Physiological Tolerance to Heat and Drought Are Decoupled Across Common Woody Urban Greening Species in North China

ABSTRACT

Urban vegetation is increasingly exposed to the compound heat and drought stress due to global warming and urban heat island effects, yet heat tolerance and its linkage with drought resistance or photosynthesis remain unclear for plants in urban environments. We evaluated corresponding traits in nine woody species common to northern China's highly urbanized megacity cluster dominated by temperate continental monsoon climate. Specifically, we assessed: (1) temperature thresholds causing declines in the maximum quantum yield of photosystem II (F _v /F _m ), (2) leaf resistance to xylem embolism and turgor loss, and (3) leaf gas exchange and biochemical efficiency of photosynthesis of field‐grown, mature plants, but under lab conditions. We also recorded the in situ F _v /F _m and leaf temperature under contrasting air temperatures to assess whether and how plants maintained functional integrity of the photosynthetic apparatus under heat. We showed that stress tolerance and photosynthetic traits differed markedly among species. The overall weak heat tolerance of these plants resulted in significantly decreased F _v /F _m in four species under high air temperature, while species with a wider thermal safety margin, primarily determined by leaf temperature, retained greater functional integrity. Traits associated with leaf heat tolerance, drought tolerance, and photosynthetic efficiency were decoupled. Moreover, correlations were identified between heat tolerance traits and climatic metrics, indicating the variability in climate of species distributional range. Findings of this study add to the limited knowledge regarding the physiological resistance of urban greening plants and may provide reference during the establishment of green infrastructure in this region.

1. Introduction

Earth's climate system is highly sensitive to changes in global energy balance (Huber and Knutti 2012; Stephens et al. 2012), as even subtle perturbations in average surface temperature can result in considerable changes in the climate (Chen et al. 2018). Global average temperature has risen by 1°C in the past century and is expected to continue rising in the future (IPCC 2023). Consequently, the frequency, duration, and intensity of climate extremes, such as heat events, are expected to increase (Domeisen et al. 2023). The prevalence of heat extremes threatens all living organisms, including plants (Kunert et al. 2022; Marchin, Backes, et al. 2022; Marchin, Esperon‐Rodriguez, et al. 2022; Slot et al. 2021; Zhu et al. 2018), subsequently undermining the structure and function of terrestrial ecosystems.

Urban ecosystems are projected to accommodate up to 70% of Earth's population by 2050, but are particularly susceptible to climate anomalies due to oversimplified ecosystem structure and fluxes (Alizadeh and Hitchmough 2019; Esperon‐Rodriguez, Tjoelker, et al. 2022). Vegetation typically accounts for a small portion of urban land cover but provides multiple benefits to urban residents (Alizadeh and Hitchmough 2019; Farrell et al. 2022). However, heat extremes, compounded by urban heat island (UHI) effects, are increasingly challenging the survival and performance of urban vegetation (McCarthy et al. 2010; Esperon‐Rodriguez, Tjoelker, et al. 2022; Esperon‐Rodriguez, Rymer, et al. 2022), with severe canopy dieback or mortality recorded for urban plants during heatwave events in many cities (Marchin, Esperon‐Rodriguez, et al. 2022; Zhang et al. 2024). It has been estimated that more than half of urban species in major cities worldwide are at risk of heat injury, with this proportion expected to reach 76% by the middle of this century (Esperon‐Rodriguez, Tjoelker, et al. 2022). Hence, understanding the heat resistance of urban greening species has become an urgent need for sustaining ecosystem services provided by vegetation (Farrell et al. 2022).

Plant physiology primarily regulates how vegetation responds to the environment, thus offering tools for understanding and quantifying plant stress resistance. Plant heat resistance can be considered as the maintenance of metabolic functions that are fundamental for plant fitness under extreme heat (Feeley et al. 2020; Slot et al. 2021; Zhu et al. 2018), which can be quantified by various metrics (Geange et al. 2021). Here, we define plant heat resistance as the ability to sustain the functional integrity of photosynthetic apparatus under heat, and specifically use the temperature dependence of the maximum quantum yield of PSII (F _v /F _m ) to reveal heat tolerance thresholds (Slot et al. 2021; Zhu et al. 2018). Recent work has demonstrated considerable interspecific variation in these thresholds across species occupying thermally contrasting habitats (Kitudom et al. 2022; Feeley et al. 2020; Slot et al. 2021; Zhu et al. 2018), and the high risk of photosynthetic failure in plants exposed to heat (Kunert 2024; Kunert et al. 2022; Manzi et al. 2024). Notably, many of these studies focus predominantly on natural vegetation, while knowledge regarding the heat tolerance of urban plants is still confined to limited cities (e.g., Mantova et al. 2024; Marchin, Esperon‐Rodriguez, et al. 2022; Zhang et al. 2024), despite the detrimental influence of heat stress on urban vegetation globally (Esperon‐Rodriguez, Tjoelker, et al. 2022). In addition, empirical studies have linked the heat sensitivity of PSII with leaf necrosis and crown dieback during heatwave events (Marchin, Esperon‐Rodriguez, et al. 2022; Valliere et al. 2023; Zhang et al. 2024). Nonetheless, it remains unclear how heat tolerance contributes to species adaptation from a physiological perspective.

Maintaining the functional integrity of photosystems under high temperatures entails multiple biochemical or biophysical solutions (Wahid et al. 2007). From a trade‐off perspective, an increase in stress tolerance may demand extra metabolic input and therefore influence the allocation of metabolic functions among organs or processes (Tiwari et al. 2026). For example, heat tolerance can be facilitated by the synthesis of heat shock proteins (HSPs) (Altschuler and Mascarenhas 1982; Aspinwall et al. 2019), or by increasing the proportion of saturated fatty acids in the thylakoid membrane (Sage and Kubien 2007; Zhu et al. 2018). Both adjustments can stabilize membrane structure under heat, but may compromise biochemical efficiency of photosynthesis, because synthesis of nitrogen‐rich proteins such as HSPs reduces nitrogen available for allocation to Rubisco (Heckathorn et al. 1996a, 1996b), and increased saturated fatty acid content will lower membrane fluidity, thereby decreasing the rate of electron transport (Sage and Kubien 2007). Likewise, dissipating heat through transpirational cooling can facilitate adaptation to hot environments by avoiding leaf overheating, if plant water access is not limited by soil water availability (Gauthey et al. 2024; Marchin, Backes, et al. 2022). Therefore, high heat tolerance thresholds may be less necessary in species that can maintain leaf temperature within a safe range through latent heat dissipation. However, excessive water loss due to high transpiration rates under heat can lower water potential, thereby increasing the risk of hydraulic failure that may cause organ or whole‐plant death (Choat et al. 2018). As a corollary, the existence of a trade‐off between heat‐tolerance thresholds and hydraulic traits conferring drought tolerance can be expected (Valliere et al. 2023). The relationships between heat tolerance, biochemical efficiency of photosynthesis, and plant drought resistance can be informative for species screening during the establishment of urban forests, given the prominent influence of these processes on growth and stress tolerance; yet, these relationships have rarely been tested empirically (but see Posch et al. 2024).

Given that higher heat resistance can contribute to plant performance by minimizing heat‐induced tissue injury, it can be inferred that natural selection will act on plant thermal adaptation, resulting in higher heat tolerance thresholds in species primarily occupying hot environments. Indeed, evidence from transect studies has shown that temperature triggering 50% loss of physiological function, such as quantum yield of PSII, tends to increase with rising average air temperature (O'sullivan et al. 2017; Slot et al. 2021; Zhu et al. 2018; Kitudom et al. 2022; Fadrique et al. 2022). Physiological mechanisms underpinning the spatial variation of heat tolerance thresholds along temperature gradients are complicated, and largely unclear, but can involve adjustments in both leaf anatomical structure (Li et al. 2022) and biochemistry (Aspinwall et al. 2019; Zhu et al. 2018; Way et al. 2011). Nonetheless, a quantitative understanding about the environmental control over plant heat tolerance remains incomplete. With respect to urban greening species, it has been shown that species planted beyond their climatic envelope (i.e., distributional range) are more susceptible to heat stress (Esperon‐Rodriguez, Rymer, et al. 2022; Esperon‐Rodriguez et al. 2024). Hence, a climate‐of‐origin analysis for selecting the most tolerant species for urban greening is appropriate given that relevant climatic metrics are commonly easy to obtain from open‐source databases (Hanley et al. 2021). Yet, relationships between stress resistance and the climate envelope are mixed (Hanley et al. 2021; Han et al. 2022; Li et al. 2022; Esperon‐Rodriguez et al. 2024; Kibria et al. 2024), suggesting that we need further evidence to make appropriate decisions.

The Jing‐Jin‐Ji region of China is dominated by a monsoon‐influenced humid continental climate, with hot, humid summers due to East Asian monsoons and cold, dry winters caused by the Siberian anticyclone. This megacity cluster represents the most densely populated region in north China, accounting for ca. 2.3% of China's land cover but approximately 7.2% of the total population; thus, this is representative of high levels of urbanization (Li et al. 2017). The negative consequences of intense urban development are also evident, especially in Beijing, where the UHI effect can raise air temperatures by up to 5°C in the city compared with the surrounding suburbs (Liu et al. 2007; Wang et al. 2020). UHI‐related urban warming is further magnified by heat extremes, which are often accompanied by drought, leading to extreme hot dry events that compromise the function of urban vegetation (Zhang et al. 2014). Taking advantage of the heat event that swept across northern China in the early summer of 2024, we studied the leaf heat tolerance of nine common urban greening woody species of the metropolitan region. Traits related to leaf heat tolerance, drought resistance, and photosynthesis were measured ex situ under lab conditions, while leaf F _v /F _m was assessed in the field during periods of both mild and high ambient air temperature. Moreover, leaf temperature and water potential were additionally assessed in the field during hot days to evaluate the risk of physiological injury under heat. Our overarching aim was to assess the heat tolerance of these species and its role in sustaining the functional integrity of the photosynthetic apparatus under heat. Moreover, correlations among heat tolerance, drought resistance, and photosynthesis, as well as the relationship between heat tolerance and metrics of species climate envelopes, were assessed to reveal the interaction among ecological strategies and the potential climate drivers of heat tolerance.

Specifically, we hypothesized that: (1) in situ quantum yield of PSII would be compromised under high growth temperature, with the negative effect ameliorated by either higher heat tolerance or leaf gas exchange rate; (2) there would be trade‐offs between traits conferring heat tolerance and biochemical efficiency of photosynthesis because of resource allocation, while heat tolerance would also be negatively related to leaf drought tolerance due to strategic redundancy; and (3) leaf heat tolerance underpins species distributional range; hence, showing correlations with variables representing the species climate envelope.

2. Materials and Methods

2.1. Study Site and Plant Materials

The study site is located in Beijing (39°57′36″ N, 116°17′54″ E), which is one of the most densely built and most populous cities in China. Mean annual temperature (MAT) and precipitation (MAP) of the city are 12.7°C and 557 mm, respectively. Specifically, the mean maximum temperature for the hottest month (July) is ca. 31.8°C based on the 30‐year mean (1991–2020). The variation of relative humidity is also seasonal, ranging from an average of ca. 45% in spring and winter to ca. 72% in summer, with the corresponding range of vapor pressure deficit (VPD) being 0.2–1.32 kPa. The impact of climate change has been increasingly significant in recent years, leading to prevailing hot days in late spring and early summer characterized by consecutive days with maximum temperature > 35°C, with heat events projected to be more common in the future (Qian et al. 2024). The topography is relatively flat across the urban region, with elevations ranging from 40 to 60 m above sea level.

Species targeted in the current study were representative of common species used in urban greening in the city, as well as the Jing—Jin—Ji region, encompassing two distinct growth forms with five shrubs and four trees. The shrub species were Euonymus japonicus Thunb., Forsythia viridissima Lindl., Ligustrum lucidum W.T.Aiton, Rosa xanthina Lindl., and Syringa oblata Lindl., while the tree species included Acer truncatum Bunge, Cotinus coggygria Scop., Ginkgo biloba L., and Populus × tomentosa Carrière. These species are naturally distributed across temperate East Asia and occupy diverse habitats ranging from mesic deciduous forests to forest margins and open woodlands, often experiencing pronounced seasonal water limitation, with marked differences in climatic variables of species distributional range (Table 1; Wu et al. 1994; Fang et al. 2011).

TABLE 1.

Species names, assigned colors, family, growth form, and climatic variables for species climate envelopes including mean annual temperature (MAT), mean annual precipitation (MAP), mean temperature of warmest month (TWM) for species selected in this study. Also shown are plant height (H) for shrubs and diameter at breast height (DBH) for trees.

Species	Color	Family	Growth form	MAT (°C)	MAP (mm)	TWM (°C)	H/DBH (cm)
E. japonicus		Celastraceae	Shrub	11.61	1040	23.7	176.5 ± 4.3
F. viridissima		Oleaceae	Shrub	13.25	1152	30.6	157.2 ± 4.5
L. lucidum		Oleaceae	Shrub	17.09	934	29.7	102.5 ± 2.2
R. xanthina		Rosaceae	Shrub	10.16	555	29.4	215 ± 13.1
S. oblata		Oleaceae	Shrub	10.2	621.5	28.6	262.5 ± 35.6
A. truncatum		Sapindaceae	Tree	11.76	709	28.8	41.2 ± 3.1
C. coggygriaca		Anacardiaceae	Tree	11.19	825	25.8	362.5 ± 7.5
G. biloba		Ginkgoaceae	Tree	12.17	1183	29.3	53.5 ± 5.4
P. tomentosa		Salicaceae	Tree	12.88	659	30.6	52 ± 5.9

Note: Data are shown as mean ± standard error of mean (SE, n = 4–5).

All species in the study were street or landscape plants, which were growing in the ground, on the Haidian campus of Minzu University of China. All plants were > 15 years old, with shrubs and trees approximately 1.5–3 m and 10–20 m tall, respectively. Plants received natural sunlight, but were irrigated regularly throughout the growing season. The historical ranges of monthly mean air temperatures and precipitation are −3.1°C to 26.4°C and 3–185 mm (1991–2020), respectively, at the site of sampling (i.e., Haidian District). For each species, 4–5 fully sunlit individuals of similar size were selected for sampling plant materials. Within each species, we chose individuals that grew apart to avoid sampling bias due to shared growth environments or competition.

Traits associated with leaf heat tolerance, pressure–volume characteristics and leaf photosynthesis were measured on plant samples collected in the field, in mid‐May 2024, which represented the peak growth season of these species. Ambient air temperature (T _air) during the period of measurement was relatively mild (22.3°C on average) and close to the multi‐year average (20.0°C). Note that the leaf vulnerability curves were sourced from our previous study (see below). In addition, leaf maximum quantum yield of PSII (F _v /F _m ) was measured in the field on 19–23 May and 10–17 June, when T _air varied substantially. Moreover, midday leaf water potential and leaf temperature were measured on plants growing in the field during the period of high T _air (10–17 June; Figure 1).

FIGURE 1.

Climatic variables including daily maximum air temperature (°C), daily maximum vapor pressure deficit (VPD, kPa) and daily cumulative precipitation (mm) at the experimental site from April to July, 2024. Sampling and measurements for leaf physiological traits, including in situ leaf maximum quantum yield of PSII (F _v /F _m ), leaf heat tolerance, photosynthesis, and pressure—volume traits were performed during 19th–23rd May, when ambient air temperature was relatively mild. Leaf in situ F _v /F _m , leaf temperature and water potential were additionally recorded from 10th to 17th June, to capture the physiological status under high growth temperatures. Periods for these measurements are highlighted by the shaded regions in the figure. During these periods, average daily maximum temperature was 28.1°C and 34.9°C, respectively.

2.2. Leaf Heat Tolerance and In Situ Quantum Yield of PSII

Leaf heat tolerance was assessed by measuring the temperature dependence of F _v /F _m for heat‐treated leaves. For each species, 1–2 leafy branches, approximately 70 cm long, were excised from the south‐facing, upper canopy of each targeted individual in the field at predawn, and then transported to the laboratory within 30 min. Fully expanded, undamaged leaves were detached from branches of the same individual, and a square leaf lamina (2 cm on each side) was cropped from each leaf while avoiding the midrib for species with large leaves (e.g., P. tomentosa and F. viridissima ) to facilitate treatment (Zhang et al. 2024). In total, 30–40 leaf samples were obtained for each targeted plant, allowing assessment of individual heat tolerance, with 4–6 samples allocated to each temperature treatment in the thermal gradient. Following Curtis et al. (2014), seven treatment water baths (35°C, 40°C, 45°C, 50°C, 55°C, 60°C, and 70°C) and one reference water bath (25°C) were used. For each temperature gradient, leaf samples were placed in zip lock bags humidified with a moist paper towel and were incubated in the water bath for ca. 15 min (15 ± 1.3 min on average), followed by 24 h dark adaptation in the bag under lab conditions (ca. 25°C). We acknowledge that the heat tolerance thresholds derived from the temperature response of leaf F _v /F _m can be sensitive to the duration of heat treatment, such that longer heat exposure (e.g., 30 min) can potentially lead to lower thresholds compared with a short treatment duration (e.g., 15 min; Didion‐Gency et al. 2025). Here, the duration of heat treatment was methodologically consistent with previous work (e.g., Ahrens et al. 2021; Zhang et al. 2024), thereby facilitating comparisons of traits across studies. Leaf F _v /F _m was measured using a handheld chlorophyll fluorometer (MultispeQ V 2.0, PhotosynQ Inc). Leaf samples were clamped in the device, with the adaxial surface facing the light guide. In the case of small leaves (e.g., R. xanthina ), multiple leaves were clamped without overlap to ensure the light guide was fully covered (https://www.photosynq.com/). For each leaf sample, the measurement typically took ca. 18 s, and chlorophyll fluorescence characteristics, including maximum, minimum fluorescence, and F _v /F _m were recorded. Temperature response curves of F _v /F _m were generated by plotting measured F _v /F _m against corresponding treatment temperature, and were fitted with a two‐parameter Weibull function (Duursma and Choat 2017), which generally produces better fits than the sigmoidal model. Temperature thresholds triggering 15% and 50% loss of quantum yield of PSII relative to the maximum were considered as the critical temperature thresholds at the incipient (T _crit, ^oC) and substantial loss (T ₅₀, ^oC) of photochemical efficiency, respectively (Zhang et al. 2024).

Moreover, for each targeted individual, 1–2 fully expanded leaves from the mid to upper canopy were randomly selected for measuring in situ F _v /F _m under different T _air. Measurements were consistently conducted with the same device (i.e., MultispeQ) before sunrise, thereby allowing the leaves to dark‐adapt overnight to achieve adequate openness of PSII. For tall trees, in which canopy leaves were difficult to reach, a ca. 15 cm long twig with fully expanded, undamaged leaves were collected using a pole saw and measurements were taken within 1 min of leaf excision. During the two rounds of measurement, average T _air at predawn was 20.2°C ± 0.7°C and 22.7°C ± 0.9°C, respectively.

2.3. Leaf Hydraulic Traits

Leaf xylem vulnerability to embolism was assessed by constructing leaf vulnerability curves with the optical visualization technique (Brodribb et al. 2016). Leaf xylem embolism resistance was quantified by calculating the water potential triggering 50% xylem embolism (P ₅₀, MPa). Note that P ₅₀ was determined based on the presence of xylem area with distinct optical characteristics that represent embolism during dehydration, rather than water potential at 50% loss of leaf water transport efficiency. In the current study, leaf P ₅₀ data were sourced from our previous work conducted at the same site, using branches with mature leaves collected from the same targeted individuals of our current study (Han et al. 2022). Details for experimental protocols were described therein.

For leaf pressure‐volume (PV) traits analysis, 1–2 fully expanded, undamaged leaves were collected from the south‐facing, upper canopy of each targeted individual of each species at predawn. Leaf samples were quickly transported to the laboratory in humidified zip lock bags and rehydrated in darkness by submerging the excised petiole in purified water for up to 3 h. Leaf PV curves were generated using the standard method described in Lenz et al. (2006). Briefly, leaf samples were bench dried at room temperature (i.e., 25°C), with leaf water potential (Ψ_leaf, MPa) and fresh mass (FM, g) measured periodically until samples were visually wilted and Ψ_leaf < −2.5 MPa. Leaf samples were then oven dried at 75°C until constant dry mass was obtained (DM, g). Leaf PV data were analyzed using the “Pressure volume analysis spreadsheet tool” (https://prometheusprotocols.net/). Leaf turgor loss point (P _tlp, MPa) was calculated by plotting the inverse of leaf water potential (1/Ψ_leaf) against leaf relative water content (RWC_leaf, %) and was taken as the point where the line became non‐linear.

2.4. Leaf Photosynthetic Traits

For each species, 1–2 terminal branches that were approximately 70 cm in length were collected from the upper canopy of each selected individual between 9 and 11 am. The cut end of the excised branch was immediately submerged in a bucket filled with water, and a ca. 15 cm segment was removed by recutting the branch 3–4 times under water. Branch samples were then sent to the laboratory and were allowed to rehydrate for 1 h at room temperature and ca. 700 μmol m⁻² s⁻¹ photon flux density (PFD) provided by an LED growing light. Leaf gas exchange measurements were performed using a portable photosynthesis system (Model 6800, Li—Cor) equipped with a red—blue LED light source and an external CO₂ injector. For each branch, one recently mature, fully expanded leaf was chosen and was placed into the cuvette supplied with 1200 μmol m⁻² s⁻¹ PFD and 420 μmol mol⁻¹ CO₂, with the temperature and vapor pressure deficit (VPD) inside the cuvette maintained at 25°C and 1.2–1.5 kPa (ca. 60% relative humidity), respectively. Leaf gas exchange traits, including light‐saturated net photosynthetic rate (A _net, μmol m⁻² s⁻¹), stomatal conductance (g _s , mol m⁻² s⁻¹) and transpiration rate (E, mmol m⁻² s⁻¹), were logged when these readings were visually stable, which usually took ca. 15 min. Thereafter, the same leaf was used for measuring the photosynthetic CO₂ response (AC_i) curve, which describes the relationship between A _net and intercellular CO₂ concentration (C _i ). This was generated by recording A _net under the following CO₂ concentrations in the cuvette: 400, 300, 200, 100, 50, 400, 600, 800, 1000, 1200 μmol mol⁻¹, with 60–180 s intervals allowed for stabilization under each CO₂ concentration. During the ACi curves, values for PFD, temperature and VPD inside the cuvette were kept at 1200 μmol m⁻² s⁻¹, 25°C, and 1.2–1.5 kPa, respectively. Leaf AC_i curves were fitted to the mechanistic model of C₃ photosynthesis (i.e., FvCB model) following the procedure described by Sargent et al. (2024) to obtain maximum carboxylation rate of Rubisco (V _cmax, μmol CO₂ m⁻² s⁻¹) and electron transport rate (J _max, μmol e⁻ m⁻² s⁻¹).

2.5. In Situ Leaf Temperature and Water Potential During Heat Event

Leaf temperature (T _leaf, ^oC) and Ψ_leaf under field conditions were measured between 12 and 2 pm on days 160, 164, and 169, with the daily maximum T _air being 31°C, 36.5°C and 37.2°C, and daily maximum VPD being 2.20, 4.20, and 4.69 kPa, respectively; thus, these measurements were representative of hot dry conditions during the growth season of 2024. For each targeted individual, T _leaf was measured on sunlit leaves with a thermal camera (H16, Hikmicor). The thermal camera was radiometrically calibrated prior to measurements following the manufacturer's instructions (https://www.hikmicrotech.com/). In addition, temperature readings were verified and corrected using a calibrated K‐type thermocouple placed in direct contact with the leaf surface during a preliminary test. On each day of measurement, the device was allowed to warm up in situ for at least 15 min before use. For leaf temperature measurements, emissivity and reflected (background) temperature of the device were set at 0.96 and ambient temperature, respectively (Harrap and Rands 2021). For shrubs, thermal images were consistently captured at ca. 15 cm from the surface of targeted leaf; for tall trees where the upper canopy was difficult to access, thermal images were obtained in long‐distance mode of the device (up to 8 m), with adjacent buildings being used as elevated platforms to obtain suitable viewing angles. Sunlit outer canopy leaves near the upper crown were targeted. For each image, at least five regions were manually delineated on fully sunlit outer canopy leaves, and the mean leaf temperature was calculated from these regions. Leaf margins and canopy gaps were excluded to reduce mixed‐pixel effects and background interference. Targeted leaves were always positioned at least 3 m apart from building surfaces to avoid potential thermal interference. During measurements, the angle for image capture was always maintained ≤ ca. 45°, thereby ensuring the visibility of most leaf surfaces. In addition, 2–3 fully sunlit leaves were collected from the mid to upper canopy of selected individuals for each species using a pole saw. Leaves were kept in humidified zip‐lock bags and were quickly transported to the lab. Leaf water potential was determined using a Scholander‐type pressure chamber (PMS Instruments, Corvallis). The highest four T _leaf and lowest four ψ_leaf recorded for each species during the period were taken as the species‐specific in situ maximum leaf temperature (T _max) and minimum leaf water potential (ψ_min).

2.6. Leaf Thermal and Hydraulic Safety Margins

We defined thermal safety margin (TSM) as the difference between T _crit and T _max (i.e., T _crit–T _max, TSM_Tmax, ^oC), which quantifies the risk for inception of photosynthetic failure under heat, as has been done in previous studies (Zhang et al. 2024). Moreover, it has been shown that temperature thresholds at early decline in F _v /F _m (e.g., T _crit) can be less sensitive to the duration of heat exposure (Didion‐Gency et al. 2025); hence, it is a robust indicator of leaf heat tolerance. In addition, we calculated the decline width (T _dw, ^oC) as the difference between T ₅₀ and T _crit (i.e., T ₅₀–T _crit), which was considered as a surrogate for the rate of decline in F _v /F _m as the temperature increases, with wider T _dw indicating slow development of photosynthetic dysfunction, once T _crit is exceeded. Likewise, we defined hydraulic safety margin (HSM) as the difference between Ψ_min and P ₅₀ (HSM_P50; P ₅₀–Ψ_min) or P _tlp (HSM_tlp; P _tlp–Ψ_min), which quantifies the risk of increase in xylem embolism level or loss of leaf turgor under high T _air. Of note, these indices quantify the risk of hydraulic dysfunction under specific urban environments, and are not indicative of the probability of physiological injury in natural vegetation.

2.7. Data Analysis

Species climate envelope (i.e., climate niche) data, including MAT, MAP, and air temperature of the warmest month (TWM), with their 5th and 95th quantiles were obtained by matching the species occurrence data, including both native and managed systems, from the Global Biodiversity Information Facility (GBIF, https://www.gbif.org/; Table 1) with averaged climate data from 1970 to 2000 of WorldClim (version 2.1; https://www.worldclim.org/) at 30 arcseconds resolution (accessed in July, 2024). Additionally, ranges of variation for MAT (MAT_range), MAP (MAP_range) and TWM (TWM_range) were calculated as the difference between the 95th and 5th quantiles of these metrics, and were used to quantify the variability in temperature and precipitation of species climate niches.

Variation of traits quantifying heat sensitivity of PSII across species or leaf thermal conditions (i.e., T _crit, T ₅₀ and T _max), species drought tolerance and water status (i.e., P ₅₀, P _tlp and ψ_min), as well as photosynthetic traits (i.e., A _max, g _s , E, V _cmax and J _max) across species, was tested using one—way ANOVA with a general linear model using the “lm()” function and the “Anova()” function of the car package, followed by Tukey's HSD post hoc with the “HSD.test()” function in the agricolae package, after data were validated for normality and homogeneity of variance. Moreover, variation of in situ F _v /F _m under contrasting T _air was analyzed using two‐way ANOVA, while t‐test was additionally used to test the difference in F _v /F _m under different T _air within species. Bivariate relationships among traits, and between traits and climatic metrics were analyzed using linear regression with the “lm()” function. Statistical significance was considered when p ≤ 0.05. Furthermore, we employed principal component analysis (PCA) to reveal the multivariate correlation among traits related to leaf heat tolerance, drought tolerance and photosynthesis, using the “prcomp()” function. All statistical analyses were performed in R computing environment (v4.4.0, R Development Core Team 2014).

3. Results

3.1. Species‐Specific Variation in Leaf Heat Tolerance, Hydraulics, Photosynthetic Traits and In Situ Leaf Hydrothermal Status

Across species, considerable variation was found for traits quantifying the heat sensitivity of PSII, including temperatures at the incipient (T _crit) and 50% reduction (T ₅₀) in maximum quantum yield of PSII (F _v /F _m ; p < 0.001 for both traits; Figure 2). Additionally, the decline width for temperature response of F _v /F _m (T _dw) varied between 8.75°C and 16.45°C across species. Maximum leaf temperature (T _max) recorded under high growth temperature (T _air) also varied significantly among species (p < 0.001), ranging from 33.7°C ± 1.24°C to 40.1°C ± 1.22°C. This led to variation in TSM_Tmax from −3.94°C for F. viridissima to 9.81°C for G. biloba.

FIGURE 2.

Responses of quantum yield of PSII (F _v /F _m ) for heat treated leaves to treatment temperature of the nine urban greening woody species targeted in the current study. Temperature dependence of F _v /F _m was fitted using a 2–parameter Weibull function, with fitted lines and corresponding confidence intervals (shaded region) being shown. Vertical solid, dashed, and dotted lines indicate temperature thresholds triggering the incipient (T _crit), 50% loss (T ₅₀) of F _v /F _m relative to the maximum, and in situ maximum leaf temperature (T _max) recorded during the heat event in early June 2024, respectively, with their corresponding values also given in the figure. Superscripted letters following trait values indicate the results of post–hoc comparisons across species. Also given are the decline width for temperature response of F _v /F _m , quantified by the difference between T ₅₀ and T _crit, and the leaf thermal safety margin calculated as the difference between T _crit and T _max (TSM_Tmax).

Traits quantifying leaf drought tolerance, including leaf water potential at 50% of xylem embolism (P ₅₀) and leaf turgor loss point (P _tlp), also varied markedly across species (p < 0.001 for both traits; Figure 3). The in situ minimum Ψ_mid ranged from −0.52 ± 0.06 MPa in A. truncatum to −1.83 ± 0.25 MPa in F. viridissima , leading to the hydraulic safety margin for considerable xylem embolism (HSM_P50) and loss of turgor in leaves (HSM_tlp) which varied between −0.75 to 5.08 MPa and −0.21 to 1.37 MPa across species, respectively. Also, species differed markedly in photosynthetic traits (p < 0.001 in all cases; Table 2), including light saturated net leaf photosynthetic rate (A _net, μmol m⁻² s⁻¹), stomatal conductance (g _s , mol m⁻² s⁻¹), leaf transpiration rate (E, mmol m⁻² s⁻¹), maximum Rubisco carboxylation rate (V _cmax, μmol CO₂ m⁻² s⁻¹) and electron transport rate (J _max, μmol e⁻ m⁻² s⁻¹).

FIGURE 3.

Response of leaf xylem embolism level to the reduction in water potential for the nine urban greening woody species. Data were fitted using a two‐parameter Weibull function, with fitted lines and corresponding confidence intervals (shaded region) shown. The vertical solid, dashed, and dotted lines indicate water potential triggering 50% of leaf xylem embolism level (P ₅₀), leaf turgor loss point (P _tlp), and minimum leaf water potential (ψ_min) recorded during the heat event, respectively, in early June 2024. Species‐specific values for P ₅₀, P _tlp, and ψ_min are provided in each panel figure, with the subscripted letters indicating the results of post hoc comparisons across species. The hydraulic safety margin (HSM) was calculated as the difference between ψ_min and P ₅₀ (HSM_P50) or P _tlp (HSM_tlp).

TABLE 2.

Variation of photosynthetic traits across the nine urban greening woody species.

Species	A net	g s	E	V cmax	J max
E. japonicus	13.61 ± 0.74abc	0.19 ± 0.01abc	2.79 ± 0.21abcd	50.82 ± 2.91bcde	111.39 ± 8.15a
F. viridissima	15.63 ± 1.68a	0.22 ± 0.03ab	3.97 ± 0.54a	74.54 ± 4.58ab	113.71 ± 8.8a
L. lucidum	8.78 ± 0.97bcde	0.11 ± 0.02bcd	2.02 ± 0.28abcd	39.33 ± 3.19de	63.92 ± 8.58b
R. xanthina	5.23 ± 0.85e	0.14 ± 0abcd	2.68 ± 0.08abcd	25.09 ± 2.7e	53.17 ± 0.75b
S. oblata	13.69 ± 1.42ab	0.19 ± 0.03abc	3.44 ± 0.46abc	43.3 ± 2.02cde	87.34 ± 12.97ab
A. truncatum	7.26 ± 0.88de	0.07 ± 0.01d	1.18 ± 0.14d	37.31 ± 4.02de	57.61 ± 5.74b
C. coggygriaca	16.35 ± 1.16a	0.23 ± 0.05a	3.77 ± 1.01ab	76.46 ± 6.93a	120.46 ± 7.52a
G. biloba	9.28 ± 0.41bcde	0.08 ± 0.01cd	1.42 ± 0.1cd	48.2 ± 1.33cde	84.06 ± 6.56ab
P. tomentosa	9.45 ± 0.59bcde	0.1 ± 0.01cd	1.78 ± 0.16bcd	61.23 ± 6.43abcd	87.66 ± 5.71ab

Note: Traits shown in table are light saturated net leaf photosynthetic rate (A _net, μmol m⁻² s⁻¹), stomatal conductance (g _s , mol m⁻² s⁻¹), leaf transpiration rate (E, mmol m⁻² s⁻¹), maximum Rubisco carboxylation rate (V _cmax, μmol CO₂ m⁻² s⁻¹) and electron transport rate (J _max, μmol e⁻ m⁻² s⁻¹). Data are shown as mean ± standard error of mean (SE). Difference in traits across species was tested using general linear model followed by Turkey’ HSD post hoc, with significant difference at p ≤ 0.05 level being indicated by superscripted letters.

3.2. Variation of In Situ F_v /F _m Under Contrasting Air Temperature and Its Correlation With Heat Tolerance and Gas Exchange

Species differed significantly in in situ F _v /F _m measured under both mild and high ambient air temperatures (T _air; p < 0.001 for both cases; Figure 4), with values ranging from 0.72–0.81 and 0.70–0.81, respectively. Overall, F _v /F _m was significantly lower under high T _air (χ ² = 22.58, p < 0.001), although the magnitude was small, with the range of decreasing F _v /F _m under high T _air (ΔF _v /F _m ) being 0–0.06. Yet significant interactive effects of species and T _air were found for the in situ F _v /F _m (Figure 4), with a significant decrease in F _v /F _m under high T _air being observed for C. coggygria , F. viridissima , R. xanthina and S. oblata (Figure 4). For these species, the maximum fluorescence of the dark‐adapted leaf (F _m ) was generally unchanged across different T _air values. However, there was a tendency for increased minimum fluorescence of dark‐adapted leaves (F ₀) under higher T _air in these species, although the difference was not statistically significant (data not shown).

FIGURE 4.

Variation of in situ maximum quantum yield of photosystem II (F _v /F _m ) under mild (solid bar) and high (hatched bar) ambient air temperatures (T _air) for the nine urban greening woody species. Error bars indicate standard error of mean (n = 4–5). Effects of species, T _air and their interaction on F _v /F _m were tested using two‐way ANOVA and the results are given in the figure. Variation of F _v /F _m within species was additionally analysed using t‐test and significant differences were indicated by the asterisk above the bars.

Across species, leaf TSM_Tmax was negatively related to T _max (R ² = 0.51, p = 0.03; Figure S1a), when G. biloba was excluded from the linear regression. While a positive correlation was found between TSM_Tmax and T _crit across species (R ² = 0.81, p < 0.001; Figure S1b), the relationship was less convincing because it was driven largely by a single species, G. biloba . Moreover, TSM_Tmax was negatively correlated with ΔF _v /F _m (Figure 5a, R ² = 0.46, p = 0.02), and a negative correlation was also detected between TSM_Tmax and T _dw (Figure 5b, R ² = 0.76, p < 0.001). Additionally, ΔF _v /F _m was found to be positively correlated with g _s (Figure 5c, R ² = 0.54, p < 0.01). Positive correlations were also found between T _max and gas exchange traits, including g _s (R ² = 0.61, p < 0.01; Figure S2a) and E (R ² = 0.55, p = 0.01; Figure S2b).

FIGURE 5.

The relationships among leaf photochemical efficiency, heat sensitivity of PSII, and leaf gas exchange traits. Variables shown in figures are the change of in situ maximum quantum yield of PSII (F _v /F _m ) measured during mild and high ambient air temperatures (ΔF _v /F _m ), leaf thermal safety margin (TSM) defined by the difference between the temperature threshold at the incipient loss of F _v /F _m (T _crit) under heat treatment and the in situ maximum leaf temperature during high ambient temperature (TSM_Tmax), decline width for temperature dependence of F _v /F _m (T _dw), quantified by the difference between temperature threshold triggering 50% loss in F _v /F _m relative to the maximum (T ₅₀) and T _crit, and light saturated leaf stomatal conductance (g _s ). Colors denote species as indicated in Table 1. Error bars denote standard error of mean (n = 3–5). Also shown are the fitted lines of linear regression, corresponding adjusted R ² and p value.

3.3. Correlations Among Traits Defining Leaf Heat Tolerance, Drought Tolerance and Biochemistry of Photosynthesis

No correlation was found between heat tolerance traits (i.e., T _crit, TSM_Tmax, T _dw) and traits associated with biochemical efficiency of photosynthesis, including V _cmax and J _max, and drought tolerance traits, including P ₅₀ and P _tlp (Figure 6). Likewise, no correlation was found between leaf thermal safety margins and hydraulic safety margins. Relationships among different sets of traits were further validated using multivariant analysis. The first two axes identified by principal component analysis cumulatively explained 75.47% of the total variation across species (Figure S3). The first axis (PC1), which accounted for 44.27% of the total variation, was positively related to leaf gas exchange traits including A _max, g _s , E, V _cmax, J _max, and was negatively related to TSMT_max and P _tlp; while the second axis (PC2) accounted for 31.20% of the total variation, and was positively correlated with T _dw, P ₅₀, HSM_P50, HSMP_tlp, but negatively associated with T _crit and T ₅₀.

FIGURE 6.

Correlation heatmap showing the relationships among traits associated with leaf heat tolerance, photosynthesis and drought tolerance for the nine urban greening woody species. Traits used for correlation analysis include temperature threshold triggering the inception of loss in maximum quantum yield of photosystem II (T _crit), thermal safety margin defined by T _crit and in situ maximum leaf temperature recorded during high ambient air temperature (TSM_Tmax), decline width (T _dw) for temperature dependence of maximum quantum yield of PSII, quantified by the difference between T _crit and temperature at 50% loss of F _v /F _m , light saturated net photosynthetic rate (A _net), stomatal conductance (g _s ), transpiration rate (E), maximum Rubisco carboxylation rate (V _cmax), maximum electron transport rate (J _max), water potential at 50% of leaf xylem embolism level (P ₅₀), leaf turgor loss point (P _tlp) as well as hydraulic safety margin defined by in situ minimum water potential and P ₅₀ (HSMP₅₀) or P _tlp (HSM_tlp). Asterisks in the squares indicate level of significance (***p < 0.001; **p < 0.01; *p < 0.05).

3.4. The Relationships Between Climatic Metrics and Leaf Heat Tolerance

Traits related to leaf heat tolerance (i.e., T _crit, T _dw and TSM_Tmax) were unrelated to climatic variables of the species climate envelope, including mean annual air temperature (MAT), mean annual precipitation (MAP) and mean air temperature of warmest month (TWM; Table S1). However, correlations were found between heat tolerance traits and metrics representing the climate variability of species distributional range. In particular, T _crit was negatively related to MAT_range, MAP_range and TWM_range (Figure 7a–c, R ² = 0.76, 0.76 and 0.45, respectively, p < 0.05 in all cases). Likewise, negative correlations were also found between TSM_Tmax and these variables (Figure 7g–I, R ² = 0.67, 0.37 and 0.40, respectively, p < 0.05 in all cases). Additionally, a positive relationship was found between T _dw and MAT_range (Figure 7d, R ² = 0.71, p < 0.01). Nonetheless, it appeared that most of the significant correlations were largely driven by G. biloba . When this species was excluded from the linear regression analysis, a significant correlation was observed only between T _crit and TWM_range (R ² = 0.42, p = 0.04).

FIGURE 7.

Correlations between heat tolerance traits including temperature threshold inducing the incipient (T _crit) loss of maximum quantum yield of PS II (F _v /F _m ) for heat treated leaves, decline width (T _dw) calculated by the difference between thermal threshold at 50% loss of F _v /F _m relative to the maximum (T ₅₀) and T _crit (i.e., T ₅₀–T _crit), thermal safety margin defined by the difference between T _crit and in situ maximum leaf temperature (i.e., T _crit–T _max; TSM_Tmax), and climatic variables including the range of mean annual air temperature (MAT_range, panel a, d, g), mean annual precipitation (MAP_range, panel b, e, h) as well as mean air temperature of warmest month (TWM_range, panel c, f, i) of species climate envelope. Ranges of climatic variables are calculated as the difference between their 95th and 5th quantiles. Relationships between traits and climatic variables were analysed using linear regression and fitted lines, corresponding R ² and p value are given if regression is statistically significant. Colors represent species as indicated in Table 1. Error bars denote standard error of mean (n = 3–5).

4. Discussion

4.1. Species‐Specific Leaf Heat Tolerance and Responses of Maximum Quantum Yield of PSII to Air Temperature

In the current study, the range of T _crit of the studied species was similar to that obtained from urban greening species in cities across south China using the same technique (Zhang et al. 2024). Yet it appeared that the average T _crit observed here (36.9°C) was lower than that reported in Zhang et al. (2024), probably reflecting a systematic difference in thermal tolerance of urban vegetation for cities located along the latitudinal gradient. Moreover, the mean T _crit identified for these species was low (in the ca. 1.7th percentile of global variation) relative to those reported previously, although the index in O'Sullivan et al. (2017) was based on more conservative characteristics of chlorophyll fluorescence (i.e., F ₀), thus indicating overall susceptibility to heat stress in these plants. Consequently, these species will be subject to increased risk of physiological impairment if their growth conditions become hotter. Indeed, decreased F _v /F _m under high T _air was observed for some species, even if the hot days in the current study could not be considered as a typical heatwave event. The variation in F _v /F _m observed here was unlikely to be associated with leaf developmental stage, given that traits were consistently measured on mature leaves. The decline in leaf quantum yield was in accordance with the findings in many studies (Fang‐Yuan and Guy 2004; Húdoková et al. 2022), demonstrating that high T _air can compromise photochemistry; nonetheless, increased F _v /F _m for species subjected to heat has also been reported (Ahrens et al. 2021). The decreased F _v /F _m observed here can be partially attributed to the changes in minimum fluorescence (F ₀), where higher values can be associated with structural damage to the photosystem (Baker 2008).

4.2. Wider Thermal Safety Margin but Not High Gas Exchange Capacity Alleviates the Decrease in Photochemical Efficiency Under High Growth Temperature

TSM_Tmax was primarily related to in situ T _max for most of our study species, suggesting that thermal safety margin primarily governed the ability to adjust temperature, consistent with the findings of Kitudom et al. (2022). Yet, the relationship between TSM_Tmax and T _crit clearly needs verification with more species. The negative TSM_Tmax recorded for most species indicates that temperature had exceeded the heat tolerance threshold in leaves, which has also been shown in some urban greening species under heat (e.g., Zhang et al. 2024), but may be less common in natural vegetation (Feeley et al. 2020). In addition, F _v /F _m for species characterized by wider safety margins (i.e., large TSM_Tmax) was less suppressed by high T _air, therefore highlighting the functional significance of heat tolerance in mitigating the negative effect of heat on the functionality of photosynthesis (Slot et al. 2021). Nonetheless, thermal tolerance thresholds reflect constitutive structural limits of PSII stability, while the observed shifts in F _v /F _m under contrasting growth temperatures represent short‐term physiological acclimation. Hence, these correlations should not be viewed entirely as the outcome of long–term adaptation. Furthermore, the negative correlation between TSM_Tmax and T _dw demonstrated that species that maintain wider safety margins have a more heat sensitive photochemistry, thereby undergoing a rapid loss in F _v /F _m once T _crit is surpassed. This suggests that wider safety margins and slower development of thermal injury with rising temperature cannot coexist; hence, there is a trade‐off between damage avoidance and damage tolerance in coping with heat across these species (Tiwari et al. 2021).

Limiting the rise in T _leaf through evaporative cooling has been considered an effective strategy to minimize thermal injury during heat events (Lin et al. 2017; Marchin et al. 2023; Sadok et al. 2021; Valliere et al. 2023; Wahid et al. 2007). Therefore, species characterized by high rates of evaporation are expected to be less physiologically damaged under heat. By contrast, higher ΔF _v /F _m was found in species characterized by higher capacity of leaf gas exchange (i.e., high g _s and E) in the current study. Such a counterintuitive relationship was unlikely to be explained by the limitation of gas exchange due to insufficient soil water availability because of routine irrigation during the study period, and it has been shown that limited plant water access does not necessarily restrict gas exchange under heat in some cases (Marchin et al. 2023). Alternatively, this observation indicates that the steady‐state capacity of gas exchange cannot be translated into transpirational cooling under heat, given that the indirect effect of high T _air, particularly increased VPD, can lead to substantial reduction in stomatal conductance (Grossiord et al. 2020). This will consequently weaken the mitigating effects of transpiration on leaf thermal state (Grossiord et al. 2020, Slot et al. 2024), although sustained evaporation under high VPD has been recorded in some species (e.g., Drake et al. 2018). In the present study, the daily maximum VPD reached > 4.5 kPa during a period of high T _air. Most likely, high VPD may have strongly suppressed in situ leaf gas exchange, especially for species with strong gas exchange capacity, which are known to be highly susceptible to atmospheric drought (Grossiord et al. 2020). Nonetheless, such inference should be further validated by field observation.

4.3. No Correlations Among Leaf Heat Tolerance, Leaf Drought Tolerance and Photosynthetic Biochemistry

We hypothesized that thermal tolerance thresholds could increase at the expense of the biochemical efficiency of photosynthesis. In support, Heckathorn et al. (1996b) reported that large amounts of nitrogen used for synthesizing HSPs in response to heat originated from enzymes involved in photosynthesis, including Rubisco, which has long been known to act as nitrogen storage in addition to its role in CO₂ carboxylation (Warren et al. 2003). Likewise, the negative impacts of membrane rigidity on electron transport have been widely recognized (Hirano et al. 1981; Yamori et al. 2008). Yet, no relationship between T _crit and V _cmax or J _max was observed in the current study. However, the lack of correlative relationships among these traits does not explicitly preclude the existence of proposed trade‐offs, given that V _cmax and J _max were assessed prior to the occurrence of hot days, when increased synthesis of HSPs and modifications in membrane composition may have not occurred or were minor. Indeed, it has been shown that maximum synthesis of HSPs and saturated fatty acid generally require high temperature to be induced (e.g., > 40°C) (Altschuler and Mascarenhas 1982). Therefore, the trade‐off between thermal safety and biochemical efficiency of photosynthesis is possibly conditional (Roff and Fairbairn 2007; Garland 2014).

On the other hand, no correlations were detected between traits quantifying heat and drought tolerance. This was inconsistent with recent studies identifying correlations between heat and drought tolerance (Fadrique et al. 2022; Kunert and Hajek 2022; Posch et al. 2024), but was in line with the finding of Münchinger et al. (2023). Noticeably, the correlation between heat and drought tolerance, or lack thereof, observed in these studies largely reflects the strategy to cope with the environmental pressures where plants naturally occur. Given that urban species are often selected with the primary objective of optimal landscaping effects, traits therefore may not necessarily be correlated as they might in natural landscapes. Alternatively, decoupling of these traits in heat may occur if excessive water loss from transpiration is restricted by VPD, leading to a reduction in gas exchange, such that water potential is largely stabilized without threatening the hydraulic integrity. Collectively, these findings suggest that leaf heat tolerance, drought tolerance, and leaf photosynthetic traits are unrelated across the targeted woody species. Again, it should be stressed that such findings do not completely reject the notion of potential interaction among these sets of traits, given that the number of species in the current study may not have sufficient statistical power for detecting robust trait relationships, and trait correlation may only emerge under specific conditions. Therefore, further studies with an increased number of urban greening species and the measurement of traits under field conditions, particularly under heatwaves, are warranted to verify the findings of our study.

4.4. Climate Control Over Leaf Drought and Heat Tolerance

Our third hypothesis was rejected by the absence of correlation between heat tolerance traits and climatic variables of species' climate envelope, including MAT, MAP, and TWM. This is generally inconsistent with the observations in previous work (Li et al. 2022), but it does agree with the findings of Hanley et al. (2021), who identified no relationship between leaf heat tolerance and climate niche across a wide range of urban species; hence, climatic variables may be poor predictors of heat tolerance for urban vegetation.

Nonetheless, correlations were found between traits and metrics quantifying the environmental variability of species distributional range. Specifically, species occupying regions characterized by variable thermal or hydro environments tended to have weaker capacity to tolerate heat. Intuitively, variable environment should select for increased stress tolerance because of the high frequency of climate extremes (Colombo et al. 1999). Nonetheless, physiological dysfunction can be mitigated by slower development of thermal injury as temperature rises for species characterized by lower heat tolerance thresholds, as evidenced by the negative relationship between T _crit and T _dw. It is also noteworthy that most of these correlations occurred because of G. biloba , which was the only gymnosperm species investigated in the current study. The distinct physiological attributes of G. biloba could have facilitated the observed relationships by maximizing traits variation. However, given the existence of a correlation between T _crit and TWM_range, even when G. biloba was removed, the correlations between climate variability and characteristics of species heat tolerance are evident, but need further testing. Specifically, heat tolerance for species originating from stable climates (i.e., narrow temperature or precipitation range) can be used for validating these patterns within the local community of urban greening species.

5. Conclusion

By assessing leaf physiology in nine urban greening woody species common in northern China, we demonstrated generally weak heat tolerance in these plants. High heat susceptibility led to decreased photochemical efficiency under relatively high T _air, while the negative effect of heat on photochemistry could be mitigated by a wider thermal safety margin. Across species, there were trade‐offs between damage avoidance and tolerance when exposed to heat stress, while no correlation was detected among leaf heat tolerance, drought tolerance, and photosynthetic traits. Furthermore, correlative relationships were identified between leaf heat tolerance traits and climate variability of species distributional range.

Heat tolerance characteristics will be informative when designing and selecting the appropriate species for green infrastructure in cities. For instance, species characterized by poor heat tolerance can be preferentially planted in places where heat stress is less intense, such as sites with low building density and access to open ground. The decoupling of traits defining physiological tolerance to heat and drought, as well as photosynthesis, although requires further validation with increased sample size and suitable measurement conditions, could potentially allow for selecting species with resistance to both drought and heat without compromising survival and productivity.

An additional limitation of this study was that leaf heat tolerance was not assessed following the heat event, so we do not know whether decreased efficiency of photochemistry or heat tolerance will recover, and if so, the rate of recovery. Additional studies monitoring these traits over seasonal or interannual scales, and under repeated heat and recovery events, will solidify our knowledge regarding how plants adapt to heat under urban environments. Overall, this study contributes to the understanding of the ecophysiology of urban vegetation, and offers insights into the challenges of urban greening in the context of global climate change.

Author Contributions

Ximeng Li: conceptualization. Jiawei Sun, Ruike Huang, Zitong Zhang, Yilin Yang, Jinyan Yang, and Runxi Luo: methodology. Ximeng Li, Jiawei Sun, and Ruike Huang: formal analysis. Jiawei Sun, Ruike Huang, Zitong Zhang, Yilin Yang, Jinyan Yang, and Runxi Luo: investigation. Ximeng Li, Jiawei Sun, and Ruike Huang: writing – original draft. Ximeng Li and David Tissue: writing – review and editing. Ximeng Li, Benye Xi, and David Tissue: supervision. Ximeng Li: project administration. Benye Xi: funding acquisition.

Funding

This work received funds from the Minzu University of China (2024JCYJ02), as well as “5.5” Engineering Research and Innovation Team Project of Beijing Forestry University (BLRC2023C05).

Disclosure

AI generative statement: No AI tools were used during the preparation of this manuscript.

Supporting information

Figure S1: The relationships between leaf thermal safety margin calculated as the difference between T _crit and T _max (TSM_Tmax), as a function of in situ maximum leaf temperature (T _max) which was recorded during high growth temperature, or critical leaf temperature (T _crit) which was determined as the temperature generating maximum quantum yield of PSII for heat treated leaves. Colors represent the different species as described in Table 1. Also shown are the fitted lines of linear regression, corresponding adjusted R ² and p value. Note that G. biloba is excluded from the regression analysis in panel (a).

Figure S2: The relationships between in situ maximum leaf temperatures (T _max) recorded during high growth temperature, as function of light saturated leaf stomatal conductance (g _s ) and transpiration rate (E). Colors stand for species as indicated in Table 1. Error bars denote standard error of mean (n = 3–5). Linear regressions were fitted to the data, with corresponding adjusted R ² and p value.

Figure S3: A principal component analysis (PCA) indicating relationships between leaf heat tolerance, leaf hydraulics and leaf photosynthetic traits for the nine urban greening woody species, with different sets of traits indicated by the color of arrows (heat tolerance, grey; hydraulics, blue; photosynthesis, red). Traits used for the analysis include heat traits: temperature thresholds triggering the incipient (T _crit) and 50% loss (T ₅₀) of maximum quantum yield of PSII (F _v /F _m ) for heat treated leaves, leaf width (T _dw) for temperature dependence of F _v /F _m , leaf thermal safety margin calculated as the difference between T _crit and in situ maximum leaf temperature (TSM_Tleaf); hydraulic traits: water potential at 50% of xylem embolism level in leaves (P ₅₀), leaf turgor loss point (P _tlp), hydraulic safety based on in situ minimum leaf water potential and P ₅₀ (HSM_P50) or P _tlp (HSM_tlp); photosynthetic traits: light saturated leaf net photosynthetic rate (A _net), stomatal conductance (g _s ), maximum leaf transpiration rate (E), maximum Rubisco carboxylation rate (V _cmax) as well as electron transport rate (J _max). Numbers in axis labels indicate the proportion of variance explained by the corresponding principal components.

Table S1: Correlation matrix showing the relationships between heat tolerance traits, including temperature thresholds triggering the incipient loss (T _crit) of quantum yield of PSII (F _v /F _m ), decline width (T _dw) for temperature dependence of F _v /F _m , leaf thermal safety margin calculated as the difference between T _crit and in situ maximum leaf temperature recorded during high growth temperatures (TSM_Tmax), and climatic variables of species climate envelope, including mean annual air temperature (MAT), mean annual precipitation (MAP) and mean air temperature of warmest month (TWM). Numbers shown in the table are coefficient of correlation, with levels of significance being represented by superscripted asterisks (***p < 0.001; **p < 0.01; *p < 0.05).

Sun, J. , Huang R., Zhang Z., et al. 2026. “Physiological Tolerance to Heat and Drought Are Decoupled Across Common Woody Urban Greening Species in North China.” Physiologia Plantarum 178, no. 4: e71010. 10.1111/ppl.71010.

Data Availability Statement

Data sharing is not applicable to this article as all newly created data is already contained within this article.