On the relationship between Photosystem II thermotolerance and irreversible heat damage in leaves of Rhizophora mangle

Abstract

• Global temperatures are on the rise and may, together with more frequent and more intense heat waves, adversely affect plant performance and survival. Measurements of Photosystem II (PSII) chlorophyll a fluorescence are a widely used tool to assess the high‐temperature limits of PSII, but information on how decreases in PSII competence relate to permanent leaf tissue damage and plant survival is limited.

• Temperature response curves of dark‐level F₀ fluorescence were established for the tropical mangrove Rhizophora mangle. Three fluorescence parameters were determined as follows: (i) the temperature at which F₀ began to sharply rise (T_crit); T_crit is the most studied and often the only studied parameter in F₀ heat‐stress tests; (ii) the temperature at which maximal F₀ fluorescence occurred (T_max); and (iii) the temperature at which F₀ was 50% of T_max = T₅₀.

• T_crit was 49.6 °C, T₅₀ 51.3 °C and T_max 54.5 °C. Leaf heating experiments that included 2 weeks of post‐culture of leaves showed that severe tissue damage and sustained reductions in F _v /F _m , a measure of maximum quantum yield of PSII chemistry in the dark‐adapted state, were associated with temperatures close to T_max, but not T_crit and T₅₀.

• The results improve our ability to interpret F₀‐temperature response curves in the context of heat‐induced leaf mortality, draw attention to the significance of T_max in assessing permanent heat damage in leaves of Rhizophora mangle and highlight the fact that in F₀‐temperature response experiments, the widely used heat tolerance parameter T_crit may grossly underestimate the capacity of leaves to prevent irreversible heat damage.

Temperature‐F₀ fluorescence curves of Rhizophora mangle leaves demonstrate that T_crit, the temperature at which F₀ begins to rise, is unrelated to tissue necrosis; irreversible leaf damage starts to occur at significantly higher temperatures where fluorescence emission is maximal.

INTRODUCTION

Global temperatures are on the rise, and in many regions, heat waves are becoming more frequent and more intense (Perkins‐Kirkpatrick & Lewis ²⁰²⁵; Martinez‐Villalobos et al. ²⁰²⁵). There is concern about potentially adverse effects of rising temperatures on plant performance and survival (Ashraf 2021; Bohn et al. ²⁰²⁴; Chopra et al. ²⁰²⁵; Zhao et al. ²⁰²⁵). Tropical forests are believed to be particularly vulnerable because they have evolved to function within narrow seasonal temperature ranges (Wright et al. ²⁰⁰⁹). Furthermore, as high‐temperature biomes, they may already operate closer to their high‐temperature survival thresholds than plants from other regions (Doughty & Goulden ²⁰⁰⁸; Sastry & Barua ²⁰¹⁷; Sullivan et al. ²⁰²⁰; Perez & Feeley ²⁰²¹; Slot et al. ²⁰²¹; Tiwari et al. ²⁰²¹; Carter et al. ²⁰²⁵; Reis de Brito et al. ²⁰²⁵).

Over the years, leaf thermotolerance assays have undergone considerable change, from the quantification of heat‐stress‐induced leaf necrosis (Sapper 1935; Lange ¹⁹⁶¹; Larcher & Wagner ¹⁹⁷⁶), to measurements of chlorophyll a fluorescence that assess heat‐stress‐induced reductions in Photosystem II (PSII) activity. PSII is considered the most heat‐sensitive component of the photosynthetic apparatus (Berry & Björkman ¹⁹⁸²). Sophisticated yet easy‐to‐use portable instrumentation is now available enabling even non‐specialists to collect data on chlorophyll a fluorescence (Schreiber et al. ¹⁹⁸⁶). Most recently, differential scanning calorimetry has been added to the repertoire of methods to determine heat‐stress effects on leaves (Bertel & Neuner ²⁰²⁵). Classical necrosis tests are time‐consuming (up to 3 weeks) and typically require large quantities of leaves. By contrast, temperature response curves of F₀ fluorescence can be performed on single leaves and can yield information on PSII thermotolerance within less than 60 min (Schreiber & Berry ¹⁹⁷⁷).

In F₀‐temperature response experiments, F₀ fluorescence is continuously monitored in leaves or leaf discs heated at a constant rate, usually 1 °C min⁻¹, from room temperature to ≥60 °C. The temperature at which F₀ fluorescence begins to markedly rise is referred to as T_crit and has been considered the critical temperature for heat‐induced damage (Schreiber & Bilger ¹⁹⁸⁷). Upon further temperature increase, F₀ fluorescence continues to rise to reach its highest value at T_max and then declines rapidly. Although any experimental thermal tolerance assay cannot capture the complexities of extreme heat events in nature (Arnold et al. ²⁰²¹), F₀‐temperature curves have yielded important insight into the plasticity of PSII thermotolerance and the acclimation potential of T_crit to growth temperature and heat waves (e.g. Armond et al. ¹⁹⁷⁸; Downton et al. ¹⁹⁸⁴; Zhu et al. ^{2018, 2024}; Andrew et al. ²⁰²³; Posch et al. ²⁰²⁶). Yet, T_crit alone is of limited value when trying to predict survival and mortality of plants facing high‐temperature stress. For example, it is still largely unknown how T_crit, T_max and T₅₀ (the temperature at which F₀ reaches 50% of its maximum (Knight & Ackerly ²⁰⁰²)) relate to irreversible tissue damage. Current heat‐stress research using F₀‐temperature response curves focuses almost entirely on the rapid determination of T_crit and pays little attention to the potential reversibility of fluorescence signals and the question whether short‐term F₀ responses eventually lead to irreversible tissue damage. Bilger et al. (1984) reported for 27 species from the Würzburg Botanic Garden a close correlation between T_crit values from F₀‐temperature response curves and T₅₀ values determined with the necrosis method. Although the two methods led to similar heat tolerance rankings of species, different heat treatments were applied (heating at a rate of 0.7 °C min⁻¹ versus 30‐min exposure to different temperatures), and no information was provided on the extent of necrotic area that developed during post‐culture of leaf samples heated to T_crit.

Early studies revealed that T_crit was associated with a strong decrease in the quantum yield of whole‐leaf photosynthesis (Berry & Björkman ¹⁹⁸²). Furthermore, at T_crit light‐saturated photosynthetic net CO₂ uptake began to decline in a ‘time‐unstable manner’ (Downton et al. ¹⁹⁸⁴; Seemann et al. ¹⁹⁸⁴); that is, photosynthesis rate began to decline continuously at that temperature. The authors referred to unpublished results when stating that once a leaf had been exposed to temperatures above T_crit, the rate of photosynthesis could not be fully recovered upon returning to a lower temperature. Schreiber & Bilger (1987) on the contrary, noted that ‘when evaluating F₀‐T curve information, one must realize that the method applied monitors heat‐induced changes of thylakoid membrane conformation at elevated temperatures. Upon recooling of the leaves, the increase in F₀ is mostly reversible’. By contrast, T_max is believed to be the temperature above which ‘catastrophic cell membrane failure’ and ‘irreversible membrane architecture damage’ occur (Bryant et al. ²⁰²⁴), but studies directly addressing the reversibility/irreversibility of fluorescence signals at T_max are scarce (Frolec et al. ²⁰⁰⁸).

In a study on Phaseolus vulgaris (Hüve et al. ²⁰¹¹), T_crit of F₀‐fluorescence rise was ~47 °C. Whereas no cellular lesions were observed at this temperature as demonstrated by the absence of Evans blue‐stained cells, enhanced areas of stained cells (5–20%) occurred above 48–49 °C. Likewise, in a range of species, cell membrane thermostability determined through ion leakage experiments was higher than PSII thermostability determined as T_crit (Ilík et al. ²⁰¹⁸).

In recent experiments on F₀‐temperature relationships of the tropical tree species Calophyllum inophyllum, leaves heated to T_crit did not develop tissue necrosis (Winter et al. ^2025b). Measurements did not extend to T₅₀ and T_max. To further advance our understanding of how PSII heat tolerance relates to irreversible tissue damage and plant survival, F₀‐temperature response curves were established for leaves of the mangrove Rhizophora mangle. All three heat tolerance parameters were determined, that is, T_crit, T₅₀ and T_max. We demonstrate that only leaves heated to T_max and above showed sustained reductions in PSII quantum efficiency and significant leaf tissue damage during post‐culture, whereas no long‐term damage occurred at T_crit and T₅₀.

MATERIAL AND METHODS

Plant material

Propagules (viviparous seedlings) of Rhizophora mangle L. (DeYoe et al. ²⁰²⁰), also known as the red mangrove, were collected in March 2025 near the Punta Galeta Marine Laboratory (9.40°, −79.86°) of the Smithsonian Tropical Research Institute on the Caribbean coast of the Republic of Panama. R. mangle trees are typically up to 10 m tall (Méndez‐Alonzo et al. ²⁰¹⁵) and tolerate a wide range of sea water salinities. Plants are known to be well adapted to full‐strength seawater (i.e. 35 ppt salinity) but can, under controlled conditions, also grow without added NaCl to the culture solution (Lin & Sternberg ¹⁹⁹³; Pereira Silva et al. ²⁰²³). Other species like R. stylosa cannot grow in the absence of NaCl (Kanai et al. ²⁰¹⁴). Propagules were planted into FT24‐24 Cavity 3.9 Deep Forestry Trays (Stuewe & Sons, Tangent, Oregon) filled with a 1:1 mix (v:v) of sea sand (Novey, Panama) and relatively nutrient rich natural topsoil from a local property. Trays with propagules were kept in a foldable pool for kids (1.83 m diameter, 38 cm high; El Costo, Albrook Mall, Panama) filled with tap water. The upper part of the trays was about 3 cm below the water surface. The water was changed once per week. The pool with trays and plants was located underneath a naturally ventilated 8 m long, 4 m wide and 3 m high glass‐roofed rain shelter without side walls at the Gamboa (9.119870°, −79.705060°) outdoor plant physiology facilities of the Smithsonian Tropical Research Institute. After 3 months, in June 2025, plants were transferred to larger pots (20 cm wide, 20 cm high; one plant per pot) filled with the same mix of sand and soil as indicated above. Plants were maintained in the same pool, with pots submerged in the water. Three weeks prior to experiments, 5 g sea salt (Instant Ocean, Amazon.com) was added per litre of water; that is, the salinity level was raised from 0 (no salt added) to 5 ppt.

Heat treatments and F₀ temperature response

Experiments were conducted from mid‐July to mid‐August 2025. Mature leaves (second leaf pair from bottom), randomly collected from 38 individuals, were used. They had a specific leaf mass of 164 ± 10 g m⁻² and a leaf water content of 384 ± 35 g m⁻² (mean ± SD, n = 5). Leaves were collected at about 9 AM and transported inside a darkened, moist zip lock bag to the plant physiology laboratory at the Earl S. Tupper Research, Library and Conference Center (8.96°, −79.54°) in Panama City. Detached leaves were weighed for fresh mass determination and then placed horizontally into a 3010 GWK1 temperature‐controlled gas‐exchange chamber (Walz GmbH, Effeltrich, Germany). The position of leaves was fixed with nylon strings. After closing the top of the chamber with a custom‐made glass plate inside an aluminium frame, the end of the fibre‐optic cable of a PAM‐2500 chlorophyll fluorometer (Walz) was positioned perpendicularly to the glass plate. During heat treatments and chlorophyll fluorescence measurements, leaves were in complete darkness. The internal fan speed (0–5) was set at 4. Leaf temperature was monitored with a fine‐wire thermocouple with a stated accuracy of ±0.2 °C. Using the Walz GSF‐Win software, the gas‐exchange chamber was programmed to increase the leaf temperature in steps of 1 °C min⁻¹ from 25 °C to 60 °C. Heating of 1 °C min⁻¹ was used in Schreiber & Berry (1977) and has been employed in many F₀‐temperature experiments since then (Downton et al. ¹⁹⁸⁴; Knight & Ackerly ²⁰⁰²; O'Sullivan et al. ²⁰¹⁷; Zhu et al. ^{2018, 2024}; Middleby et al. ²⁰²⁴; Posch et al. ²⁰²⁶), facilitating comparisons across studies, because cumulative effects of thermal stress related to different heating rates matter (Arnold et al. ^{2021, 2025}; Neuner & Buchner ²⁰²³; Cook et al. ²⁰²⁴); slower heating than the recommended 1 °C min⁻¹ (Posch et al. ²⁰²⁶) increases the heat dose but on the other hand can provide more time to induce heat‐protective processes. Dark‐level F₀ fluorescence was continuously monitored with the PAM‐2500 system, and values were recorded at the end of each 1‐min interval. At the onset of each 1‐min interval, a 1 s pulse of weak far‐red light was applied to relax electron acceptors downstream of PSII and thus help to ensure that changes in the F₀ signal are attributable to experimental heating (Schreiber et al. ¹⁹⁷⁵; Bilger et al. ¹⁹⁸⁴; Posch et al. ²⁰²⁶).

A thermal ramp rate of 1 °C min⁻¹ means that although samples heated to, for example, 50 °C remained at this temperature for only 1 min, it took 25 min to get there. Thus, temperature treatments were not stand‐alone 1‐min thermal shocks but represented substantially higher heat doses because a leaf heated to, for example, 50 °C had already experienced 1 min at 49 °C, 1 min at 48 °C, 1 min at 47 °C, etc.

We have used detached instead of attached leaves, because leaf temperatures are easier to control in detached leaves where transpirational cooling is largely suppressed. In our heating experiments, leaf temperatures were close to air temperatures. By contrast, in a study with fully developed attached leaves of seedlings of two tropical tree species, Calophyllum longifolium and Ficus insipida, inserted into a gas‐exchange cuvette, cuvette air temperature had to be increased up to 70 °C to reach leaf temperatures >50 °C, owing to strong transpirational cooling, both in the light and the dark (Krause et al. ²⁰¹⁵).

Post culture, visible leaf damage and sustained F_v /F _m decrease

In a second series of experiments, leaves were heated in a stepwise fashion as described above, but instead of heating all leaves to 60 °C, final temperatures varied. They were 30, 38, 40, 42, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59 and 60 °C. Leaves from each temperature treatment were kept in humid, transparent plastic boxes at room temperature (~23 °C) and a continuously low PFD of ~3 μmol m⁻² s⁻¹ for 14 days of post‐culture. After 14 days, leaves were examined for visible tissue necrosis (dark areas) and photographed with a Sony a7 IV digital camera equipped with a Sony FE 2.8/24–70 GM II lens. Using four leaf clips (Walz GmbH) placed between the midrib and edge of the upper and lower part of the right and left half of leaf laminas (two on each side), the chlorophyll fluorescence parameter F _v /F _m was determined after 20 min of dark adaptation with a MINI‐PAM‐II system (Walz GmbH); that is, on each leaf, four F _v /F _m measurements were made and averaged. The ratio of variable to maximum fluorescence emission, F _v /F _m , is a measure of the maximum quantum yield of PSII chemistry in the dark‐adapted state. On a scale from 0 to 1, F _v /F _m is ~0.83 in healthy, non‐stressed leaves (Murchie & Lawson ²⁰¹³). Upon heat stress, the ratio declines (Krause et al. ²⁰¹⁰; Winter et al. ^2025a) and approaches 0 when PSII becomes inactive.

RESULTS AND DISCUSSION

F₀‐temperature response curves of five R. mangle leaves revealed T_crit of 49.6 °C ± 0.6 °C, T₅₀ of 51.3 °C ± 0.4 °C and T_max of 54.5 °C ± 0.2 °C (Fig. 1). There was little variation between the 5 replicates and differences between T_crit, T₅₀ and T_max were statistically significant at P < 0.001. T_crit and T_max values are in the range of those reported for 12 mangrove species along a salinity gradient in Australia (Bryant et al. ²⁰²⁴). The most significant result of our study is that heating to T_crit and T₅₀ did not result in sustained irreversible leaf damage: after 2 weeks of post‐culture, in leaves heated to T_crit and T₅₀, no tissue necrosis was observed and F _v /F _m was above 0.8, indicative of high maximum PSII quantum efficiency typical of healthy leaves (Figs 2 and 3). Heating to T_max led to a 10% reduction in F _v /F _m and caused 10% of leaf area to become necrotic. Heating to ~1 °C above T_max was required to induce 50% tissue damage and a 50% decrease in F _v /F _m after 2 weeks of post‐culture. Our research confirms and extends previous observations with Calophyllum inophyllum (Winter et al. ^2025b) and demonstrates that T_crit of F₀‐temperature curves, a widely used parameter in leaf thermotolerance assays, may severely underestimate the capacity of leaves to cope with high‐temperature stress, and that T_max is more closely correlated with sustained tissue damage than T_crit and T₅₀ (see also Winter ²⁰²⁴). Our observations support previous suggestions by Neuner & Buchner (2023) that functional impairment of PSII occurs at lower heat temperatures than permanent heat damage.

Fig. 1.

Temperature response of F₀ fluorescence of five leaves of Rhizophora mangle. Different leaves are highlighted with different colours. Leaf temperature was increased in steps of 1 °C per minute.

Fig. 2.

F _v /F _m chlorophyll fluorescence (top) and tissue necrosis (bottom) in leaves of Rhizophora mangle 2 weeks after they had been heated at a rate of 1 °C per minute up to different temperatures. Each data point refers to a different leaf, and temperatures are those after which the heat treatment was stopped. After heat treatments and prior to F _v /F _m measurements and the evaluation of the percentage of necrotic leaf area, leaves were kept for 2 weeks in a moist chamber (see Materials and Methods section).

Fig. 3.

Rhizophora mangle leaves heated to different temperatures at a rate of 1 °C per minute. Photos were taken after 2 weeks of post‐culture. The numbers above the leaves refer to the final temperatures in °C reached during each heat treatment.

Although the focus of the current study is primarily on methodological and interpretive aspects of F₀‐temperature curves and only secondarily on mangrove biology, our study adds to the notion of high thermotolerances of mangroves compared with other woody species (Li et al. ²⁰²²). Thermal safety margins – the difference between the temperature a plant can tolerate and the maximum air or leaf temperature a plant is experiencing in its habitat – would increase by ~5 °C in the R. mangle seedlings studied by us were the calculations based on T_max instead of T_crit. Exceptionally high thermotolerance margins have already been reported for 13 mangrove species from the south‐western Indian coast using T₅₀ from F _v /F _m ‐temperature curves to quantify PSII heat tolerance (Naseef et al. ²⁰²⁴). Future research needs to consider the effect of salinity level on F₀‐temperature curves in mangroves and must explore whether the relationships between T_crit, T₅₀ and T_max and irreversible tissue damage observed in this study are general plant responses. Studies on T_crit‐necrosis relationships at heating rates different from the conventional 1 °C min⁻¹ temperature steps are also warranted.

We used detached leaves in a well‐ventilated chamber for heat treatments which resulted in a decline of leaf fresh mass during heat treatments. Samples heated to T_crit lost ~5% of fresh mass and those heated to T_max lost ~10% of fresh mass (Fig. 4). Some researchers prevent dehydration during F₀‐temperature experiments by immersing samples in water, but this may enhance ‘anaerobic fluorescence’ (Schreiber et al. ¹⁹⁷⁵; Bilger et al. ¹⁹⁸⁴). Dehydration of mangrove leaves on a laboratory bench in dim light by 0.725 g water per g dry mass prior to F₀ measurements has been shown to lead to increases in T_crit and T_max of up to ~3.4 °C and ~1.8 °C, respectively (Bryant et al. ²⁰²⁴; see also references therein). A water loss of 0.725 g g⁻¹ would have corresponded to a fresh mass loss of ~22% in our samples which is a much larger fresh mass reduction than the 5 and 10% fresh mass loss observed by us. Furthermore, in Bryant et al. (2024), dehydration was almost certainly significantly slower than in the experiments reported here, providing more time for metabolic adjustments underlying the increases in T_crit and T_max. Even if there would have been a small dehydration‐related increase in T_crit and T_max in our study, this would not affect our conclusions because the heating procedure was the same for leaf samples used for T_crit and T_max determination and leaf samples used for F _v /F _m analysis and necrosis tests after post‐culture.

Fig. 4.

Percentage decrease of leaf fresh mass during heat treatments. Leaf fresh mass at the onset of heat treatments was 100%. The change in leaf fresh mass was determined immediately after the heat treatments had stopped.

CONCLUSION

Modern plant heat tolerance research has become largely synonymous with research on PSII thermotolerance because PSII is highly heat‐sensitive and because its activity and short‐term acclimation potential can be easily and rapidly assessed through chlorophyll fluorometry, especially when the methodology is refined for high throughput screening. However, rapidly performed fluorescence assays alone – either during ongoing heat treatments of F₀‐temperature curve experiments like the ones presented in this paper, or soon after heat treatments as in F _v /F _m ‐temperature curve experiments discussed elsewhere (Winter et al. ^2025a) – may severely underestimate the resistance of leaves to thermal stress. If the principal goal of heat‐stress tests is to predict plant survival and mortality in the face of rising temperatures and increased heat wave frequency and intensity, one must clearly separate irreversible from reversible effects. Identifying irreversible heat damage requires patience, because sufficient time after heat treatments must be given for damage repair and potential reversal of fluorescence signals, emphasizing – as shown here – the importance of remeasuring PSII fluorescence, combined with conventional necrosis tests, after prolonged post‐culture of plant samples.

AUTHOR CONTRIBUTIONS

K.W. planned the research, conducted the experiments and wrote the manuscript. M.G. wrote the program for the stepwise increase in leaf temperature. C.R.K.N. and J.A. collected the mangrove propagules in the field and grew the plants. A.V. prepared the graphs. All authors read, provided suggestions and approved the final version of the manuscript.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT

The data supporting Figs. 1, 2, and 4 are available on Smithsonian Figshare https://doi.org/10.25573/data.32038065.