Climatic Factors Influencing Aleppo Pine Sap Flow in Orographic Valleys Under Two Contrasting Mediterranean Climates

Abstract

Global climate change projections highlight the Mediterranean Basin as one of the most susceptible areas to the effects of intense and prolonged droughts, as well as increasing air temperatures. Accordingly, the productivity and survival of forests in this area will depend on their ability to resist and adapt to increasingly drier conditions. Different climatic conditions across the Mediterranean Basin could drive differences in forest functioning, requiring trees to acclimate to them. Sea breeze dynamics along orographic valleys can also influence climatic conditions, accentuating differences between inland and coastal forests. However, there is limited information on whether the climatic factors regulating tree transpiration in Aleppo pine forest in orographic valleys vary according to climate. This study aims to identify and compare the climatic factors that regulate tree transpiration along a gradient and determine the thresholds at which these factors affect transpiration rates. This study was carried out by means of sap flow gauges, since this technique is a key feature for quantifying and understanding tree transpiration. It was conducted in two Aleppo pine dry sub-humid forests (inland and coastal, 750 and 675 trees ha⁻¹, respectively) and in two pine semi-arid forests (inland and coastal, 600 and 400 trees ha⁻¹, respectively) in the western Mediterranean basin during January–November of 2021. No significant rainfall events or droughts were recorded during the period of study, indicating a standard climatic condition in these areas. The main findings demonstrated that the variability in sap flow could be attributed to the interaction between soil water content and vapour pressure deficit in all the forests studied. However, the highest threshold values of these climatic factors in relation to the increase or decrease in maximum sap flow (i.e., less sensitivity) were exhibited in semi-arid forests, highlighting the adaptability of Aleppo pine to more limiting climatic conditions. These findings are relevant for the consequences of the predicted increase in harsh climatic conditions and the balance among vapour pressure deficit, temperature and soil water availability. Future research will be essential to confirm forest acclimatisation in the transitional dry to semi-arid forest ecosystems predicted by global climate change projections, given their potential to strongly alter ecosystem function and water cycles.

1. Introduction

1.1. Previous Considerations and Objectives

Mediterranean forests have been historically conditioned by summer droughts, which represent a limitation on tree growth and survival due to soil water shortage [1,2,3]. These forests are also expected to experience rapid and drastic global climate change [4,5]. Global climate change predicts scenarios of severe and extended drought, as well as high temperature [6,7,8], which will compromise ecosystem functions and tree survival [9,10,11]. The adverse effects of both drought and temperature rise have been reported at the leaf level [12], in which alterations in biochemical and physiological processes were described (e.g., alterations in transpiration), even causing leaf necrosis due to damage in the membranes during periods of extremely high temperatures [13,14]. Among the impacts on ecosystem functioning, alterations in tree transpiration may affect not only ecosystem sustainability but also global water balance [15,16]. Understanding forest water consumption offers insights into forecasting future forest water balance and guiding forest management actions to guarantee the survival of the forest [17,18].

Forest water consumption is crucial for understanding forest functionality and it is central to establishing the ecosystem water balance [19,20]. Moreover, forest water consumption also plays an important role in trees’ ability to store atmospheric carbon and facilitate energy exchange [21,22]. Among other things, forest water consumption will condition the recycling of precipitation water into the atmosphere, the formation of storms, and the water available to aquifers [2,23]. Forest water consumption can be estimated from tree sap flow measurements, which measure the water movement through xylem tissues [24,25,26]. The sap flow method is commonly used due to the precision of the measurements, and the ability to record continuously at the individual tree level [27,28,29].

Several studies reported that tree sap flow is controlled by several factors: the vegetation structure, the physiological characteristics of tree individuals (stomatal control and tree hydraulic conductance), climatic factors such as rain, temperature, vapor pressure deficit, net solar radiation and soil water content, soil properties, and finally, geomorphological factors [30,31,32,33,34]. Regarding the climatic factors that control tree transpiration, soil water content (SWC) is one of the principal limiting factors for forest sap flow on daily time scales, especially in dry regions [35,36]. On the other hand, the atmospheric vapour pressure deficit (VPD) is another key climatic factor in determining sap flow [37,38]. It is widely recognized that sap flow increases because of the increase in VPD; nevertheless, when the VPD threshold is reached, sap flow decreases by stomatal regulation [39,40]. However, the current variability in climatic conditions makes it crucial to achieve a more comprehensive integration of the different climatic conditions determining Aleppo pine (Pinus halepensis Mill.) sap flow response. These factors will ascertain how sensitive Aleppo pine forests are in dry regions such as the Mediterranean basin.

Within the climatic factors influencing coastal areas, the circulation of the sea breeze is a significant phenomenon that exerts a considerable impact on both the atmosphere and the land surface [41]. This circulation is driven by the movement of air from the sea to inland areas during the day, and in the opposite direction during the night. During daylight hours, the temperature gradient between the water and land generates a sea breeze, which transports cooler, more humid air towards the coast. This results in a cooling effect on the land surface and an inrush of humid air, leading to a reduction in VPD [42,43,44]. In the Mediterranean, this phenomenon is especially notable due to the region’s high temperatures, making it a crucial component of climate regulation, especially in summer [24,45]. Moreover, the influence of the sea breeze on the local climate depends on the local orography and distance from the sea [46]. On the Mediterranean coastline, the sea breeze has been observed to penetrate more than 50 km inland [24,46]. This inland reach is facilitated by the presence of relatively gentle topographic features and the abundance of orographic corridors such as mountains and valleys; nonetheless, coastal forests are more exposed to sea breeze than inland forests due to their proximity to the sea [44,47]. Accordingly, studies on forests at different positions in a valley provide insights into the influence of climatic conditions, shaped in part by sea breeze circulation, on the water balance of forests. Such insights are valuable for understanding forest water dynamics, with implications for ecosystem resilience and forest management under changing climatic conditions.

Aleppo pine, the most representative species around both the eastern and western Mediterranean basins, shows a water-saving strategy compared to other Mediterranean forest species [48,49]. In this study, Aleppo pine transpiration and climatic factors were measured during the year 2021 in four Aleppo pine forests located in contrasted climatic locations along a coastal–inland air mass trajectory in two contrasted Mediterranean climates (dry sub-humid and semi-arid). The main aim of this study is to analyse climatic factors associated with climate areas (dry sub-humid vs. semi-arid) which affect the sensitivity of transpiration modulation in Aleppo pine. With this proposal, the specific objectives of the present study were the following: (i) to quantify the relative influence of SWC and VPD on sap flow dynamics across different climatic zones, (ii) to identify the climatic conditions at which maximum sap flow values are reached or they decreased, and plants need to control gas exchange, and (iii) how climatic factors associated with coastal–inland locations (sea-breeze effect) and ombroclimate conditions (semi-arid, dry sub-humid) may influence the former factors and thresholds. These findings may assist in implementing management strategies to address climate change and forest decline in Mediterranean forests. Furthermore, they may indicate whether the adaptations to dry environments are due to climate or sea-breeze exposure.

1.2. Review of the Impact of Climate Change on the State of Forests

The global trends predicted by climate change [5], coupled with extreme climatic events such as heatwaves and extreme droughts, create uncertainties about the functioning and survival of forests worldwide [50,51,52], with potentially diverse and even contrasting impacts depending on the climate and ecosystem type.

Numerous studies have documented tree mortality events directly caused by extreme events, as well as indirectly by their lingering effects [53,54,55]. Although these impacts are global, certain ecosystems are markedly affected, such as Mediterranean ecosystems [56,57,58]. In contrast, temperate and boreal forest ecosystems seem to be less significantly impacted, except in specific locations in Interior Alaska, Canadian prairies, and the southern boreal zone of central North America [51,59]. In particular, research has frequently highlighted the negative consequences of increasing drought frequency and severity on various aspects of Mediterranean forests. These include constraints on tree phenology and growth, carbon sequestration, and transpiration [60,61,62,63]. Consequently, the decline and mortality of Mediterranean forests have risen over the past decade and continue to do so [64,65,66].

In contrast, ecosystems where water scarcity is generally not the main limiting factor have experienced different consequences of climate change. For instance, boreal forests generally lack water stress and experience extended periods of permanent snow cover. Under normal conditions, snowmelt and rising temperatures trigger the growing season for these forests [67,68]. However, the warming induced by global climate change has facilitated the northward expansion of plant species composing boreal forests [69,70,71]. This shift in land use—marked by the replacement of wetlands or snowy tundra areas with shrub vegetation typical of tundra ecosystems by species characteristic of boreal forests—has implications for the water cycle [72,73,74].

In addition to the contrasting impacts observed in Mediterranean and boreal forests, other ecosystems, including tropical rainforests and temperate forests, also face significant alterations due to climate change, with notable impacts on their biodiversity and ecological functions [75,76].

2. Materials and Methods

2.1. Study Area

This study was performed at four experimental sites in the Valencian Community (southeast Spain) in 2021 (Figure 1a,b). Two sites were located in a dry sub-humid climate and the other two in a semi-arid climate. For each specified climate condition, the forest plots were selected following the most typical air-mass trajectory when well-developed sea-breezes form in coastal–inland valleys (Figure 1c). This spatial position along the valleys enables the observation of differences in relation to historical climatic conditions (dry sub-humid vs semi-arid), as well as variations in air climatic conditions (temperature and relative humidity) that are modulated by sea breeze circulation (inland vs. coastal).

The four forests studied are the consequences of extensive reforestation programs conducted during the last century in Europe, aged approximately 50–60 years old [77]. The forests exhibited similarities in terms of species composition, understory, soil type and vegetation structure, consistently falling within the normal range of variability in Mediterranean regions (Table 1). All locations were on gentle north-facing slopes to avoid differences due to terrain orientation.

The dry sub-humid forest was located in the Turia Valley, in Valencia Province. The dry sub-humid inland forest plot (DSH-I, hereafter) was located in the municipality of Aras de los Olmos (39°57′45″ N 1°8′31″ W) at ~100 km from the coastline, while the dry sub-humid coastal site (DSH-C) was located in the municipality of Lliria (39°36′45″ N, 0°37′59″ W), ~30 km away from the coastline (Figure 1d). The overstory was predominantly represented by conifers, especially by Aleppo pine. The understory at the DSH-I site mainly consisted of Juniperus oxycedrus, Salvia rosmarinus and Juniperus phoenicea; meanwhile, the DSH-C site’s understory featured Rhamnus alaternus, J. oxycedrus, Erica multiflora, Rhamnus lycioides, S. rosmarinus and Brachypodium retusum [78].

The semi-arid forests were located in Alicante Province, particularly in the Maigmó-Hoya de Castalla valley. The territory of the province of Alicante is characterized by its complex topography, where the valleys are not exactly perpendicular to the coastline, but rather curve as they rise in elevation [79]. Consequently, the semi-arid inland site (SA-I) was located in the municipality of Xixona (38°38′55.99″ N 0°27′2.42″ W) at ~50 km from the coastline in a particular valley which forms an L-shape from the coast to its head (Figure 1e). The semi-arid coastal site (SA-C) was located at ~25 km from the coastline in the municipality of Tibi (38°30′40.34″ N, 0°37′18.11″ W, Figure 1e). The overstory was dominated by Aleppo pine, representing 90% of the forested area [80]. The understory was represented by R. officinalis, Cistus albidus, B. retusum and Ulex parviflorus.

Soil type determination at dry sub-humid sites was described in [78], while at semi-arid sites, the soil type was determined in the laboratory according to [81]. For the determination, five samples were taken per area from spots below pine, below shrubs, and an open soil with minimal vegetation. Soil type was similar among sites, being fine loam and fine clay in dry sub-humid sites, and sandy loam in semi-arid sites (Table 1). The stoniness was measured by sieving the same samples into gravel (>2 mm diameter) and fine soil (≤2 mm diameter). Gravel values were notably varied, from 17.6% to 57.0% in DSH-C and SA-C, respectively (Table 1).

Tree density and tree diameter class of the whole plot were measured. Tree density varied among sites, with DSH-I presenting values of 750 trees ha⁻¹, while SA-C had values of 400 trees ha⁻¹ (Table 1). The diameter class of all trees in the plot also varied among sites (Figure S1), but the diameter class of the monitored pine individuals was similar (diameter class between 20 and 25, Table 1). Additionally, all the monitored pine individuals were dominant or co-dominant and were located within a large homogeneous forest stand. The sapwood area of the monitored pine was also measured, ranging from 181 to 372 cm² (Table 1).

Table 1. Geography, climate, soil and vegetation characteristics for each site. Aridity index calculated by Aridity Index (AI: Precipitation/(Temperature + 10)) of 1981–2010 period [82]. Tree density and basal area correspond to all the trees in the plots. Average and standard errors of diameter at breast height (DBH) and sapwood area were calculated only for the monitored individuals. Diametric class was represented as a graphic (Figure S1).

Ombroclimate	Dry Sub-Humid		Semi-Arid
Position in the Valley	Inland (DSH-I)	Coastal (DSH-C)	Inland (SA-I)	Coastal (SA-C)
Altitude (m)	1200	200	875	700
Distance to the sea along the valley’s trajectory (km)	100	30	50	25
Thermoclimate	Meso-Mediterranean	Thermo-Mediterranean	Meso-Mediterranean	Thermo-Mediterranean
AI	18.9 ± 0.2	17.3 ± 0.2	14.7 ± 0.2	13.5 ± 0.2
Soil type	Fine loam	Fine loam–fine clay	Sandy loam	Sandy loam
Gravels (>2 mm diameter; %)	31	18	39 ± 3	57 ± 2
Fine soil (≤2 mm diameter; %)	69	82	61 ± 3	43 ± 2
Tree density (trees ha−1)	750	675	600	400
Basal area per ground area (m2 ha−1)	36.72	17.95	18.06	14.64
DBH (m)	0.21 ± 0.01	0.17 ± 0.01	0.23 ± 0.01	0.24 ± 0.01
Sapwood area (cm2)	273.7 ± 41.0	181.2 ± 20.7	327.6 ± 30.1	372.1 ± 41.1

Table 1. Geography, climate, soil and vegetation characteristics for each site. Aridity index calculated by Aridity Index (AI: Precipitation/(Temperature + 10)) of 1981–2010 period [82]. Tree density and basal area correspond to all the trees in the plots. Average and standard errors of diameter at breast height (DBH) and sapwood area were calculated only for the monitored individuals. Diametric class was represented as a graphic (Figure S1).

Ombroclimate	Dry Sub-Humid		Semi-Arid
Position in the Valley	Inland (DSH-I)	Coastal (DSH-C)	Inland (SA-I)	Coastal (SA-C)
Altitude (m)	1200	200	875	700
Distance to the sea along the valley’s trajectory (km)	100	30	50	25
Thermoclimate	Meso-Mediterranean	Thermo-Mediterranean	Meso-Mediterranean	Thermo-Mediterranean
AI	18.9 ± 0.2	17.3 ± 0.2	14.7 ± 0.2	13.5 ± 0.2
Soil type	Fine loam	Fine loam–fine clay	Sandy loam	Sandy loam
Gravels (>2 mm diameter; %)	31	18	39 ± 3	57 ± 2
Fine soil (≤2 mm diameter; %)	69	82	61 ± 3	43 ± 2
Tree density (trees ha−1)	750	675	600	400
Basal area per ground area (m2 ha−1)	36.72	17.95	18.06	14.64
DBH (m)	0.21 ± 0.01	0.17 ± 0.01	0.23 ± 0.01	0.24 ± 0.01
Sapwood area (cm2)	273.7 ± 41.0	181.2 ± 20.7	327.6 ± 30.1	372.1 ± 41.1

2.2. Aleppo Pine Sap Flow

Aleppo pine sap flow was monitored with a 40 mm HRM probe constructed according to [83] and following the HRM theory based on [84]. The Heat Ratio Method requires two temperature probes and a heater, all located inside three hollow needles. Previous studies have evaluated the Heat Ratio method (HRM), a sap flow method for sap flow monitoring in Aleppo pine, which obtained accurate results [27,85,86].

The needles were perpendicularly inserted into the tree trunk of each tree at breast height after removing the bark to expose the outer surface of the sapwood. All needles were installed on the north side to avoid direct solar heating and shielded from precipitation and radiation with aluminium reflectors. The heater situated equidistantly between the temperature probes released a heat pulse. As the vertical separation between each of the probes and the heater was never exactly 0.6 cm, as the theory requires, the misalignment correction proposed in [85] and validated in [86] was applied (Table S1).

The sap flux velocity was recorded on a half-hourly basis by averaging one-second instantaneous measurements taken during a 41 s interval using the two temperature probes connected to a Campbell Sci data logger and an Analog Multiplexer (Campbell Scientific Ltd., Logan, UT, USA). The systems used 12-V batteries (Lead Acid Battery, 12 V 55 Ah, AGM, RS Pro, China). For each individual tree, sap flux velocity (Vs, cm h⁻¹) was scaled to the stand level multiplying by the conducting sapwood area to obtain pine sap flow (SF_half-hourly, cm cm² h⁻¹ or L h⁻¹ tree⁻¹). Sapwood was determined by visual inspection of the cores and by subtracting the area represented by the hardwood portion [87]. Sapwood represented ~80% of the basal area of pine trees. Sap flow probes were replaced only in case of disruption; it was assumed that probes are reliable for mid- to long-term measurements due to slow growth in Aleppo pine [88]. For each of the individual trees, the SF_half-hourly values during daylight hours were hourly averaged (SF_hourly, L h⁻¹ tree⁻¹). Daylight hours were estimated by means of the “maptools” R package using the latitude and longitude coordinates of the study sites. The sum of the resulting hourly sap flow was defined as the SF_day (L day⁻¹ tree⁻¹). Transpiration (T, L day⁻¹ m⁻² ground) was estimated as the mean of SF of monitored individuals per tree density (Table 1).

2.3. Climatic Data

Air temperature (°) and relative humidity (%) were recorded half-hourly using a register situated in an open area at each site (U23 Pro v2, Onset Computer Corporation, Bourne, MA, USA). VPD (kPa) was calculated every half-hour from air temperature and relative humidity [89]. Sea breeze circulation is considered one of the modulators of VPD and, consequently, of climatic conditions. Specifically, sea breeze circulation lowers temperature and increases relative humidity, resulting in a reduction in VPD [44].

SWC was determined by three probes in each plot, which were distributed at three different positions: (i) below pine, (ii) below shrubs, and (iii) in an area of open soil with minimal vegetation; representing the different plot conditions. The probes were installed 20–25 cm deep in the soil due to the Aleppo pine’s active root system being in the upper 40 cm of the soil [90]. Data were recorded in m³ m⁻³ on a half-hourly basis, and stored in battery-powered data-loggers. At the dry sub-humid sites, probes and data logger corresponded to the model SSND M005 and HOBO Micro Station (Onset Computer Corporation, Bourne, MA, USA), respectively; while for the semi-arid sites, the probes and data logger models chosen were ECH20 10HS METER Group (Pullman, WA, USA) and Em5b data logger DECAGON devices (Pullman, WA, USA), respectively. The slope in all the forests is similar and without steep inclinations (slope of 5–20°), so there was no significant influence of this on the soil water balance.

Precipitation (mm) was registered by a rain gauge situated in an open area with minimal vegetation cover in each forest. At the dry sub-humid sites, precipitation was sampled every minute by 0.2 mm-resolution rain gauges (RGR-M, Onset Computer Corporation, Bourne, MA, USA) and registered on a half-hourly basis as an accumulated value by the installed data logger (CR800 Campbell Scientific Inc., Logan, UT, USA). Meanwhile, at the semi-arid sites, for the SA-I site, precipitation values were obtained from meteorological open-access data at Ibi (Torretes/Font Roja, 38°38′11.40″ N, 0°32′18.24″ W; [91]) and the Climatology Laboratory of the University of Alicante (38°30′35.43″ N, 0°36′19.50″ W) in SA-I and SA-C, respectively.

2.4. Data Management and Analysis

The division of the year into its four distinct seasons occurs as follows: winter includes the months of January and February (Day of year (DOY): 1–59); spring spans from the months of March, April, and May (DOY: 59–151); summer occurs during the months of June, July, and August (DOY: 151–243); and autumn extends through the months of September, October, and November (DOY: 243–334). Throughout the year of 2021, data was recorded from January to November.

To comprehend the climatic factors that influence the Aleppo pine sap flow and compare these factors across sites, a linear mixed-effects model using the lme4 package and lmer function [92] was fitted in R [93] with SF_hourly as the response variable, and the climatic factors (VPD and SWC) as the continuous factors based on the characteristics of the data (

Table 2. Compilation of linear mixed-effects models applied in this study. Abbreviations: cumulative sap flow per hour (SF_hourly). Soil water content (SWC). Vapour pressure deficit (VPD). Hourly value at which the maximum sap flow is reached (SWC and VPD threshold). Fractional day of the year (DOYcont).

Response	Fixed	Random
SFhourly	VPD, SWC (hourly and continuous)	Trees, DOYcont
SWC threshold	Site (category)	Trees
VPD threshold	Site (category)	Trees

). Random effects were introduced by incorporating pine trees and a continuous variable of time (DOYcont, Equation (1)) separately. This approach allowed for the integration of variability among the monitored pine trees, and throughout the recorded period. Precipitation was not included in the model according to the parsimony principle because the prediction ability did not increase substantially. Marginal and conditional R² values of the linear mixed-effects model were calculated using the piecewiseSEM R package version 2.0 [94] to assess the proportion of variance explained by the fixed, and both the fixed and random effects [95].

$$\mathrm{DOYcont}=Day of the year+{\displaystyle \frac{Hour}{24}}+{\displaystyle \frac{Minute}{3600}}$$

(1)

In this study, the SWC value at which the maximum sap flow is reached is referred to as the SWC threshold. Variations in SWC threshold among forests could reflect differing acclimatization strategies based on climate and position along the air-mass trajectory. To determine the SWC threshold, a quadratic plateau model was designed using the stats package and nls function [96,97]. This model assumes a quadratic relationship between the maximum sap flow and SWC, with a maximum point at the SWC threshold value. Then, the resulting SWC threshold value for each individual was compared across sites using linear mixed-effects models. The SWC threshold was considered as the response variable, with the site as the categorical factor given the nature of the variable (Table 2). The variability among the monitored pine trees was integrated into the model as a random effect (Table 2). The analysis of the VPD threshold followed the same procedure, but aimed to obtain the VPD value from which the SF maximum decreases.

3. Results

3.1. Climatic Conditions and Pine Sap Flow Throughout the Year

Total precipitation was similar in all sites during 2021, from 351 mm year⁻¹ to 392 mm year⁻¹ (Figure 2a, Table S2). The mean annual total precipitation by climate was 366 mm year⁻¹ in both areas, which is below the average associated with a dry sub-humid climate (Table 1). Climatic variables such as SWC fluctuated seasonally, with the lowest SWC values during summer (Figure 2b). The dynamics of SWC were largely consistent across sites (Figure 2b). Nonetheless, daily mean SWC values were significantly higher in SA areas than in DSH areas (p-value < 0.05,

Table 3. Annual daily means of soil water content (SWC), air temperature and relative humidity, vapour pressure deficit (VPD), sap flow (SF) and transpiration (T) at the four study sites during the year 2021. Minimum and maximum values are respectively indicated in brackets. Different letters mean significant differences between sites at p-value < 0.05. n.s. means non-significant differences between sites.

	Site		Climate
	Inland	Coastal	Dry Sub-Humid	Semi-Arid
SWC (m3 m−3)	0.21 a	0.20 b	0.20 a	0.21 b
	(0.12, 0.33)	(0.12, 0.30)	(0.12, 0.31)	(0.12, 0.30)
Air Temperature (°C)	14.1 a	16.49 b	14.9 a	15.6 b
	(−4.5, 32.0)	(−0.1, 32.9)	(−4.5, 31.4)	(−0.1, 32.9)
Air Relative Humidity (%)	71.6 a	71.0 b	71.9 n.s.	70.7 n.s.
	(23.2, 100)	(25.8, 100)	(24.85, 100)	(23.2, 100)
VPD (kPa)	0.77 a	0.86 b	0.76 a	0.88 b
	(<0.1, 3.80)	(<0.1, 4.11)	(<0.1, 3.41)	(<0.1, 4.11)
SFday (L day−1 tree−1)	11.99 a	6.43 b	4.80 a	13.70 b
	(0, 53.59)	(0, 27.84)	(0, 14.01)	(0, 53.59)
Tday (L day−1 m−2 ground)	0.75 a	0.34 b	0.34 a	0.75 b
	(0, 3.22)	(0, 1.11)	(0, 0.95)	(0, 3.22)

) and higher in inland compared to coastal areas (p-value < 0.05, Table 3).

VPD also fluctuated seasonally, with the highest VPD values during summer (Figure 2c). Daily mean VPD values also showed similar dynamics among sites but were lower in the DSH climate relative to the SA, and in inland sites compared to coastal sites (p-value < 0.05, Table 3). Variations between climates and valley positions were also observed for air temperature. In contrast, relative humidity exhibited minor differences, observed only in relation to valley position (Table 3; see also Table S2, Figure S2 for specific differences among the four sites).

In terms of historical climatic conditions (from 1961 to 2021), SA-C exhibited the highest number of events with Standardized Precipitation Evapotranspiration Index (SPEI) values below −1.5, and it was also the site with the second highest number of events with SPEI values below −2 (Figure S3).

Water consumption variables such as SF_day and T_day also fluctuated seasonally and reached their lowest values during summer (Figure 2d,e), presenting dynamics quite similar between sites (Figure 2d,e). In terms of accumulated values, there is a variation in both SF_day and T_day among sites, with higher values in inland sites (p-value < 0.05, Table 3) and in SA climate compared to DSH (p-value < 0.05, Table 3). The specific differences between the four sites are shown in the Supplementary Materials (Table S2), where SA-I presented the significant highest SF_day and T_day values compared to the other areas.

3.2. SWC and VPD as Determinants of the Aleppo Pine Sap Flow

The SWC_hourly ranged approximately between 0.10 m³ m⁻³ and 0.32 m³ m⁻³ at all sites, with a slightly higher minimum of 0.17 m³ m⁻³ at SA-C (Figure 3). The VPD ranged from <0.1 kPa and 0.49 kPa in dry sub-humid sites (Figure 3a,b), while it occasionally reached values between 5 kPa and 6 kPa in semi-arid sites (Figure 3c,d). The results indicated that the higher the SWC, the higher the SF_hourly, always when combined with moderate evaporative air demand (VPD = 1–2 kPa, Figure 3). As a result, the interaction of VPD and SWC played a significant role in determining the Aleppo pine sap flow throughout the year in the four studied sites (

Table 4. Summary statistics of the half-hourly sap flow linear mixed-effects models in response to climatic factors for all sites. Marginal and conditional R² of the linear mixed-effects model indicated the proportion of variance explained by the fixed and by both the fixed and random effects. Interaction between variables was indicated by a cross (x). VPD and SWC were scaled by the “scale” R function to facilitate compression of the resulting coefficients. c.i. (2.5, 97.5) means a confidence interval from 2.5% to 97.5% of the interaction between VPD and SWC. Abbreviations: VPD, vapour pressure deficit; SWC, soil water content. (*) denotes p-value lower than 0.05.

	Dry Sub-Humid		Semi-Arid
Fixed Effects	Inland (DSH-I)	Coastal (DSH-C)	Inland (SA-I)	Coastal (SA-C)
Intercept	0.301 ± 0.054 *	0.450 ± 0.110 *	1.600 ± 0.133 *	0.541 ± 0.067 *
VPD	0.226 ± 0.003 *	0.170 ± 0.004 *	1.084 ± 0.010 *	0.367 ± 0.005 *
SWC	0.098 ± 0.003 *	0.200 ± 0.004 *	0.396 ± 0.010 *	0.412 ± 0.005 *
VPD x SWC	0.145 ± 0.002 *	0.168 ± 0.004 *	0.657 ± 0.001 *	0.287 ± 0.005 *
c.i (2.5, 97.5)	0.141, 0.147	0.160, 0.175	0.643, 0.670	0.278, 0.297
Random effects
ID Tree	0.134	0.270	0.265	0.023
DOYcont	0.121	0.270	0.534	0.055
Residual	0.100	0.251	0.336	0.060
R2 marginal	0.48	0.33	0.68	0.67
R2 conditional	0.88	0.76	0.92	0.91

). This positive interaction between VPD and SWC was consistently observed across all models, showing a high level of predictability concerning the climatic conditions (Table 4). The R² conditional values, which indicate the accuracy of the models, were always consistently above 0.76 (range 0.76–0.92; Table 4).

While the variability of SF_hourly was influenced by the interaction of VPD and SWC across all sites, the observed response patterns differed (Figure 3). The contribution of this interaction varied by site, with no significant differences observed for DSH-I and DSH-C (Table 4, confidence interval). The confidence interval revealed that the VPD × SWC interaction had the most significant impact on sap flow in SA-I, followed by SA-C, and then by DSH-I and DSH-C.

3.3. Climatic Threshold of Maximum Sap Flow

The relationship between maximum SF_half-hourly and climatic variables SWC and VPD displayed a bell-shaped curve across all sites (Figures S4 and S5). There were significant differences in the threshold values of the climatic variables determining the maximum values of SF_half-hourly between the sites, particularly in semi-arid sites (Figure 4; Tables S3 and S4). Specifically, the SWC threshold in the SA-C forest was significantly higher than those of the other sites, exceeding 0.25 m³ m⁻³, while the other sites averaged around 0.21 m³ m⁻³ (Figure 4a; Table S3). The VPD threshold value was significantly higher in semi-arid sites compared to dry sub-humid sites, measuring around 2.2 kPa as opposed to around 1.5 kPa in the dry sub-humid sites (Figure 4b; Table S4). No significant difference was observed in the VPD threshold value based on the distance to the sea along each valley’s trajectory.

4. Discussion

4.1. Climatic Factors Determining Aleppo Pine Sap Flow

The effect of climatic factors on tree transpiration is central for understanding the consequences of the predicted increase in intensive and prolonged droughts and evaporative demand in a global climate change context [5]. The highly dynamic daily pattern of sap flow observed in Aleppo pine during the study period follows the same trend of transpiration patterns in other semi-arid Aleppo pine forests [19,98]. This pattern reflected the water-saving strategy of Aleppo pine, with early stomatal closure and sap flow rates close to zero for SWC values around 0.1 m³ m⁻³. The early stomatal closure decreases water consumption, which not only affects the water balance but also decreases the needle water potential, which could help avoid damage to the hydraulic system (e.g., xylem cavitation). However, as several studies pointed out, this strategy could have implications for the carbon cycle and tree health because it reduces non-structural carbohydrate reserves and diminishes carbon dioxide fixation if drought conditions persist [12,99,100].

Our results indicated that the fluctuations in SF_hourly were associated with the interaction between VPD and SWC, showing a remarkably high model accuracy in both dry sub-humid and semi-arid sites. These results agree with previous research which highlighted the impact of these climatic factors on sap flow by regulating the stomatal activity in trees and other plants [40,101,102], and studies realized throughout the Mediterranean Basin [103,104,105]. In contrast, for Aleppo pine forests in the eastern Mediterranean Basin (Greece), SWC was not a limiting factor, and transpiration was primarily determined by VPD and solar radiation [106].

Moreover, our findings underscore the importance of considering both VPD and SWC in an integrated manner as a response to ecosystem abiotic characteristics, which has been reported on a global scale and also at the local scale in Aleppo pine forests [3,19,30,78,98,101], as well as in juvenile Aleppo pine individuals under laboratory-controlled conditions [86]. This behaviour aligns with the isohydric strategy described for the species, due to its strict stomatal control, which may also be enhanced by vulnerability to hydraulic failure (i.e., xylem embolism) associated with shallow root systems [105,107]. Under conditions of high or moderate water availability, VPD drives transpiration regulation up to a threshold where stomatal closure occurs [37,108]. However, under low soil water availability, this climatic factor becomes the primary determinant of transpiration. The integration of soil water availability with VPD has been also observed in western subalpine forests of P. contorta and P. flexilis, which displayed a curvilinear relationship between pine transpiration and VPD during periods of high soil water content (SWC ~ 0.35 m³ m⁻³; [109]). This pattern was also noted in pioneering studies of P. contorta forests in the United States [110,111,112]. Therefore, continued research is essential to comprehensively understand and prepare for potential shifts in forest ecosystems functioning in a context of global climate change.

4.2. Climatic Threshold for Maximum Aleppo Pine Sap Flow

Concerns are growing regarding the survival of pine forests in the face of climate change [55,113,114]. Specifically, changes in the sensitivity of transpiration to soil water content and evaporative demand may be crucial for responding to global climate change [115,116]. In this study, we found that sap flow in all the examined forests was regulated by the interaction of SWC and VPD. However, the threshold of these climatic conditions, which determine both the maximum sap flow and its subsequent decrease, varied among the different forests. Notably, semi-arid pine forests displayed the highest thresholds, which represent lower sensitivity to climatic conditions, which indicates the acclimation to lower water availability conditions in dry areas by regulation of water consumption. These findings agree with the plasticity of Aleppo pine in adapting to different climatic conditions [3,98,117]. Additionally, previous research has suggested that the range of climatic sensitivity may differ significantly between inter- and intra-species due to (i) distinct adaptive strategies and physiological mechanisms that govern stomatal functions [32,118], and (ii) the effects of current and past climatic conditions, such as high temperatures or drought frequency [10,119].

Particularly, it is worth noting that SA-C exhibits a significantly higher threshold compared to the other sites, but this different pattern was not evident when examining the daily values of SWC throughout the study period. The SA-C forest showed an SWC threshold value of 0.25, which surpassed the reported values for pine species and other species such as oak [35,120,121]. These findings emphasize strict control of transpiration at SA-C, promoting conservative use of soil water reserves [122,123].

Our results clearly state that climatic conditions at each site were influencing tree transpiration to different degrees according to VPD and SWC threshold. A general analysis of factors related to sea breeze, such as relative humidity or main wind directions, showed that both could influence coast sites due to increments in atmospheric water content [42,43,44]. However, our results are not clear in this sense, as no strong differences in relative humidity and VPD were observed between coastal and inland sites. This lack of effect could also be mediated by the fact that all sites were located close to the coast, where sea breezes were influential. Generally, higher VPD creates greater pressure on plants to transpire and match the increased evaporation rate from the stomata up to a specific VPD threshold [37,108,124]. However, the relationship between sap flow and VPD varies depending on tree species and the season [125,126]. Our findings indicate that semi-arid sites exhibited a higher VPD threshold than dry sub-humid sites. These results are consistent with previous studies that found a lower sensitivity of sap flow to VPD in drier and warmer sites compared to humid and mild sites [6,127,128]. In this study, the VPD threshold values in dry sub-humid sites were close to ~1.5 kPa. These values were close to the common behaviour reported up to 1 kPa in global studies [37], and the interval of sensitivity reported of 1–1.5 kPa agreed with studies in Aleppo pine forest [98,129]. Additionally, the recorded value of 2.2 kPa in semi-arid sites was similar to values recorded in other conifer species such as Pinus edulis and Juniperus monosperma, which were approximately 2 kPa [130]. Thus, the sensitivity to VPD would enable Aleppo pine trees in the arid site to sustain transpiration for a longer duration as environmental conditions become warmer and drier. This may be essential for maintaining their functionality, and avoiding xylem embolism, in the face of aridification projections for the upcoming decades, but could compromise their functioning and survival due to the reduction of non-structural carbohydrate reserves [12,49,131]. Thus, in addition to individual acclimatisation, managing Mediterranean pine forests to ensure their long-term survival is recommended [55,132], as this may be essential for maintaining their functionality in the face of climate change projections for the coming decades.

4.3. Limitations of the Study and Future Research Directions

This study analyses sap flow values in relation to climatic variables using data recorded at a half-hourly scale from January to November. This temporal resolution of sap flow technique, along with the climatic and position variety, not only strengthens the robustness of the study’s results but also allows for potential extrapolation to other Mediterranean pine forests. Nevertheless, it is important to note that the study is limited to a single year with standard precipitation, approximately 360 mm. More data from years with different environmental conditions would be useful for understanding the temporal projection of these results.

The study’s findings are based on measurements from four to six trees per plot recorded by sap flow technique. This provides a limited but sufficient intraspecific representation, though it is offset by the high resolution of the heat ratio sap flow technique [86] and the near-year-long temporal series recorded across the four studied pine forests. Although replication by individual and plot is sufficient, increasing replication by climate and valley position would be beneficial, as would adding a greater number of plots.

The four forests studied exhibit somewhat variable densities, from 400 to 750 trees ha⁻². This is the consequence of different climatic and abiotic conditions such as small orographic conditions, all plots were north-facing, small variations in soil types, and climate which was the main driver (semi-arid vs dry sub-humid; meso- and thermo-Mediterranean. Nevertheless, there are three fundamental pillars to support comparisons between stands of varying densities when the objective is to study climatic responses: (i) similar physiological conditions, (ii) homogenization of initial conditions within the dominant and codominant category, and (iii) species homogeneity. Studies [133,134] demonstrate the following: (i) Codominant P. halepensis trees exhibit similar physiological characteristics, making them comparable across sites, highlighting their representativeness for studying climatic responses without significant interference from intraspecific competition. (ii) Differences in stand density in P. halepensis have a minimal effect on dominant trees, as they have preferential access to resources such as water and light, even under high competition conditions. (iii) The adaptability of P. halepensis to diverse Mediterranean environments, attributed to its phenotypic plasticity and drought tolerance, is extensively documented. This ensures consistent physiological responses to drought and VPD gradients, regardless of stand density. Furthermore, its ability to maintain productivity under climatic stress underscores its use as a model species for studying climate responses.

Moreover, future studies could focus on a more detailed analysis of the influence of sea breezes entering the valleys and their quantitative impact on variations in climatic variables such as air humidity and temperature, as well as their relationship with vapor pressure deficit.

5. Conclusions

A deeper understanding of the Aleppo pine sap flow patterns in Mediterranean forests under varying climatic conditions is a key factor for knowing the dynamics of the water cycle under different climatic conditions and projected global climate change. Our findings in this study supported the integration of SWC and VPD as relevant climatic factors determining sap flow and therefore, tree transpiration, with remarkably high model accuracy. Pine trees presented different thresholds for obtaining and decreasing maximum sap flow in response to climatic conditions, with higher thresholds observed in the semi-arid forests, supporting the plasticity of Aleppo pine to adapt to more limiting climatic conditions [98,135]. Nevertheless, the exacerbation of aridity and temperature rise resulting from climate change may further influence the susceptibility of these pine forests, compromising the viability of pine stands situated in more semi-arid regions in the short term and the dry sub-humid regions in the long term. Therefore, forest management strategies are needed to promote the survival of both forest types in a future scenario [55,113,132]. However, the exact role of sea breeze in modulating climatic factors deserves deep research in the future that was outside of the scope of this study.