Effect of climate change on the spatial distribution and cork production of Quercus suber L., the risk of exclusion by the Aleppo pine expansion, and management practices to protect Q. suber habitat: A review

Climate change represents an important challenge for forest management and the silviculture of stands and it is known that climate change will have complex effects on cork oak forest ecosystems. North Africa and the Mediterranean basin are especially vulnerable to climate change. Under the effect of climate change, cork oak will disappear from a large area in the future, and the rest will migrate to higher altitudes and latitudes. This study aimed to evaluate the effect of climate change on the spatial distribution of Quercus suber L. and cork production in the Mediterranean area, and the risk of its exclusion by the Aleppo pine (Pinus halepensis Mill.) expansion. The literature review showed that up to 40% of current environmentally suitable areas for cork oak may be lost by 2070, mainly in northern Africa and the southern Iberian Peninsula. Temperature directly influences atmospheric evaporative demand and should affect cork productivity. Precipitation is the main factor that positively influences cork growth and several authors have confirmed the negative effect of drought on this growth. Currently, cork oak habitats are colonized in several places mainly by the Aleppo pine. Under climate change, Aleppo pine is projected to occupy higher altitude sites and several authors have predicted that current and future global warming will have a positive influence on Aleppo pine growth in wet sites. In the future and under climate change, there is a strong possibility that the Aleppo pine will colonize cork oak habitat. Finally, we proposed management practices to protect cork oak against climate change and Aleppo pine expansion.


Introduction
Climate change represents an important challenge for ecologists, biologists, and modelers, whose research interest is the study of the potential effect of climate change on ecosystem services provided by forests (Deal et al., 2017;Fréjaville et al., 2019). A rise in the mean temperature and variation in rainfall patterns could modify the current distribution of plant species (Lopez-Tirado et al., 2018). Over the last few decades, the scientific community has become increasingly concerned with climate change. More extreme climatic conditions, such as rising mean temperatures and variations in rainfall patterns, are expected within the present century (IPCC, 2014). The Mediterranean basin is especially vulnerable to climate change (Goubanova and Li, 2007). Recent climatic projections for the Mediterranean basin predict an air temperature increase and a decline in rainfall, mainly during the summer period, resulting in extended periods of soil water deficit (Gratani et al., 2016;IPCC, 2014). Prolonged dry periods result in a negative soil water balance as the amount of water that evaporates overcomes the amount of water intercepted by the soil through rainfall and summer humidity (Brunetti et al., 2002;Moretti et al., 2015). Species might undergo a marked risk of local extinction with a consequent fragmented distribution, especially in North Africa and Iberia (Benito Garzón et al., 2008;Klausmeyer and Shaw, 2009). If the habitable area is reduced for this species, or the potential for its permanence is predicted to be limited in a given location, other species may establish within the ecosystem (Pons and Pausas, 2006;Lopez Tirado and Hidalgo, 2016a). Tree species may undergo shifts in distribution within a short time owing to the effect of global change. This is especially important for species growing at the boundaries of their current distribution Lopez-Tirado and Fidalgo (2016a). Despite the massive impact of global warming predicted by the models in the southern Mediterranean, some areas would maintain their suitability for species occurrence, thus assuming an important role in terms of biodiversity conservation. These areas might be worthy of a deeper scientific investigation and public awareness for preserving cork oak from extinction. Global climate change affects ecosystems worldwide (Parmesan and Yohe, 2003;Parmesan, 2006). Forest ecosystems are particularly susceptible to shifts in natural disturbance regimes induced by climate change (Dale et al., 2001;Trumbore et al., 2015). Given this, the niche modeling approach could be considered as a promising tool for planning, monitoring, and managing native populations and their habitats (Araujo and Williams, 2000;Costa et al., 2010;Lemes et al., 2014). Climate change will inevitably affect species geographical distribution, change species composition and richness, impact biodiversity, and influence the structure and function of ecosystems (Thuiller et al., 2011;Scheffers et al., 2016). Dyderski et al. (2018) quantified changes in the habitats of 12 European forest tree species, demonstrating that most of these species would face severe habitat contraction due to climate change. Several authors have worked on the decline and dieback of cork oak (Quercus suber L.) forests in the Mediterranean basin: (i) Climate change impacts on cork forest decline in the Mediterranean basin: Cases of drought and/or heat-induced forest mortality, elevated CO2 concentration, high temperatures, water deficit, and (ii) Other factors affecting cork oak forest decline: Anthropic pressure and wildfire insect epidemics and disease problems of cork oak forest regeneration. These authors ignored the phenomenon of the expansion of cork oak forests by other species (colonization), mainly the Aleppo pine in North Africa and in the Mediterranean area, which happens to be the main factor in the degradation of cork oak forests. In this study, we show the following: (i) Several studies have shown that under the effect of climate change, cork oak will disappear from a large area in the future and the rest will migrate northward as well as to higher altitudes.
(ii) Other studies have shown that the Aleppo pine is too plastic a species and will move in the future to areas with a humid climate.
(ii) Other authors have shown that currently cork oak habitats is colonized in several places mainly by the Aleppo pine.
The questions posed in the present study are as follows: 1) In the future, will Aleppo pine colonize the cork oak habitat under the effect of climate change? 2) Will the studies and modeling of the spatial distribution under the effect of climate change of cork oak and Aleppo pine justify existing hypotheses?

3
We aimed to answer the above questions based on current localities and in several others that seem apt for the Aleppo pine to invade cork oak forests under the effect of climate change. We intended to better understand if this is a true threat to cork oak forests in Northern Africa and the Mediterranean basin.
Potential impact of climate change on the spatial distribution of cork oak forest The genus Quercus comprises 531 species worldwide (Govaerts and Frodin, 1998). Oaks are widely represented in the northern hemisphere, where they constitute one of the most important groups in terms of ecology, biodiversity, and economy (Axelrod, 1983;Menitsky 2005;Cañellas et al., 2005Cañellas et al., , 2007Nixon, 2006). Cork oak (Q. suber L.) is a strictly Mediterranean species distributed in the western Mediterranean basin, between southwestern Europe and northern Africa (Figure 1), a climatic, ecological, and socioeconomic sensitive region (Costa et al., 2014). Cork oak woodland is considered a keystone ecosystem (Vicente and Alés, 2006) while enhancing other important provisioning and regulating ecosystem services . Moreover, cork oak woodland is an important biodiversity hotspot where cork oak is harvested throughout their lifetime for their bark (Oliveira and Costa, 2012), the cork being a valuable global non-timber forest product (FAO, 2013). Cork oak is an evergreen tree characterized by a thick bark, which grows in Mediterranean sclerophyllous forests (Hidalgo et al., 2008). Cork oak woodlands are agro-silvopastoral systems of high socio-economic and conservation value, typical of the western Mediterranean Basin ( Figure 2). They cover approximately 1.5 million hectares across Portugal, Spain, Italy, and France, and 1 million hectares in North Africa between Morocco, Algeria, and Tunisia (Diázetal, 1997;Bugalho et al., 2011). North-facing slopes, abundant annual rainfall, and the original stones that produce acidic soils were the main explanatory variables for cork oak distribution in southwestern Spain (Hidalgo et al., 2008). Most of the studies carried out on oak decline in Italy and Spain reported drought as the main driving factor. Although this phenomenon seems generalized and strictly related to site-specific conditions, drought, and in general, ongoing climate warming is identified as the main threat to declining oak stands (Gentilesca et al., 2017). The increase in temperature values and the decrease in precipitation expected for the southern part of the Mediterranean would reduce the habitable areas for cork oak (Vessella et al., 2017). The future distribution of evergreen oaks is most likely driven by climate change (Schirone et al., 2015). Under climate change, the first species that disappears is Q. suber a species that does not withstand very low temperatures in continental areas (Costa Tenorio et al., 2005), followed by Quercus coccifera in the inner continental areas (Lopez-Tirado et al., 2018). Using maximum entropy modeling (MaxEnt), Correia et al. (2017) predicted environmentally suitable areas for cork oak (Q. suber) woodlands, a socio-economically important forest ecosystem protected by the European Union Habitats Directive. The authors showed that up to 40% of current environmentally suitable areas for cork oak may be lost by 2070, mainly in northern Africa and the southern Iberian Peninsula. Almost 90% of new cork oak stands are predicted to lose suitability by the end of the century, but future plantations can take advantage of increasing suitability in the northern Iberian Peninsula and France.
According to Lopez-Tirado et al. (2018) in the predicted future scenarios, cork oak is the species with poleward migration, barely reaching France (only a small suitable area on the Atlantic coast). The authors showed that under climate change, Q. suber dominates mainly in Portugal and along a belt following the south Atlantic coast of Spain. Vessella et al. (2017), in a study of cork oak in the face of climate change in the Mediterranean area, showed that the results of eight combined ecological modeling techniques and two global circulation models highlight a broad contraction of the species potential range over the twenty-first century, both under intermediate and high emissions scenarios. The authors found that coupled northward and upward shifts were predicted, mostly pertaining to Iberia and North Africa and the potential areas detected at Levantine will likely undergo disappearance. The Mediterranean region is projected to be extremely vulnerable to global change, which will affect the distribution of typical forest types such as native oak forests (Acácio et al., 2016). Climatic extremes are also important predictors of oak forest changes, namely extreme temperatures for evergreen oak forests and deficit of precipitation for deciduous oak forests (Acácio et al., 2016). In fact, according to Lopez-Tirado et al. (2018), evergreen oaks could benefit from global change, making up larger mixed forests than the current ones. The influence of global warming is also imprinted into the upward shift in elevation, which would be experienced by cork oak, mostly in the southwestern part of its potential suitable area (Vessella et al., 2017).

5
The increased concentration of greenhouse gases in the atmosphere is acknowledged as one of the main reasons for the observed global climatic change. This phenomenon significantly affects the species geographical distribution and changes their richness distribution pattern (Sun et al., 2020). Costa Tenorio et al. (2005) showed that although holm oak and cork oak forests are estimated to have spanned up to 30 million hectares in the past, their current distribution is limited to about 2-3 million hectares worldwide. However, according to Vericat et al. (2012), Mediterranean oak formations presently occupy around 7.3 million hectares. Drought and temperature instability could pose a threat to the distribution of Q. suber L., inducing shifts and reductions of distribution ranges, thus eventually changing the richness distribution pattern (Sun et al., 2020). The authors showed that the distribution center of Q. suber L., is migrating towards the northeast in response to climate change. This has also been confirmed by numerous studies of different species in different regions (Peñuelas and Boada, 2003;Vessella and Schirone, 2013;Dyderski et al., 2018). Under climate change, the southern areas of the current distribution of oak corks will be the most affected, including Alentejo and Algarve in Portugal, Extremadura, and Andalucía in Spain, and most of North Africa (Correia et al., 2017). In the southernmost region, which includes North Africa and parts of Iberia, the expected decrease in suitability is so great that resources should be concentrated in the preservation of the few refuges where microclimatic conditions will maintain long-term suitability for the species (Dobrowski, 2011). In the Mediterranean climatic region, a warming condition is forecasted, and the distribution of native oak forests could be affected (Lopez-Tirado et al., 2018). In the period between 1966-2006, a given area of cork oak forest in Portugal was altered mostly to pine and eucalyptus (Acácio et al., 2016). Pecchi et al. (2020) showed that the analyses reported an unchanged amount of total land suitability for forest growth in mountain areas, while smaller values were predicted for valleys and floodplains compared to high-elevation areas. The authors showed that pure woods were predicted as the most influenced when compared with mixed stands, which are characterized by greater species richness. Therefore, a higher level of biodiversity and resilience reduces the threats of climate change on a forest. In southern Spain, Lopez-Tirado and Fidalgo (2016a) showed that under climate change effects, cork oak and gall oak underwent a drastic potential reduction. On the other hand, the Pyrenean oak and Algerian oak might find shelter at higher elevations. By exception, holm oak exhibited the opposite trend and was favored by projected global warming. This projection is rather adverse for biodiversity and oakdependent ecosystems. Acácio et al. (2016) showed that under increasing human pressure and forecasted climate change, evergreen oak forests will continue to decline and deciduous oak forests will be replaced by forests dominated by more xeric species. According to Sun et al. (2020), climate change affects the richness distribution pattern of oaks (Q. suber L.), and these species may migrate to higher altitudes or higher latitudes.
The high percentage of species lost is the reason for the higher turnover values in the mountainous areas, while other regions are mostly influenced by the high percentage of species gained associated with the northward shift of species. Predicting changes in the richness distribution pattern of Q. suber L. as a result of climate change can help to understand the biogeography of Q. suber L. and enact conservation strategies to minimize the impacts of climate change (Sun et al., 2020).
The cork production of Quercus suber L.

Climate-growth relationships
Cork oaks are well-known for cork production (Figure 3), a non-wood forest product of high industrial value that economically strengthens the otherwise fragile production regions (Oliveira et al., 2016). Cork is a biological material with a rather unique set of properties that gave it worldwide recognition as a wine sealant and insulator (Fortes et al., 2004;Pereira 2007Pereira , 2015. During summer droughts, Mediterranean oak transpiration is mainly supplied by direct groundwater and subsoil water uptake through deep roots (David et al., 2007(David et al., , 2013. On the other hand, Besson et al. (2014) reported that the accumulated tree transpiration matched the precipitation in spring at the stand level and suggested that the cork oak trees rely on water sources 6 from precipitation during the peak of the growing season that is made available to the many superficial roots. Temperature directly influences atmospheric evaporative demand and should affect cork productivity (Rambal et al., 2003). Cork growth recorded a high climate signal, with highly significant and coherent responses to the yearly climate-related sources of variation . The authors concluded that the high mean sensitivities and inter-series correlations found for cork ring chronologies combined with the significant variance explained by climate variables suggest that climate is likely one dominant signal that affects cork growth, but local environmental stresses can decisively affect this climate signal. Oliveira et al. (2016) showed that precipitation is the main factor positively influencing cork growth in the Coruche region, one of the major cork production areas in Portugal. Temperature had a positive influence on growth at the beginning of post-dormancy phellogen activity (until April) but exerted a negative influence during the growth period (from May to August). The authors suggest that cork growth is strongly hindered by drought conditions related to spring precipitation. Precipitation and temperature have been recently found to be clearly imprinted in cork rings (Oliveira et al., 1994;Caritat et al., 1996Caritat et al., , 2000Costa et al., 2003Costa et al., , 2015. Unlike the wood rings, which are often faint and indistinct (Natividade, 1950;Sousa et al., 2009), cork rings have clear boundaries and climate seems to be the dominant signal. Ghalem et al. (2018) suggested that cork growth encodes a climatic signal and drought-driven cork growth reduction is a threshold function of the precipitation-temperature ratio, and with an expected increase in drought occurrence under changing climate conditions, cork growth is likely to be similarly affected in Mediterranean regions. Lack of precipitation effectively reduces soil water availability and lowers groundwater levels and, thus, directly constrains the corkring widths that are formed in each growing season, with important implications for cork yield (Mendes et al., 2016). Leite et al. (2019) confirmed the outcomes of several studies on the negative effects of drought on cork growth (Caritat et al., 1996;Costa et al., 2016;Oliveira et al., 2016), reinforcing that spring and winter rain strongly influence phellogen activity. Ghalem et al. (2018) found precipitation-temperature index thresholds below which cork growth should be constrained in drier climatic conditions. The climate/growth relationships are remarkably strong in young trees (< 30 years) of Q. suber (Leal et al., 2008). A strong direct cork growth response has been found to be connected to the previous year's winter precipitation (Caritat et al., 1996(Caritat et al., , 2000Costa et al., 2001), together with a generally inverse relationship with monthly (summer) temperature (Caritat et al., 2000). Costa et al. (2016) showed that climate is likely the dominant signal affecting cork growth in mature trees under cork exploitation. However, a similar direct effect of precipitation on cork growth across Mediterranean environments may not be straightforward, because limiting local environmental stresses, related to (soil) water availability, would affect the strength of climate signals on cork rings. Most of the studies carried out on oak decline in Italy and Spain reported drought as the main driving factor (Gentilesca et al., 2017). The authors concluded that although this phenomenon seems generalized and strictly related to site-specific conditions, drought, and in general, ongoing climate warming is identified as the main threat to declining oak stands. Declining oak trees are often characterized by a reduction in ring width, mostly due to the absence or low production of late-wood in ring-porous species (Gentilesca et al., 2017).

Cork-ring growth
The average cork-ring widths are between 2.2 and 4.8 mm yr −1 in southwestern Portugal reported at the national (Portugal) level (Ferreira et al., 2000). In Algeria growth is between 0.8 mm yr −1 and 3.2 mm yr −1 (Dehane, 2012) (Table 1). In Spain, the cork ring width averages are between 2.2 and 4.8 mm yr -1 and ranged between 1.85 and 5.25 mm yr -1 Caritat et al., 2000). The annual average cork growth (2.43 mm) was in line with that mentioned in the European Research Project (CORKASSESS, 2001). In Spain (Andalusia and Catalonia), the annual growth was an average of 3.1 mm, ranging between 2.0 and 4.8 mm (Table 1). In Sardinia (Italy) and Corsica (France), the averages were 2.6 and 3.4 mm, respectively, ranging between values of 2.3 and 2.8 mm and 2.8 and 4.4 mm, accordingly (Chorana et al., 2019). Considering only the growth of cork, the accumulated radial increment of cork during the 10 years of the production cycle ranged 7 between 18 and 42 mm among individual trees, with an average of 26.50 mm Chorana et al. (2019). This value was lower than the values shown in previous studies (Ferreira et al., 2000;Aloui et al., 2006): in Portugal (33. Eight mm), Spain (31.7 mm), Morocco (31.5 mm), and Tunisia (28.9 mm). In addition, it was also noted that in Algeria, the thicknesses of cork in the northwest were much lower than those of the center (32.85 mm), and East (32.21 mm) of the country (Dehane and Ghefar, 2017). In a study of the average annual cork-ring width, Leite et al. (2019) showed that the mean value of the entire sample was 3.30 ± 1.44 mm. The values found for cork ring width by different authors are in the range of 2.8 to 3.6 mm (3.30 mm and 3.56 mm by Caritat et al. (2000); 3.8 mm by Costa et al. (2002); 3.5 mm by Pereira (2007); 3.3 mm by Oliveira et al. (2016); 3.6 mm, 3.1 mm, and 2.8 mm by Costa et al. (2016)). In a large cork sampling covering all the production regions in Portugal, a mean value of cork ring width of 3.6 mm was reported ranging from site means of 1.6 mm to 4.6 mm (Lauw et al., 2018).  Potential impact of climate change on the spatial distribution of Aleppo pine In one study, the potential distribution of three Mediterranean pine species (Pinus pinea L., Pinus halepensis Mill., and Pinus pinaster Aiton) in southern Spain in response to the forecasted increase in aridity was explored. Lopez-Tirado and Hidalgo (2016b) showed that the results predict a wider distribution of stone pine, which could expand its potential area in southern Spain. In contrast, Aleppo pine, especially cluster pine, would reduce their present distribution, with cluster pine occupying sites at higher altitudes. Pine forests are mainly concentrated in sites with poorly developed soils and under harsh climatic conditions (Cabezudo Artero and Perez Latorre, 2004). Largescale reforestation, however, has led to the introduction of pines in potential areas for oaks (Lopez-Tirado and Hidalgo, 2016b). The thermophilous nature of Aleppo pine will probably hinder migration to higher altitude sites (Lopez Gonzalez, 2007). Similar poleward and upward migration trends have also been projected by other authors (Parmesan and Yohen, 2003;Lenoir et al., 2008;Bertrand et al., 2012;Rabasa et al., 2013) and in addition, Aleppo pine is reported as the most suitable species for the most desertification prone zones (Lopez-Tirado and Hidalgo 2016b;Fons-Esteve and Paramo, 2003). Garah and Bentouati (2019) showed that the annual thermal amplitude followed by altitude appears to be the main factor affecting the spatiotemporal distribution of Aleppo pine in the study area. The authors showed that future predictions expect an extension of the areas classified as "moderately favorable" to the Aleppo pine. The authors concluded that in response to climate change, the Aleppo pine may display two contrasting tendencies: a progressive evolution in the north, and a regressive evolution in the south of the Aurès region (Algeria), as well as a displacement of suitable areas for Aleppo pine to the north. In this context, this result coincides with the findings of Vennetier et al. (2011), who predicted that current and future global warming will have a positive influence on Aleppo pine growth in wet sites and negative effects in dry sites. In such cases, the Aleppo pine would reduce its present distribution in the lowest altitude areas (Garah and Bentouati, 2019). This was also predicted by Vennetier et al. (2005) and observed by Tirado and Hidalgo (2016) for pine species located in the southern shore of the Mediterranean.
Aleppo pine expansion in cork oak habitat Biological invasions are dynamic processes whose extent and impact commonly increase over time, while the chances of effective intervention are reduced. The invasion of plants in mountainous areas is a particular challenge that has been increasing due to climate change and their use for tourism, among other reasons (Becker and Bugmann, 2001). Invasive species are recognized as one of the main causes of erosion of global biodiversity, and also threaten economies and human health (Millennium Ecosystem Assessment, 2005). Invasive species can profoundly affect ecological communities and processes, potentially leading to rapid or irreversible changes in trophic complexity and structure (Rodewald and Arcese, 2016). Historically, human activities have caused the accidental or planned dispersal of many species beyond their original distribution ranges, and this process is presently at its highest expression (Zalba et al., 2008). Invasion changes the diversity of ecosystems and affects the provision of certain services . In general, environmental factors that prevent or facilitate invasions are predictable, but other factors may be present that ultimately determine the success or failure of Pinaceae invasion, such as biotic interactions (Nunez et al., 2017). Mechanical control of pine trees requires a high level of diligence to be successful. All invasive pines need to be cleared from a site and all green 9 foliage (when trees are felled or cut) must be removed, or trees need to be completely removed from the soil (when hand pulled) to avoid regrowth (Ledgard, 2009). Understanding the dispersal ability of an invasive species is a major challenge in attempting to predict its future potential spread (Munzbergova et al., 2010).
The impacts of pine invasions have been widely recognized (Langdon et al., 2010). Pines show similar invasion processes throughout the world, involving two main stages: (a) immigration and establishment and (b) population growth/expansion (Richardson et al., 1994). Other impacts related to pine invasions include the reduction of structural diversity, the increase in biomass, the disruption of prevailing vegetation dynamics, and changes in nutrient cycling (Mack et al., 2000;Peterken, 2001;Richardson, 2001). The successful establishment of pines and their ability to invade natural ecosystems are influenced by climate, topography, and disturbance regimes such as fires and herbivory (de Villalobos et al., 2011). By definition, expansive species are species that increase their distribution and colonize new habitats in a geographical area where they are native (Prach and Wade, 1992;Pyšek et al., 2004). Although the effects of alien invasive plants on ecosystems and biodiversity lead to economic losses (Didham et al., 2005) and have a strong negative effect on global biodiversity (Woziwoda et al., 2014;Van der Meersch et al., 2020), expansion by indigenous species can have detrimental consequences on biodiversity through the reduction in ecosystem productivity, leading to a loss of various ecological goods and services, loss of genetic diversity, and decline in the abundance of populations of affected species (Pyšek et al., 2004;Simberloff, 2010;Kelbel and Adamcikova, 2011).
In Tunisia, we observed the phenomenon of Aleppo pine expansion in cork oak forests in several habitats (Figures 4 and 5). Intensive regeneration of Aleppo pine under cork oak was also observed (Figures 6   and 7). According to Nathan et al. (1999) and Nathan and Ne'eman (2004), mean relative humidity and maximum temperature are good predictors of the release of P. halepensis seeds in the Mediterranean basin. The earlier release of seeds in the high areas could result in higher growth rates in this habitat, at least during the initial stages of colonization of new sites (Lavi et al., 2005) and in greater risks of long-distance dispersal, considering that the peaks act as take-off sites for seeds (Ledgard, 1988;Buckley et al., 2005). de Villalobos and Schwerdt (2020) showed that the effects of grazing of feral horses promote the invasion of P. halepensis and have provided some key information that allow the elaboration of management recommendations to avoid the propagation of P. halepensis. Q. suber is a Mediterranean species, and while it is widespread around the world, these areas are within the natural distribution of P. halepensis, and thus, the danger for the species replacement from P. halepensis invasion is high. This is the same case for the invasion of P. pinea by P. halepensis mentioned by several authors (Athanasiadis and Gerasimidis, 1986;Barbero et al., 1998;Tapias et al., 2004). Richardson and Higgins (1998) showed that Aleppo pine invasion can result in substantial changes in the invaded ecosystems, a decrease in biodiversity and complexity of ecosystems Higgins et al., 1999;Richardson and Higgins, 1998), increased biomass and consequently in fire events , negative impact on water resources (le Maitre et al., 1996), changes in soil properties, and the availability and flow of nutrients in the ecosystem . Unlike early successional species, oaks might show considerable time lags in adjusting their ranges due to extremely slow migration rates (Delzon et al., 2013;Gerber et al., 2014). Current cork oak populations will therefore mostly rely on existing phenotypic plasticity, genetic diversity, and adaptation to respond to the expected scenario (Ghalambor et al., 2007;Aitken et al., 2008;Crispo, 2008;Alberto et al., 2013;Kremer et al., 2014;Bussotti et al., 2015). Several provenance tests demonstrated the occurrence of the species local variability at several phenotypic and ecophysiological traits (Varela, 2001;Aranda et al., 2005Aranda et al., , 2007Gandour et al., 2007;Ramırez-Valiente et al., 2010, 2011. A sustainable forest management strategy may reduce the potential impact of climate change on forest ecosystems (Pecchi et al., 2020). Silvicultural practices should be aimed at increasing species richness and favoring hardwoods currently growing as dominant species under conifer canopy, stimulating natural regeneration, gene flow, and supporting (spatial) migration processes (Pecchi et al., 2020). The models indicate that climate change will cause major shifts in the global distribution of cork oak woodlands. It is necessary to start addressing such shifts with appropriate mitigation and adaptation measures, as this ecosystem is particularly important for the rural economy and conservation of biodiversity in much of the western Mediterranean Basin (Bugalho et al., 2011). The following practices proposed by Gentilesca et al. (2017) could enhance the resilience of susceptible oak stands in drought-prone Mediterranean areas: (i) controlling the understory vegetation through targeted cuts and controlled grazing; (ii) reducing aboveground biomass through selective thinning; (iii) anticipating regeneration cuts to stimulate seed dispersal and promote sexual regeneration; and (iv) shortening the rotation period of coppice stands. 12 Figure 7. Concurrency between cork oak and Aleppo pine in Tunisia Management practices to limit Aleppo pine expansion into cork oak habitat Since the habitat of Q. suber is of high socio-ecological importance, several authors propose that appropriate silvicultural measures should aim to conserve species habitat. These measures should include : (i) gradual transformation of even-aged Q. suber stands to more diversified structures (uneven, multi-aged, and multistorey stands) in order to favor the natural regeneration of the species; (ii) grazing control; (iii) careful control of the understory maquis species through the improvement of the moisture conditions in order to favor seedling establishment and growth (Ciancio et al., 1986); and (vi) removal of P. halepensis trees in the Q. suber stands where the species has invaded, as early as possible, and preferably at the age before the trees reach fullreproduction. To limit the Aleppo pine expansion into cork oak habitat, it is necessary to integrate programs of development in the forest for sustainable development, and for territories and people and to introduce thinning to adapt to climate change to help reduce water competition in the forest, by controlling the understory vegetation, anticipating the regeneration cuts to stimulate seed dispersal, and promote sexual regeneration.

Conclusions
Climate change represents an important challenge for researchers studying forest silviculture and management. Several studies have shown that, under the effect of climate change, cork oak will disappear from a large area in the future and the rest will migrate to wet areas. In addition, the Aleppo pine is a plastic a species and will move into areas that are projected to have a humid climate. Other authors have shown that current cork oak habitat has been colonized in several places mainly by the Aleppo pine. In the future, it is possible that Aleppo pine will colonize cork oak habitat under the effect of climate change in North Africa and the Mediterranean basin, and thus, it is necessary to implement management and silviculture practices that help to alleviate the effects of climate change and thus the expansion of Aleppo pine into cork oak habitat as well.