Canopy Transpiration and Stomatal Conductance Dynamics of Ulmus pumila L. and Caragana korshinskii Kom. Plantations on the Bashang Plateau, China

Abstract

Constructing protective forests to control water and soil erosion is an effective measure to address land degradation in the Bashang Plateau of North China, but forest dieback has occurred frequently due to severe water deficits in recent decades. However, transpiration dynamics and their biophysical control factors under various soil water contents for different forest functional types are still unknown. Here, canopy transpiration and stomatal conductance of a 38-year-old Ulmus pumila L. and a 20-year-old Caragana korshinskii Kom. were quantified using the sap flow method, while simultaneously monitoring the meteorological and soil water content. The results showed that canopy transpiration averaged 0.55 ± 0.34 mm d⁻¹ and 0.66 ± 0.32 mm d⁻¹ for U. pumila, and was 0.74 ± 0.26 mm d⁻¹ and 0.77 ± 0.24 mm d⁻¹ for C. korshinskii in 2020 and 2021, respectively. The sensitivity of canopy transpiration to vapor pressure deficit (VPD) decreased as soil water stress increased for both species, indicating that the transpiration process is significantly affected by soil drought. Additionally, canopy stomatal conductance averaged 1.03 ± 0.91 mm s⁻¹ and 1.34 ± 1.22 mm s⁻¹ for U. pumila, and was 1.46 ± 0.90 mm s⁻¹ and 1.51 ± 1.06 mm s⁻¹ for C. korshinskii in 2020 and 2021, respectively. The low values of the decoupling coefficient (Ω) showed that canopy and atmosphere were well coupled for both species. Stomatal sensitivity to VPD decreased with decreasing soil water content, indicating that both U. pumila and C. korshinskii maintained a water-saving strategy under the stressed water conditions. Our results enable better understanding of transpiration dynamics and water-use strategies of different forest functional types in the Bashang Plateau, which will provide important insights for planted forests management and ecosystem stability under future climate changes.

1. Introduction

Afforestation is an important ecological measure for restoring fragile and degraded land. China has the largest afforested area in the world, with the area of planted forests accounting for about 23% of the global plantation area [1]. The area of planted forests in China now stands at about 69.33 × 10⁶ ha. Planted forests can effectively contribute to improving the functions of water and soil conservation, sand-fixing, carbon storage, and adjusting the microclimate of weak ecological environment systems [2,3]. However, under the influence of global climate change-induced drought and human activities, these planted forests, especially in arid and semiarid regions, have suffered severe canopy dieback and mortality approximately 30–35 years after planting [4,5]. Immense afforestation may consume too much water and cause a severe water deficit, and, compared with native plantation species, introduced plantation species cause a greater reduction in soil water availability, negatively impacting ecosystem services and functions [6,7]. Transpiration and related water-use strategies are fundamental to understanding the physiological processes of plantations and play a vital role in their survival and growth, especially in semiarid and arid regions, where water availability is greatly affected by the increased frequency and intensity of droughts [8,9]. However, previous studies have shown that the effects of excessive water loss caused by transpiration and its environmental driving factors differ significantly, resulting in different growth performances among different functional types [10,11,12]. Therefore, it is imperative to understand the transpiration processes and dynamics of planted tree and shrub species and how these processes change from their normal functions.

Transpiration, the plant process in which water is consumed from the soil and released into the atmosphere, is an important parameter in understanding forest hydrological processes and further determining the forest water balance [13,14]. The accurate quantification of forest transpiration is essential for tree physiology and ecohydrology [15,16]. Based on their advantages of simple operation, relatively low cost, fewer limitations in environmental conditions, and increasingly robust features, sap flow measurements have been considered the most practical method to estimate tree transpiration at both the individual and stand scales [17,18]. Environmental factors, including solar radiation, air temperature, wind speed, vapor pressure deficit (VPD), and soil water content, have significant impacts on the transpiration process. These abiotic factors can be divided into two aspects: evaporative demand (potential evapotranspiration, PET) and water supply [19]. In semiarid and arid regions, soil water availability is a critical factor for plantation growth and vegetation productivity due to its influence on plant transpiration [20]. Previous studies showed that soil water stress in plantations occurs when relative extractable water content (REW) is below a threshold of 0.4 [21,22]. A linear or nonlinear relationship, including exponential and polynomial functions, was found between transpiration and evaporative demand when REW > 0.4, whereas, when REW < 0.4, increased soil water stress influenced transpiration, which was expressed following exponential and polynomial functions for evaporative demand for different planted forests [23]. Moreover, transpiration is regulated via canopy stomatal conductance (G_c) in response to variations in VPD and soil water content. High VPD causes a decrease in canopy stomatal conductance. In order to prevent irreversible damage to tree hydraulic traits, transpiration then gradually increases nonlinearly with increasing VPD, and is maintained or decreases when the canopy stomata begin to close. However, different specific nonlinear functions, such as exponential functions [24], logarithmic functions, and polynomial functions [25], have been reported for different plantation types in different artificial ecosystems. Additionally, reduced soil water availability may decrease hydraulic conductance from soil to leaves, causing canopy stomatal closure to avoid embolism and hydraulic failure, and thereby reducing tree transpiration. Denham et al. quantified the collaborative influence of soil water content and VPD along a hydro-climatological gradient, and indicated that a more sensitive relationship between canopy stomatal conductance and soil water content was found in dry than in wet sites [26]. Most climate models predict that soil droughts will be more frequent and more severe in semiarid and arid regions [27]. However, few studies have explored the impact of environmental factors on transpiration and G_c under different soil water availability conditions, and influenced by climate change-induced drought. Moreover, it is also unclear whether these relationships exhibit interspecific differences among different forest functional types.

The Bashang Plateau is one of the most vulnerable and sensitive portions of northern China, and is a major source of dust and sand storms that affect Beijing and its surrounding areas. Land degradation and desertification are the most serious environmental problems in this area. A series of afforestation projects have been implemented by China’s government during the past four decades, e.g., the Grain-for-Green Program, the Three-North Shelterbelt Program, and the Beijing-Tianjin Sand Source Restoration Project, to address land degradation and improve the regional ecological environment [28,29]. Many fast-growing and stress-tolerant species, such as Populus simonii Carr, Pinus sylvestris var. mongolica Litv, Ulmus pumila L, and Caragana korshinskii Kom were planted to improve vegetation cover and control wind and sand erosion. The degree of desertification on the Bashang Plateau has improved significantly during the past four decades. However, these plantations have frequently suffered canopy dieback and mortality in recent years. The main reason for this is related to water deficiency, which affected plantations’ transpiration dynamics and water-use strategies. U. pumila is a major native tree species and C. korshinskii is an introduced shrub species. Nevertheless, the transpiration processes of the two functional types are still not fully understood, and the biophysical control factors of transpiration under various soil water availability conditions are still unknown. The objectives of this study were to explore: (1) canopy transpiration and stomatal conductance processes during the growing season for U. pumila (native species) and C. korshinskii (introduced species); and (2) the biophysical mechanisms of canopy transpiration control under different soil water content conditions. The results can provide a deep understanding of the water-use dynamics of planted tree and shrub species, and thus provide scientific guidance for afforestation species selection and optimal allocation.

2. Materials and Methods

2.1. Site Description

Our study was conducted in the Kangbao pasture region, Kangbao County, Zhangjiakou City, Hebei Province, China (Figure 1). This area is located in the temperate climate zone. The average annual air temperature is about 2.3 °C, and the average annual precipitation is 330 mm, with over 85% falling between May and September. The mean annual potential evapotranspiration is 880 mm and the mean wind speed is 3.15 m s⁻¹. The major soil type is chestnut soil. The main growing season runs from May to September. The dominant tree species here are U. pumila, P. simonii, and Pinus sylvestris var. mongolica Litv, and the dominant shrub species are C. korshinskii, Hippophae rhamnoides Linn, and Armeniaca sibirica (L.) Lam.

This experiment was conducted in stands of U. pumila (114°47′ E, 42°06′ N, altitude 1325 m) and C. korshinskii (114°48′ E, 42°07′ N, altitude 1305 m) in 2020 and 2021. The two sites were approximately 3.0 km apart. A plot having an area of 20 by 20 m was set up in each of the U. pumila and C. korshinskii stands. The U. pumila were about 38 years old and the C. korshinskii were about 20 years old. The corresponding maximum leaf area indexes (LAI) were 0.31 and 0.43 m² m⁻², respectively. The mean diameter at breast height (DBH, 1.3 m) of U. pumila was 16.60 cm and the mean stem basal diameter (SBD, 10–15 cm above the ground) of C. korshinskii was 2.38 cm. The physical characteristics of the soil differed between the U. pumila and C. korshinskii stands. The soil particle size distribution was 39.64% sand (0.05–2 mm), 50.85% silt (0.002–0.05 mm), and 9.69% clay (<0.002 mm) in the U. pumila stand, and 52.89% sand, 39.25% silt, and 7.86% clay in the C. korshinskii stand. The soil organic matter (SOM) in the U. pumila stand at the 0–100 cm depth was 1.34%, and that of the C. korshinskii stand was 0.85% (

Table 1. Detailed information about vegetation characteristics and soil properties of U. pumila and C. korshinskii stands.

Species	Vegetation Characteristics				Soil Properties (%)
	Age	DBH/SBD (cm)	SA (cm2)	LAI (m2 m−2)	Soil Particle-Size Distributions			SOM
					Sand	Silt	Clay
U. pumila	38	16.60	120.63	0.31	39.46	50.85	9.69	1.34
C. korshinskii	20	2.38	2.65	0.43	52.89	39.25	7.86	0.85

Note: DBH = diameter at breast height; SBD = stem basal diameter; SA = sapwood area; LAI = leaf area index; SOM = soil organic matter.

).

2.2. Measurement of Sap Flux Density and Estimation of Canopy Transpiration

Measurements of sap flow were carried out using Granier-type thermal dissipation probes (TDPs) between 1 May and 30 September in 2020 and 2021 [30]. The sensors consisted of a pair of probes having a 20 mm length and 2 mm diameter, each of which had a cooper constant thermocouple. Based on the distributions of DBH and SBD, six sampled trees in the U. pumila stand and six sampled stems in the C. korshinskii stand were selected for sap flux measurements. The probes were inserted into the north-facing side of the truck (stem) at a height of 1.3 m (sampled U. pumila) and 15 cm (sampled C. korshinskii) above ground level to minimize exposure to the sun. In order to protect the probes against physical damage, solar radiation, temperature fluctuations, and rain, they were mounted in waterproof silicone and covered with an aluminum cover. The temperature differences between heated and reference probes were recorded on a data logger at 10 min intervals (CR1000, Campbell Scientific Inc., Logan, UT, USA). The sap flux density (F_d) was estimated using the standard calibration for the TDP method:

$$F_d=0.714\times {\left(\frac{\mathsf{\Delta }{T}_{max}-\mathsf{\Delta }T}{\mathsf{\Delta }T}\right)}^{1.231}$$

(1)

where F_d (mL cm⁻² min⁻¹) is the sap flux density, ΔT (°C) is the temperature difference between the two probes at any given time, and ΔTmax (°C) is the maximum temperature difference between sensors when the sap flux density is close to zero. Due to the fact that we only used one pair of probes per sample tree, azimuthal and radial variations in sap flux within trees were ignored In calculating canopy transpiration per area of ground (E_c, mm h⁻¹); we multiplied the mean F_d for each sample tree (F_d-avg) by the total sapwood area of the study plot per unit area of ground:

$$E_c=F_{d-avg}\times \frac{A_{si}}{A_g}\times 600$$

(2)

where F_d-avg is the average sap flux density of sampled trees (or shrubs), A_g is the ground surface area of the studied plots (400 m²), and A_si is the total sapwood area in the studied plot. In order to measure the sapwood area of U. pumila, the staining method was applied to 19 trees both inside and outside the plot. Exponential growth function relationships between sapwood and DBH ($A_s=27.10\times e^{0.09DBH},n=19,R^2=0.83$) were established. Furthermore, we used the stem basal cross-sectional area as the scalar to estimate the E_c and stand transpiration for the C. korshinskii stand. The total sapwood area in the plot was 6016.04 cm² and 6877.70 cm² for the U. pumila and C. korshinskii stands, respectively.

2.3. Meteorological Data and Soil Water Measurement

Precipitation (P, mm), air temperature (T_a, °C), relative humidity (RH, %), wind speed (u₂, m s⁻¹), and photosynthetically active radiation (PAR, μmol m⁻² s⁻¹) were measured continuously with an Onset HOBO U21 automatic weather station (Onset Computer Corp., Bourne, MA, USA) located in the open areas neighboring the U. pumila plantation stand. Air temperature and relative humidity were measured at a height of 1.5 m, and precipitation, wind speed, and photosynthetically active radiation were measured at a height of 2.0 m (Figure 1). These variables were recorded every 10 min as averages using data loggers of the CR1000 series (Campbell Scientific, Logan, UT, USA). VPD (kPa) was derived from air temperature and the relative humidity:

$$VPD=0.611\times \exp \left(\frac{17.502T_a}{T_a+240.97}\right)\times \left(1-RH\right)$$

(3)

The soil volumetric water content (θ, %) was continuously monitored at a single location in U. pumila and C. korshinskii stands with an EC–5TE sensor (Decagon, Inc. Decagon, USA). Both EC–5TE sensors were installed at depths of 0–10 cm, 10–20 cm, 20–40 cm, 40–60 cm, and 60–100 cm below the ground surface. Measurements were taken every 10 s, and the 30 min averages were recorded by CR1000 data loggers (Campbell Scientific, Logan, UT, USA). The soil water content values used in the paper were averaged for the depths of 0–100 cm for each plot. Relative extractable soil water (REW) was calculated using averaged θ across 0–100 cm:

$$REW=\frac{\theta -{\theta }_{min}}{{\theta }_{max}-{\theta }_{min}}$$

(4)

where θ_min and θ_max are the minimum and maximum daily average soil water content, respectively, during May–September in 2020 and 2021.

To account for climate impacts on stand canopy transpiration, daily potential evapotranspiration (PET, mm d⁻¹) was calculated by the Penman–Monteith equation based on FAO–56 [31]:

$$PET=\frac{0.408\mathsf{\Delta }\left(R_n-G\right)+\gamma \left(\raisebox{1ex}{$900$}\!\left/ \!\raisebox{-1ex}{$\left(T_a+273\right)$}\right.\right)u_{2}VPD}{\mathsf{\Delta }+\gamma \left(1+0.34u_2\right)}$$

(5)

where R_n is the net radiation (MJ m⁻² d⁻¹), G is the soil heat flux density (G ≈ 0 for daily scale), γ is the psychrometric constant (kPa °C⁻¹), and Δ is the slope of the vapor pressure curve (kPa °C⁻¹).

2.4. Canopy Conductance and Decoupling Coefficient

G_c (mm s⁻¹) was calculated by multiplying the conductance coefficient (K_G) by E_c and dividing the result by VPD [32]:

$$G_c=\frac{K_G\times E_c}{VPD}$$

(6)

where K_G is the conductance coefficient as a function of air temperature (115.8 + 0.4236 T_a, kPa m³ kg⁻¹), which accounts for the temperature effect on the psychometric constant, the latent heat of vaporization, and the specific heat and density of air.

We estimated the decoupling coefficient (0 < Ω < 1) as a means of quantifying the degree of coupling between the canopy and atmosphere. Canopies are aerodynamically well coupled to the atmosphere when Ω approaches 0, and canopy transpiration is determined primarily by stomatal opening. As Ω approaches 1, stomatal control on canopy transpiration becomes weaker, and canopy transpiration is more influenced by solar radiation. The Ω was estimated as follow [33]:

$$\mathsf{\Omega }=\frac{1}{1+\left[\raisebox{1ex}{$\gamma $}\!\left/ \!\raisebox{-1ex}{$\left(\mathsf{\Delta }+\gamma \right)$}\right.\right]\left(\raisebox{1ex}{$g_a$}\!\left/ \!\raisebox{-1ex}{$G_c$}\right.\right)}$$

(7)

where g_a is the aerodynamic conductance (m s⁻¹), calculated as follows:

$$g_a=\frac{1+0.54U_z}{{\left[\ln \raisebox{1ex}{$\left(z-d\right)$}\!\left/ \!\raisebox{-1ex}{$z_0$}\right.\right]}^2}$$

(8)

where U_z is the wind speed above the canopy (m s⁻¹), which is derived from the measured wind speed at a 2.0 m height (u₂). z is usually equal to the canopy height (m), z₀ is the roughness height (usually 0.1 * H, where H is the canopy height) and d is the displacement height (0.75 * H).

2.5. Data Analysis

In the present study, significant differences in daily E_c, G_c, and Ω between the two growing seasons and the two planted forests were tested using two-way repeated measures ANOVA. The differences in the micrometeorological variables during the two growing seasons of 2020 and 2021 were testing using a paired samples t-test. Then, the relationships between PAR, VPD, and E_c were mainly analyzed using an exponential threshold function after partitioning the data into two classes (stressed water conditions: REW < 0.40 and non-stressed water conditions: REW > 0.40), as follows:

$$E_c=a\left(1-e^{-bx}\right)$$

(9)

where a and b are the fitting parameters, and x is the corresponding meteorological variable.

The response of G_c to VPD was quantified using a linear logarithmic function [34]:

$$G_c=-m\ln VPD+G_{ref}$$

(10)

where m is the canopy conductance’s sensitivity to VPD and G_ref is the reference canopy conductance when VPD = 1 kPa. An upper boundary line was derived based on the data of E_c at least one standard deviation greater than the mean E_c of each VPD (0.5 kPa) interval, from which the parameters of m and G_ref were determined.

All statistical analyses were performed using the SPSS 21 software program (SPSS Inc., Chicago, IL, USA) and all the figures were created using Sigmaplot 11.0 software (Hearne Scientific Software Plc, Melbourne, Australia).

3. Results

3.1. Environmental Variables

During the growing season (May–September), the total precipitation was 256 and 283 mm in 2020 and 2021, respectively (Figure 2), accounting for 88.90% and 98.31% of the long-term average precipitation (288 mm, 1980–2020). The photosynthetically active radiation was 402.47 ± 124.65 μmol m⁻² s⁻¹ and 442.56 ± 126.41 μmol m⁻² s⁻¹, whereas PET averaged 3.65 ± 1.36 mm d⁻¹ and 3.72 ± 1.30 mm d⁻¹ during the growing season in 2020 and 2021, respectively (Figure 2a). During both growing seasons, daily air temperatures displayed marked seasonal variations, with a range of 5.1 to 22.6 °C and 1.5 to 23.1 °C, and mean values of 15.14 ± 3.88 °C and 15.02 ± 4.28 °C, in 2020 and 2021, respectively (Figure 2b). The average VPD was 0.63 ± 0.35 kPa and 0.61 ± 0.32 kPa, whereas u₂ was 3.11 ± 1.56 m s⁻¹ and 3.31 ± 1.77 m s⁻¹ in 2020 and 2021, respectively (Figure 2c). The mean soil water content at the 0–100 cm soil layer was 8.03% for U. pumila and 5.09% for C. korshinskii in 2020, and 10.42% for U. pumila and 5.33% for C. korshinskii in 2021. The soil water content in the 0–40 cm layer increased quickly after precipitation events, whereas that of the 40–100 cm layer was stable during the growing season. In the vertical profile, the soil water content in the 40–100 cm layer was lower than that in other profiles in U. pumila, whereas the lowest soil water content occurred at 20–40 cm in C. korshinskii (Figure 3).

3.2. Canopy Transpiration per Unit Ground Area

During the measurement period, E_c for the U. pumila stand ranged from 0.08 to 1.77 mm d⁻¹, with a mean value of 0.55 ± 0.34 mm d⁻¹ in 2020; whereas, in 2021, E_c ranged from 0.08 to 2.57 mm d⁻¹, with a mean value of 0.66 ± 0.32 mm d⁻¹. For C. korshinskii, E_c ranged from 0.10 to 1.41 mm d⁻¹ and from 0.21 to 1.30 mm d⁻¹, with mean values of 0.74 ± 0.26 mm d⁻¹ and 0.77 ± 0.24 mm d⁻¹ in 2020 and 2021, respectively (Figure 4a,c). The accumulated E_c over the entire growing season was 83.72 and 113.65 mm for U. pumila in 2020 and 2021, respectively, accounting for 32.70% and 40.14% of the precipitation over the same period. The accumulated E_c was 101.29 and 117.77 mm for the C. korshinskii stand in 2020 and 2021, respectively, accounting for 39.57% and 41.60% of the precipitation over the same period (Figure 4b,d).

3.3. Response of Canopy Transpiration to Environmental Variables

Statistical analysis indicated that the positive effect of environmental variables on E_c had a ranking of θ > T_a > PAR > u₂ > VPD > PET in 2020 and PAR > u₂ > θ > T_a > P > PET in 2021 for U. pumila. The positive effect of environmental variables on E_c had a ranking of PAR > θ > u₂ > T_a > VPD > P in 2020 and P > θ > T_a > VPD > PAR > u₂ > PET in 2021 for C. korshinskii (

Table 2. Correlation coefficients between canopy transpiration (E_c) and environmental factors for U. pumila and C. korshinskii in 2020 and 2021.

		PAR	Ta	RH	VPD	u2	p	PET	θ
U. pumila	Ec (2020)	0.522 **	0.615 **	-	0.366 **	−0.404 *	-	0.277 **	0.691 **
	Ec (2021)	0.513 **	0.369 **	-	-	−0.508 **	−0.253 **	0.176 *	0.502 **
C. korshinskii	Ec (2020)	0.475 **	0.362 **	-	0.354 **	−0.462 **	−0.262 **	-	0.474 **
	Ec (2021)	0.349 **	0.402 **	-	0.389 **	−0.330 **	−0.497 **	0.185 *	0.444 **

Note: PAR = photosynthetically active radiation; T_a = air temperature; RH = relative humidity; VPD = vapor pressure deficit; u₂ = wind speed; P = precipitation; PET = daily potential evapotranspiration; θ = soil water content. ** indicates p < 0.01, * indicates p < 0.05.

). This suggests that PAR, θ, and VPD were major determinants of transpiration for both U. pumila and C. korshinskii.

The E_c responded to REW following a quadratic polynomial function for U. pumila both in 2020 and 2021 (Figure 5a,b). E_c reached the maximum values when REW = 0.4–0.5 in 2020 and REW = 0.4 in 2021. The E_c responded to REW following a saturated exponential function for C. korshinskii both in 2020 and 2021 (Figure 5c,d). E_c rapidly increased with rising REW when REW < 0.4, E_c then increased slowly when REW > 0.4, and, finally, E_c tended to be saturated at 1.31 and 1.14 mm when REW was close to 1 in 2020 and 2021, respectively. Additionally, the response of E_c to PAR and VPD was affected by different soil water conditions. The E_c increased with an increasing PAR and VPD for U. pumila under REW < 0.40, while it tended to be saturated when VPD reached 2.0 kPa (Figure 6a,b). PAR and VPD explained 28% and 22% of the variation in E_c for U. pumila under REW < 0.40, respectively, whereas E_c maintained a high value and stability when REW > 0.4. The E_c increased significantly with an increasing PAR and VPD for C. korshinskii under different soil water conditions (Figure 6c,d). The E_c exhibited an exponentially saturating response to VPD and tended to level off at 1.5 kPa under REW < 0.4, and there was no obvious threshold value of Ec under REW > 0.4. VPD explained 16% and 81% of the variation in E_c for C. korshinskii under REW < 0.40 and REW > 0.4, respectively, whereas PAR explained 23% and 52% of the variation in E_c for C. korshinskii under REW < 0.40 and REW > 0.4, respectively.

3.4. Response of Canopy Stomatal Conductance to VPD

The G_c showed similar seasonal variations for U. pumila and C. korshinskii during 2020 and 2021. The maximum monthly G_c of U. pumila was 1.98 ± 1.07 mm s⁻¹ and 2.02 ± 1.42 mm s⁻¹ in July during 2020 and 2021, respectively. However, in both years, the G_c of C. korshinskii was significantly higher in July–September than in May–June (F = 15.517, p < 0.001 in 2020 and F = 11.477, p < 0.001 in 2021) (Figure 7). The G_c significantly decreased with an increase in VPD for both afforestation species (Figure 8). VPD explained 15% and 41% of total variation in G_c for U. pumila in 2020 and 2021, respectively (Figure 8a,b), and 45% and 51% of total variation in G_c for C. korshinskii in 2020 and 2021, respectively (Figure 8c,d). In addition, m had a significantly higher effect in 2021 than in 2020 for U. pumila (3.01 > 2.50) and C. korshinskii (2.83 > 1.96) after applying boundary line analysis, whereas no significant difference in the reference canopy stomatal conductance (G_ref) was observed for the two planted species. Moreover, the higher G_c of U. pumila and C. korshinskii when REW > 0.4 than when REW < 0.4 (U. pumila: 2.46 > 0.78, F = 26.383, p < 0.001 in 2020; C. korshinskii: 2.02 > 1.36, F = 14.359, p < 0.001 in 2021) indicates the significant effect of soil water content on canopy stomatal conductance. m was also higher under REW > 0.4 conditions than under REW < 0.4 conditions for both of the planted species (U. pumila: 3.30 > 0.87 in 2020 and 3.59 > 2.98 in 2021; C. korshinskii: 2.26 > 1.97 in 2020 and 3.70 > 2.08 in 2021) (Figure 8e–h).

Throughout the measurement period, the daily Ω of U. pumila ranged from 0.0001 to 0.0112, with mean values of 0.0015 ± 0.0014 and 0.0020 ± 0.0019 in 2020 and 2021, respectively (Figure 7). The daily Ω of C. korshinskii ranged from 0.0003 to 0.0145, with mean values of 0.0024 ± 0.0016 and 0.0024 ± 0.0018 in 2020 and 2021, respectively. No differences were found for the value of Ω between 2020 and 2021 for both U. pumila (F = 5.896, p = 0.016) and C. korshinskii (F = 0.021, p = 0.884).

4. Discussion

4.1. Canopy Transpiration of U. pumila and C. korshinskii

The E_c of the U. pumila stand in this study was 0.55 and 0.66 mm d⁻¹, which is similar to the values reported for Mongolian pine planted in the southeastern Keerqin Sandy Land [35] and half-mature Mongolian pine planted in the Rare Psammophytes Protection Botanical Base [36], but significantly lower than those of Chinese pine (0.7 mm d⁻¹), Mongolian pine (1.1 mm d⁻¹), and Populus × xiaozhuanica plantations (1.2–1.5 mm d⁻¹) planted in the Keerqin Sandy Land [16]; Populus simonii (175.2 kg d⁻¹) planted on the Loess Plateau [37]; and Populus euphratica Oliv. planted in the lower reaches of the Heihe River Basin [38]. The E_c of the C. korshinskii stand in this study was 0.74 and 0.77 mm d⁻¹ in the two seasons, respectively. These values are lower than those for the same species planted in the Loess Plateau under different precipitation regions [37,39]. Compared with other major planted shrubs, a similar range of transpiration was reported for a Vitex negundo L. plantation (107.21 mm cumulative in the growing season), but a higher transpiration of Hippophae rhamnoides L. was found in the Loess Plateau (122.33–180.28 mm cumulative in the growing season) [11]. This may be related to the higher evaporative demand in drier and colder regions and the lower soil water availability in the Bashang Plateau. The mean annual precipitation in our study site was 100–150 mm lower than that for planted forests in the Keerqin Sandy Land and the Loess Plateau. In addition, a relatively lower soil water content may also increase the hydraulic resistance of the soil–root system, preventing water movement between soil and plant leaves, which ultimately decreases plantations’ transpiration rate [40]. Finally, short heights and small sapwood areas, resulting from severe canopy dieback and planted forest degradation, were found for the U. pumila stand in this study (Table 2), and canopy transpiration was thus exhibited at a lower level.

E_c responded to REW as a saturated exponential function for C. korshinskii in this study (Figure 5c,d), which was in agreement with results reported for broadleaved and coniferous trees under different climates [41], a larch plantation in the semiarid Northwest China [42], a black locust plantation [43], and C. korshinskii [44] on the Loess Plateau. However, the REW threshold in this study is similar to or relatively higher than those in the studies reported above. Li et al. indicated that the REW threshold is generally higher under dry conditions than under wet conditions [23]. Averaged soil water content of the C. korshinskii stand was 5.09% in 2020 and 5.33% in 2021, which was lower than those in the plantations reported above (Figure 3b). E_c presented a polynomial pattern in REW for U. pumila in this study (Figure 5a,b), which was consistent with that in Haloxylon ammodendron and Calligonum mongolicum in Northwest China [45]. The reason that E_c was lower during the higher REW period is linked to the much lower VPD and T_a when compared to the dry period (Figure 2 and Figure 6b). The daily E_c of both the U. pumila and C. korshinskii stands increased with increasing PAR and VPD, and E_c tended to be saturated at high VPD values. These results are similar to those reported for poplar trees [16], Mongolian pine [46], and other planted species [12]. In both plantations, however, the VPD threshold decreased as soil water content decreased, ranging from non-stressed to water stressed. An obvious VPD threshold was observed at approximately 2.0 kPa for the U. pumila stand and 1.5 kPa for the C. korshinskii stand when REW < 0.4. Liu et al. indicated that the utilization of surface soil water gradually increased with the increase in the degree of degradation of the agroforestry shelterbelts [47]. We then concluded that the U. pumila and C. korshinskii stands primarily obtained water from the shallow and middle soil layers because of the root distribution pattern and the degree of canopy dieback. As soil water decreased, the increasing whole-tree hydraulic resistance triggered canopy stomatal closure, and hence limited the response of transpiration to major environmental variables. In addition, the Ω values (Figure 6) were also lower than those reported for mature Mongolian pine (0.041), half-mature Mongolian pine (0.15), and young Mongolian pine (0.18) [36], and Schima superba (0.22) [48]. The smaller LAI of the U. pumila (0.31) and C. korshinskii (0.43) plantations in this study, compared to that of the Mongolian pine (0.96–1.54) in the Horqin desert and the Mu Us desert, indicates that decreasing the canopy resistance resulted in strong canopy–atmosphere coupling. Therefore, under high VPD conditions, canopy stomata must be closed to prevent water potential from dropping below the threshold and thus inhibiting tree transpiration. Moreover, a relatively lower VPD threshold under water stress was observed for both C. korshinskii (1.5 kPa) and U. pumila (2.0 kPa), indicating that the transpiration process of the two species was affected significantly by soil drought [49]. Planted species usually have a shallower fine root distribution and lower stem and leaf hydraulic conductivity [38,50]. As concluded above, the transpiration process and water-use strategy of U. pumila and C. korshinskii were revealed in the Bashang Plateau, and further research on the water-use pattern of typical planted forests should be undertaken in the future. However, in the present study, we did not investigate seasonal variations in LAI and its effects on canopy transpiration during the growing season for U. pumila and C. korshinskii. Additional studies on canopy transpiration and stomatal conductance combined with different plantation ages and densities would provide further insights into the ecohydrological processes and afforestation management on the Bashang Plateau.

4.2. Canopy Stomatal Conductance of U. pumila and C. korshinskii

The G_c values were significantly higher under non-stressed conditions than under stressed conditions for both plantations, indicating the strong limitations of soil water on canopy stomatal conductance. In addition, the significantly nonlinear negative relationship between G_c and VPD in both the U. pumila and C. korshinskii plantations suggested that stronger stomatal regulation of canopy transpiration responded to higher VPD. Stomatal closure thus played an important role in maintaining water status to avoid a catastrophic loss of xylem function [51]. This drought adaptation mechanism can be characterized as a water-saving strategy, consistent with those reported for P. sylvestris plantations [36,52,53].

We also found that the parameters of m and G_ref for the U. pumila plantation were higher than those for the C. korshinskii plantation (Figure 8). This was contrary to the findings of Song et al. [46], who observed that, in comparison with Mongolian pine (an introduced species), Chinese pine (a native species) had relatively higher values of m and G_ref. We further suggest that m and G_ref for both the U. pumila and C. korshinskii plantations were higher under REW > 0.4 than under REW < 0.4 (Figure 8e–h), indicating that the sensitivity of canopy stomatal conductance to VPD decreased with soil drought; this was consistent with the findings of Novick et al. [54], Jiao et al. [19] on the Loess Plateau, and Song et al. [35] in the Horqin desert. Soil drought can negatively affect the hydraulic conductance of soil-to-plant leaf transporting paths and decrease m values, thus limiting the transpiration rate. Grossiord et al. indicated that reduced soil water availability negated the benefits of stomatal and hydraulic adjustments and resulted in reduced transpiration in juniper [55]. Similar results were found for grapevines, both in pot grown and field-grown experiments [56]. Therefore, further decreasing hydraulic conductance to prevent xylem cavitation as soil water stress increases results in a less sensitive response of G_c to VPD. This may be one of the important reasons why planted forests are vulnerable to dieback and degradation. Further research on the hydraulic traits of typical planted forests should be undertaken in the future.

5. Conclusions

In the present study, canopy transpiration and stomatal conductance were estimated for an U. pumila and a C. korshinskii stand. During the growing season, canopy transpiration was 83.72 and 113.65 mm in 2020 and 2021, respectively, for the U. pumila stand, and 101.29 and 117.77 mm in 2020 and 2021, respectively, for the C. korshinskii stand. Transpiration of U. pumila and C. korshinskii were better correlated to soil water content, photosynthetically active radiation, and VPD. Both planted species showed reduced VPD sensitivity of canopy transpiration as soil water decreased, indicating the transpiration process was affected significantly by soil drought. The VPD threshold was 1.50 and 2.0 kPa for U. pumila and C. korshinskii, respectively. Furthermore, the transpiration of both planted species was mainly regulated by stomatal opening due to low values of the decoupling coefficients (0.0015 and 0.0020 for U. pumila; 0.0024 for C. korshinskii). Both U. pumila and C. korshinskii gradually displayed a reduced canopy stomatal conductance with increasing VPD, but decreasing stomatal sensitivity to VPD in response to soil drought was also observed. Based on the results above, we conclude that both U. pumila and C. korshinskii plantations exhibited water-saving strategies under a cold and arid environment on the Bashang Plateau. These findings provide a deeper understanding of transpiration dynamics of different forest functional types on the Bashang Plateau, which may be applied for afforestation management in semiarid regions.