Seasonal Variations in the Leaf Temperature of Four Common Tropical Woody Plants

1. Introduction

Leaf temperature (T_leaf) plays a pivotal role in the microenvironment of plants [1], significantly influencing their ecophysiological processes, such as photosynthesis and respiration, which determine the carbon assimilation rate [2,3]. It also governs the transpiration rate through stomatal conductance [4]. Beyond the individual plant level, T_leaf is a crucial factor in ecosystems and regional water, carbon, and energy budgets [5,6]. Its role is prominently recognized within the frameworks of dynamic global vegetation models [7,8]. This underscores the importance of understanding and accurately modeling T_leaf in predicting and managing ecosystem responses to environmental changes.

T_leaf can significantly differ from air temperature (T_air), leading to substantial impacts on carbon and water fluxes [9]. Different studies have shown that the difference between T_leaf and T_air can reach 20°C and that this thermal decoupling varies greatly across species and environments [1,10]. In certain environments, such as alpine regions and humid tropics, T_leaf can be as much as 20°C higher than T_air, whereas in warm deserts, it can be 15–20℃ lower [2,9]. These deviations highlight the complex relationships between T_leaf and T_air and their ecological implications.

The thermal response of a leaf to changes in air temperature is crucial for its physiological function, and this response is characterized by the slope of the regression line between T_leaf and T_air [11,12]. A slope less than 1 indicates limited homeothermy, where leaves remain cooler than the air, which is especially useful for carbon uptake at higher T_air [3,13]. A slope equal to 1 represents poikilothermy, where T_leaf closely follows T_air, showing little to no thermal regulation [7]. Conversely, a slope greater than 1 suggests megathermy, where leaves heat up more than the air does, potentially leading to thermal stress [14]. The implications of these responses for plant performance are significant but not fully explored, and the mechanisms behind each type of response require further investigation.

Trees and shrubs may have different physiological and morphological adaptations to their environment, which can affect how their leaves respond to temperature changes. Previous studies have reported that shrubs have exhibited greater sensitivity to climate warming than trees have in recent decades [15,16]. However, Gazol and Camarero [17] reported that shrub species might actually be less affected by temperature fluctuations than trees are, as they are more significantly influenced by microclimatic conditions, topography, and soil temperature. This microclimatic connection allows shrubs to grow at sites that are more sheltered from regional atmospheric temperatures, potentially increasing their survival in areas with more extreme climate conditions [18].

Previous studies have shown that increases in T_leaf are associated with decreased transpiration rates [3,19], which are a consequence of stomatal closure in response to environmental stressors, such as water scarcity [20]. In the rainy season, when soil moisture is plentiful, plants reach their zenith in terms of liquid photosynthesis, stomatal conductance, transpiration, and leaf water potential [21]. This optimal physiological state effectively shields plants from high-temperature damage [22], facilitating normal growth. In contrast, during the dry season, leaf stomata are predominantly impacted by drought conditions [23,24]. A reduction in stomatal density and the closure of stomata lead to a significant decline in photosynthesis and an increase in T_leaf [25].

T_leaf is intricately linked to the leaf’s energy balance, which is influenced by a suite of leaf traits that modulate the absorption and dissipation of energy [7,26]. Specific leaf characteristics, such as the leaf chlorophyll content (LCC) [27], leaf area (LA) [9], leaf thickness (LT) [28], leaf mass per area (LMA) [29], and leaf water content (LWC) [1], play pivotal roles in this context. As the LA increases, the size of the leaf’s boundary layer also correspondingly increases [30]. This increase in boundary layer thickness leads to a decrease in the rate of sensible heat exchange between the leaf and its surrounding air [31]. Consequently, larger leaves typically exhibit higher maximum surface temperatures than smaller leaves do [32,33], assuming that all other factors are equal. When T_leaf increases, the activity of chlorophyll synthesis enzymes increases, leading to increased chlorophyll content [29]. However, once a certain T_leaf (>30°C) is reached, the enzyme activity rapidly decreases, and the chlorophyll content also rapidly decreases [34]. A higher LWC results in greater thermal inertia due to the high specific heat capacity of water. This allows leaves to absorb more heat with less significant temperature increases, thus protecting them from rapid temperature fluctuations and potential damage [35]. Thicker leaves can accumulate more energy and maintain a higher thermal mass [28], which helps stabilize T_leaf with minimal fluctuations. This is facilitated by the positive correlation between LT, LMA, and the content of heat-retaining substances within the leaf, such as water and macromolecules [30]. Consequently, an increase in LT and LMA not only enhances the leaf’s water retention capacity but also slows the rate at which T_leaf increases. Despite the acknowledged influence of leaf traits on T_leaf, the primary determinants that drive T_leaf remain elusive.

In this study, we selected four common tropical woody species in South China on a university campus and examined the seasonal variations (rainy and dry seasons) in T_leaf between different growth forms (shrubs and trees) and their dominant drivers. We hypothesized that (1) T_leaf consistently exceeded T_air, suggesting limited homeothermy; (2) the pattern of T_leaf between shrubs and trees would be significantly different between the dry and rainy seasons; and (3) leaf traits and environmental temperature could strongly affect T_leaf among species.

2. Materials and Methods

2.1. Site Description

The study area was located at Hainan Normal University Longkun South Campus in Haikou, Hainan Province, China (19.59°N, 110.20°E). The area has an annual average sunshine duration of 1752 hours and an average annual temperature of 23.8°C. The climate is characterized by a tropical maritime monsoon with an average annual rainfall of 1724.5 mm, which is highly unevenly distributed. The year is divided into two seasons: the rainy season, which spans from May to October, and the dry season, which spans from November to April of the following year. The rainfall during the rainy season accounts for approximately 80% of the annual total [36]. The average annual evaporation is 1834 mm, with an average relative humidity of 85%. The area is predominantly influenced by northeast and east winds, with an average annual wind speed of 3.4 m s⁻¹.

2.2. Measurement of Leaf Temperature

We selected four common woody plants on the campus for experimentation, including two tree species, Bauhinia blakeana and Fagraea ceilanica, and two shrubs, Ligustrum × vicaryi Rehder and Hibiscum rosa-sinensis Linn. From December 2022 to October 2023, leaf temperature (T_leaf) measurements were conducted on clear days in the middle of each month. Between 12:00 and 14:00 under full sunlight conditions, the leaf canopy temperature was measured remotely via an infrared thermometer gun (Fluke-MT4, Fluke Corporation, Everett, USA) [37]. Canopies emit longwave infrared radiation as a function of their temperature. The infrared thermometer senses this radiation and converts it to an electrical signal, which is displayed as temperature [38]. For each tree species, three individuals were selected for replication, and for each individual, eight random leaves were repeatedly measured, and the average value was taken.

2.3. Leaf Trait Measurements

From each of the three individual plants, twenty fully mature and robust green leaves were collected at random from the sunlit central portion of the canopy. The leaf chlorophyll content (LCC, SPAD) was measured via a chlorophyll meter (SPAD-502, Konica Minolta Holdings, Tokyo, Japan). The fresh leaf mass (LFM, g) was then measured via an analytical balance with a precision of one part in ten thousand. The leaves were subsequently placed into a portable leaf area scanner (LI-3000C, LI-COR, Nebraska, USA) to scan and determine the single leaf area (LA, cm²) [39]. Finally, the leaves were placed in envelopes and dried in a 60°C oven for 48 hours until they reached a constant weight, after which they were weighed again, and the leaf dry mass (LDM, g) was recorded. The leaf mass per area (LMA, g cm⁻²) and leaf water content (LWC, %) were calculated as follows:

$$\mathrm{L}\mathrm{M}\mathrm{A}\left(\mathrm{g}{\mathrm{c}\mathrm{m}}^{-2}\right)=\mathrm{L}\mathrm{D}\mathrm{M}/\mathrm{L}\mathrm{A}$$

$$\mathrm{L}\mathrm{W}\mathrm{C}=\left(\mathrm{L}\mathrm{F}\mathrm{M}-\mathrm{L}\mathrm{D}\mathrm{M}\right)/\mathrm{L}\mathrm{D}\mathrm{M}\times 100\mathrm{\%}$$

2.4. Environmental Conditions Measurement

Infrared radiation thermometers were used to measure the temperature of the cloudless open sky near each target plant (T_air). Portable soil moisture and temperature meters (TDR 350, Spectrum Technologies Inc., Aurora, USA) were employed to determine the soil temperature (T_soil) and soil water content (SWC). For each tree species, three individuals were selected for replication, with measurements taken in a circular pattern three times around each plant, and the average value was then calculated [40].

2.5. Data Analysis

We calculated the slope of the regression line between T_air and T_leaf for different growth types (shrub vs. tree), seasons (rainy vs. dry), and species. Analyses of covariance (ANCOVAs) were performed to test whether the slopes among those regression lines were significantly different [41]. To examine the effects of growth type and season on leaf traits, we used linear mixed effect models with sampling date as a random effect to account for temporal autocorrelation.

To estimate the relative importance of the predictor variables, we performed multimodel inference via Akaike’s information criterion (AIC). The relative importance value, regarded as the overall support for each predictor across all models, was estimated as the sum of the Akaike weights for the models in which the predictor appeared. A cutoff relative importance value of 0.8 was set to differentiate between the important and unimportant predictors. Partial regression plots were created for each important predictor to illustrate its relationship with T_leaf while holding all the other variables constant. All the statistical analyses were performed via the R statistical software package (ver. 4.3.0). All the statistical tests were considered significant at P < 0.05.

3. Results

3.1. Leaf Temperature Across Different Groups

T_leaf was consistently greater than T_air with respect to growth type, season, and species (Table 1). In general, T_leaf increased with T_air across the different groups (Figure 1). Across the species, the slope of the regression line between T_air and T_leaf was 0.45 (< 1), indicating limited homeothermy (Table 1 and Figure 1). For different growth types and species, the slopes were lower than one and did not differ among them (Table 1 and Figure 1). During different seasons, the slopes significantly differed (Table 1). The plants in the dry season displayed megathermy (slope >1), whereas the plants in the rainy season presented a slope < 0, suggesting a negative relationship between T_leaf and T_air (Figure 1).

3.2. Leaf Traits According to Growth Type and Season

The LCC was significantly greater in the rainy season than in the dry season (P < 0.05, Figure 2). However, no significant differences between seasons were found for LA, LDM, T_leaf, LMA, or LWC (P > 0.05, Figure 2). Compared with shrubs, trees had greater LCC, LA, LDM, T_leaf, and LMA (P < 0.05), whereas growth form had no significant effect on LWC (P > 0.05, Figure 2). We found no significant interactive effects between season and growth form on any of the traits (P > 0.05, Figure 2).

3.3. The Impact of Environmental Conditions and Leaf Traits

Model selection analysis showed that the best model describing T_leaf included LCC, T_soil, LMA, and T_air as important predictors (Figure 3). Partial regression analysis indicated that T_leaf increased significantly with increasing LCC and T_soil but decreased with increasing LMA and T_air (P < 0.05, Figure 4).

4. Discussion

4.1. Thermal Regulation Capacity of Leaves

Consistent with our initial hypothesis, we found that T_leaf consistently surpassed T_air, which aligns with certain prior studies [1,42,43]. In numerous tropical studies, T_leaf can exceed ambient T_air by more than 10°C in sunlit leaves[44,45]. This temperature disparity is attributed to variations in microclimatic conditions [10,28]. For example, wind speeds below 0.5 m s⁻¹ are common in nature [46], and even transient lulls can cause T_leaf to rise by more than 5°C within just a few seconds [47]. Furthermore, the leaf energy balance indicates that the leaf-to-air temperature difference is contingent upon the net energy provided (or lost) through radiation and the energy expended via transpiration [10]. Zhou et al. [1] reported that the effects of leaf physical warming are stronger than those of leaf transpirational cooling in hot environments. Consequently, the T_leaf of all the species was found to be greater than the ambient T_air.

The mechanistic relationship between T_leaf and T_air has received considerable attention in the literature [2,48]. Some studies have argued that leaves exhibit either limited homeothermy [11,49], poikilothermy [50], or megathermy [13,51]. In this study, the regression line slopes between T_air and T_leaf for the overall dataset, each different growth type, and each species were all less than one, indicating limited homeothermy. Energy balance theory suggests that limited homeothermy occurs under conditions of high stomatal conductance and low convective resistance, which implies that transpirational cooling outweighs physical warming [13]. However, our research did not observe a greater transpirational cooling effect over physical warming, as T_leaf consistently exceeded T_air. This unexpected result could be explained by the possibility that, under extremely hot conditions, plants may mitigate physical warming to alleviate heat stress. For example, Zhou et al. [1] reported that savanna woodland species possess the lowest absorptivity and the highest reflectivity, thereby reducing the radiation load and resulting in low levels of physical warming. Our findings indicate that the relationships between T_air and T_leaf, as well as the interactions between leaf physical traits and environmental parameters, are too intricate to be accurately simulated by a simplistic model.

Contrary to our initial hypothesis, plants during the dry season exhibited megathermy, characterized by a slope greater than 1, whereas those in the rainy season presented a slope less than 0, indicating a negative correlation between T_leaf and T_air. The capacity of plants to mitigate heat stress at the leaf level appears to decrease with increasing water stress [52,53]. In light of the anticipated rise in T_air due to global warming, the evaporative demand is projected to increase, resulting in the escalation of the air vapor pressure deficit (VPD) across many regions [54]. Prolonged exposure to elevated temperatures, in conjunction with high leaf-to-air VPD, is anticipated to elicit stomatal closure as a strategy to conserve water in hot-dry environments [43,55]. A reduction in stomatal aperture can lead to a faster increase in T_leaf than in T_air during the dry season, as diminished transpiration results in less cooling of the leaf surface [56]. Conversely, a negative relationship between T_leaf and T_air during the hot–wet season should increase overall carbon uptake by keeping leaves within optimal photosynthetic temperature ranges and below damaging temperatures [9].

4.2. Effects of Growth Form and Seasonal Variation on T_leaf

In this study, we found that the T_leaf of shrubs was significantly lower than that of trees, which may be interpreted in three ways. First, smaller leaves, as typically found on shrubs, have a thinner boundary layer [57], which allows for more efficient cooling through convection rather than evaporation of water [30,58]. Second, shrubs often have a relatively high surface area-to-volume ratio because of their relatively small stature and large number of small leaves [59]. This can lead to increased transpiration rates, which in turn can result in greater cooling of the leaf surface through the evaporation of water [60]. Finally, shrubs, which are closer to the ground, are more susceptible to the influence of soil temperature and moisture [61], which can significantly impact T_leaf. In contrast, trees, which are more exposed to atmospheric conditions, are less affected by ground-level microclimates. The soil temperatures and microclimates beneath shrubs can differ from those under trees, with shrubs often having denser canopies that shade the ground [62], potentially leading to lower soil temperatures and, consequently, reduced T_leaf [63]. As a result, the T_leaf of trees is generally greater than that of shrubs. This differential T_leaf response has implications for our understanding of plant thermal regulation and its ecological consequences for different growth forms.

Our findings revealed that there was no significant seasonal variation in leaf T_leaf across all the species studied. This outcome contradicts the results of a previous study by Rey-Sánchez et al.[45], which indicated that tree canopies in a tropical forest presented higher T_leaf values during the dry season than during the wet season. Additionally, Rey-Sánchez et al. [45] reported a correlation between T_leaf and photosynthetic photon flux density in the wet season, a relationship that was absent in the dry season. This discrepancy was hypothesized to be due to seasonal variations in wind speed, plant physiology, and canopy phenology [64,65]. In contrast, our study demonstrated a consistent T_leaf between the dry and wet seasons, which we attributed to the thermal regulation strategies employed by the plants. These strategies include transpirational cooling, physical warming, and species-specific adaptations that enable them to maintain optimal temperatures for photosynthesis and other physiological processes [1,66,67]. These adaptive mechanisms assist plants in buffering against extreme temperature fluctuations, thereby sustaining relatively stable T_leaf values irrespective of seasonal changes.

In contrast to our second hypothesis, we found that trees presented significantly higher T_leaf values than shrubs did across the rainy and dry seasons. Trees and shrubs may have evolved different adaptive strategies to cope with environmental conditions [68,69]. Trees, as the dominant species in a forest canopy, may have adapted to maintain a relatively high T_leaf as a way to optimize photosynthesis and growth under full sunlight conditions [70]. Shrubs, on the other hand, may have adapted to lower light and temperature conditions characteristic of the understory, leading to lower T_leaf [71]. These factors combined can result in trees maintaining higher T_leaf values than shrubs do, regardless of the seasonal changes in rainfall.

4.3. T_leaf Regulated by LCC and LMA

In agreement with our third hypothesis, we found that T_leaf increased with increasing LCC. This finding aligns with the research of Winter and Koniger [27], who noted significant reductions in LCC and photosynthetic capacity in Gossypium hirsutum L. as T_leaf decreased from approximately 33°C to 27°C under high light conditions. The higher the LCC is, the greater the leaf’s capacity to absorb light energy, leading to increased heat production and, consequently, an increase in T_leaf [72,73]. Conversely, Khan et al. [74] reported that the T_leaf of Triticum aestivum at temperatures ranging from 20°C to 35°C did not affect the LCC or nitrogen content per unit leaf area. Phylogenetic history can significantly influence plant traits, including LCC and T_leaf [34]. Our study indicates that different plant species have evolved distinct adaptive strategies to cope with environmental conditions, affecting their T_leaf and LCC.

In this study, we found a negative correlation between T_leaf and LMA, which is in line with several previous studies [75,76]. Owing to the significant positive correlation between LT and LMA with the substances contained within the leaf, such as water and macromolecules, an increase in LT and LMA not only enhances the water retention of the leaf but also slows the rate of increase in T_leaf [77].

4.4. T_soil and T_air Mediated T_leaf

In agreement with our third hypothesis, the multivariate analysis revealed that T_leaf was positively regulated by T_soil and negatively mediated by T_air.

Similar findings were also reported in several previous studies [76,78,79], which reported a positive correlation between T_leaf and T_soil. The relationship between epidermal cell area and stomatal density follows an inverse pattern, as noted by Rogiers et al. [80]. Consequently, an increase in T_soil can lead to a reduction in stomatal density, which may be attributed to the increase in epidermal cell size and leaf expansion influenced by T_soil[81]. This reduction in stomatal density, coupled with stomatal closure, could significantly curtail photosynthesis and consequently increase T_leaf [82]. Furthermore, the stresses induced by elevated T_soil led to a decrease in the relative water content of the leaves, which in turn resulted in increased Tleaf and cellular rupture [83].

Additionally, Manzi et al. [9] reported a positive correlation between the T_leaf of sun-exposed leaves and T_air across various species. In contrast, our study revealed an inverse relationship, which we attributed to the enhanced transpirational cooling effect under conditions of warming. Several mechanisms have been proposed to explain the stimulation of plant transpiration by elevated temperatures. First, elevated temperatures lead to increased expression of aquaporins in the mesophyll and heightened activity of potassium (K+) channels, which are crucial for stomatal opening, facilitating greater transpiration rates [84]. Second, warming enhances plant transpiration by increasing membrane fluidity, which is essential for water transport, and by reducing water viscosity, thereby lowering resistance to water flow within the plant [85]. Third, our research indicates that under warmer conditions, fine roots adapt by becoming thinner and growing deeper into the soil, which improves their capacity for water absorption [86]. These observations underscore the intricate and heterogeneous responses of diverse vegetation types to climate change, which has profound implications for the field of forest ecology and the formulation of adaptive management strategies.

5. Conclusions

Our study demonstrated that T_leaf consistently surpassed T_air, indicating limited homeothermy in tropical woody species. Trees have higher T_leaf values than shrubs do, potentially due to their distinct microclimatic adaptations. Notably, leaf traits, with a focus on LCC and LMA, substantially influence T_leaf. T_soil positively modulates T_leaf, whereas T_air exerts a negative regulatory effect. These insights highlight the complex interplay between leaf thermal dynamics and environmental factors. Future studies should broaden the species and climate ranges to obtain more generalizable findings. Examining additional environmental factors and their interactions with leaf traits will deepen our understanding of T_leaf dynamics. Long-term studies will also be crucial for tracking T_leaf responses to climate change, informing ecosystem management strategies.