Adaptive Strategies of Seedlings of Four Mediterranean Co-Occurring Tree Species in Response to Light and Moderate Drought: A Nursery Approach

: In Mediterranean environments, light and water are clearly dominant ecological drivers of seedling growth and survival, and their interaction could modify productivity and forest composition. We examine the early response of seedlings of four co-occurring Mediterranean tree species ( Pinus pinea, Juniperus thurifera , Quercus ilex and Quercus faginea ) grown in the nursery to differing light and water availability conditions. Morphological (survival, height, biomass) and physiological (shoot midday water potential, net photosynthesis, transpiration rate, stomatal conductance, Fv/Fm, PSII and ETR) traits were studied, along with biomass allocation traits and drought resistance and plasticity indices. A signiﬁcant effect of both factors was mainly shown for growth and morphological traits, while survival, physiological traits and allometry were affected by either water availability or light, being more remarkable the effect of water availability over light. Drought severely limited Q. faginea ’s survival under both light intensities. The high plasticity of J. thurifera and Q. ilex seedlings to different light and water availability environments will confer these two species with an adaptive advantage in the early growth stages in comparison with P. pinea and Q. faginea seedlings. Thus, to maintain the codominance of the four species, silvicultural interventions should focus on the joint management of all of them.


Introduction
Light and water availability at stand level are environmental factors that greatly conditioned survival, growth and regeneration dynamics of forest species [1][2][3]. Their interactive effect occurs in a great variety of site conditions [4], and could modify productivity and forest composition [5]. The complex interactions and trade-offs between light and water availability have helped to determine the optimal regeneration niche of a wide range of forest species [6][7][8][9]. Understanding how such interactions affect seedling performance will be especially relevant in those regions that show a marked drought season and that are highly sensible to the effects of climate change [10]. This is the case of the Mediterranean continental environments, which show higher vulnerability to climate change than other regions in Europe [11].
Mediterranean continental forests are highly constrained by dry conditions and high solar loads, mainly in the summer (e.g., [12]). These forests are characterized by the presence of only two to four species commonly distributed in contiguous patches dominated by one or other of the species. Thus, the loss of a species in these forests may have serious consequences as regards the functional ecology, since the species lost cannot be replaced by another species with a similar function [13]. This is aggravated by the fact that seedling survival and establishment in these forests are major bottlenecks to species recruitment. Thus, seedling abundance by itself may not guarantee the regeneration that the studied species display contrasting strategies to cope with drought, as well as different light preferences. To test this hypothesis, we performed a greenhouse experiment and used a trait-based approach to characterize seedlings responses to the combined effect of different water and light conditions in terms of survival, growth, biomass allocation and physiology. We first determined seedling survival and growth of the four species when exposed to the combined effect of two soil water conditions (W+: field capacity; W−: medium soil moisture) and two light regimes (high and medium) for 12 months. Then, we further reduced the soil moisture in W−during two more months in order to mimic the effect of summer drought, and different physiological and biomass allocation traits were assessed.
The quantification of trait variation across different watering and light levels allowed us to determine phenotypic responses of seedlings. Specifically, our aims were to (1) quantify the variability in the functional response among the coexisting species in Mediterranean continental forests in central Spain; and (2) to study the morphological mechanisms that underlay the response of the species in environments with differing light and water availability, when seedlings are grown under identical, controlled environmental conditions. The knowledge of the combined effects of light and soil moisture on seedling survival and performance is important to define guidelines for planning regeneration of these species in the mixed forest stands where they coexist. The particular characteristics of the environmental gradient in which these species live should also be considered. Thus, this knowledge will be of great importance for assessing possible changes in the distribution of these species under climate change scenarios.

Plant Material and Methods
The study was conducted with seedlings of four Mediterranean tree species (P. pinea, J. thurifera, Q. ilex and Q. faginea) that coexist in the mixed forests within the limestone plain areas of the Spanish Northern Plateau. Seedlings were grown from seeds in the Central Nursery belonging to the Castilla-León regional government under standard growing conditions (FP containers 250 cc; soil was a mixture of peat, vermiculite and clay; seedlings were fertirrigated; oaks were grown in a greenhouse with no heating, while conifers were kept in the nursery under a shade cloth). Seeds were collected from 20-25 trees per species in a mixed natural forest in the Spanish Northern Plateau (5 • 15 W, 41 • 28 N). After one year, seedlings were transferred in June to a greenhouse where they were grown for the duration of the experiment (467 days). In this case, 192 seedlings (48 seedlings per species) were transplanted to 3-L pots filled with a mixture of peat and vermiculite (3:1, v:v). Seedlings were kept at field capacity for one month inside the greenhouse (26.0 • C ± 2.4 • C day/20.7 • C ± 1.4 • C night, 52 ± 10 per cent humidity, natural photoperiod) prior to initiating the experiment. Heights at the moment of initiating the experiment were 24.4 ± 1.8 cm in Q. ilex, 36.2 ± 0.8 cm in P. pinea, 19.7 ± 1.0 cm in J. thurifera and 27.5 ± 2.6 cm in Q. faginea.
After the 1-month acclimation period, seedlings were arranged in a split-plot experiment with two factors (light and water), each with two levels, and the treatments were replicated in 12 blocks. Seedlings were randomly divided into two groups, according to two light environments: (L+) 100% sunlight inside the greenhouse and (L−) 40% sunlight. Seedlings were grown under 12 open metal frames (50 × 80 × 150 cm, 6 blocks per light environment). Six of these metal frames had no shade (treatment L+) while the other six metal frames were covered with a neutral shade white cloth to create the L− environment. The average photosynthetic flux density under each light environment during a sunny day was 25.0 mol m −2 day −1 in L+ and 7.8 mol m −2 day −1 in L−. Light and temperature conditions throughout the experiment are shown in Figure 1.

Long-Term Drought Experiment
A long-term drought experiment was set up from day 0 to day 394 to asses of moderate drought on plant survival and growth. For this purpose, seedling under each light treatment were divided into two groups and assigned to a w ment (16 seedlings per block: 2 plants per species and water regime under eac vironment) ( Figure 2): (W+): well-watered seedlings, soil humidity was mainta

Long-Term Drought Experiment
A long-term drought experiment was set up from day 0 to day 394 to assess the effect of moderate drought on plant survival and growth. For this purpose, seedlings growing under each light treatment were divided into two groups and assigned to a water In day 394, plants in W− treatment were subjected to a two-month (up to day 467) severe drought experiment, aiming to mimic artificially the effect of summer drought. For this purpose, soil moisture in the W− treatment was further reduced for two months until reaching 3.5 ± 0.9%. During this period, plants under W+ treatment were maintained on similar well-watered conditions (25.6 ± 0.6%).
During the whole experiment, seedlings were irrigated individually and soil-water moisture was monitored for each seedling every week using a portable time domain reflectometer (TDR) equipped with three 16 cm probes (TRIME FM 3, IMKO GmbH, Ettlinger, Germany) ( Figure 1). Analysis of variance of soil water moisture at every measuring date indicated no significant differences (p-value > 0.05) between light levels for seedlings in both watering treatments.

Plant Growth and Survival in the Long Term
During the growing period, seedling survival was monitored weekly. Plant height was measured nine times (day 0, day 55, day 96, day 145, day 225, day 328, day 344, day 394 and day 467). Relative shoot growth (RSG) was calculated as Y = [log (x2) − log(x1)]/[t2 − t1], where Y is RSG, x2 and x1 represent the final and initial height and t2 − t1 is the time elapsed between these two measurements.

Physiological Responses to Severe Drought Experiment
In day 394, at the beginning of the severe drought experiment, a random subset of five plants per species and watering x light environment combination was selected to determine morphological and physiological resistance to severe summer drought.
Chlorophyll (Chl) a fluorescence, gas-exchange parameters and water status were measured three times, at the beginning, middle and end of the two months length severedrought experiment, on a 5-7cm long branch. The three measurements were made in the In day 394, plants in W− treatment were subjected to a two-month (up to day 467) severe drought experiment, aiming to mimic artificially the effect of summer drought. For this purpose, soil moisture in the W− treatment was further reduced for two months until reaching 3.5 ± 0.9%. During this period, plants under W+ treatment were maintained on similar well-watered conditions (25.6 ± 0.6%).
During the whole experiment, seedlings were irrigated individually and soil-water moisture was monitored for each seedling every week using a portable time domain reflectometer (TDR) equipped with three 16 cm probes (TRIME FM 3, IMKO GmbH, Ettlinger, Germany) ( Figure 1). Analysis of variance of soil water moisture at every measuring date indicated no significant differences (p-value > 0.05) between light levels for seedlings in both watering treatments.

Plant Growth and Survival in the Long Term
During the growing period, seedling survival was monitored weekly. Plant height was measured nine times (day 0, day 55, day 96, day 145, day 225, day 328, day 344, day 394 and day 467). Relative shoot growth (RSG) was calculated as Y = [log (x2) − log(x1)]/[t2 − t1], where Y is RSG, x2 and x1 represent the final and initial height and t2 − t1 is the time elapsed between these two measurements.

Physiological Responses to Severe Drought Experiment
In day 394, at the beginning of the severe drought experiment, a random subset of five plants per species and watering x light environment combination was selected to determine morphological and physiological resistance to severe summer drought.
Chlorophyll (Chl) a fluorescence, gas-exchange parameters and water status were measured three times, at the beginning, middle and end of the two months length severedrought experiment, on a 5-7cm long branch. The three measurements were made in the same five seedlings per treatment and species throughout the drought experiment. A portable infrared gas analyser (IRGA LCPro+ Analytical Development Corporation, BioScientific Ltd., Hoddesdon, UK) was used to measure net photosynthesis (An), transpiration rate (E) and stomatal conductance (gs). Gas exchange parameters were measured in the morning at maximum irradiance light of 1500 µmol (photon) m −2 s −1 . Temperature was set at 25 ± 2 • C and CO 2 at 400 ± 10 ppm, vapour pressure deficit inside the leaf chamber was 10.1 ± 0.2 mbar. A leaf chamber with broad leaf jaws was used for the two oaks, while the conifer jaws were used for the two conifers. Photosynthesis values were expressed on a leaf area basis. Chl a fluorescence was measured in the same twig with a portable pulse-modulated fluorometer (FMS2, Hansatech Instruments Ltd., Norkfold, UK). For each measuring date, two sets of measurements were recorded, dark-adapted fluorescence (leaves were acclimated to dark for 30 min with dark-adaption leaf clips) and morning light fluorescence, thus measuring the maximum photochemical efficiency of photosystem II (Fv/Fm), the yield of photochemistry in PSII (ΦPSII) and the apparent photosynthetic electron transport rate (ETR). Shoot midday water potential was measured using a pressure chamber (Scholander PMS 1000, Instruments Co., Corvallis, OR, USA).
At the end of the experiment different indices of plasticity and drought resistance were calculated. An index of phenotypic plasticity ranging from 0 to 1 was calculated for each morphological (PIm), physiological (PIph)and allometry (PIa) traits and species as the difference between the maximum and the minimum mean values between the four treatment combinations (L+W+, L+W-, L-W+, and L-W-) divided by the maximum mean value [28]. This index estimates the level of response of the species to the treatment combinations. Once these indices were calculated for all traits and species, an average phenotypic plasticity index for morphological traits (dry weights and growth), allometry (biomass ratios) and physiological traits (gas exchange, fluorescence and water potential) was computed for each species.
We quantified the species' drought resistance (D) based on [29] in the dry treatment (W−) relative to the irrigated treatment (W+) for each morphological, physiological and allometry traits, plus survival, as follows: For gas exchange parameters and water potential, K(W−) and K(W+) represent the final values of the measured traits after the two months reduction in soil moisture relative to values at the beginning of the drought period in the dry (W−) and wet (W+) treatment, respectively. For the other traits, K(W−) and K(W+) represent the final values of the measured traits relative to values at the beginning of the experiment, under the dry (W−) and wet (W+) treatment. An average drought resistance index for morphological traits (dry weights and growth) (Dm), allometry (Da) physiological traits (gas exchange and water potential) (Dph) and survival (Ds) was calculated.

Data Analysis
The dataset was checked for normality, homoscedasticity and outlier detection A logistic function was fitted to study the effect of water availability and light on survival over time for the different species. A repeated measures analysis of variance was used to analyze the effect of light and water availability on height over time for each species. Height at the moment of transplanting was considered as a covariate. To simultaneously detect the effects of light and water availability for each species and their interaction on the different morphological, physiological and allometry traits, a two-way ANOVA was performed. Tukey tests were performed for comparison of means. Finally, differences between species in terms of plasticity and drought resistance indices were evaluated using a one-way ANOVA. The SAS software was used for all the analyses.

Plant Survival and Growth in the Long-Term
The reduction in water availability (treatment W−) significantly decreased (p-value < 0.0001) seedling survival in Q. faginea seedlings (survival of 29% at the end of the experiment). Q. faginea seedlings grown under high light environment (L+) died even under moderate rates of drought, while seedlings under shaded environments (L−) only died after they were submitted to the 2-months severe drought experiment. Survival rates of the other three species were high (100% in P. pinea, 98% in J. thurifera and 94% in Q. ilex), and were not significantly affected by the treatment (Figure 3).
Forests 2022, 13, x FOR PEER REVIEW 7 of 17 moderate rates of drought, while seedlings under shaded environments (L−) only died after they were submitted to the 2-months severe drought experiment. Survival rates of the other three species were high (100% in P. pinea, 98% in J. thurifera and 94% in Q. ilex), and were not significantly affected by the treatment (Figure 3). Results of the repeated measures ANOVA showed a significant effect of light and water treatments over time on height (p-value = 0.0008). At the end of the experiment, seedlings grown under shade conditions and high water availability (W+L−) were the tallest for all species except for Q. faginea, which was tallest at full light conditions and high Results of the repeated measures ANOVA showed a significant effect of light and water treatments over time on height (p-value = 0.0008). At the end of the experiment, seedlings grown under shade conditions and high water availability (W+L−) were the tallest for all species except for Q. faginea, which was tallest at full light conditions and high water availability (W+L+) (Figure 3). A head start in early growth of seedlings from the large-seeded species (both oak species and P. pinea) was maintained throughout the experimental period, in contrast to the slow-growing pattern observed in J. thurifera. For instance, the effect of treatments on this species was only apparent almost a year after the beginning of the experiment (day 328) (Figure 3).

Morphology and Allometry Traits
Two-way ANOVA over different biomass and allometry traits revealed significant effects of water and light availability depending on the species. Light and water availability significantly affected relative shoot growth (RSG) showing higher relative shoot growth (RSG) values under moderate shade and high water availability (W+L−), except for Q. faginea (Table S1). Total dry weight significantly increased with high light availability for J. thurifera and Q. ilex, and with high water availability for Q. faginea. Shoot, root and total biomass significantly increased with high water availability and high light (W+L+) in the two conifers, while biomass traits were affected either by light (Q. ilex) or by water availability (Q. faginea) in the oak species. Significant interactions between light and water availability were observed in P. pinea for three biomass traits measured, indicating that the positive effect of high light was only shown under high water availability conditions. Seedlings showed greater investment in shoots than in roots under low light environments (L−) and high water availability (W+) (Table S1), although significant differences were only detected for P. pinea and Q. ilex (light) and J. thurifera (water availability). Irrespective of treatment, the two oak species showed greater investment in roots than in shoots compared to conifers (S/R, p-value < 0.0001, results not shown). For instance, 66% of the total biomass in the two oaks was allocated to roots, while for the conifers this figure was only 38%.
Average leaf mass fraction (LMF) decreased significantly with high irradiance (L+) in Q. ilex and Q. faginea; while it increased with high water availability (W+) in the two conifers ( Figure 4, Table S1). Stem mass fraction (SMF) decreased significantly in the two oak species with high light (L+), while light had no effect on SMF in the two conifers. Root mass fraction (RMF) increased very consistently under high light (L+) in P. pinea and Q. ilex. In P. pinea, SMF increased at the expense of RMF. In the case of high water availability (W+), on average, allocation shifted from shoot to root, except in Q. faginea ( Figure 4).
Higher specific leaf area (SLA) was related to moderate shade (L-) in Q. faginea and J. thurifera (Table S1). For all treatment combinations, Q. faginea had higher SLA than the three evergreen species (the two conifers and Q. ilex), while the lowest SLA was recorded in J. thurifera seedlings.

Physiological Responses to Severe Drought
During the 2-months severe drought period, water potential and gas exchange parameters were mainly affected by water availability (higher values in W+ treatment), with no effect of light or the interaction light x water availability. Meanwhile, Fv/Fm changed with the light treatment (higher values for all species in L− treatment) (Table S2).
Water potential decreased significantly over the 2-months severe drought period in the water-stressed seedlings of P. pinea (p-value = 0.0068), J.thurifera (p-vale = 0.0064) and Q. faginea (p-value = 0.0095). At the end of this period, water potential for all species was significantly lower in seedlings submitted to water stress (W− compared to well-watered seedlings (W+) (Figure 5). At this moment, values ranged between −0.9 and −1.1 MPa in well-watered seedlings (W+) and between −1.5 and −1.9 MPa in water stressed seedlings (W−. Higher water potential was measured in water stressed seedlings (W−) in P. pinea (−1.5 MPa), compared to Q. faginea (−1.9 MPa), with intermediate values of −1.6 MPa for J. thurifera and −1.8 MPa for Q. ilex, although non-significant differences between species were detected (p-value = 0.1466).
Net photosynthesis (An) of water-stressed seedlings significantly decreased over the 2-months severe drought period in all species except J. thurifera that maintained a relatively constant An (9.2 μmol CO2 m −2 s −1 ). Especially significant was the decrease shown

Physiological Responses to Severe Drought
During the 2-months severe drought period, water potential and gas exchange parameters were mainly affected by water availability (higher values in W+ treatment), with no effect of light or the interaction light x water availability. Meanwhile, Fv/Fm changed with the light treatment (higher values for all species in L− treatment) (Table S2).
Water potential decreased significantly over the 2-months severe drought period in the water-stressed seedlings of P. pinea (p-value = 0.0068), J.thurifera (p-vale = 0.0064) and Q. faginea (p-value = 0.0095). At the end of this period, water potential for all species was significantly lower in seedlings submitted to water stress (W− compared to well-watered seedlings (W+) (Figure 5). At this moment, values ranged between −0.9 and −1.1 MPa in well-watered seedlings (W+) and between −1.5 and −1.9 MPa in water stressed seedlings (W−. Higher water potential was measured in water stressed seedlings (W−) in P. pinea (−1.5 MPa), compared to Q. faginea (−1.9 MPa), with intermediate values of −1.6 MPa for J. thurifera and −1.8 MPa for Q. ilex, although non-significant differences between species were detected (p-value = 0.1466).
Net photosynthesis (An) of water-stressed seedlings significantly decreased over the 2-months severe drought period in all species except J. thurifera that maintained a relatively constant An (9.2 µmol CO 2 m −2 s −1 ). Especially significant was the decrease shown by P. pinea and Q. faginea water-stressed seedlings over the 2 months, compared to well-watered seedlings. Meanwhile, Q. ilex was able to maintain a significantly higher An under water stress at the end of the period, compared to the other three species (p-value = 0.0049) ( Figure 5).  Table S2 for statistical significance. Transpiration (E) and stomatal conductance (gs) of water-stressed seedlings decreased over the 2-months drought period in water-stressed seedlings of P. pinea and Q. faginea. Meanwhile, J.thurifera and Q. ilex maintained relatively constant values (Table S2) (E values: 3.3 mmol H 2 0 m −2 s −1 in J. thurifera, 6.7 mmol mmol H 2 0 m −2 s −1 in Q. ilex; gs values: 0.19 µmol CO 2 m −2 s −1 in J. thurifera, 0.67 µmol CO 2 m −2 s −1 in Q. ilex). Significantly higher transpiration and stomatal conductance were recorded in waterstressed Q. ilex seedlings compared to the other species along the 2-months drought period (p-value < 0.05). Similar to photosynthesis, after the 2-months period, transpiration was significantly lower in seedlings submitted to water stress (W−) compared to well-watered seedlings (W+), except for J. thurifera. Regarding stomatal conductance, significant differences were shown between the water availability treatments in P. pinea (p-value = 0.0022) and Q. faginea (p-value = 0.0002) ( Figure 5).

Drought Resistance and Trait Plasticity
At the end of the experiment significant differences in the drought resistance indices between species were observed ( Table 1). Significant higher drought resistance index for physiological traits was shown in J. thurifera compared to P. pinea and Q. faginea (p-value = 0.0474). Regarding drought resistance index for morphological traits, higher drought resistance indices were shown in Q. ilex compared with the other three species (p-value < 0.0001). The lowest drought resistance in terms of survival was recorded in Q. faginea (p-value < 0.0001). No significant differences between species were shown for drought resistance in terms of allometry. Differences between species were significant for the phenotypic plasticity indices related to allometry (p-value = 0.0005), thus Q. faginea showed higher allometric plasticity than the two conifers, while no significant difference between species were observed on neither physiological nor morphological plasticity indices (Table 1).

Discussion
Seedlings of the four Mediterranean tree species studied developed different mechanisms of tolerance and avoidance, together with different functional and morphological traits to deal with the interactive effect of water shortage and light availability during the early phases of growth. Such interactive effect of water availability and light was both trait-and species-dependent. Thus, a significant interactive effect was mainly shown for growth and morphology traits, while survival, physiological traits and allometry were affected either by water availability or by light depending on the trait considered but being more remarkable the effect of water availability over light. The most outstanding species-dependent effect was that of drought on Q. faginea's survival, as drought severely limited the survival of the species under both light intensities (Table 2). Table 2. Schematic representation of the main significant effects of treatments (water availability (W) in blue, light (L) in yellow, water availability and light (W,L) in green, no effect in white) on the different traits studied on the four species, at the end of the experiment.

Q. ilex
moisture after isolated summer rainfall events [36]. This is an advantage for Q. ilex, as the typical sandy soils at the Northern Plateau suffer a rapid water loss due to drainage, so the optimal moisture conditions are only present for a few days in spring and autumn, just after the rainfall events. In these drought-prone environments, the multistratified root system of P. pinea [37] may also confer a competitive advantage to this species, although it does not benefit from sporadic summer rainfall events to the same degree as Q. ilex [38]. This fact, together with its higher plasticity to morphological traits, its active physiological performance during drought and its wider thermal amplitude, must be seen as an advantage for Q. ilex in mixed pine-oak-juniper stands in drought-prone environments. Table 2. Schematic representation of the main significant effects of treatments (water availability (W) in blue, light (L) in yellow, water availability and light (W,L) in green, no effect in white) on the different traits studied on the four species, at the end of the experiment. Regarding Q. faginea, their seedlings showed low gas exchange activity as drought conditions aggravated, even endangering its survival; and such effects were independent on light conditions. For instance, Q. faginea showed the lowest drought resistance to drought in terms of survival. Similar to Q. faginea, P. pinea showed low gas exchange activity under drought conditions, irrespective of the light environment. However, P. pinea was able to maintain the highest water potential after the 2-months drought period, while Q. faginea's water potential decreased the most. eedlings growth, morphology and allometry in Q. faginea and P. pinea were also negatively affected by drought. Thus, seedlings of P. pinea and Q. faginea seemed to show a higher performance in terms of growth, morphology, allometry and gas exchange traits Q. faginea moisture after isolated summer rainfall events [36]. This is an advantage for Q. ilex, as the typical sandy soils at the Northern Plateau suffer a rapid water loss due to drainage, so the optimal moisture conditions are only present for a few days in spring and autumn, just after the rainfall events. In these drought-prone environments, the multistratified root system of P. pinea [37] may also confer a competitive advantage to this species, although it does not benefit from sporadic summer rainfall events to the same degree as Q. ilex [38]. This fact, together with its higher plasticity to morphological traits, its active physiological performance during drought and its wider thermal amplitude, must be seen as an advantage for Q. ilex in mixed pine-oak-juniper stands in drought-prone environments. Table 2. Schematic representation of the main significant effects of treatments (water availability (W) in blue, light (L) in yellow, water availability and light (W,L) in green, no effect in white) on the different traits studied on the four species, at the end of the experiment. Regarding Q. faginea, their seedlings showed low gas exchange activity as drought conditions aggravated, even endangering its survival; and such effects were independent on light conditions. For instance, Q. faginea showed the lowest drought resistance to drought in terms of survival. Similar to Q. faginea, P. pinea showed low gas exchange activity under drought conditions, irrespective of the light environment. However, P. pinea was able to maintain the highest water potential after the 2-months drought period, while Q. faginea's water potential decreased the most. eedlings growth, morphology and allometry in Q. faginea and P. pinea were also negatively affected by drought. Thus, seedlings of P. pinea and Q. faginea seemed to show a higher performance in terms of growth, morphology, allometry and gas exchange traits P. pinea moisture after isolated summer rainfall events [36]. This is an advantage for Q. ilex, as the typical sandy soils at the Northern Plateau suffer a rapid water loss due to drainage, so the optimal moisture conditions are only present for a few days in spring and autumn, just after the rainfall events. In these drought-prone environments, the multistratified root system of P. pinea [37] may also confer a competitive advantage to this species, although it does not benefit from sporadic summer rainfall events to the same degree as Q. ilex [38]. This fact, together with its higher plasticity to morphological traits, its active physiological performance during drought and its wider thermal amplitude, must be seen as an advantage for Q. ilex in mixed pine-oak-juniper stands in drought-prone environments. Table 2. Schematic representation of the main significant effects of treatments (water availability (W) in blue, light (L) in yellow, water availability and light (W,L) in green, no effect in white) on the different traits studied on the four species, at the end of the experiment. Regarding Q. faginea, their seedlings showed low gas exchange activity as drought conditions aggravated, even endangering its survival; and such effects were independent on light conditions. For instance, Q. faginea showed the lowest drought resistance to drought in terms of survival. Similar to Q. faginea, P. pinea showed low gas exchange activity under drought conditions, irrespective of the light environment. However, P. pinea was able to maintain the highest water potential after the 2-months drought period, while Q. faginea's water potential decreased the most. eedlings growth, morphology and allometry in Q. faginea and P. pinea were also negatively affected by drought. Thus, seedlings of P. pinea and Q. faginea seemed to show a higher performance in terms of growth, morphology, allometry and gas exchange traits J. thurifera moisture after isolated summer rainfall events [36]. This is an advantage for Q. ilex, as the typical sandy soils at the Northern Plateau suffer a rapid water loss due to drainage, so the optimal moisture conditions are only present for a few days in spring and autumn, just after the rainfall events. In these drought-prone environments, the multistratified root system of P. pinea [37] may also confer a competitive advantage to this species, although it does not benefit from sporadic summer rainfall events to the same degree as Q. ilex [38]. This fact, together with its higher plasticity to morphological traits, its active physiological performance during drought and its wider thermal amplitude, must be seen as an advantage for Q. ilex in mixed pine-oak-juniper stands in drought-prone environments. Table 2. Schematic representation of the main significant effects of treatments (water availability (W) in blue, light (L) in yellow, water availability and light (W,L) in green, no effect in white) on the different traits studied on the four species, at the end of the experiment. Regarding Q. faginea, their seedlings showed low gas exchange activity as drought conditions aggravated, even endangering its survival; and such effects were independent on light conditions. For instance, Q. faginea showed the lowest drought resistance to drought in terms of survival. Similar to Q. faginea, P. pinea showed low gas exchange activity under drought conditions, irrespective of the light environment. However, P. pinea was able to maintain the highest water potential after the 2-months drought period, while Q. faginea's water potential decreased the most. eedlings growth, morphology and allometry in Q. faginea and P. pinea were also negatively affected by drought. Thus, seedlings of P. pinea and Q. faginea seemed to show a higher performance in terms of growth, morphology, allometry and gas exchange traits In the long-term, the four species showed higher shoot growth under moderate shade and high water availability (W+L−). This result is consistent with theories and empirical data on habitat-related traits, which suggest that species adapted to high-resource environments show higher growth, high potential rates of resource capture and high tissue metabolic rates (and associated morphological traits) relative to species characteristic of low-resource environments (e.g., [30]).
The higher plasticity of oaks (mainly Q. faginea) in biomass partitioning patterns (shoot to root ratio, leaf, stem and root mass fractions, and specific leaf area) constitute one of the most outstanding divergences between oaks and conifers. Under moderate shade and high water availability conditions, a common pattern of biomass allocation to stems and leaves at the expense of roots in the two oak species [31] suggests a relationship between shoot mass fraction (SMF) and the leaf habit [32]. The small-seeded J. thurifera and the resprouter Q. ilex were able to maintain relatively constant values (or with small decreases) of gas exchange parameters along the 2-months drought period, irrespective of the light environment; meanwhile, P. pinea and Q. faginea suffered remarkable decreases in these measured traits. Additionally, the relatively low decrease in gas exchange parameters in J. thurifera and Q. ilex ran in parallel to intermediate midday water potentials in these two species. Consequently, both species, but mainly J. thurifera, showed higher drought resistance indices in the physiological traits compared to P. pinea and Q. faginea. In the case of J. thurifera, such effect was even more remarkable when seedlings were grown under shade.
The higher drought resistance of J. thurifera seedlings to the combination of light and water availability in terms of physiological traits could confer an adaptive advantage to this species. As reported by [27], J. thurifera is colonizing the abandoned agricultural land in central Spain, where seedling survival and establishment is assured as long as sporadic recruitment windows with wet episodes are maintained. This ability to colonize recently abandoned land is intimately related to lower competition in these areas, as the distribution of J. thurifera has historically been considered to be limited by competition with species which have higher growth rates and faster life cycles [33].
Apart from maintaining an active physiological performance during the 2-months severe drought period, Q. ilex showed also higher drought resistance for the morphological traits. Additionally, the species showed higher plasticity in the morphological traits under shade. The pattern of biomass allocation to stems and leaves at the expense of roots typical in oak species [31] allows higher relative shoot growth compared to the two conifers, thus conferring an advantage under drought conditions, particularly to the evergreen Q. ilex. Increased allocation to roots under conditions of low water availability has been previously reported in Q. ilex (e.g., [34]). For instance, when growing in drought-prone environments under high irradiance, Q. ilex tends to develop large taproots to reach deeper soil layers [35], thus facilitating the exploitation of even small increases in soil moisture after isolated summer rainfall events [36]. This is an advantage for Q. ilex, as the typical sandy soils at the Northern Plateau suffer a rapid water loss due to drainage, so the optimal moisture conditions are only present for a few days in spring and autumn, just after the rainfall events. In these drought-prone environments, the multistratified root system of P. pinea [37] may also confer a competitive advantage to this species, although it does not benefit from sporadic summer rainfall events to the same degree as Q. ilex [38]. This fact, together with its higher plasticity to morphological traits, its active physiological performance during drought and its wider thermal amplitude, must be seen as an advantage for Q. ilex in mixed pine-oak-juniper stands in drought-prone environments.
Regarding Q. faginea, their seedlings showed low gas exchange activity as drought conditions aggravated, even endangering its survival; and such effects were independent on light conditions. For instance, Q. faginea showed the lowest drought resistance to drought in terms of survival. Similar to Q. faginea, P. pinea showed low gas exchange activity under drought conditions, irrespective of the light environment. However, P. pinea was able to maintain the highest water potential after the 2-months drought period, while Q. faginea's water potential decreased the most.
Seedlings growth, morphology and allometry in Q. faginea and P. pinea were also negatively affected by drought. Thus, seedlings of P. pinea and Q. faginea seemed to show a higher performance in terms of growth, morphology, allometry and gas exchange traits when water was not a limiting factor, being the effect of water availability mediated by light mainly in P. pinea. For instance, P. pinea showed higher morphological performance under low resource availability (L-W-), compared to the other species, especially as regards shoot development (higher leaf, stem and shoot mass fractions and shoot to root ratio). Under such environment, this species showed high net photosynthesis (An), while maintaining a relatively high water potential, well beyond the risk of catastrophic xylem failure. For instance, the optimal regeneration niche for the species in the Spanish Northern Plateaus (in terms of growth and survival) is found on mid-shaded locations, suggesting that, under drier locations, it is benefited by shelter from excessive radiation [39].
Although Q. faginea developed a deeper root system (greater root dry weight and root mass fraction), a greater SLA to adjust leaf morphology and a high An than the other species under L-W-conditions, these environmental conditions had a negative effect on Q. faginea survival. This is intimately related to the lower water potential measured, which is a direct consequence of extreme dependence of the species on soil water reserves. There is strong evidence that a moderately shaded environment is beneficial for oak seedlings, at least during the early stages [40]. However, the benefits of shade should be associated with higher soil water content, which is not the case. Thus, the shade conditions in this experiment seemed to ameliorate the impact of drought on P. pinea while they negatively affected the performance of Q. faginea, due its low plastic response to the light environment as regards leaf and root allocation, which is consistent with the low response observed in more shade tolerant species [41].
In particular, there is a continuum of water use strategies to cope with water stress, representing P. pinea (the highest midday water potential) and Q. faginea (the lowest midday water potential) the opposite ends of this spectrum. Such performance could confer a physiological adaptive advantage to J. thurifera and Q. ilex over the other two species, enabling them to colonize high-water availability early-successional sites.
In summary, according to our results, despite the low growth of J. thurifera seedlings in early growth stages that could affect its ability to compete for soil moisture, the species showed a high plasticity to light and water availability in the physiological traits. This fact, joined to its ability to use the water from the sporadic summer and autumn rainfalls [38] and its high resistance to frost events [42] will confer the species with an adaptive advantage in these early growth stages compared with the other three species. For instance, survival of J. thurifera seedlings is higher than that of P. pinea in the mixed pine-oak-juniper stand of Central Spain [43]. Similar to J. thurifera, Q. ilex seedlings also show a high plasticity to treatments, mainly related to the allometric traits and some physiological traits (photosynthesis rate and water potential). The large taproot that facilitates the exploitation of deeper soil layers, joined to its resprouting ability and its marked anysohidric behaviour benefited the species in these mixed stands. The high sensitivity of Q. faginea to drought will restrict its presence to the most mesic sites within the mixed pine-oak-juniper stands, where water and shade are available. However, while the shade conditions in this experiment seemed to ameliorate the drought impact on P. pinea seedlings, they aggravated Q. faginea performance.

Conclusions
The four species studied are able to develop different mechanisms of tolerance and avoidance to deal with the interactive effect of water and light availability during the early phases of growth which will facilitate their coexistence and allow complementarity. However, regeneration dynamics in mixed pine-oak-juniper stands in the Northern Plateau, formerly dominated by P. pinea as a result of management to promote pine nut production, will be affected in the mid-term. For instance, a modelling approach on sapling recruitment in these stands, revealed successional trajectories towards J. thurifera and Q. ilex in the long-term [44]. The high plasticity of J. thurifera (physiological traits) and Q. ilex seedlings (physiological and allometric traits) to different light and water availability environments, and their ability to benefit from sporadic summer rainfall events will confer these two species with an adaptive advantage in the early growth stages in comparison with P. pinea and Q. faginea seedlings. The high vulnerability of Q. faginea seedlings to an increment in aridity will restrict its presence to the most mesic sites within the mixed pine-oak-juniper stands, where water and shade are available. In contrast, the low competitive capacity of P. pinea seedlings could restrict its presence to drought-prone shaded environments.
If future climate predictions in Mediterranean continental areas are met, these mixed stands could be likely transformed in the long-term into Q. ilex coppice forests, with sparse