourable variations, respectively, in percentage, relative to pre-treatment values. · 12 12 0% 9 -25% 13 13 0% 10 -23% 14 14 0% 11 -21% · 10 10 0% 7 -30% 11 11 0% 8 -27% 12 12 0% 9 -25% · 7 7 0% 4 -43% 8 8 0% 5 -38% 9 9 0% 6 -33% 60 60 0% 30 -50% 114.9 121.7 6% 97.4 -15% 12 110.5 110.5 0% 86.7 -22% 2.4.3. Fire behaviour The pre-treatment model generates surface fire under the mildest meteorological scenario (P50). In the more severe meteorological scenarios (P75 and P99), it produces a conditional crown fire type, meaning that under the simulated within-stand conditions, a surface fire cannot transition into the canopy. However, an active crown fire could still occur if it spreads into the overstory from an adjacent area and burns into the stand. The surface fire behaviour variables are presented in Table 7, while the variables corresponding to the potential active crown fire in the pre-treatment scenario are shown in Table 8. The possible active crown fire predicted in the pre-treatment scenario under P75 and P99 exhibits the highest values across all fire behaviour variables (i.e. rate of spread, heat per unit area, fireline intensity and flame length), indicating the maximum potential fire severity among the simulations. Table 7. Simulated fire behaviour parameters for each treatment (PRE, SOH, SOVH, WTH, WTVH) and meteorological scenario (P50, P75, P99). FT, fire type; ROS, rate of spread; HPUA, heat per unit area; FLI, fireline intensity; FL, flame length; SH, scorch height. S, surface fire; CC, conditional crown fire. Red, yellow, and green indicate unfavourable, neutral, or favourable variations relative to pre-treatment values. S S S=S S S=S S S=S S S=S · 0.8 0.4 -50% 0.8 0% 0.3 -63% 0.5 -38% · 4952 4231 -15% 4510 -9% 3363 -32% 2964 -40% · 62 26 -58% 63 2% 19 -69% 26 -58% 0.5 0.3 -40% 0.5 0% 0.3 -40% 0.3 -40% 1.4 0.5 -64% 0.6 -57% 0.4 -71% 0.3 -79% CC S CC>S S CC>S S CC>S S CC>S · 1.2 0.6 -50% 1.5 25% 0.5 -58% 0.9 -25% · 5161 4368 -15% 4656 -10% 3487 -32% 3056 -41% · 104 43 -59% 113 9% 31 -70% 47 -55% 0.7 0.4 -43% 0.7 0% 0.4 -43% 0.5 -29% 1.8 0.6 -67% 0.8 -56% 0.4 -78% 0.4 -78% CC S CC>S S CC>S S CC>S S CC>S · 1.8 0.9 -50% 2.6 44% 0.8 -56% 1.6 -11% · 5476 4555 -17% 5313 -3% 3648 -33% 3454 -37% · 167 68 -59% 228 37% 50 -70% 93 -44% 0.8 0.5 -38% 0.9 13% 0.5 -38% 0.6 -25% 2.8 0.9 -68% 1.4 -50% 0.6 -79% 0.6 -79% 13 All post-treatment fuel models generate surface fire across all three meteorological scenarios (P50, P75, and P99). In all cases, except for SOVH, potential fire severity is lower than in the surface fire of the pre-treatment scenario (Figure 4). The most effective treatments under all meteorological conditions are those involving whole-tree harvesting. Table 8. Simulated fire behaviour parameters for the PRE scenario. Values correspond to active crown fire cases under conditional crown fire conditions in meteorological scenarios P75 and P99. Abbreviations: see Table 7. AC AC ·-1 16.4 26.4 ·-2 26961 27276 ·-1 7350 11979 10.1 13.9 Figure 4. Relationship between heat per unit area and rate of spread across preand posttreatment scenarios. The lower graph zooms in on the lower-left section of the upper graph, focusing exclusively on surface fire scenarios. AC denotes scenarios where the fire type is active crown fire. The remaining scenarios represent surface fires. 14 2.5.1. Forest structure changes All the proposed treatments are expected to affect significantly overstory and understory variables due to the inherent intensity of the fire prevention silvicultural treatments. The thinning operations carried out are considered very intensive, with the extraction of 44% (H) and 67% (VH) of the initial basal area. This is due to the abandonment of wood production objective in favour of pasture improvement and fire prevention. We created detailed custom fuel models to initialize the fire behaviour models as accurately as possible, though it is a time-consuming method (Keane, 2015). The wholetree harvesting scenarios were created through the parameterization of the effects of the treatments on understory structure. Mitsopoulos & Dimitrakopoulos (2017) used a similar approach. The significant alteration of the canopy and the increase in dead fuel loads are consistent with other observations (Piqué et al., 2022; Piqué & Domènech, 2018; Vaillant, FitesKaufman, Reiner, et al., 2009). Most of the pre-treatment fuel model estimated parameters (i.e., litter load, fine woody debris load, live total woody load and fuel bed depth) show great variability. The high-intensity thinning from below treatments described in Piqué & Domènech (2018) are comparable to our high-intensity thinning treatments (SOH and WTH). In our case, the canopy base height varies by +14.5% (1 m), and the canopy bulk density decreases by -44.7% (0.084 kg·m⁻³), which is similar to the +14.5% (1 m) and -60% (0.14 kg·m⁻³) changes observed in the previous study. The microclimatic changes affecting fuel moisture after thinning are not evident (Banerjee, 2020). In this study, we attempted to reflect the canopy's protective effect by assigning lower fuel moisture values in the very-high-intensity treatments (SOVH and WTVH), as reduced canopy cover allows more solar radiation to reach and dry the surface fuels. It should be noted that estimating surface fuel loads, particularly harvesting residues, is a laborious and complex task that may require destructive sampling. For this study, we used the line-intersect method by Brown (1974) to estimate fine woody fuel load (1h, 10h, 100h), as it is a simple and quick method to apply. However, this method has a limitation in characterizing recent silvicultural residues, as it does not include needles present on branches. In this study, we considered estimating needle loads through the allometric equations of Ruiz-Peinado et al. (2011) as a possible solution. In any case, this estimation should be validated with data obtained from destructive sampling. 2.5.2. Fire prevention expected effectivity Before treatment, fire behaviour simulations result in a low-severity surface fire in all meteorological scenarios. In the P75 and P99 meteorological scenarios, there is also the possibility of a crown fire, which corresponds to the highest potential severity among those obtained. Both thinning intensities substantially reduce the likelihood of an active crown fire. This is due to a significant reduction in canopy bulk density and, consequently, an increase in the critical active crown rate of spread, which is consistent with Crecente- 15 Campo et al. (2009), Piqué et al. (2022), Piqué & Domènech (2018), Stephens & Moghaddas (2005), among others. The probability of fire transitioning from the surface to the canopy is very low in all simulated cases, especially in the high-intensity thinning treatments (SOH and WTH), due to the sheltering effect of the canopy, which influences wind speed and fuel moisture. Fire behaviour has been compared across different meteorological scenarios defined by percentiles, as is common in similar experiments (Agee & Lolley, 2006; Piqué & Domènech, 2018; Vaillant, Fites-Kaufman, & Stephens, 2009). In the simulated cases, fire behaviour remains largely consistent across meteorological scenarios, except for the potential occurrence of an active crown fire in the pre-treatment scenario (P75 and P99). This homogeneity is largely due to the high compactness of the fuel bed, which is the most determining factor in these simulations. Consequently, fire behaviour variables do not increase proportionally with fuel load, in agreement with Piqué & Domènech (2018). According to Rothermel’s (1983) fire suppression capacity classification, all simulated surface fire scenarios are low-intensity and can be suppressed using hand tools. Conversely, the active crown fire type in the pre-treatment P75 and P99 scenarios exhibit extreme intensity and would exceed suppression capacity (Costa et al., 2011). 2.5.3. Practical implications for management The designed silvicultural treatments are expected to modify resource availability and microclimatic conditions in the understory, thereby improving pasture productivity (Pachas et al., 2023), as well as its composition and quality (Ducherer, 2005). However, the accumulation of logging residues hinders pastoral use (Kaufmann, 2011). Therefore, the best management option for pastoral use should include whole-tree harvesting. Wind and snow episodes can cause severe damage to Pinus sylvestris forests (del Río et al., 2017). The pre-treatment stand height-to-diameter ratio (Hm·Dm⁻¹) is approximately 45, which is below the critical range (70-90) reported by del Río et al (1997). In any case, the possibility of windthrow damage in the stand should be assumed due to the highintensity treatments required to achieve the objectives in the short term. In the medium to long term, a significant emergence of shrub and tree regeneration is expected due to the reduction of canopy cover (CC) below 70% (Piqué, Beltrán, et al., 2011). Targeted grazing is considered an economical and synergistic strategy for maintaining fuel break areas (Varela et al., 2018). This should be combined with mechanical treatments or periodic prescribed burns to ensure low fuel loads (Oikonomou et al., 2023), thus achieving management objectives over an extended period. Relevant highlights regarding the potential of treatments performance include: • All tested treatments help reduce potential fire severity and improve suppression opportunities under the most adverse weather conditions. In the most favourable weather scenario, the treatments maintain a similar level of fire severity and suppression opportunities as the untreated stand. 16 • Whole-tree harvesting (WTH and WTVH) could be the best option to achieve our dual management objectives of fire prevention and pasture improvement. Slash accumulation in stem-only harvesting hinders performance in both directions. • The high compactness of resulting slash is of great importance, significantly mitigating the adverse effect of weather conditions. • Grazing combined with future mechanical interventions or prescribed burns will serve to maintain this Strategic Management Point. 17 Agee, J. K., & Lolley, M. R. (2006). Thinning and prescribed fire effects on fuels and potential fire behavior in an eastern Cascades forest, Washington, USA. Fire Ecology, 142–158. Agee, J. K., & Skinner, C. N. (2005). Basic principles of forest fuel reduction treatments. Forest Ecology and Management, 211(1–2), 83–96. https://doi.org/10.1016/j.foreco.2005.01.034 Arilla, E.; Bachfischer, M.; Castellarnau, X.; Cespedes, J.; Castellnou, M.; Castellví, J.; Dalmau, E.; Estivill, L.; Ferragut, A.; Larrañaga, A.; Miralles, M.; Nebot, E.; Pagès, J.; Pallàs, P.; Rosell, M.; Ruiz, B. (2023). Piloting the adaptation of methodology of forest fire potential polygons. D1.3. Deliverable D1.1 FIRE-RES project. 20 pages. DOI: 10.5281/zenodo.7991283 Augusto, L., Beaumont, F., Nguyen, C., Fraysse, J. Y., Trichet, P., Meredieu, C., Vidal, D., & Sappin-Didier, V. (2022). Response of soil and vegetation in a warm-temperate Pine forest to intensive biomass harvests, phosphorus fertilisation, and wood ash application. Science of the Total Environment, 850(May). https://doi.org/10.1016/j.scitotenv.2022.157907 Ballart, H., Pagès, J., & Canaleta, G. (2020). Vulnerabilitat dels perímetres de protecció prioritària (PPP) de les muntanyes d’Ordal (B6) i el Garraf (B7) al foc forestal . Identificació de les oportunitats d’extinció. In VIII Trobada d’Estudiosos del Garraf i d’Olèrdola (pp. 86–99). Diputació de Barcelona. Àrea d’Espais Naturals. Servei de Parcs Naturals. Banerjee, T. (2020). Impacts of forest thinning on wildland fire behavior. Forests, 11(9). https://doi.org/10.3390/F11090918 Bergmeier, E., Capelo, J., Di Pietro, R., Guarino, R., Kavgacı, A., Loidi, J., Tsiripidis, I., & Xystrakis, F. (2021). ‘Back to the Future’—Oak wood-pasture for wildfire prevention in the Mediterranean. Plant Sociology, 58(2), 41–48. https://doi.org/10.3897/pls2021582/04 Brown, J. K. (1974). Handbook for inventorying downed woody material. (General Technical Report INT-16). USDA Forest Service. Bütler, R., Lachat, T., Krumm, F., Kraus, D., & Larrieu, L. (2020). Field Guide to Tree-related Microhabitats. Descriptions and size limits for their inventory. Swiss Federal Institute for Forest, Snow and Landscape Research WSL. www.wsl.ch/fg-trems Castellnou, M., Nebot, E., Estivill, L., Miralles, M., Rosell, M., Valor, T., Casals, P., Duane, A., Piqué, M., Górriz-Mifsud, E., Coll, L., Serra, M., Plana, E., Colaço, C., Sequeira, C., Skulska, I., Moran, P. (2022). FIRE-RES Transfer of Lessons Learned on Extreme wildfire Events to key stakeholders. Deliverable D1.1 FIRE-RES project. 119 pages. DOI: 10.5281/zenodo.10260790 Castellnou, M.; Nebot, E.; Saavedra, J.; Miralles, M.; Estivill, L.; Rosell, M.; Tapia, G.; Alegría, D.; Medina, V.; Arilla, E.; Bachfischer, M.; Castellarnau, X.; Cespedes, J.; 24 3.3.2. Strategic Management Points (SMP) In Gran Canaria, there are 8 High Wildfire Risk Areas (ZARI, for its acronym in Spanish), legally defined as those areas where the frequency or severity of wildfires, as well as the significance of the threatened values, make special fire protection measures necessary (Ley 43/2003 de Montes). Given that these areas account for more than 40% of the island’s surface (64,488 ha), it is essential to define strategic management points (SMP or ZEG for its acronym in Spanish) on which to focus resources and efforts. These priority intervention areas are locations where either the protection of environmental or social values is considered a priority or where, with the appropriate preventive treatment, wildfire spread can be slowed down, creating windows of opportunity for emergency services. Their identification requires a thorough study of the territory and an analysis of historical fires behaviour. The Strategic Management Points identified in Gran Canaria include: • Main ridge axes: these act as the main propagation axes of wind-driven wildfires. Particularly critical are the ridge nodes, which can bifurcate the flame front, and mountain passes which mark the endpoint of topographic fires. The generation of low-load areas in these zones aims to generate firefighting opportunities, limit the projection of sparks, compartmentalise the terrain and reduce the fire severity. • Main gully axes: these act as topographic fire propagation axes. In these areas, the aim is to reduce the fuel load or generate green barriers (areas made of laurel or riparian vegetation, permanently humid and hygrophilous). Ravine nodes become particularly critical since they act as potential bifurcation points for the flame front. • Auxiliary roadside strips: roads within ZARI areas constitute the main access routes for firefighting teams as well as the main evacuation routes for citizens. They are also a key element to create extinguishing opportunities and ensuring safety for both firefighting resources and citizens. • Strategic agricultural areas, which are managed agricultural zones that generate discontinuities in the territory and therefore modify the fire behaviour, and wildland-urban-interface (WUI) defensive belts, which act as protection buffers between urban settlements and the forest. At present, the Cabildo of Gran Canaria utilises several silvicultural tools and techniques to manage and reduce the fuel load in these strategic areas, as part of a comprehensive land management strategy. As previously mentioned, the present work focuses on the implementation of prescribed burns and preventive grazing to reduce wildfire risk in specific fire-prone locations in the island of Gran Canaria. 25 3.3.3. Treatments implemented Prescribed burning Prescribed burning has become nowadays a key tool for fuel management on the island of Gran Canaria. The Cabildo of Gran Canaria initiated the use of this technique in 2002, treating, in just five years, an area of 166.5 ha. This achievement positioned the island as one of the leading European territories in the use of prescribed technical fire (Fababú et al., 2007). Over the years, the application of prescribed burns has been refined and has evolved alongside the growing specialization of the operational teams. From a fire prevention perspective, prescribed burns aim to create low fuel load areas in strategic locations across the island, serving as potential opportunity points for fire control and contributing to the reduction of the overall fire risk. In addition to fire prevention, this technique has several secondary objectives, including the regeneration of pasture areas and the removal of agricultural and forestry debris, as well as the training of firefighting personnel. The Rothermel fuel models on which prescribed burns are typically applied range from grass models to promote regeneration, to tree-covered without understory models (Models 8 and 9) and shrub under canopy models (Models 4 and 6). The latter represent the most complex burns, since they require initial structural modifications through silvicultural treatments, generating wood debris (Models 11 and 12) that will be subsequently burned. The prescribed burns are predominantly conducted on Canarian pine forests (Pinus canariensis), as they have proved to be the most effective and economical tool for reducing the accumulation of fine dead fuel in these habitats. Regarding the ecological impact, prescribed burns can affect the composition and diversity of plant and animal communities, as well as water resources and soil properties depending on the habitat and the manner in which they are applied (Fernandes et al., 2013). However, several studies suggest a neutral or even positive effect of low-intensity fires on soil health and biodiversity, as well as a rapid recovery after treatment (Alcañiz et al., 2018; Fernandes et al., 2013). As for the case of Gran Canaria, several studies conducted in the island found no significant effects and observed a rapid recovery on floristic composition, soil properties and ground-dwelling invertebrate communities after prescribed burns (Arévalo et al., 2023; Arévalo et al., 2014; García-Domínguez et al., 2010). Nonetheless, further research may be necessary to assess the impacts of different levels of fire intensity, seasonality, and frequency, as well as the effects of burns on rare or sensitive species, given the high level of endemism in Canarian ecosystems. Prescribed burning: Study sites In light of this and within the framework of this Innovative Action, a total of 10.13 ha of low intensity prescribed burns were conducted between January and May 2024 (Figure 5 and Figure 6). Sites were located at elevations ranging from 1,250 to 1,900 m.a.s.l. on moderate to steep slopes. The dominant tree species in these areas was Pinus canariensis, combined with non-native conifers such as Pinus radiata and Pinus pinea in the Llanos de Ana López site. The main fuel models present at the treatment sites included shrub under canopy and tree-covered without understory models. More details about the sites are provided in Table 9. 26 The treated sites were located within three of the Strategic Management Points outlined in the Fire Prevention, Surveillance, and Extinction Annual Plan developed by the Cabildo of Gran Canaria for 2024. Specifically: • Llanos de Ana López Site: Located at the Strategic Management Point Cruz de Los Llanos – Cruz de Tejeda, which corresponds to a main ridge axis. The main objective was to maintain a low fuel load safety strip adjacent to a recreational area to maximize safety during potential fire events. Given the large number of visitors that normally frequent the area, the absence of appropriate preventive measures could lead to serious challenges for Civil Protection in the event of an emergency. Prior to the burns, silvicultural treatments were applied to remove the understory and reduce vertical fuel continuity, thereby ensuring that prescribed burns would be carried out at low intensity. • Pico de la Gorra Site: Located at the Strategic Management Point Pico de Las Nieves-Pico de La Gorra, area of particular interest since it is the main ridge axis that separates the northern and southern slopes of the island. The objectives here are multiple. Firstly, to create low fuel load areas that prevent the fire from jumping from one slope to the other and thereby significantly reducing its spread potential. Secondly, to protect the military air traffic control in the area (EVA 21). These prescribed burns were conducted in two days: one in March and the second in May. Prior to the May burns silvicultural treatments were carried out to reduce fuel vertical continuity. It is worth noting that the burns conducted in May were part of a training program aimed at enhancing fire suppression capabilities. Therefore, they had an additional objective: to improve the skills of the firefighting teams. • Artenara Base 2 Site: Here, the objective was to create a protection zone for the defence of a critical infrastructure; the Artenara heliport, where the helicopterborne brigades (PRESA teams) of the Operational Unit for Forest Fires of the Cabildo of Gran Canaria are based. Since this infrastructure is one of the main forest bases in the island, its operability in the event of a fire is essential and any damage on it could severely affect the response capacity of the emergency services. To illustrate the effectiveness of these treatments in wildfire prevention, a later section presents the case of the Cortijo de Huertas wildfire in 2023, which affected an area previously treated with prescribed burns. It is important to note that the plots executed in Pico de la Gorra in 2024 are located in the same Strategic Management Point where the fire occurred but were not affected by it. 27 Table 9. Details of the prescribed burn sites. 11/01/2024 Low Yes 2,97 1611 20% N-NE 9, 10 2223/04/2024 Low No 3,21 1250 30% E-NE 9 13/03/2024 21/05/2024 Low Yes (21/05/2024) 3,95 1900 20% SW-SSE 7, 9, 11 Figure 5. Location of the study sites managed with prescribed burns (red polygons). 28 Figure 6. Prescribed burns carried out in the selected sites. Source: Cabildo de Gran Canaria – UOFF. Prescribed grazing In response to the increasing threat posed by EWEs, extensive grazing and particularly prescribed grazing, has emerged as an effective and powerful strategy for modifying fuel structure and reducing the risk of high-intensity wildfires. In Gran Canaria, extensive grazing was first employed for wildfire prevention in 2005, when the Cabildo de Gran Canaria started to grant permission for grazing in public forest areas. By 2013, these authorizations were expanded to include ravine beds. However, it was not until 2018 that the significance of this practice in landscape management and wildfire prevention was officially acknowledged, following the signing of an agreement between the Cabildo de Gran Canaria and local shepherds. This agreement led to the establishment of the “Gran Canaria Pastorea” project, which formalised the allocation of grazing areas within public forests and ravines. It also introduced economic compensation for shepherds through Payments for Ecosystem Services aimed at fire prevention efforts. The first payment was made in 2022, and participation has steadily increased ever since. In 2024, 36 livestock families benefited from the project, with 45 areas of public forests and watercourses allocated, which covered a total of 1,681 hectares (Figure 7). 29 Figure 7. Prescribed grazing in a ravine (right) and in a main ridge axis (left). Source: Cabildo de Gran Canaria. Payments typically range from 40 to 180 euros per hectare and year, varying based on the vegetation type, the distance to be covered by the shepherds and the strategic significance of the area for fire prevention. Higher payments are allocated to the most strategically important or difficult-to-access locations. Each year, two payment cycles are established: the first in June, prior to the fire season, and the second in November, coinciding with the transhumance to the lower regions of the island. The “Gran Canaria Pastorea” project (https://grancanariamosaico.com/pastorea/) aims to achieve the following objectives: • To establish low-load areas within the territory, paying special attention to Strategic Management Points (SMPs). • To collaborate in the management and control of invasive exotic species. • To ensure that sustainable grazing practices align with biodiversity conservation efforts. • To promote the training of new shepherds, emphasizing fire prevention strategies. • To support generational renewal and contribute to the development of rural communities. • To investigate the response of island ecosystems to different grazing regimes. However, it is worth noting that this project is not based in the application of traditional extensive grazing but rather focuses on the so-called prescribed grazing. Unlike traditional practices, prescribed grazing involves the controlled introduction of livestock in carefully selected areas, based on specific technical guidelines established in advance (Figure 8). 30 Figure 8. Parameters conditioning the inclusion of a site in “Gran Canaria Pastorea” and preventive measures to be taken. Source: Cabildo de Gran Canaria. In general, allocated sites usually correspond to Strategic Management Points, such as main ridge axes or ravines, although traditionally grazed areas are also considered. Grazing in Protected Natural Areas is included as well, but only in locations where the corresponding planning instruments allow livestock activity. The proposed stocking rates are set at less than 0.5 LSU per hectare and year, except for reed bed areas in ravines, where higher stocking densities are permitted to manage the growth of the invasive species Arundo donax. These low stocking rates are adequate to meet fire prevention objectives (such as reducing accumulated biomass and modifying fuel structure) without causing significant ecological disruption to the landscape. Recent studies conducted in the Canary Islands found no significant negative effects of lowintensity grazing on biodiversity or the spread of invasive alien species (Arévalo et al., 2011; Fernández-Lugo et al., 2011; Bermejo et al., 2012). In any case, ongoing monitoring since 2021 continues to assess the effects of grazing on flora, invertebrate communities, and avifauna, with preliminary results suggesting that grazing influences population structures and creates opportunities for the growth of other native species while maintaining overall biodiversity (Fababú et al., 2024). In addition to the primary objectives of the “Gran Canaria Pastorea” project, several complementary actions are being implemented as well. These include the installation of perimeter or individual fences in reforestation areas to avoid possible damage caused by livestock and, thus, ensuring the compatibility of both activities, or silvicultural treatments, such as clearing and prescribed burns as preparatory steps before grazing. 31 In reference to the latter, the integration of different treatments turns out to be, in many cases, the most effective strategy. Mechanical clearing enables the removal of highly lignified vegetation that cannot be assimilated by livestock. Subsequently, the resulting plant debris is eliminated through prescribed burning. The regrowth of vegetation is then managed through grazing to maintain the functionality of the low-load areas generated. This combination of treatments is particularly beneficial in the giant reed beds located in many of the island's ravines. The tender sprouts are highly palatable, and the continuous presence of grazing herds transforms the giant reed beds into productive pastures, simultaneously reducing the risk of wildfires. Currently, approximately ten ravines across the island are being managed using this integrated approach. Finally, the herds are continuously monitored using GPS devices, allowing for precise tracking of both grazed areas and grazing intensity. Additionally, since 2022, mobile shelters have been set up in the most remote locations to accommodate shepherds during the transhumance season. It is important to highlight that the “Gran Canaria Pastorea” project is part of the “Gran Canaria Mosaico” strategy, promoted by the Cabildo of Gran Canaria. This initiative aims to prevent large wildfires by fostering the regeneration of diverse and inhabited mosaic landscapes in which the balance between the natural environment and agricultural activities allows the creation of fire-resilient ecosystems. The strategy emphasises the active involvement of society, particularly rural communities, in land management and in addressing the evolving challenges posed by EWEs. Prescribed grazing: Study sites The preventive grazing treatments outlined in this Innovative Action were implemented across two areas included in the Gran Canaria Pastorea project. These treatments were carried out by local shepherds who participated in the tender process in 2024. The selected sites were located within two Strategic Management Points, specifically: • Barranco de Las Goteras (BRI_004): This ravine is located in the Tafira Protected Landscape, in a location categorised as “Zone of Traditional Use” where, according to the Management Plan of the Natural Protected Area, grazing is allowed. Although the site also falls within the Natura 2000 Network, the action focuses on the bed of the ravine, where almost monospecific populations of giant reed (Arundo donax), an invasive and highly flammable exotic species, currently predominate. Thus, the objective of this treatment was twofold: to generate a low fuel-load area in a key location for fire risk management (due to the islands orography, ravines are often the axis of propagation of topographic fires) while controlling the spread of an invasive alien species. Prescribed goat grazing was implemented in a total of 7.47 ha of the ravine (Figure 11) and, prior to this, mechanical clearing was carried out to facilitate the accessibility to regrowth. • La Abejerilla (MAT_002): Located at the Strategic Management Point Pico de Las Nieves-Pico de La Gorra, area of particular interest since it is the main ridge axis that separates the northern and southern slopes of the island. Prescribed sheep grazing was implemented in 10.56 ha of an area that had been affected by the Cortijo de Huertas wildfire in 2023 (Figure 12). Despite the short time elapsed since 32 the fire, prescribed grazing was introduced to maintain the low fuel-load in a grassland already recovered from the fire. Figure 9. Prescribed grazing treatments in a ravine dominated by the invasive alien species Arundo donax. Figure 10. Prescribed grazing treatments in a grassland located within the perimeter of Cortijo de Huertas wildfire occurred in 2023. 33 3.4.1. Case study: Cortijo de Huertas 2023 wildfire In July 2023, a wildfire that started in the municipality of Tejeda (Gran Canaria) burned 431 ha of forested land dominated by Pinus canariensis. Although the wildfire was controlled after two days, it exceeded extinction capacity in several moments, especially during the first day, and produced a considerable number of spot fires that multiplied its spread rate. In the years preceding this event, prescribed burns had been implemented in areas that would later be affected by the fire. To assess the effectiveness of these prescribed burns, those conducted within the fire perimeter and within a period of 5 to 7 years prior to the event were selected. This can be considered as a reasonable time frame to expect treatments to be functional and effective (estimate based on the expertise of Gran Canaria firefighting teams). Thus, a total of 19 plots managed between 2020 and 2022 were selected (Figure 11), 14.53 ha of which were affected by the wildfire. The severity of fire damage within the prescribed burn areas was compared to that in adjacent untreated areas (Figure 12). Table 10 summarises the proportion of the managed area affected by the wildfire at different severity levels. Table 10. Severity of the Cortijo de Huertas wildfire over the prescribed burn area and the percentage relative to the total treated area affected. Source: Balguerías (2024). Figure 11. Perimeter of the Cortijo de Huertas wildfire and sites with prescribed burns in the previous years. 0.71 5% 2.64 18% 11.18 77% 40 remains largely unknown. The accelerating impacts of climate change make it increasingly urgent to advance knowledge on the best strategies for implementing this forest management measure to support adaptation (Martín-Alcón et al., 2022). Any post-fire restoration management strategy, including assisted migration, should consider both the expected capacity of the system to recover by itself as well as its resistance and resilience capacity to further hazards. In a context of harsher climate and increased probability of EWEs, the use of assisted migration as a post-fire restoration strategy represents an opportunity to enhance landscape resilience to more severe fires and associated hazards. The main objective of this innovation action was to restore vegetation cover after an EWE to minimize associated post-fire hazards such as flooding, soil erosion, landslides and alien plant invasions. Specifically, the objectives of this action were to: • Establish demonstrative stands of post-fire regeneration strategies in Aleppo pine forests (Habitat type 9540 of the Habitats Directive) • Recover forest cover after an EWE by implementing assisted migration actions with different Aleppo pine provenances with the aim to explore their adaptation to the anticipated local climate aridification. • Advance in the understanding of assisted migration techniques in post-fire restoration actions of Aleppo pine stands. 4.3.1. Study site Aleppo pine seedlings were planted in 4 ha located on the northeastern side of the Santa Coloma de Queralt wildfire perimeter (Figure 17). The site is located on a northwesternfacing gentle slope at 500 meters above sea level, with a maximum inclination of 25%, and exhibits a Mediterranean climate characterized by mild winters and hot, dry summers. Mean annual temperature is 13.2 ºC and mean total annual precipitation is 572 mm. The soil developed from rocks of different lithologies, all rich in carbonates. The soils are shallow, drained, and rich in coarse particles. Before the fire, the area was a fragmented landscape consisting primarily of forests dominated by Aleppo pine (Pinus halepensis), along with shrublands and agricultural areas. Forest understory was abundant and rich in species, with Salvia rosmarinus, Thymus vulgaris, Viburnum tinnus, Genista scorpius, Dorycnium pentaphyllum, Quercus coccifera, Quercus ilex, and Quercus faginea. 41 Figure 17. Location of the study site (green polygon) within the wildfire perimeter (red line). The inset shows the location of the Santa Coloma de Queralt wildfire in Catalonia, NE Spain. 4.3.2. Treatments design A total of 2,688 Aleppo pine seedlings from six different regions of origin were planted in the study site: two local provenances (one from an area next to the study site and one from the closest climatic region), two provenances from regions with a climate similar to the expected climate in the study site by the end of the century under a moderate emissions scenario, and two provenances from regions with a climate similar to the expected climate in the study site by the end of the century under a severe emissions scenario (Figure 18, Table 11). Additionally, seedlings from the Alaquàs tree nursery in Valencia, where forest genetic improvement is conducted, were planted due to their outstanding performance in previous studies. Unlike the rest of seedlings, the Alaquàs material does not come from a single provenance region but from a mix of three regions. Aleppo pine seedlings were planted along with their common companion understory woody species: Quercus ilex, Quercus faginea, Sorbus domestica, and Acer monspessulanum. All planted seedlings were sown in 2023 at the Forest Nursery of the General Council of Aragon, Zaragoza. 42 Table 11. Regions of origin of Aleppo pine provenances planted Local Local Moderate emissions scenario Moderate emissions scenario Severe emissions scenario Severe emissions scenario Genetic improvement The seedlings were planted in 8 blocks of 0.5 ha, each with all provenances represented. To ensure the representation of potential sources of variability in the growth of Aleppo pines, the distribution of blocks, provenances, and seedlings was carefully designed: blocks were arranged considering the main source of environmental variability (i.e., the slope, Figure 19), provenances were distributed randomly within each block (Figure 20, Figure 21), and a minimum number of seedlings was planted by provenance within each block. The target planting density was 800-1000 seedlings per hectare (seedlings separated by 3 to 3.5 m), corresponding to 75% of pines and 25% of companion species. Thus, the final design includes a total of 336 pines per provenance divided in 8 blocks. Figure 18. Provenance regions selected: two from local climate (Interior Catalonia, Coastal Catalonia), two from moderate emissions scenario (Bárdenas-Ribagorça, Interior Levante), two from severe emissions scenario (Monegros-Ebro valley, Meridional Betica). 43 Figure 19. Experimental design with random distribution of the blocks within the study site. Figure 20. Experimental design with random distribution of the six Aleppo pine provenances and the seed orchard seedlings. Each point represents a seedling (total of 2,500 seedlings were planted). 44 4.3.3. Plantation description In February 2024, two and a half years after the EWE, 2,688 Aleppo pine seedlings were planted in optimal planting conditions, following the winter frost and the onset of spring rains. The terrain was prepared 10 days before planting using an excavator with a 60 x 60 x 60 cm bucket to dig the holes, which were spaced 3.5 meters apart. The hole was dug, and the extracted soil was returned to the hole without turning it, to improve water retention and facilitate planting. To prevent mixing of provenances, each seedling was carefully tagged with a colour corresponding to its provenance. After planting, each seedling was georeferenced with a submeter precision GPS (Figure 20) to facilitate the monitoring of its performance after planting, as part of another European project. Planting was done manually, and the seedlings were not watered afterwards (Figure 22). Figure 21. Example of the planted seedlings within a block. The tape defines the areas where different provenances are planted. Figure 22. Aleppo pine seedling planted. 45 Seedling survival was assessed in June 2024 to evaluate survival following planting, and in October 2024 to assess survival after enduring harsh environmental conditions during summer. In June, most of the seedlings had survived, with survival rates ranging from 78.7% to 86.5%, indicating that the soil preparation techniques, seedling selection, and nursery management were appropriate to mitigate planting stress (Figure 23). However, by October, the survival rate had significantly decreased, ranging from 55.5% to 62.7%, reflecting the challenges of growing and establishing under high temperatures and water deficit (Figure 24). Overall, seedlings from Monegros-Ebro valley, representing the severe climate scenario, and from Alaquàs, were the best performers. However, it is essential to evaluate the response of the provenances in the long term, since long-term survival not only depends on the initial establishment conditions but also on the adaptive capacity of the seedlings to face new climatic conditions (Voltas, 2023). Survival Provenance region Provenance region Figure 23. Survival rate of provenances in June 2024. 46 Figure 24. Survival rate of provenances in October 2024. Assisted migration is a promising strategy for forest restoration in the face of climate change. However, its success requires a comprehensive approach that combines the proper selection of provenances with adaptive management practices and long-term monitoring to assess its effectiveness and sustainability. Provenance region Survival Provenance region 47 Broncano, M. J., & Retana, J. (2004). Topography and forest composition affecting the variability in fire severity and post-fire regeneration occurring after a large fire in the Mediterranean basin. International Journal of Wildland Fire, 13(2), 209-216. CFRS Fire report (2022). https://interior.gencat.cat/web/.content/home/030_arees_dactuacio/bombers/foc_ forestal/consulta_incendis_forestals/informes_incendis_forestals/20202029/2021/20210724_I_RET_Santa_Coloma_de_Queralt_ENGLISH.pdf Chamorro, D., Luna, B., & Moreno, J. M. (2013). Germination response to various temperature regimes of four Mediterranean seeder shrubs across a range of altitudes. Plant ecology, 214, 1431-1441. Eugenio, M., & Lloret, F. (2006). Effects of repeated burning on Mediterranean communities of the northeastern Iberian Peninsula. Journal of Vegetation Science, 17(6), 755–764. https://doi.org/10.1111/J.1654-1103.2006.TB02499.X Lucas-Borja, M. E., Plaza-Álvarez, P. A., Ortega, R., Miralles, I., González-Romero, J., Sagra, J., Moya, D., Zema, D. A., & de las Heras, J. (2020). Short-term changes in soil functionality after wildfire and straw mulching in a Pinus halepensis M. forest. Forest Ecology and Management, 457, 117700. https://doi.org/10.1016/J.FORECO.2019.117700 Mendez-Cartin, A. L., Coll, L., Valor, T., Torné-Solà, G., & Ameztegui, A. (2024). Post-fire growth of Pinus halepensis: Shifts in the mode of competition along a precipitation gradient. Forest Ecology and Management, 554, 121693. https://doi.org/10.1016/J.FORECO.2024.121693 Martín-Alcón, S., Coll, L., Ameztegui, A., & Tomé Morán, J. (2022). Proyecto LIFE ADAPTALEPPO: selvicultura para la adaptación de los pinares mediterráneos al cambio climático. 1–6. https://adaptaleppo.eu/wp-content/uploads/2023/08/8CFE-937_LIFEAdaptaleppo.pdf Moreira, F., Arianoutsou, M., Corona, P., & De las Heras, J. (2012). Post-fire management and restoration of southern European forests, managing forest ecosystems. International Forestry Review, 14, 1. Neilson, R. P., Pitelka, L. F., Solomon, A. M., Nathan, R., Midgley, G. F., Fragoso, J. M. V., Lischke, H., & Thompson, K. (2005). Forecasting regional to global plant migration in response to climate change. BioScience, 55(9), 749–759. Paneghel, M., Torné, G., Morin, X., Alday, J. G., & Coll, L. (2024). Increasing temperature threatens post-fire auto-successional dynamics of a Mediterranean obligate seeder. Journal of Ecology, 00, 1–15. https://doi.org/10.1111/1365-2745.14433 Pausas, J. G., Ribeiro, E., & Vallejo, R. (2004). Post-fire regeneration variability of Pinus halepensis in the eastern Iberian Peninsula. Forest Ecology and Management, 203(1– 3), 251–259. https://doi.org/10.1016/j.foreco.2004.07.061 48 Ste-Marie, C., Nelson, E. A., Dabros, A., & Bonneau, M. E. (2011). Assisted migration: Introduction to a multifaceted concept. Https://Doi.Org/10.5558/Tfc2011-089, 87(6), 724–730. https://doi.org/10.5558/TFC2011-089 Vallejo, V. R., Arianoutsou, M., & Moreira, F. (2012). Fire ecology and post-fire restoration approaches in Southern European forest types. Post-fire management and restoration of southern European forests, 93-119. Voltas, J. (2023). Tres siglos de Genética Forestal: situación actual, retos e incertidumbres Three centuries of Forest Genetics: current situation, challenges and uncertainties Cuadernos de la Sociedad Española de Ciencias Forestales Conferencias y Ponencias del 8 0 Congreso Forestal Español. Cuad. Soc. Esp. Cienc. For, 49(2), 101–116. https://doi.org/10.31167/csef.v0i49.19939