Stem sap flow of Haloxylon ammodendron at different ages and its response to physical factors in the Minqin oasis-desert transition zone, China

QIANG Yuquan, ZHANG Jinchun*, XU Xianying, LIU Hujun,DUAN Xiaofeng

1 Minqin National Station for Desert Steppe Ecosystem Studies, Minqin 733300, China;

2 Gansu Desert Control Research Institute, Lanzhou 730000, China;

3 College of Forestry, Gansu Agricultural University, Lanzhou 730000, China

Abstract: Haloxylon ammodendron, with its tolerance of drought, high temperature, and salt alkali conditions, is one of the main sand-fixing plant species in the oasis-desert transition zone in China.This study used the TDP30 (where TDP is the thermal dissipation probe) to measure hourly and daily variations in the stem sap flow velocity of H. ammodendron at three age-classes (10, 15, and 20 years old,which were denoted as H10, H15, and H20, respectively) in the Minqin oasis-desert transition zone,China, from May through October 2020.By simultaneously monitoring temperature, relative humidity,photosynthetically active radiation, wind speed, net radiation, rainfall, and soil moisture in this region, we comprehensively investigated the stem sap flow velocity of different-aged H. ammodendron plants (H10,H15, and H20) and revealed its response to physical factors.The results showed that, on sunny days, the hourly variation curves of the stem sap flow velocity of H. ammodendron plants at the three age-classes were mainly unimodal.In addition, the stem sap flow velocity of H. ammodendron plants decreased significantly from September to October, which also delayed its peak time of hourly variation.On rainy days, the stem sap flow velocity of H. ammodendron plants was multimodal and significantly lower than that on sunny days.Average daily water consumption of H. ammodendron plants at H10, H15, and H20 was 1.98, 2.82, and 1.91 kg/d, respectively.Temperature was the key factor affecting the stem sap flow velocity of H. ammodendron at all age-classes.Net radiation was the critical factor influencing the stem sap flow velocity of H.ammodendron at H10 and H15; however, for that at H20, it was vapor pressure deficit.The stem sap flow velocity of H. ammodendron was highly significantly correlated with soil moisture at the soil depths of 50 and 100 cm, and the correlation was strengthened with increasing stand age.Altogether, our results revealed the dynamic changes of the stem sap flow velocity in different-aged H. ammodendron forest stands and its response mechanism to local physical factors, which provided a theoretical basis for the construction of new protective forests as well as the restoration and protection of existing ones in this region and other similar arid regions in the world.

Keywords: Haloxylon ammodendron; stem sap flow; stand age; soil moisture; water consumption; Minqin oasis-desert transition zone

1 Introduction

The oasis-desert transition zone is the most intensive and prominent area of oasis and desert inter-conversion activities (Pan, 2000; Cao et al., 2021).This type of area is therefore extremely fragile and sensitive, being vulnerable to human activities and quite difficult to restore once degraded or destroyed (Ding, 2018; Lu et al., 2023).Moreover, this zone was also the interface area with the most frequent material cycle, energy flow, and information exchange between oasis and desert ecosystems (Thomas et al., 2006; Chang et al., 2007).

Haloxylon ammodendronis a deciduous shrub and small tree in the family Amaranthaceae,being a C4plant species native to Central Asian desert habitats (Liet al., 2015; He et al., 2018).Further, it is a key functional species for establishing windbreaks and sand-fixation forests in the Minqin oasis-desert transition zone, China (Zhu and Jia, 2011; Zhang et al., 2019).Due to the impact of various factors, such as climate change and human activities, both the groundwater level and soil moisture in this transition zone are declining (Wang et al., 2015; Wang et al., 2023).Nearly 60% of the existing 44,600 hm2of theH.ammodendronshelterbelt has suffered from poor growth or degradation, which greatly impairs its windbreak and sand fixation functioning.This ongoing degradation ofH.ammodendronvegetation threatens the survival and development of the Minqin oasis (Yang et al., 2020).

The water consumed via plant transpiration is a fundamental component of the water cycle and energy balance of forest ecosystems, being a crucial indicator reflecting the water status of plants as well as a key contributing factor affecting regional and even global climate dynamics (Ninget al., 2020).Recently, Poyatoset al.(2021) summarized the existing body of research findings on the relationship between stem sap flow and transpiration; they believed that under normal circumstances, the stem sap flow could accurately reflect the transpiration efficiency and water use status of a single plant in each time period.Accordingly, studying the responses in the stem sap flow of different-agedH.ammodendronplants to physical factors is of great significance to better understand how the water consumption ofH.ammodendronchanges through ontogeny.In China, research onH.ammodendronin the Minqin oasis-desert transition zone has focused on leaf photosynthesis, transpiration, and water use efficiency (Hu et al., 2021), drought tolerance(Lü et al., 2019), and root morphological characteristics (Jiang et al., 2018).However, a systematic investigation of water consumption per plant and its relationship with external meteorological factors and soil moisture during the whole growing season ofH.ammodendronat different ages in the Minqin oasis-desert transition zone has yet to be reported.By analyzing the dynamic changes in the stem sap flow of different-agedH.ammodendron, not only can we grasp the water consumption law ofH.ammodendronduring its growth process but also can we reveal its response mechanism under self-thinning when adapting to local changes in arid environments;the knowledge gains can be useful for effectively carrying out protection and restoration work for different-agedH.ammodendronplants.

To that end, taking a typical sand-fixing plant in the Minqin oasis-desert transition zone in China as the object of research, this study continually measured the stem sap flow ofH.ammodendronat different ages across the entire growing season.External physical factors were simultaneously measured to analyze the changes in the stem sap flow of different-agedH.ammodendronand the correlation between stem sap flow and physical factors, to better understand the age-dependent characteristics of water consumption by this species and how this responds to those abiotic factors.This study's findings are expected to provide a theoretical basis for the restoration and protection of degradedH.ammodendronplants and the establishment of newH.ammodendronshelterbelt.

2 Materials and methods

2.1 Study area

The study area (Minqin oasis-desert transition zone; 39°08′56′′-39°09′02′′N(xiāo),103°36′54′′-103°38′01′′E; Fig.1) is located near the Minqin Desert Control Research Station at the southeastern edge of the Badain Jaran Desert, China, at an elevation of 1380 m.This region has a typical continental climate, with an annual average temperature of 7.3°C and an effective cumulative temperature greater than 10.0°C of 3289.1°C.Average annual precipitation is 110.0 mm, mainly falling from July to September, with at least 2640.0 mm of potential evaporation.The frost-free period lasts 168 d and there are 3181 h of sunshine per year, with 630 kJ/cm2of solar radiation; the year-round prevailing northwest wind has an annual average speed of 4.1 m/s.Soil here is mostly sandy, with poor nutrients and serious wind erosion.The landform is mostly semi-fixed sandy land interspersed with sand dunes at 3-10 m in height.From 2019 to 2022, the buried depth of groundwater in the study area was between 22.95 and 23.41 m.Vegetation of the study area is dominated by xerophyte shrubs, semi-shrubs, and annual and perennial herbs,namely consisting of the shrubsH.ammodendron,Nitraria tangutorum, andArtemisia arenaria,and the herbsPhragmites australis,Kali collinum,Halogeton glomeratus, andAgriophyllum squarrosum.

Fig.1 Overview of the Minqin oasis-desert transition zone at the southeastern edge of the Badain Jaran Desert and locations of sampled H. ammodendron forest stands as well as the field photos of H10 (b), H15 (c) and H20(d).H10, H15, and H20 are H. ammodendron forest stands with different afforestation ages of 10, 15, and 20 years old, respectively.The Landsat 8 OLI-TIRS images were downloaded from Geospatial Data Cloud.

2.2 Experimental design

According to the elapsed time since they were planted, we selected three artificialH.ammodendronforest stands with different afforestation ages of 10, 15, and 20 years old (hereon namedH10,H15, andH20, respectively) in the Minqin oasis-desert transition zone for study.Four representativeH.ammodendronindividuals (samples) that were growing well and looked healthy were chosen at each age class (see Table 1 for their specific parameters).The plant height ofH.ammodendronwas measured with a tower ruler, the crown width and distance between plants and rows were determined with a tape measure, and the stem ground diameter was measured with a vernier caliper.Using the TDP30 (where TDP is the thermal dissipation probe)(Rainroot Scientific Limited, Beijing, China), we measured the stem sap flow velocity for each individual from 1 May to 30 October, 2020.The TDP30 was applied to a smooth spot at 40 cm aboveground on the stem's east side.Specifically, this measuring spot was first sanded with sandpaper, and then the probe was drilled and inserted into the polished spot with a drill needle matching the probe.Once installed, the probe was wrapped with a special plastic fixing sponge,and the latter was wrapped in silver radiation-proof aluminum platinum paper; thermal insulation cotton was placed on the ground around the probe to reduce the influence of ground temperature on its measurements.A CR1000 data collector (CR1000, Campbell Scientific Inc., Logan, UT,USA) was connected to the probe to collect data every 1 min every day throughout the observation period.

Table 1 Basic information of the sampled Haloxylon ammodendron at different ages (H10, H15, and H20)

2.3 Determination of the stem sap flow and water consumption of H. ammodendron

Stem sap flux density (i.e., stem sap flow velocity) was calculated on the basis of the empirical formula in Granier (1987):

whereFdis the stem sap flux density (cm3/(cm2·h));kis a dimensionless unit;αandβare the coefficients, depending on the thermal coefficient; ΔTmaxdenotes the maximum temperature difference between the two probes when there was no stem sap flow (°C); and ΔTis the temperature difference between the two probes when there existed stem sap flow (°C).The measurement time without stem sap flow was set to 00:00-01:00 (LST) in May, July, August,September, and October, and to 03:00-04:00 in June.

The daily transpiration of a single individual (sample) was calculated as follows:

whereQis the daily transpiration (cm3);Asis the sapwood area (cm2);Trefers to 24 h of time in a given day; andxis theH.ammodendrondiameter (cm).The sapwood area was calculated using the exponential function described in Zhang et al.(2017).

Finally, a mass formula (m=ρ×V/1000, wheremis the water consumption (kg),ρis the density of water (g/cm3), taking 1 g/cm3in this study,andVis the daily transpiration (cm3), equivalent toQ) was used to calculate water consumption, by which the daily transpiration (Q; cm3) was converted to water consumption.

2.4 Determination of meteorological factors

Three portable weather stations were deployed synchronously at the sampled forest stands to monitor local meteorological factors.These stations recorded daily average temperature (°C),relative humidity (RH; %), photosynthetically active radiation (PAR; μmol/(s·m2)), surface net radiation (Rn; W/m2), daily average wind speed at 2 m above the ground (m/s), and rainfall (mm)every 10 min, from May through October 2020.The calculation method of vapor pressure deficit(VPD) was referred to Zheng and Wang (2015) and Zhanget al.(2016).

2.5 Determination of soil moisture

A series of EC-5 soil moisture sensors (Ecotek Technology Limited, Beijing, China) were installed at the depths of 5, 20, 50, 100, and 150 cm belowground at eachH.ammodendronforest stand and connected to EM50 data collectors (Ecotek Technology Limited, Beijing, China).Data were downloaded from EM50 data collectors by using ECH2O Utility software (Ecotek Technology Limited, Beijing, China).After the installation was completed, near to where each sensor was positioned, soil bulk density was taken and the actual soil moisture was measured to calibrate the moisture probe error.The measurement period spanned from April to November 2020, with the data measurement interval of 1 h.

2.6 Data processing and calculations

Under sunny conditions, based on the stem sap flow velocity data ofH.ammodendronat each age-classes (H10,H15, andH20), we selected a day with more stable changes from consecutive days that showed similar trends for each month, to represent the stem sap flow variations ofH.ammodendronin that month.Under rainy conditions, a day with high rainfall and significant changes in the stem sap flow velocity ofH.ammodendronfor each stand age was selected for analysis.Next, the average value of fourH.ammodendronindividuals at each age class was used to express the age-specific trend in changed stem sap flow velocity (Table 1).The hourly and daily variation data of the stem sap flow velocityofH.ammodendronwere obtained by adding up the monitoring data every 10 min.SPSS 16.0 software (SPSS Inc., Chicago, IL, USA) and Microsoft Excel 2019 software were used for the statistical analysis.

2.7 Response of the stem sap flow of H. ammodendron to physical factors

A structural equation model (SEM) was used to determine the relative response of the stem sap flow of different-agedH.ammodendronindividuals to meteorological factors and the size of the weights accounted for.The model-building process was divided into five steps: constructing a theoretical model, formulating research hypotheses, testing the reliability and validity, assessing the model fitness, and adjusting and revising the model as needed.Daily values of the stem sap flow velocity ofH.ammodendronindividuals atH10,H15, andH20 were used as the criterion layer, and temperature, RH,Rn, wind speed, and VPD were used as the indicator layer, assuming that all these physical factors could exert a driving effect on the stem sap flow.MPLUS 8.0 software (Muthen &Muthen, Los Angeles, CA, USA) was used to test the discriminant validity of the SEM.The data were assessed for reliability by applying Kronbach and Kaiser-Meyer-Olkin (KMO) tests, to determine whether the data could be used for structural equation modeling.The model fit was evaluated by the chi-square free ratio (χ2/df), comparative fit index (CFI), Tucker-Lewis index(TLI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR) metrics.Based on the path coefficients in the path diagram of the SEM, we used the Analytic Hierarchy Process (AHP) to obtain the weighted results of each meteorological factor.Correlation analysis between soil moisture andstem sap flow velocity was done using SPSS 16.0 software.

3 Results

3.1 Dynamics of meteorological factors

The changes in each meteorological factor (temperature,RH,PAR,wind speed, VPD,rainfall,andRn) during the growing season (1 May-31 October, 2020) were presented statistically (Fig.2).For temperature, its maximum value (30.8°C) occurred in early July and its minimum value(-4.5°C) in October.For RH, it fluctuated between 10% and 97% from May to July, and between 26% and 91% from August to October.For PAR, its overall trend was one of increasing at first and then decreasing, peaking at 940 μmol/(s·m2) in June.For wind speed, its highest daily average reached 3.7 m/s (in May), and its lowest value was 0.3 m/s (in May).The wind frequency increased from August to September, and generally the daily average wind speed fluctuated greatly.For VPD, its overall trend was of increasing then decreasing, peaking at 2.6 kPa in June.The total rainfall was 55.5 mm during the whole growing season in 2020 and the highest daily rainfall of 7.1 mm occurred on 30 September.Rnfluctuated with multiple peaks evident; i.e.,172.9 W/m2in June, 179.2 W/m2in July, and 175.0 W/m2in September.

Fig.2 Dynamics of seven meteorological factors during the growing season (from 1 May to 31 October) of 2020.(a), temperature; (b), relative humidity (RH); (c), photosynthetically active radiation (PAR); (d) wind speed;(e), vapor pressure deficit (VPD); (f), surface net radiation (Rn); (g), rainfall.

3.2 Dynamics of the stem sap flow velocity in H. ammodendron at different ages

3.2.1 Hourly dynamics of the stem sap flow velocity of different-agedH.ammodendronon sunny days

Hourly dynamics in the stem sap flow velocity ofH.ammodendronon six typical sunny days from May to October (20 May, 17 June, 15 July, 20 August, 16 September, and 21 October) were analyzed to reveal hourly variation characteristics of the stem sap flow velocity ofH.ammodendronat different ages (H10,H15, andH20) (Fig.3).In June, the peak value of the stem sap flow velocity ofH.ammodendronatH10 andH15 arose between 08:50 and 09:00, whereas that atH20 was delayed to 14:10.In July, the start-up time of the stem sap flow ofH.ammodendronatH10 andH15 was 05:40-06:00, and the peak value appeared at 08:50-09:20.The start-up time of the stem sap flow for the plants atH20 in July was 08:00, for which the peak occurred at 11:20.In August, the start-up time of the stem sap flow ofH.ammodendronat all three age-classes happened between 07:00 and 07:30, whose peak appeared between 11:40 and 12:00 forH15 andH20 but later at 13:30 forH10.In September, the start-up time of the stem sap flow ofH.ammodendronatH10 andH15 was 02:30, whose peaks both appeared between 13:20 and 13:30, whereas that was much later (at 07:30) forH20 though it peaked sooner at 11:10.The start-up time of the stem sap flow ofH.ammodendronin October was 02:10-02:30 forH10 andH15, and 07:40 forH20; forH.ammodendronplants at the three age-classes (H10,H15, andH20), the corresponding peak values differed considerably, appearing at 15:50, 10:40, and 12:30,respectively.

Fig.3 Hourly dynamics in the stem sap flow velocity of H. ammodendron at different ages (H10, H15, and H20)on sunny days of 20 May (a), 17 June (b), 15 July (c), 20 August (d), 16 September (e), and 21 October (f).Bars mean standard errors.

3.2.2 Hourly dynamics of the stem sap flow velocity of different-agedH.ammodendronon rainy days

In each month from May through August, four typical rainy days with high rainfall amounts (i.e.,26 May, 20 June, 10 July, and 24 August) were selected to analyze the effect of rainfall on the stem sap flow velocity ofH.ammodendron, by examining its hourly variation (Fig.4).Evidently,the stem sap flow velocity ofH.ammodendronon rainy days was significantly lower than that on sunny days, and there was a significant time lag in its response to rainfall, of about 1 h, for all age-classes.Yet the magnitude of the response of the stem sap flow ofH.ammodendronto rainfall was inconsistent amongH10,H15, andH20, for which the overall response by age could be ranked as follows:H15＞H10＞H20.

Fig.4 Hourly dynamics in the stem sap flow velocity of H. ammodendron at different ages (H10, H15, and H20)on rainy days of 26 May (a), 20 June (b), 10 July (c), and 24 August (d).

3.3 Daily dynamics in the stem sap flow velocity of H. ammodendron at different ages

Daily dynamics in the stem sap flow velocity were compared among the three age-classes ofH.ammodendronduring the growing season (from 1 May to 31 October) (Fig.5).Thestem sap flow velocity ofH.ammodendronatH10 showed an overall trend of initially rising and then declining,which peaked at 112.28 cm3/(cm2·d) on 28 August, and decreased sharply afterwards in October.The stem sap flow velocity ofH.ammodendronatH15showed a multi-peak pattern; these peaks appeared on 17 June, 19 July, and 18 September, with corresponding values of 227.11, 188.70,and 288.79 cm3/(cm2·d), respectively.The stem sap flow velocity ofH.ammodendronatH20 fluctuated greatly in May and its overall trend was one of increasing then decreasing, which peaked at 119.01 cm3/(cm2·d) on 11 August and then decreased sharply in September.

3.4 Variation in water consumption of H. ammodendron at different ages

Figure 6 shows the average daily and total water consumption ofH.ammodendronat three age-classes (H10,H15, andH20) during the growing season of 2020.The average daily water consumption was highest in May, irrespective of the age-class, with maximal values of 5.6, 5.2 and 3.6 kg/d forH10,H15, andH20, respectively.Across the entire growing season, the average daily water consumption ofH.ammodendronforH15 differed significantly from that forH10 orH20, with the values of 1.9, 2.8, and 1.9kg/d forH10,H15, andH20, respectively.The total water consumption ofH.ammodendronin the growing season was 357.3, 509.1, and 344.3 kg forH10,H15, andH20, respectively.

3.5 Correlations between the stem sap flow velocity of H. ammodendron at different ages and physical factors

3.5.1 Relationships between the stem flow sap velocity of different-agedH.ammodendronand meteorological factors

Fig.5 Daily dynamics in the stem sap flow velocity of H. ammodendron at different ages (H10, H15, and H20)during the growing season from 1 May to 31 October.(a), H10; (b) H15; (c), H20.

During the growing season, daily variation data of the stem flow sap velocity ofH.ammodendronat different ages (H10,H15, andH20) for four consecutive days of each month (19-22 May,Fig.6 Variations in the average daily water consumption ofH.ammodendronat different ages (H10,H15, andH20) in each month of the growing season (from May to October) as well as the total water consumption.AV is the average daily water consumption averaged throughout the growing season, and TWC is the total water consumption throughout the growing season.In the analysis of variance, different lowercase letters indicate significant difference atP＜0.05 level among theH.ammodendronat different ages for the same period.Bars mean standard errors.16-19 June, 14-17 July, 19-22 August, 14-17 September, and 19-22 October) were selected,along with corresponding meteorological data (temperature, RH,Rn, wind speed, and VPD), to fit the SEM.As seen in Table 2, both the Kronbach and KMO test coefficients of the variables forH10,H15, andH20 surpassed the minimum acceptable values; hence, the correlations between the variables were deemed reliable.The second-order validation factor analysis of the SEM theoretical model was performed (using MPLUS 8.0 software), and these results demonstrated that the SEM combined model fitted well: χ2/df=1.202, RMSEA=0.049, SRMR=0.073,CFI=0.984, and TLI=0.927.According to previous research (Sun et al., 2007), the smaller the value of χ2, the better the effect; the applicable criteria for CFI and TLI are greater than 0.900,and the minimum acceptable criteria for RMSEA and SRMR are lower than 0.090.Therefore, all metrics satisfied the critical value requirements for establishing the SEM model's goodness-of-fit.

Table 2 Reliability and validity test statistics of sample data for the structural equation model (SEM) for H.ammodendron at different ages (H10, H15, and H20)

In Figure 7, meteorological factors were clearly correlated with the stem sap flow velocity ofH.ammodendronat different ages.According to Table 3, in considering the respective effects of meteorological factors on the stem sap flow velocity ofH.ammodendron, temperature had the largest weight; hence, temperature was the key factor affecting the stem sap flow ofH.ammodendron.The effects of the five meteorological factors upon the stem sap flow velocity of different-agedH.ammodendronplants can be ranked as follows: temperature＞Rn＞VPD＞wind speed＞RH forH10 andH15, and temperature＞VPD＞Rn＞wind speed＞RH forH20 (Table 3).

Fig.7 Path diagram of the structural equation model (SEM) showing the response of the stem flow velocity of H. ammodendron to meteorological factors during the growing season.** indicates statistically significant correlation at P＜0.01 level; * indicates statistically significant correlation at P＜0.05 level; e1 to e6 represent residual variables; the number 1 represents the factor load; the value in parentheses represents the root mean square of the residuals ; the variable pointed by the arrow is an endogenous variable.

Table 3 Calculation results of the meteorological factors' weights for H. ammodendron at different ages (H10,H15, and H20)

Moreover, the stem sap flow velocity ofH.ammodendronat the three age-classes showed correlations with various meteorological factors under different weather conditions (Table 4), and these correlations were significantly greater on sunny days than on rainy days.Under sunny conditions, VPD was the main factor influencing the stem sap flow velocity.The correlations between stem sap flow velocity and meteorological factors changed across months and shifted among different ages.Rainfall showed a very significant correlation with stem sap flow velocity(P＜0.01), for which the ranking of correlation coefficients wasH15＞H20＞H10.

Table 4 Correlation analysis of hourly-scale stem sap flow velocity of H. ammodendron at different ages (H10,H15, and H20) and meteorological factors

3.5.2 Relationships between the stem sap flow velocity of different-agedH.ammodendronand soil moisture

On sunny days, the stem sap flow velocity ofH.ammodendronat each age-class had very significant correlations with soil moisture at soil depths of 50 and 100 cm (Table 5).Specifically,stem sap flow velocity and soil moisture were negatively correlated in the upper soil depth, and positively correlated in the deeper soil depth.Moreover, the correlation coefficients on sunny days were ranked asH10＞H15＞H20.On rainy days, in contrast, the correlations between stem sap flow velocity and soil moisture were weaker and not significant.

Table 5 Correlation analysis between the stem sap flow velocity of H. ammodendron at different ages (H10,H15, and H20) and soil moisture at different depths

4 Discussion

4.1 Variation in the stem sap flow velocity of H. ammodendron at different ages

In May, compared withH10 andH20, the stem sap flow velocityofH.ammodendronplants atH15 was larger and their stem sap flow started earlier but ended later in a day.The reason for this is that the crown canopy diameter ofH.ammodendronplants atH15 was relatively large (Table 1).Moreover, the temperature at the forest stand was suitable in May (Fig.2), and winds were more frequent then (Guo et al., 2016).These winds reduced the humidity around the leaves ofH.ammodendron, affecting the stem sap flow velocity by increasing the VPD, and May's suitable natural conditions resulted in greater stem sap flow velocityH.ammodendronplants atH15 in that month.Recently, Xia et al.(2019) reported that the stem sap flow velocity of 'Hanfu' apple trees showed a single-peak "Z" shape change and its value was higher in the daytime and lower at night.In this study, we found that the hourly variation curves of the stem sap flow velocityofH.ammodendronat the three age-classes under sunny conditions were mainly unimodal, with weak peak at night and significant hourly differences.A similar conclusion was also drawn forElaeagnus angustifolia(Liu et al., 2021).In May, the stem sap flow ofH.ammodendronplants atH20 started late and ended early, and their flow velocity peaked later, with the values significantly smaller than that ofH.ammodendronplants atH10 andH15 at the germination stage of their newly assimilated branches.A plausible explanation for this is the low rainfall and soil moisture in May and June.Furthermore,H.ammodendronplants atH20 showed signs of branch dieback and rodent damage (Luo et al., 2017), pests, and diseases (Ma et al., 2012), as well as the natural dieback of their own branches in this water-scarce environment.Collectively, this led to smaller crown diameter size (Table 1), decreased surface accumulation, poor water-holding capacity, and reliance on groundwater for most of the water sources these plants used.

So now he was king, but no one could persuade him to give up Wallis. Not even Prime Minister Stanley Baldwin who spoke6 for the nation when he said Britain did not want an American divorcee for a queen. Ministerial car shuttled between Westminster and Buckingham Palace but the king could not be budged7. He was forced to abdicate8（，） and all over the country flags flew at half-mast.

In October, the temperature dropped sharply, the self-metabolism ofH.ammodendronplants slowed, and their absorption and transpiration of soil moisture decreased.The stem sap flow velocity ofH.ammodendronplantswas starkly different amongH10,H15, andH20.On sunny days, their hourly variation curves mainly displayed a single peak, with weak stem sap flow at night, and the difference between daytime and night was significant.On rainy days, the stem sap flow velocity of different-agedH.ammodendronplants was significantly lower than that on sunny days, and could be ranked asH15＞H10＞H20.The reason for this phenomenon is that the canopy ofH.ammodendronplants can directly absorb water during rainy days and then contribute to the stem sap flow.The larger the crown canopy diameter, the more water it can absorb.This result is consistent with earlier findings of Xia et al.(2014) and Yueet al.(2020) on the water consumption dynamics ofH.ammodendronin the Gurbantunggut Desert, China.

4.2 Variation in the water consumption of H. ammodendron at different ages

In May, the average daily water consumption ofH.ammodendronplantsat different ages (H10,H15, andH20) reached the maximum, likely because of a confluence of factors.In the study area,the plants' branches started to sprout in April (Zhang et al., 2021), the temperature varied between 5.0°C and 25.0°C, and the weather was windiest in May.This windy weather increased the transpiration rate ofH.ammodendron, which together with the stronger photosynthetic effective radiation and frequent rainfall created suitable environmental conditions, allowing for the rapid development and growth of the rhizomes and assimilated branches ofH.ammodendron(Fig.3).Similarly, Jia et al.(2020) argued that due to the timing of the rainy season and drought stress, the maximum water consumption of some shrubs actually occurs in May.However, work by Cao et al.(2013) and Zhang et al.(2016) concluded that the maximum water consumption byH.ammodendronoccurs in July.That discrepancy with our study can be explained by the different rainfall and temperature regimes during the observation period among studies, given that temperature and rainfall are key factors known to determine the water consumption ofH.ammodendronplants.We found that temperatures were highest in June and July during the observation period when rainfall was scarce.Accordingly, this higher temperature would have promoted the stomata closure ofH.ammodendron, thereby reducing photosynthesis and slowing the material conversion and energy metabolism of plants, which entails consuming less water.In June and July, compared withH10 orH20,H.ammodendronplants atH15 consumed much more water, because they had larger canopy diameter and transpired strongly at high temperatures.The most rainfall occurred in September, when the average daily water consumption ofH.ammodendronat the three age-classes was rather low.Because the temperature decreased and diurnal temperature difference was so large in September, photosynthetic effective radiation was diminished and the abscisic acid content ofH.ammodendronwas at its greatest level (Ma et al.,2012).This would have cuedH.ammodendronplants to enter dormancy (Zhao et al., 2017; Gao et al., 2020), thus leading to low water consumption.

By comparing the water consumption ofH.ammodendronin different regions (Table 6), we found that the water consumption ofH.ammodendronplants atH10 andH20 in our study area was similar to that ofH.ammodendronplants native to the Gurbantunggut Desert, China, while the water consumption ofH.ammodendronplants atH15in our study area was on par with the water consumption ofH.ammodendronplants native to the Badain Jaran Desert, China.

Table 6 Comparison of the water consumption of H. ammodendron plants studied in the different regions

Previous research has shown that theH.ammodendronplantations may undergo a pronounced self-thinning process after more than 10 years (Song et al., 2021; Zhao et al., 2023).ForH.ammodendronplantation atH10 in this study, the number of surviving plants decreased, the shrub belt structure disappeared, and the soil water consumption of the entire plant community gradually declined (Song et al., 2021; Liu et al., 2022).Consequently, forH.ammodendronplantation atH15, there were only a few plants remaining but soil water consumption was better balanced.Further experimental findings showed that, even thoughH.ammodendronplants atH20 can use groundwater, their mortality rate in the plantation was still high.Hence, the death rate ofH.ammodendronplants increased with stand age.

4.3 Responses of the stem sap flow velocity of H. ammodendron at different ages to physical factors

Temperature was the key factor governing the stem sap flow velocity ofH.ammodendron.In the effective temperature range, as the temperature rose, the growth rate ofH.ammodendronplants accelerated.When the temperature reached the light saturation point,H.ammodendronreached its maximum water consumption (Sun et al., 2010; Mahdavi et al., 2022), so temperature had a high weight among the meteorological factors (Table 3).Rninfluenced the photosynthetic rate ofH.ammodendron, and as a heat source, this net surface radiation warmed or cooled the surface and advective atmosphere, with heat consumed by evaporation and transpiration (Dong, 2013).In this way, the influence ofRnupon the stem sap flow velocity ofH.ammodendronplants atH10 andH15 took on a high weight.The leaf stomata were pores consisting of a pair of guard cells, which functioned as key links in regulating the exchange of materials and energy such as water and CO2between plants and their external environment; thus, stomatal changes profoundly affected transpiration and photosynthesis rates (Xia et al., 2014).VPD reflected the degree of atmospheric aridity and evaporative stress of plants (Yang et al., 2018).When the air humidity decreased, the stomata closed due to the increase of VPD, which altered the stem sap flow velocity ofH.ammodendron.In this way, VPD indirectly improved photosynthetic performance, so it took on a high weight forH.ammodendronplants atH20.Although both wind speed and RH exerted some influence on thestem sap flow velocity ofH.ammodendron, theircorresponding weights were much lower.

Under sunny conditions, the stem sap flow velocityofH.ammodendronand soil moisture at the depths of 50 and 100 cm showed strong significant correlation.The younger the stand age ofH.ammodendronplants, the more significant was the correlation.The reason for this is that the soil layer at the depth of 0-100 cm in the study area consisted of sandy soil with a poor water-holding capacity (Guo et al., 2016).Underground of the study area at the 100 cm depth belowground did have a certain water-holding capacity, and the depth at which roots ofH.ammodendronconsumed water increased with stand age (Zhang et al., 2016).Therefore, the older the stand age of this species, the better was the correlation between thestem sap flow velocity and soil moisture in the deep layer.At 50 cm belowground, the stem sap flow velocity ofH.ammodendronplants at the three age-classes (H10,H15, andH20) showed a very significant negative correlation with soil moisture.This result could be due to the fact that sandy soils in the top 50 cm depth layer were involved in surface evaporation in the face of high temperatures, thus reducing water availability around the root system ofH.ammodendronand inhibiting its water consumption (Zhou et al., 2017).At the 100 cm belowground, thestem sap flow velocity ofH.ammodendronand soil moisture were significantly positively correlated, indicating that the soil in the layer below 100 cm in the study area had an ability to retain water, and that most of the water-absorbing roots ofH.ammodendronwere distributed below 100 cm ground (Xu et al.,2021).

5 Conclusions

This study's results showed that the hourly curves of thestem sap flow velocity ofH.ammodendronfeatures a single peak on sunny days and multiple peak on rainy days.The magnitude of the stem sap flow velocity is significantly lower on rainy days than on sunny days.Temperature is the chief factor affecting the shifts in the stem sap flow velocity ofH.ammodendron, whose correlation is more significant in younger stands.The older the stand age of these plants, the deeper is their use of soil moisture.The water consumptionofH.ammodendronplants in the Minqin oasis-desert transition zone reaches its maximum at an age of 15 years old(H15), compared to theH10 andH20.AlthoughH.ammodendronplants inH20 could use groundwater, its water consumption is low and its mortality rate likely very high.This may be because the water consumption ofH.ammodendronplants causes the groundwater level to drop,such that the groundwater absorbed byH.ammodendronplants is insufficient to maintain their physiological activity.Therefore, the initial density ofH.ammodendronforest stands must be reasonably planned to mitigate their future consumption of valuable soil moisture in arid regions,so as to maintain the stability ofH.ammodendronplantations there and facilitate greater ecological benefits.

Conflict of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This study was supported by the National Natural Science Foundation of China Subsidization Project (32260425,31860238) and the Natural Science Foundation of Gansu Province, China (32060246, 21JR7RA733).

Author contributions

XU Xianying and ZHANG Jinchun designed the experiment.ZHANG Jinchun provided funding for this study.QIANG Yuquan, LIU Hujun and DUAN Xiaofeng completed the experiment and QIANG Yuquan wrote the manuscript.