Climate Change Projections for Bioclimatic Distribution of Castanea sativa in Portugal

Abstract

The chestnut tree is an important forestry species worldwide, as well as a valuable food resource. Over recent years, Portugal has shown an increasing trend in chestnut tree area, as well as increases in production, hinting at the socioeconomic relevance of this agro-forestry species. In this study, bioclimatic indices are applied to analyse the spatial distribution of chestnut trees in mainland Portugal, namely growing degree days (GDD; 1900–2400 °C), annual mean temperature (AMT; 8–15 °C), summer days with maximum temperature below 32 °C (NTX), and annual precipitation (PRE; 600–1600 mm). These indices are assessed for the baseline (IBERIA01, 1989–2005) and future climates (EURO-CORDEX: 2021–2040, 2041–2060, and 2061–2080) under two forcing pathways (RCP4.5 and RCP8.5), also taking into account the chestnut tree land cover. For the baseline, the GDD showed only 10% suitability for chestnut tree cultivation in southern Portugal, whereas much higher values are found in the north of the country, and at higher altitudes (50–90%). For the AMT, higher elevation areas in northern Portugal show almost 100% suitability. Concerning NTX, the suitability reduces from the west (100–90%) to the east (40%). Regarding PRE, the suitability is heterogeneous throughout the territory, with areas under 50%. A new Chestnut Suitability Index (CSI) was then computed, which incorporates information from the four previous indices. The CSI reveals a suitability ranging from 100 to 75% in the north, while central and southern Portugal show values from 25 to 50%. For future climates, a progressive reduction in CSI was found, particularly for RCP8.5 and in the long-term period. Changes in bioclimatic conditions may restrict the 100% suitability to a narrow area in the north of the country. These reductions in chestnut bioclimatic suitability may have socio-economic and ecological implications for the management of the important agro-forestry species.

1. Introduction

Chestnut trees (Castanea spp., Fagaceae family) are important forestry species worldwide [1], as well as a valuable food resource, which has accompanied the evolution of the human population over the centuries. The Castanea Mill. genus is generally disseminated throughout southern Europe, north-eastern America, northern Africa, and in certain parts of Asia, although there is a large variability in terms of species [2]. According to the Food and Agriculture Organization data [3] in 2020, the worldwide chestnut tree area was nearly 600 thousand ha, while chestnut production reached over 2 million t. This represents an increasing trend of nearly 60 thousand t/year. According to the latest available reports, Asia is the largest chestnut (Castanea crenata Siebold & Zucc. and Castanea mollissima Blume) producer, with 2 × 10⁶ t, while Europe (Castanea sativa Miller) and America (Castanea dentata (Marshall) Borkh.) produced 333 × 10³ t and 90 × 10³ t, respectively [3,4]. In Europe, Portugal has the largest chestnut (C. sativa) producing area, with a total of 51,700 ha [3]. According to Figure 1a,b, in recent years, Portugal has shown an upward trend in the cultivation area of 803 ha/yr. Regarding chestnut production, the country also showed an upward trend of 380 t/year, resulting in 35,830 t in 2020.

In Portugal, this ancestral species is one of the most important crops for rural populations. Chestnut is broadly distributed in the northeast of Portugal, mostly in the coolest areas of the districts of Vila Real, Bragança, Viseu, and Guarda (Figure 1c), where this crop represents one of the most profitable activities [1,5,6,7]. In some other smaller areas, mainly in north-western and central Portugal, this species is also traditionally cultivated [6,7]. Chestnut production is mostly concentrated in the Bragança and Vila Real districts, with 59% and 19% national share, respectively. These districts are followed by Viseu (11%), Guarda (9%), Portalegre (1%), and Castelo Branco (1%) (Figure 1c). In mainland Portugal, Protected Denomination of Origin (PDO) areas are also defined, namely Terra Fria (Bragança district), Soutos da Lapa (Viseu and Guarda district), Padrela (Vila Real district), and Marvão (Portalegre district) (Figure 1d) [1,2].

Figure 1. Chestnut (a) land cover area (hectares) and (b) yield (t) in Portugal, between 1980 and 2020. The linear regression trends are also shown, along with the corresponding trend value (c) chestnut distribution areas in mainland Portugal; (d) location of the chestnut Protected Denomination of Origin (PDO) regions [3,8].

Figure 1. Chestnut (a) land cover area (hectares) and (b) yield (t) in Portugal, between 1980 and 2020. The linear regression trends are also shown, along with the corresponding trend value (c) chestnut distribution areas in mainland Portugal; (d) location of the chestnut Protected Denomination of Origin (PDO) regions [3,8].

Throughout Portugal, there are approximately 23 cultivars of C. sativa, such as “Longal”, “Judia”, and “Martaínha” [2,9]. The fruits from these cultivars can reveal different sizes, types, flavours, sweetness, or prices [4]. For example, fruits from the Portuguese cultivar “Judia” are the largest, with a calibre of about 50–60 fruits kg⁻¹, and fruits from “Longal” are around 70–80 fruits kg⁻¹, but are easier to peel and conserve, and have a good flavour [10]. The “Martaínha” is better adapted and is an economically important chestnut variety, due to its precocity and its technological properties [11].

Chestnut tree development and growth is largely influenced by several pedoclimatic factors, such as physical-chemical characteristics of the soil, orography, soil water distribution (more than 90% of these orchards are rainfed), and climatic conditions. In effect, this tree thrives in different environmental and bioclimatic conditions, ranging from the warm and dry Mediterranean conditions to the cool and wet conditions in the Atlantic regions [4,12]. This tree can be encountered at altitudes varying from 0 to 1800 m [4], which highlights the wide range of growing thermal conditions. C. sativa is a moderately thermophilic plant that succeeds in environments with annual mean temperatures of 8 to 15 °C [4,7], accepting maximum temperatures between 27 and 31 °C, although being highly sensitive to summer droughts. In turn, it bears minimum temperatures as low as −16 °C. Chestnut yield can be considerably reduced (over 50%) when maximum temperatures are above 32 °C over several days [7,13]. It is known that chestnut growing areas should fulfil a heat accumulation requirement of 1900–2400 degree days (°C) [4,13]. This crop is usually rainfed and can grow within rainfall levels between 600 and 1600 mm [14], though many young commercial orchards are usually irrigated. These bioclimatic indicators are useful tools for forecasting how tree species will respond to future climates, since they focus on the climatic thresholds that restrict species distributions.

Spatial distributions of bioclimatic conditions are applied in different contexts in the world as well as under climate change [15,16,17,18], although applied to chestnut trees, studies are still incipient, particularly in the Portuguese context. The chestnut tree is projected to be particularly affected by climate change. Factors such as the abandonment of traditional orchards, wildfire, and an increased incidence of pests and diseases [19] may be consequences of climate change. Modifications in crop microclimatic conditions are projected to occur in the future [12,20]. The incidence of pests and diseases threatening chestnut trees may also intensify under climate change [1,20,21]. Moreover, quality parameters and chemical composition of chestnuts may also shift [22]. Some climate change studies suggest warmer temperatures accompanied by recurrent and intensified extreme events, such as severe rainfall events, droughts, or heatwaves [23,24,25], are also expected, increasing crop damage in the upcoming decades [26]. For the chestnuts in the Iberian Peninsula, climate change may represent a major threat, leading to significant losses [12].

Portugal is characterized by a temperate climate, with Mediterranean-like characteristics [27,28], some of them being particularly suitable for the cultivation of chestnut trees. The Portuguese climate reveals large interannual variability and seasonal irregularity for temperature and precipitation, featuring typical dry warm summers and rainy cool autumn-winter periods [7]. Moreover, it is also highly affected by orographic factors, such as elevation and proximity to the Atlantic Ocean. As reported by [29], there is a disparity between the north (more mountainous) and south (flatlands) of the country (Figure 2a). The heterogeneous orography largely influences the precipitation distribution and temperature pattern [7]. The annual accumulated precipitation varies from approximately 400 to 2000 mm (Figure 2b), whereas the annual temperature fluctuates between 9 °C in the mountains and 18 °C along the south (Figure 2c). Nevertheless, future climate projections hint at significant warming and drying [27,30,31,32,33], with an increase in temperature of 1.5 to 4 °C by 2100, accompanied by a decrease in rainfall. Therefore, chestnut cultivation can be threatened by climate change, such as through advancing phenological timings, shifts in regional bioclimatic suitability, or proliferation of pests and diseases [4,34,35], among other climate-driven challenges.

In the present study, several metrics for assessing climatic influences on the chestnut tree distribution are developed, which can be used as a bioclimatic zoning tool for climate change impact assessment over Portugal, aiming to evaluate the potential impacts of cli-mate change on chestnut distribution in Portuguese territory. Therefore, the objectives of this study are: (1) to improve the bioclimatic chestnut tree zoning, by using high-resolution climate data; (2) to identify the representative climatic conditions associated with the current chestnut tree cultivation in Portugal; and (3) to assess future shifts in bioclimatic zones under climate change scenarios. Furthermore, a new and innovative Chestnut Suitability Index (CSI) that identifies the best conditions for chestnut tree growth was developed. It is expected that this study could become a decision support tool for the regional sector stakeholders, according to the scenarios envisaged for the future.

2. Materials and Methods

2.1. Land Use Maps

The present study is conducted over mainland Portugal. The characterization of the chestnut area was based on the digital inventory provided by the Geographical Institute of Portugal (Carta de Uso e Ocupação do Solo de Portugal Continental—COS 2007) (Source: Direção-Geral do Território, http://www.dgterritorio.pt/ (accessed on 15 September 2021)) [7,8,12]. Areas with the designation “Florestas de castanheiro” and “Pomares de castanheiro” were selected as a data source to identify the chestnut agro-forestry systems (Figure 1c).

2.2. Meteorological Variables and Climate Projections

A comprehensive climate change assessment is performed based on a multi-model multi-scenario ensemble to assess the potential impacts of climate change on the optimum chestnut tree cultivation areas in Portugal. Spatially, interpolated climate datasets were herein used as a tool to study both recent past and future. Climatic variables for baseline (1971–2015, IBERIA01) and future climates, with three periods: 2021–2041, 2041–2060, and 2061–2080, under RCP 4.5 and RCP 8.5, were used. For the future climates, four different experiments generated by the global climate model (GCM)—regional climate model (RCM) chains, from the EURO-CORDEX dataset (

Table 1. Ensemble of Global Climate Model—Regional Climate Model (GCM-RCM) chains selected for this study [36], along with the corresponding institute that undertook each model experiment.

Institute	Global Climate Model	Regional Climate Model
Institut Pierre-Simon Laplace	IPSL IPSL CM5A MR	WRF331F
Koninklijk Nederlands Meteorologisch Instituut	ICHEC EC EARTH	RACMO22E
Max Planck Institute for Meteorology	MPI M MPI ESM LR	CCLM4-8-17
Sveriges Meteorologiska och Hydrologiska Institut	CNRM CERFACS CNRM CM5	RCA4

), were used [27,32,33].

To perform the bioclimatic zoning of chestnut trees, the following variables are used: daily minimum, mean and maximum temperatures, and daily total precipitation [32]. The IBERIA01 dataset provides these variables on a regular grid of 0.10° latitude × 0.10° longitude, whereas the EURO-CORDEX variables are on a 0.11° × 0.11° grid [29,35,37]. The selected scenarios encompass a wide range of uncertainties regarding future climates, where temperature increases from ~2 °C in RCP4.5 to ~4 °C in RCP8.5 (global averages) [33]. Regarding these data, RCP4.5 is a trajectory that describes radiative forcing of 4.5 Wm⁻² (650 ppm CO² eq.) with stabilization after mid-century, while for RCP8.5, a radiative forcing of 8.5 Wm⁻² (1 370 ppm CO² eq.) is projected until the end of the century [38]. This pathway has the highest greenhouse gas emissions for the future [39].

2.3. Bias Correction and Downscaling of the EURO-CORDEX Dataset

Data from the EURO-CORDEX dataset were bias-corrected. Firstly, the higher-resolution IBERIA01 grid was resampled (bilinear interpolation) to the match the coarser EURO-CORDEX grid (∼12.5 km, 0.11° latitude × longitude regular grid). Secondly, the model data were bias-corrected for 1989–2005 using the IBERIA01 dataset as a baseline and following the “Empirical Quantile Mapping” methodology [40]. This methodology corrects the full empirical probability distribution of each model variable (daily temperatures and precipitation), based on its corresponding observational distribution within a pre-defined common period. Lastly, the same quantile-based bias correction was applied to the selected future periods [32].

2.4. Bioclimatic Indices

The geographical distribution of species is influenced by both biotic and abiotic factors. In the present study, climatic variables are taken as the main first-order constraints of species distribution [41]. Bioclimatic indicators preserve climate features that are physiologically relevant to plant growth and species distribution and are widely used in studying spatial distribution mapping for predicting climatic influences on plants, especially when assessing the potential impacts of climate change on species distribution [42]. Herein, 4 bioclimatic indices are considered: (i) growing degree days (GDD, in °C), (ii) annual mean temperature (AMT, in °C), (iii) the number of summer days with maximum temperature below 32 °C (NTX), and (iv) annual accumulated precipitation (PRE, in mm), for each driving model, period and scenario [43]. These four indices are complementary and reflect the main climatic factors that commonly influence the growth and development of chestnut trees. The spatial patterns of these bioclimatic indices are thereby obtained for Portugal. Although the 12.5 km spatial resolution can be considered already high for general climate change assessments, it is still very limiting for agronomical assessments in areas with strong spatial gradients, such as those resulting from complex orography. Therefore, a downscaling methodology was applied to the computed bioclimatic indices based on a kriging approach and elevation data at 25 m resolution. This methodology allows characterizing the meso-climatic influences on crops throughout Portugal.

2.5. Optimum Bioclimatic Thresholds

Specific thresholds for adequate chestnut growth, physiological development, and fruit ripeness [7] can be defined in each bioclimatic index. These thresholds establish the bioclimatic niche and are succinctly explained below:

(i)

The GDD for chestnuts are usually the sum of temperature departures above 6 °C. It is generally accepted that chestnut areas must accomplish a GDD of 1900–2400 °C between May and October [1,4];

(ii)

An AMT ranging between 8 and 15 °C is usually considered for chestnuts, as this tree is a temperate climate and a moderate thermophilic species [4,44,45];

(iii)

Regarding NTX, it may exhibit thermoinhibition (production is reduced over 50%) when the air temperature is above 32 °C [4,46];

(iv)

For PRE, rainfalls between 600 and 1600 mm are generally favourable, generally providing sufficient soil water content. Values below 600 mm may require additional irrigation, whereas excessive rainfall levels may be detrimental when occurring during spring-summer [4,14,24].

The aforementioned bioclimatic indicators were computed for each period separately (1989–2005, 2021–2040, 2041–2060, 2061–2080). For each year and bioclimatic index, if a given grid box achieved the corresponding threshold, a value of 1 was attributed to that grid box. Otherwise, a 0 score was ascribed. Subsequently, the results are presented as the percentage of years in each period fulfilling the pre-defined bioclimatic threshold. Lastly, an aggregated categorization is developed that integrates all bioclimatic indices, i.e., GDD, AMT, NTX, and PRE. This versatile index combines information about regional bioclimatic characteristics and is therefore henceforth designated as Chestnut Suitability Index (CSI). Herein, for the development of the CSI, all indices have the same weights, since there are no studies suggesting a higher impact of one index over the other. While more advanced methodologies could have been applied to define this index, we opted for a simple method that is easily replicable by sector stakeholder.

3. Results

3.1. Current Chestnut Tree Distribution vs. Optimal Zoning

The chestnut tree distribution (Figure 1c) is characterized according to current climatic (air temperature and precipitation) and orographic conditions (Figure 2). In Portugal, 61.3% of the chestnut trees are distributed within the elevation class of 751 to 1000 m, 34.5% between 501 and 750 m, 2.5% between 251 and 500 m, 1.3% between 1001 and 1250 m, and 0.3% below 250 m. Above 1251 m, the presence of chestnut trees is not observed. However, in the interval 751–1000 m, 3.7% of the total area in this elevation class corresponds to chestnut tree orchards (

Table 2. Characterization of chestnut tree distribution in Portugal (frequency in %) as a function of different classes of elevation, annual mean accumulated precipitation and temperature. The chestnut tree land cover ratios for each class are also listed.

Elevation (m)	Frequency (%)	Chestnut Cover Ratio (%)
1751–2000	<0.1	<0.1
1501–1750	<0.1	<0.1
1251–1500	<0.1	<0.1
1001–1250	1.3	0.4
751–1000	61.3	3.7
501–750	34.5	1.1
251–500	2.6	<0.1
<250	0.3	<0.1
Mean temperature (°C)	Frequency (%)	Chestnut Cover Ratio (%)
18.1–19	<0.1	<0.1
17.1–18	<0.1	<0.1
16.1–17	<0.1	<0.1
15.1–16	0.2	<0.1
14.1–15	3.0	<0.1
13.1–14	42.2	1.2
12.1–13	53.7	4.2
11.1–12	0.8	0.2
10.1–11	<0.1	<0.1
9.1–10	<0.1	<0.1
<9	<0.1	<0.1
Mean precipitation (mm)	Frequency (%)	Chestnut Cover Ratio (%)
1801–1950	<0.1	<0.1
1651–1800	<0.1	<0.1
1501–1650	0.1	<0.1
1351–1500	0.3	<0.1
1201–1350	1.8	0.2
1051–1200	13.4	0.7
901–1050	29.4	1.3
751–900	35.1	1.1
601–750	19.0	0.3
451–600	<0.1	<0.1
<450	<0.1	<0.1

). According to [4], elevations between 700 and 1000 m indeed provide the best conditions for fruit production in Portugal. The authors of [24] also highlighted that the majority of chestnut production in Portugal comes from high-elevation areas, namely from the “Terra Fria—Trás-os-Montes” region (higher elevations of the Vila Real and Bragança districts, Figure 1c). Furthermore, 53.7% of the chestnut orchard areas show AMT between 12.1 and 13 °C, whereas 42.2% are grown under temperatures from 13.1 to 14 °C. Temperatures from 14.1 to 15 °C stand out at 3.0%. Residual chestnut distributions (<1%) can be found for temperatures in the intervals from 11.1 to 12 °C, owing to the scarcity of these conditions in Portugal, or from 15.1 to 17 °C, due to the excessively warm conditions. In addition, 5.7% of the total area within this temperature class is covered with chestnut tree orchards (Table 2). According to [24], the higher elevation (coolest) regions suggest that relatively cool winters are suitable for chestnut development. This may be partially associated with the chilling requirements of chestnut trees to promote dormancy [28].

Regarding the annual precipitation, chestnut trees in Portugal are mostly distributed in the precipitation areas between 600 and 1200 mm. In total, 19.0% of the chestnut area corresponds to the precipitation class 601–750 mm, 35.1% corresponds to 751–900 mm, and 29.4% and 13.4% correspond to 901–1050 mm and 1051–1200 mm, respectively. For precipitations in the intervals 451–600 mm and 1351 to 1950, the areas are residual (less than 1%) (Figure 1c). According to the literature [4], the minimum annual rainfall for a proper chestnut tree cultivation is 600–700 mm (Table 2), which is in general agreement with these distributional values. Lastly, 3.5% of the total area in this precipitation class corresponds to chestnut tree areas.

3.2. Bioclimatic Indices

3.2.1. Current Conditions

Following the GDD requirement for the chestnut tree growth and development, there is a heterogeneous distribution (Figure 3a). In the south of the country, the percentage of suitability is lower than 10%. In the north, however, there are areas suitable for chestnut trees at high elevations (suitability of 50–90%), while at low elevations suitability remains at 10–20%. In the case of AMT (Figure 3b), most of the northern region is characterized by favourable conditions for chestnut tree cultivation (suitability near 100%), whereas the opposite is observed in the southern half of the country. The map in Figure 3c, regarding NTX, shows a percentage gradient from the west (100–90%) to the East (40%). PRE (Figure 3d) reveals heterogenous values throughout Portugal. Areas in the Viana do Castelo, Bragança, Portalegre, and Beja districts (Figure 1c) correspond to less than 50%, while along the northern coast the percentage is 100–70%. However, this region is not favourable for chestnut tree growth due to the excessive humidity and maritime influence that are characteristic of these areas [47,48]. Figure 3e presents the combination of the four conditions, the CSI, showing a percentage for the chestnut tree varying from 25 to 100%. The central area of the country mostly shows CSI values of approximately 75–50%, whilst the CSI is of nearly 25% in the southeast of the country (Portalegre, Faro, Évora, and Beja district). The PDOs of Padrela, Terra Fria, and Soutos da Lapa are in regions with a high CSI (100%), thus highlighting the correspondence between the actual growing areas and the CSI. On the other hand, the PDO of Marvão shows a lower CSI of 75%.

3.2.2. Future Conditions

Figure 4 presents the results for the period 2021–2040 under RCP4.5 (top panel) and RCP8.5 (bottom panel). The patterns of RCP4.5 and RCP8.5 show vast similarities, particularly in the two first future periods. Due to their likeness, and for the sake of succinctness, discussion will be focused on RCP8.5 (RCP4.5 results are generally less severe). Regarding GDD, more than half of the area of mainland Portugal, particularly in the south, presents only 10% of favourable conditions for chestnut tree cultivation. In the northern region, at higher elevations (>751) the percentage increases up to 90% (Figure 4a). In the north of the country, the AMT is heterogeneous, standing out at high elevations 100%. In Figure 4a,b, in the Marvão PDO, there is a reduction in areas with high suitability percentages compared to the baseline period. As expected, the number of days with a daily maximum temperature above 32 °C will increase compared to the baseline. For NTX, there is a gradual reduction from the coast (100–80%) to the innermost parts of the country (30–20%) (Figure 4c). There is also a reduction in the area concerning the most adequate precipitation levels for the development of chestnut trees, particularly in the northeast and southernmost areas (Figure 4d). Concerning the CSI, compared to the baseline, the area with 100% suitability is reduced, while the area with 25% suitability is increased. The Marvão PDO region, in the baseline, depicts a CSI of 75%, reducing to 50% suitability in 2021–2040. Although the regions at higher elevations tend to maintain the CSI levels at 100%, as is the case with the central mountain range, there is an overall reduction in the CSI for all four PDOs (Figure 4e).

In 2041–2060 under RCP8.5, for GDD, AMT, and NTX (Figure 5a–c), the optimum conditions tend to decrease, lowering the overall percentage levels of these indicators. For AMT and GDD, the northern regions at higher elevations tend to provide the best conditions for chestnut tree cultivation. In the case of NTX, conditions are preserved in coastal areas and high-elevation regions. In the case of PRE, the conditions tend to worsen, particularly in the inner areas (Figure 5d). According to the CSI pattern (Figure 5e), the favourable areas for chestnut tree cultivation continue to decline throughout the country. As an example, in the Bragança district (northeast, Figure 1c), with the largest area of chestnut trees, there is a steady decrease in CSI, particularly in its southern part, showing suitability levels lower than 25%.

In the period 2061–2080 under RCP8.5, trends in GDD, AMT, and NTX reinforce the projected significant reduction in areas favourable to chestnut tree cultivation (Figure 6c). PRE also decreases in this period, reaching minimum levels of 20% in the Beja district, southern Portugal (Figure 6d). Regarding the CSI (Figure 6e), the area corresponding to 100% is now residual, only being found at the highest elevations in the mountainous areas of central and northern Portugal. According to these projections, the production of chestnuts may become challenged in Portugal under future climates. Nonetheless, under RCP4.5, a less severe scenario, the CSI presents a higher percentage of areas for chestnut development (Figure 6j).

Figure 7 represents the distribution of CSI as a function of chestnut tree areas in Portugal, also expressing the interaction between chestnut areas (in percentage) and the corresponding CSI values for the different periods and scenarios. In the baseline period, the distribution of the CSI in Portugal shows a marked peak in the area corresponding to a CSI of around 85%. In the period 2021–2040 (Figure 7a), the distribution is significantly flattened, and the mentioned peak is no longer prominent in the distribution. In this period, the highest CSI values are 75% for RCP 4.5, and 68% for RCP8.5. In both scenarios, the areas with CSI above 50% predominate (Figure 7b). In the period 2041–2060, for RCP4.5, the largest areas are around 60%, while the largest areas are around 40% for RCP8.5. From this period onwards, it becomes noticeable that RCP8.5 depicts generally worse conditions for chestnut growth than RCP4.5 (Figure 7c), as was expected due to its stronger anthropogenic radiative forcing. In the period 2061–2080, there is even a more evident difference between the two scenarios. In the case of RCP4.5, the higher CSI corresponds to 40–60%, whereas it is between 20 and 40% in RCP8.5 (Figure 7d). Over time, there is a gradual reduction in areas, as well as in the percentage of suitable conditions for the development of chestnut trees.

4. Discussion

In the present study, the spatial-temporal changes in the climate patterns over Portugal were investigated, with emphasis on the chestnut tree development. The analysis was based on a very high-resolution dataset for Portugal in the baseline period (1989–2005) and under two future scenarios, RCP4.5 and RCP8.5, over three future periods of 2021–2040, 2041–2060, and 2061–2080 and using a four-member ensemble of GCM-RCM. High-resolution bioclimatic zoning in mainland Portugal is performed based on several chestnut tree bioclimatic indices, including a new and innovative Chestnut Suitability Index. These bioclimatic indices are useful for approaching species distribution patterns and of their likely shifts under changing climates, as well as in outlining suitable adaptation strategies, which are essential for sector decision-makers [7,42,49,50,51]. This study fundamentally assesses shifts in bioclimatic zones in Portugal, under climate change scenarios, which may have implications on chestnut tree cultivation and distribution. It is confirmed that the future climatic conditions will be challenging for chestnut sustainability.

Plant growth is primarily regulated by precipitation and temperature, while changes in these factors may shift optimal climates for species growth and development. The patterns of these climatic factors are rapidly changing, and this is becoming a challenge for plant preservation. This highlights the importance of this study, based on innovative high-resolution bioclimatic zoning applied for chestnut trees. On the other hand, factors such as soil properties or cultivation techniques can also have major effects on the plant but are not considered in this investigation [4,50,52]. Moreover, the genetic diversity and phenotypic plasticity and interaction between species are key factors to enable plants to adapt to climate change [4,27,33,53].

Air temperature increases are projected over the upcoming decades in the main areas of European forests [54], also including Portugal [29,55]. Moreover, Portugal is already very exposed to the occurrence of temperature extremes, namely heatwaves [49]. Higher temperatures result in earlier phenology, faster growing seasons, and usually, yield reduction [4,29]. The chestnut tree is mostly distributed in north-eastern Portugal, where the intricate topography triggers strong thermal differences [7]. Chestnut trees require moderate chilling accumulation and present a high sensitivity to summer droughts, thus being preferably located in Portugal in the areas with the temperatures above 8 °C and below 32 °C [7]. Overall, in Portugal, the projected warming is expected to reduce these areas more prone to chestnut tree cultivation. According to the GDD, ATM, and NTX results, until to 2080, the favourable air temperature for chestnut trees will decrease and cause negative impacts on the specie development.

In the Mediterranean, the rainfall patterns are characterized by a strong space-time variability [26]. In the future, an overall drying trend and, though with less confidence, higher frequencies of occurrence of extreme precipitation events are projected in Portugal [56,57]. For this reason, detrimental impacts on yields and increased production risks are expected in the future [53]. Not only the annual accumulated precipitation is important for chestnut tree production, but also its seasonality and annual distribution. Summer precipitation contributes to the fruit dimension growth, while winter precipitation favours soil water retention which promotes the beginning of fruit setting [57].

In response to climate change, species may progressively migrate to higher latitudes or elevations. In central and northern Europe, the expansion of bio-climatically suitable areas for chestnut trees are expected. On the contrary, in the southern areas, suitability will be reduced due to water shortage and more extreme weather events [25,27,33]. These future conditions in southern Europe may accelerate the replacement of chestnut orchards by other trees or even land uses [53,58]. Furthermore, pest and disease pressure may also intensify these climate changes impacts. Nonetheless, the decrease in precipitation and humidity in southern European counties, such as Portugal, may be beneficial for certain fungal diseases [4]. The CSI index may have several applications to support crop management and maintenance. For farmers, this index can help to predict the best zones for new chestnut plantations under future climates.

Climatic changes indeed embody an important threat to the chestnut trees in Portugal, becoming a socio-economic and ecological challenge. Adaptation tactics are crucial in mitigating the consequences of climate change. Measures that can improve adaptability and minimize susceptibility to climate change consequences while also making use of good opportunities [4]. As an illustration, irrigation (short-term measure) may increase the commercial value of the chestnut, as it increases fruit size by reducing water stress while keeping its nutritional value and sensory profile [59]. Improved soil management or the application of leaf protectors against extreme weather are among other adaptation measures that may be considered in the coming decades.

5. Conclusions

According to the results obtained herein, climate change may reduce the viability of chestnut tree cultivation in Portugal. Climate change impacts on chestnut bioclimatic suitability may have socio-economic and ecological implications for the management of the important species. Similarly to other agro-forestry species, the chestnut tree sector should also adapt to climate change, with the implementation of adequate measures to warrant the sustainability of production and species adaptability. The methodologies used in the study, particularly the new CSI, may be a relevant aid for companies and entities of the Portuguese chestnut sector. Other studies, complementary to the present, should be carried out in the future to include the use of other biological factors (e.g., chilling requirements) that may influence the development and production of the species. The application of process-based models to simulate the responses of chestnut trees to climate and environmental changes is of utmost relevance to better understand the potential impacts on species growth, development, abiotic stress, yield, and fruit quality. Such models may help relate climate change to species adaptation at different response rates. Taking into account the importance of the present work for sector stakeholders, future studies should focus on the modelling of yield and quality parameters under future climates. Lastly, but very importantly, partnerships with research units and agricultural associations are strongly encouraged to acquire more field data and outline strategies to cope with climate change and reduce its derived risks.