The phenotypic and physiological response mechanisms of Tetracentron sinense, an endangered plant and a relict from the tertiary period, to global warming

Abstract

Tetracentron sinense, an endangered relict species surviving since the Quaternary Period, was investigated to assess its adaptive responses to climate warming. A downward transplantation experiment simulated warming effects by transferring plants from a high-altitude site (2448 m) to a low-altitude site (2023 m). We analyzed seed germination traits, seedling survival dynamics, and adaptive mechanisms through phenotypic plasticity and physiological adjustments. Downward transplantation significantly enhanced germination percentage, vigor, and index, while reducing seedling survival-evidenced by cumulative survival decline and elevated mortality. Mortality peaked during the first month post-transplantation, creating a critical survival bottleneck. Physiological analyses revealed stable chlorophyll a, b, and a/b ratios, alongside stable total chlorophyll content and improved photosynthetic capacity. Plants alleviated low-altitude stress by accumulating osmoregulatory compounds: soluble sugars, proteins, and proline. Furthermore, catalase activity significantly increased, whereas peroxidase activity correspondingly decreased under this stress regime. In summary, while climate warming may compromise T. sinense early seedling survival, surviving individuals exhibit adaptive potential through enhanced phenotypic plasticity and physiological adjustments under warming-induced selective pressure.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-026-08225-2.

Introduction

The rate of global warming since the 1970s has exceeded that of any other 50-year period over the past two millennia [1]. This accelerated warming, primarily driven by elevated atmospheric CO₂ concentrations, is driving shifts in mean surface temperatures, altering hydrological cycles, and modifying solar radiation regimes [2]. These compounding environmental changes pose significant ecological challenges, particularly for species with critically declining populations, by disrupting habitat suitability, phenological synchrony, and resource availability etc [3]. Therefore, investigating the responses and adaptive capacities of vulnerable species to multifaceted climate change is crucial for predicting their long-term persistence and for elucidating the evolutionary mechanisms that underpin their resilience or susceptibility [4, 5].

Traditional research on plant responses to climate warming has primarily relied on controlled environments, such as open-top chambers and infrared heating systems [6–8]. However, these approaches inherently oversimplify environmental complexity due to their artificial constraints. In contrast, altitudinal gradients serve as a valuable natural proxy for investigating long-term in situ climate responses, as they generate a temperature continuum that facilitates space-for-time substitution [9]. In these gradients, decreasing altitude is associated with rising temperature and atmospheric pressure, and typically, declining solar radiation intensity. Downward transplantation experiments—where high-altitude plants are transplanted to lower elevations—utilize these gradients as a natural laboratory to simulate climate warming [10, 11]. Monitoring plant performance across this altitudinal continuum under natural conditions provides valuable insights into the effects of climate warming on plant growth, development, and adaptive strategies.

In response to these climate changes, plants exhibit marked phenotypic and physiological plasticity [12, 13]. Phenotypically, shifts in plant height represent a strategic trade-off between stress avoidance and competitive ability for light; reduced stature in warmer environments minimizes water loss and heat stress by decreasing the surface area exposed to the atmosphere [14]. Concurrently, thicker stems enhance mechanical stability and hydraulic conductivity. Leaf morphological adaptations include thicker or smaller leaves to reduce transpiration, as well as dissected leaf shapes that improve convective cooling [15]. Biomass allocation is also reconfigured, prioritizing root development for enhanced water and nutrient acquisition or favoring aerial parts for superior light capture [16].

Physiologically, photosynthetic efficiency is highly temperature-sensitive, and the capacity to maintain high photosynthetic rates under moderate warming confers a competitive advantage [17]. Plants also modulate chlorophyll content to optimize light harvesting while mitigating photoinhibition. To combat heat-induced oxidative damage, antioxidant enzyme activities are upregulated [17]. Furthermore, osmotic adjustment via compatible solutes (e.g., proline) helps maintain cellular turgor under drought stress [18]. However, when environmental stresses exceed the integrated compensatory capacity of these phenotypic and physiological responses, mortality risk escalates [19]. Therefore, elucidating these interconnected plastic response mechanisms is crucial for predicting future shifts in plant survival, distribution, and community dynamics, ultimately informing effective conservation management strategies.

Tetracentron sinense (Oliv.), the sole surviving tall deciduous tree species in the Trochodendraceae family, is a Tertiary relict plant of significant ecological, economic, and ornamental value [20]. Paleobotanical evidence indicates that T. sinense was once widely distributed across Europe, North America, and East Asia [21]. However, its range has undergone a dramatic contraction, likely due to past geological and climatic upheavals, and it is now confined to fragmented montane habitats in central and southwestern China, and parts of Nepal, Myanmar, and Bhutan. Ongoing environmental changes and anthropogenic pressures continue to threaten its natural regeneration, warranting its inclusion in Appendix III of CITES [22]. Given its relict status and documented vulnerability, a critical question arises: how will T. sinense respond to accelerated future warming? Its long-term survival will depend on its adaptive capacity, manifested through phenotypic and physiological plasticity.

Therefore, we implemented a downward transplantation experiment to simulate climate warming effects on T. sinense. We assessed a comprehensive suite of responses spanning seed germination, seedling survival, and key phenotypic (growth, leaf morphology, anatomy) and physiological (photosynthesis, chlorophyll, antioxidants, osmolytes) traits. This integrated approach aims to (1) unravel the mechanistic basis of T. sinense’s warming responses and (2) identify its core adaptive strategies. Our findings are expected to establish a theoretical foundation for the conservation and management of T. sinense germplasm resources.

Materials and methods

Seed collection and processing

Mature seeds of T. sinense were collected from a high-altitude population (2448 m a.s.l.) in October 2021. Ten healthy parent trees with diameters at breast height (DBH) ranging from 30 to 40 cm were selected. From each tree, 50 fruit clusters were randomly harvested. The collected fruit clusters were pooled in equal proportions to form a composite sample, transported to the laboratory, and allowed to dry naturally in a greenhouse at approximately 26 °C. Seeds were subsequently extracted from the dried fruits and stored at 4 °C in cold storage. The plant material was authenticated by Dr. Yumin Shu of Beijing Forestry University.

Downward transplantation experiment

A downward transplantation experiment was conducted in March 2022 to simulate climate warming effects, following an established methodology [24]. The experimental design consisted of two transplantation treatments: (1) seeds sown at their native high-altitude site (control, hereafter HH), and (2) seeds transplanted to the low-altitude site (warming simulation, hereafter HL). Each treatment was replicated four blocks, with each block containing 3 pots of 100 seeds, resulting in a total of 1200 seeds per site (Fig. 1).

Fig. 1.

Design of the reciprocal transplant experiment

These seeds were sown into standardized plastic pots (15 cm diameter × 20 cm depth) filled with a 3:1 mixture of native soil (collected from the species’ natural habitat) and sandy soil, which represents the optimal growth medium for T. sinense [21]. Seeds were sown at a depth of 2–3 cm, reflecting natural soil stratification observed for seedling establishment in the field. Each pot was then covered with a thin layer of moss to maintain moisture.

Analysis of seed germination characteristics

Seed germination was evaluated based on seedling emergence. A seed was recorded as germinated only when its cotyledons had visibly emerged from the soil surface [22]. Emergence was monitored daily until no new seedlings was observed for seven consecutive days, at which point the experiment was concluded. The following germination parameters (based on seedling emergence) were calculated according to established methods [25]:

1

2

3

In the above equations, Ng represents the total number of seeds that successfully emerged (germinated seedlings); Nt represents the total number of seeds sown; Np represents the number of seedlings that emerged during the peak germination period; Gt represents the number of seedlings emerged on day t; Tt represents the number of days from sowing to day t.

Dynamic life table and survival analysis

Following complete germination, seedlings were cultivated in situ within their natural habitat. Mortality was recorded when a seedling’s aboveground portion became fully chlorotic and prostrate, with no observable capacity for recovery. Thereafter, according to the method described by Lu et al. [26], seedlings were considered 0 months old when germination rate statistics ceased. Subsequently, monthly survival monitoring continued until October 2022, when seedlings were covered by snow. A dynamic life table was compiled based on seedling survival from 0 to 5 months. The dynamic life table was established based on the average number of seedlings from the 12 pots in each treatment group, using the following parameters: x represents the age of seedlings (in months); a_x indicates the number of surviving individuals at the x-age class; l_x represents the standardized number of surviving individuals at the beginning of time x (generally converted to 1,000); d_x refers to the standardized number of deaths from age x to x + 1; q_x represents the mortality rate from age x to x + 1; L_x indicates the number of individuals surviving from age x to x + 1; T_x refers to the total number of individuals from age x age and older; and e_x signifies the life expectancy of individuals at age x [27]. The life table parameters were calculated as follows:

4

5

6

7

8

9

Based on the life table data, the cumulative survival function (S_x), cumulative mortality function (F_x), death density function (f_x), and hazard rate function (λ_x) for the T. sinense seedlings were calculated as follows:

10

11

12

13

In these equations, x represents the age of seedlings (in months); p_x represents the survival rate of seedlings in each treatment, q_x represents the mortality rate from the x to x + 1 age class, and h_x represents the time span [26].

Assessment of seedling growth and leaf morphology

In April 2023, all surviving seedlings were individually transplanted into separate seedling bags. Seedlings originating from high-altitude seeds and cultivated at their native high-altitude site were designated as control seedlings (HH), whereas those from high-altitude seeds but grown at the low-altitude site were classified as experimental seedlings (HL), simulating a warming scenario (Fig. 1).

After a five-month acclimation and growth period, comprehensive measurements were conducted in August 2023. Twenty seedlings were randomly selected from each group for assessment. Plant height was measured using a height gauge (161 − 116, Guilin Guanglu), and basal diameter was determined with a digital caliper (111–102 V-10G, Guilin Guanglu).

For leaf morphological analysis, the third fully expanded leaf from the top was excised from each sampled seedling. Leaves were flattened and scanned using a universal plant image analyzer (Scan Maker i800^plus, MICROTEK), and the resulting images were used to determine leaf length, width, and area [27]. For biomass measurements, the roots, stems and leaves of the seedlings were separated, thoroughly washed, and placed in labeled kraft paper envelopes. The samples were dried in an oven at 85 ℃ (GZX-9030MBE, Boxun) until a constant weight was attained. The dry weights of roots, stems, and leaves were measured separately, and the total biomass was calculated as their sum [28]. The specific leaf area (SLA) was calculated as the ratio of leaf area to leaf dry weight. To determine leaf nitrogen content and carbon-to-nitrogen (C/N) ratio, the dried leaf samples were ground into a fine powder and passed through a 2-mm sieve. The powdered samples were then analyzed using an elemental analyzer (Vario MACRO cube, Elementar, Germany). Differences in growth and morphological traits between the HH and HL groups were assessed using an independent samples t-test. Statistical significance was set at P < 0.05.

Observation of leaf anatomical structures

Three seedlings were randomly selected from each group. From each seedling, the second fully expanded leaf was harvested, and segments (1.5 mm × 2.0 mm) were excised, avoiding major veins. The segments were immediately fixed in pre-cooled 2.5% glutaraldehyde solution and processed to ensure submersion. The prepared samples were subsequently sent to Wuhan Pronets Testing Technology Co., Ltd. for detailed analysis of the leaf ultrastructure [29]. The obtained anatomical images were qualitatively described and compared between treatments.

Determination of photosynthetic parameters

Photosynthesis measurements were performed on clear days (August 10–12, 2023) between 9:00 and 12:00 using a Li-6400XT portable photosynthesis system (LI-COR, USA). For each treatment group, nine healthy, sun-exposed leaves from three isolated individuals (three leaves per individual) were selected. The measurement protocol was as follows: the built-in red-blue light source was used with a sequence of 15 photosynthetic photon flux density (PPFD) gradients, ranging from 2000 down to 0 µmol·m⁻²·s⁻¹. The leaf chamber temperature was maintained between 23 and 26 °C, the CO₂ concentration in the reference chamber was set at 400 µmol·mol⁻¹, and the airflow rate was 500 µmol·s⁻¹. Prior to formal measurements, leaves were pre-illuminated at a PPFD of 800 µmol·m⁻²·s⁻¹ for 15 min to achieve steady-state photosynthesis [24]. The net photosynthetic rate (P_n) was recorded, and measured the stomatal conductance. The light response parameters of the seedlings were derived by fitting the obtained net photosynthetic rate values using the rectangular hyperbolic modification model in the “Photosynthesis Calculation” software, version 4.1.1 [30, 31]. For statistical comparison between groups, the P_n value obtained at the highest light intensity was used, and an independent samples t-test was applied. A P-value of less than 0.05 was considered statistically significant.

Determination of chlorophyll content

Twenty seedlings were randomly selected from each treatment group. The second fully expanded leaf from each seedling was harvested for chlorophyll content analysis. Approximately 0.1 g of fresh leaf tissue (midrib removed) was weighed, ground in extraction buffer, and brought to a final volume of 10 mL. After incubating the extract in darkness for 3 h, the supernatant was then collected, and 200 µL aliquots were transferred to a 96-well plate. Absorbance was measured at 665 nm and 649 nm using a microplate reader (MULTISKAN GO, Thermo Scientific). Chlorophyll a, b, and total chlorophyll concentrations were calculated according to the specifications of the reagent kit (Suzhou Grace Biotechnology Co., Ltd. G0601W) [32].

Determination of biochemical parameters

Twenty seedlings were randomly sampled from each of the treatment groups. The second leaf of each selected seedling was designated for measuring biochemical parameters. Fresh leaf tissue was homogenized in 1 mL of pre-cooled extraction buffer and centrifuged at 12,000 rpm for 10 min at 4 °C.

Osmotic adjustment substances

The concentrations of soluble sugars (SS, G0501W48), free proline (FP, G0111W48), and soluble proteins (SP, G0417W) were determined using commercial assay kits (Suzhou Grace Biotechnology Co., Ltd.) according to the manufacturer’s instructions [33, 34].

Antioxidant enzyme activities

The resulting supernatant was used to determine the activities of superoxide dismutase (SOD, G0101W48), catalase (CAT, G0105W48), and peroxidase (POD, G0107W48) using specific kits (Suzhou Grace Biotechnology Co., Ltd.) [35]. For all biochemical parameters, differences between the HH and HL groups were analyzed using an independent samples t-test. A P-value of less than 0.05 was considered statistically significant.

Study sites and experimental design

The experiment was conducted within the natural habitat of T. sinense in the Dafengding National Nature Reserve (102°52′-103°20′ E, 28°32′-28°50′ N), situated in Meigu County, Sichuan Province, China. Based on previous surveys indicating that T. sinense populations primarily occur between 2000 and 2500 m elevation within the reserve [23], we established two experimental sites to represent the upper and lower limits of this altitudinal gradient. The high-altitude site (2448 m a.s.l.; 103°09′10.86″ E, 28°46′44.84″ N) and the low-altitude site (2023 m a.s.l.; 103°08′22.34″ E, 28°47′17.16″ N) were selected. The two sites are separated by an approximate vertical distance of 425 m, creating a natural thermal gradient to simulate climate warming scenarios. Microclimate monitoring was conducted using an automatic weather station (QCC4, Shandong Tianhe Environmental Technology Co., Ltd., China) deployed at each site. These instruments were programmed to record measurements of air temperature, relative humidity, and intensity of illumination at 10-day intervals. The annual average values of these meteorological factors for each site were calculated, which serve as the basis for our thermal gradient assessment. The soil indicators from the two experimental sites were analyzed by Lanzhou Bolisen Ecology Technology Co., Ltd. Each site contained a delimited experimental plot measuring 4 × 4 m (Fig. 2). To prevent herbivory and anthropogenic interference, both plots were enclosed with 2-m-high galvanized wire fencing.

Fig. 2.

The study area for this experiment. A Geographic location of the study area. B Experimental plot at the high-elevation site. C Experimental plot at the low-elevation site. D Germinated seedlings at the high-elevation site. E Germinated seedlings at the low-elevation site

Results

Environmental gradient verification

The downward transplantation successfully established a distinct climatic gradient. As designed, the low-altitude site had a significantly higher mean annual air temperature than the high-altitude control site (+ 2.4 °C; P = 0.031) (Table 1), creating a warming scenario consistent with mid-century projections. Beyond this primary thermal contrast, a significant reduction in intensity of illumination at the low-altitude site was also recorded (P = 0.001). However, other environmental factors, including key soil properties such as total nitrogen and available phosphorus, showed no significant differences between sites (all P > 0.05; Table 1). Therefore, the phenotypic and physiological differences observed in T. sinense seedlings can be primarily interpreted in the context of this imposed thermal gradient.

Table 1.

Environmental factors in different months in low-altitude and high-altitude areas

Environmental Factor	High-Altitude Site	Low-Altitude Site	t	P
Air Temperature (℃)
Annual Mean	9.5 ± 2.022	11.9 ± 1.997	-1.859	0.031
Relative Humidity (%)
Annual Mean	83.3 ± 1.155	79.5 ± 1.215	2.285	0.025
Intensity of Illumination (Lux)
Annual Mean	9182 ± 721.307	6311 ± 437.892	3.403	0.001
Soil Properties
Moisture Content (%)	49.423 ± 0.517	46.253 ± 1.091	2.626	0.058
Total Nitrogen (g/kg)	5.933 ± 0.160	5.983 ± 0.115	-0.254	0.812
Rapidly Available Nitrogen (mg/kg)	582.163 ± 12.431	582.193 ± 15.524	-0.002	0.999
Total Phosphorus (g/kg)	0.763 ± 0.009	0.743 ± 0.026	0.728	0.507
Rapidly Available Phosphorus (mg/kg)	11.157 ± 0.643	10.993 ± 0.789	0.160	0.880
Total Carbon (g/kg)	139.093 ± 7.127	143.233 ± 3.313	-0.527	0.626

The low altitude is 2023 m and the high altitude is 2448 m. Differences were considered statistically significant at P < 0.05, highly significant at P < 0.01, and extremely significant at P < 0.001, with the same significance levels applying hereafter

Seed germination and seedling growth

The downward transplantation profoundly enhanced early establishment and growth in T. sinense, significantly improving both seed germination and subsequent seedling development (Table 2). Seeds in the HL treatment exhibited markedly higher germination percentage, germination potential, and germination index than those in the HH control (all P < 0.001). This initial advantage was sustained through the seedling stage, as HL seedlings developed significantly greater basal diameter and plant height compared to their HH counterparts (both P < 0.001). Consequently, biomass accumulation was substantially greater in the HL group, with root, stem, leaf, and total dry weights all being significantly elevated (all P < 0.001). In contrast, the root-to-shoot ratio remained comparable between treatments (P = 0.077), indicating that the pronounced growth stimulation did not alter the proportional allocation of resources between above- and below-ground parts.

Table 2.

Germination characteristics of transplanted seeds of T. sinense

Characteristics	HH	HL	t	P
Germination rate (%)	23.17 ± 1.461	59.17 ± 2.092	−14.109	0.000
Germination potential (%)	15.83 ± 1.100	49.33 ± 1.806	−15.841	0.000
Germination index	0.98 ± 0.062	2.53 ± 0.087	−14.530	0.000
Basal diameter (mm)	0.40 ± 0.016	1.28 ± 0.014	-41.141	0.000
Plant height (cm)	1.3 ± 0.045	3.0 ± 0.054	-24.321	0.000
Root dry weight (g)	0.0225 ± 0.001	0.0529 ± 0.002	-14.977	0.000
Stem dry weight (g)	0.0322 ± 0.001	0.0724 ± 0.001	-20.615	0.000
Leaf dry weight (g)	0.0259 ± 0.001	0.0469 ± 0.001	-10.785	0.000
Total dry weight (g)	0.0806 ± 0.002	0.1722 ± 0.003	-25.148	0.000
Root-Shoot ratio	0.3947 ± 0.024	0.4454 ± 0.014	-1.819	0.077

HH represents high-altitude control seeds; HL represents high-altitude to low-altitude transplanted seeds. The values indicate the mean ± SE

Seedling survival analysis

Downward transplantation imposed severe survival pressure on T. sinense seedlings, resulting in significantly higher mortality in the HL group compared to the HH controls over the five-month monitoring period (Table 3). This disparity was most acute during the first month post-germination: the mortality rate of one-month-old HL seedlings reached 59%, vastly exceeding the 2% observed in HH seedlings. Correspondingly, the life expectancy (e_x) of one-month-old HL seedlings was only 2.17 months, in contrast to 4.86 months for HH seedlings (Table 3). The survival curve further demonstrated this trend, showing a sharp decline for HL seedlings within the first two months, whereas HH seedlings maintained a stable, high survival rate (Figs. 3A).

Table 3.

Dynamic life table of transplanted T. sinense seedlings

Seedlings	Month old	Age class	ax	lx	dx	qx	ex
HH	0	1	23	1000	16	0.02	4.86
	1	2	23	984	0	0.00	3.93
	2	3	23	984	15	0.02	2.93
	3	4	22	969	7	0.01	1.97
	4	5	22	962	41	0.04	0.98
	5	6	21	921	-	-	-
HL	0	1	59	1000	592	0.59	2.17
	1	2	24	408	30	0.08	3.55
	2	3	23	378	18	0.05	2.82
	3	4	22	360	11	0.04	1.93
	4	5	21	349	9	0.03	0.99
	5	6	20	341	-	-	-

Here, x represents the age of seedlings (in months); a_x indicates the number of surviving individuals at the x-age class; l_x represents the standardized number of surviving individuals at the beginning of time x (generally converted to 1,000); d_x refers to the standardized number of deaths from the x to x + 1 age class; q_x represents the mortality rate from the x to x + 1 age class; and e_x signifies the life expectancy of individuals in the x age-class. HH represents high-altitude control seedlings; HL represents high-altitude to low-altitude transplanted seedlings. The definitions of HH and HL below are the same as the current ones

Fig. 3.

The (A) survival curve (B) cumulative survival rate, C cumulative morality rate, D morality density, and (E) hazard rate of transplanted T. sinense seedlings. Survival analysis were assessed for one hundred seedlings over a continuous five-month period. S(t), F(t), f(t), and h(t) represent cumulative survival rate, cumulative mortality rate, mortality density function, and hazard rate function, respectively. Asterisks indicate significance at P < 0.05 (*), P < 0.01 (**), and P < 0.001 (***) for comparisons between treatments. The same applies below

Analysis of survival functions confirmed this pattern: the cumulative survival rate was significantly lower, and the cumulative mortality rate significantly higher, in HL seedlings (Fig. 3B and C). Furthermore, the mortality density (fₓ) and hazard rate (λₓ) for HL seedlings peaked sharply at the one-month stage before declining substantially by the second month, pinpointing this initial period as a critical survival bottleneck under the warming-simulation environment (Fig. 3D and E). In contrast, HH seedlings maintained stable and low mortality density and hazard rates throughout the study period.

Leaf structural and anatomical traits

To acclimate to the warmer environment, T. sinense seedlings exhibited marked adjustments in leaf structure and anatomy. Seedlings in the HL treatment did not differ significantly from HH controls in leaf length or width (both P > 0.05; Table 4). In contrast, they developed leaves with a significantly lower specific leaf area (SLA) (P < 0.001; Table 4), indicating a shift toward structurally denser leaves under warming. Consistent with this reduced SLA, HL seedlings displayed distinct ultrastructural adaptations (Fig. 4). Their chloroplasts were well-developed, spindle-shaped, and peripherally arranged along the cell wall (Fig. 4A and D). Furthermore, leaves exhibited thinner cell walls with clear, intact chloroplast membranes (Fig. 4B and E). In addition, an increased accumulation of starch grains within chloroplasts was observed in HL seedlings, alongside a marked reduction in osmiophilic globules (Fig. 4C and F).

Table 4.

Leaf characteristics indicators of transplanted T. sinense seedlings

Characteristics	HH	HL	t	P
Leaf length (cm)	1.98 ± 0.048	1.89 ± 0.049	1.291	0.204
Leaf width (cm)	0.93 ± 0.022	0.92 ± 0.029	0.413	0.682
Leaf area (cm2)	5.5133 ± 0.264	4.9699 ± 0.259	1.471	0.150
SLA (cm− 2·g− 1)	707.4024 ± 35.758	501.7846 ± 24.846	4.722	0.000

The values indicate the mean ± SE. SLA represents specific leaf area

Fig. 4.

Anatomical structures of the leaves of transplanted T. sinense seedlings. In the image, the meanings of the uppercase abbreviations are: C: Chloroplast; CM: Chloroplast membrane; CW: Cell wall; OG: Osmiophilic globule; SG: Starch grains. The red arrows represent the areas where there are differences

Leaf functional traits and photosynthetic capacity

Seedlings in the HL treatment exhibited a major shift in leaf nutrient economy without changes in chlorophyll content (Table 5). Their leaf nitrogen content was dramatically lower than that of HH seedlings (P < 0.001), while carbon content remained stable, leading to a substantially higher carbon-to-nitrogen (C/N) ratio (P < 0.001). This strategy of reduced nitrogen investment coincided with a reprogramming of photosynthetic performance. HL seedlings maintained higher net photosynthetic rates (Pn) across light intensities (Fig. 5). Correspondingly, stomatal conductance (Gs) was also significantly higher in the HL treatment compared to the HH control (P < 0.001), indicating less restrictive gas exchange. Specifically, they achieved a significantly higher apparent quantum yield (AQY) and maximum photosynthetic rate (P_nmax) (both P < 0.001; Table 6), demonstrating enhanced photosynthetic capacity under both limiting and saturating light. This was accompanied by a significant increase in dark respiration (R_d, P < 0.001) and a lower light saturation point (LSP, P < 0.05), whereas the rise in light compensation point (LCP) was not statistically significant (P = 0.054; Table 6).

Table 5.

Leaf functional traits and photosynthetic capacity of transplanted T. sinense seedlings

Characteristics	HH	HL	t	P
Content of chlorophyll a (mg·L− 1)	1.190 ± 0.019	1.184 ± 0.025	0.200	0.842
Content of chlorophyll b (mg·L− 1)	0.414 ± 0.009	0.420 ± 0.011	-0.425	0.673
Ratio of chlorophyll a/b	2.909 ± 0.095	2.853 ± 0.095	0.414	0.681
Content of chlorophyll (mg·L− 1)	1.605 ± 0.020	1.604 ± 0.029	0.007	0.994
Content of leaf nitrogen (%)	1.671 ± 0.035	1.096 ± 0.029	12.715	0.000
Content of leaf carbon (%)	43.618 ± 0.312	43.740 ± 0.354	− 0.257	0.798
Ratio of leaf carbon/nitrogen	26.285 ± 0.487	40.499 ± 1.211	-10.893	0.000

Fig. 5.

Light response curve of transplanted T. sinense seedlings

Table 6.

Photosynthetic parameters of transplanted T. sinense seedlings

Treatments	Parameters
	AQY (mmolCO2/molphotons)	P nmax (µmol·m− 2·s− 1)	LCP (µmol·m− 2·s− 1)	LSP (µmol·m− 2·s− 1)	R d (µmol·m− 2·s− 1)	Gs (mol·m− 2·s− 1)
HH	0.061 ± 0.001	4.305 ± 0.083	3.937 ± 0.365	1242.436 ± 92.478	0.229 ± 0.020	0.0812 ± 0.004
HL	0.097 ± 0.002	5.419 ± 0.025	4.814 ± 0.179	1022.687 ± 43.031	0.436 ± 0.018	0.1319 ± 0.007
t	-14.681	-12.893	2.154	-2.159	-7.662	-6.182
P	0.000	0.000	0.054	0.046	0.000	0.000

AQY: apparent quantum yield; P_nmax: maximum photosynthetic rate; LCP: light compensation point; LSP: light saturation point; R_d: dark respiration rate; G_s: stomatal conductance

Physiological and biochemical responses

Downward transplantation triggered a coordinated physiological response in T. sinense seedlings, characterized by concurrent osmotic adjustment and a reconfigured antioxidant defense system. HL seedlings accumulated significantly higher levels of key osmotic adjustment substances, including soluble sugar, soluble protein, and free proline, compared to the HH control (Figs. 6A-C).

Fig. 6.

The contents of (A) soluble sugar, B soluble protein, and (C) free proline; and the activities of (D) SOD, (E) CAT, and (F) POD in transplanted T. sinense seedlings. SOD: superoxide dismutase; CAT: catalase; POD: peroxidase

This osmotic adjustment was accompanied by a selective modulation of the antioxidant enzyme cascade. While SOD activity remained unchanged, CAT activity was significantly upregulated in HL seedlings. In contrast, POD activity was significantly downregulated relative to HH seedlings (Figs. 6D-F). Corresponding to these enzymatic adjustments, the concentration of malondialdehyde (MDA), a marker of oxidative damage, was significantly lower in HL seedlings than in HH controls (Figure S2).

Discussion

A Stage-Dependent demographic paradox

Climate warming reshapes ecosystems by imposing novel selective pressures, to which relict species with specialized traits and limited genetic diversity are particularly vulnerable [36–38]. T. sinense, a vesselless angiosperm, exemplifies such a species facing acute challenges [39]. Our experimental warming simulation (Table 1) reveals a demographic paradox in its response: rather than uniformly enhancing or suppressing performance, warming induces a critical, stage-dependent trade-off. This finding challenges simplistic forecasts of species fate and underscores that life-stage transitions act as pivotal bottlenecks under rapid climate change [40].

Our results demonstrate that elevated temperatures strongly enhanced early life-history traits, significantly improving seed germination and subsequent seedling growth of T. sinense (Table 2). Consequently, biomass accumulation increased substantially across all organs (roots, stems, leaves, and total plant mass) compared to the high-altitude controls (Table 2). However, the root-to-shoot ratio remained stable between treatments, indicating that warming accelerated overall growth proportionally without altering the fundamental biomass allocation strategy. This pattern of proportional growth enhancement under moderate warming aligns with observations in other temperate tree species, where it is often attributed to increased metabolic and resource-use efficiency rather than strategic allocation shifts [13].

Paradoxically, this pronounced growth stimulus coincided with a dramatic survival bottleneck. Seedlings in the warming treatment suffered extreme mortality (59%) within the first month post-germination—a period that survival analysis identified as one of critically high risk (Table 3; Fig. 3). We posit that this window corresponds to the ecophysiologically sensitive transition from heterotrophic dependence on seed reserves to photosynthetic autonomy. During this vulnerable phase, high metabolic demands coupled with diminishing reserves likely amplify seedling susceptibility to combined thermal and hydric stress [41]. Thus, warming acts as a dual agent: a growth promoter for established survivors, yet a stringent filter during early establishment [42, 43].

This stage-dependent demographic trade-off implies that the future trajectory of T. sinense under climate warming may hinge less on its capacity for growth in later stages, and more on the microclimatic and biotic conditions that determine early seedling survival. Identifying this precise vulnerability provides a critical target for evidence-based conservation strategies aimed at mitigating climate change impacts on this endangered relict species.

Coordination of leaf structural and photosynthetic traits

According to the leaf economics spectrum, high specific leaf area (SLA) is generally associated with high photosynthetic capacity, reflecting a rapid resource-acquisition strategy, whereas low SLA typically indicates a conservative, resource-retentive strategy with reduced photosynthetic rates [44]. Our results for T. sinense seedlings under experimental warming, however, revealed a notable decoupling from this classic trade-off: despite a significant decrease in SLA (Table 4), photosynthetic performance was simultaneously enhanced. This was evidenced by a higher net photosynthetic rate (P_n) under a range of conditions, which was further characterized by increased apparent quantum yield (AQY), maximum photosynthetic rate (P_nmax), light saturation point (LSP), and dark respiration rate (R_d) (Fig. 5; Table 6). Critically, this enhancement in P_n was accompanied by a significant increase in stomatal conductance (G_s), indicating that greater CO₂ availability supported, rather than limited, the photosynthetic process.

This synergy was underpinned by coordinated anatomical and biochemical adjustments. At the ultrastructural level, HL seedlings exhibited thinner cell walls (Fig. 4B and E). This modification likely reduced mesophyll resistance and facilitated CO₂ diffusion to chloroplasts, representing a key adaptation for enhancing carboxylation under warmer conditions [45]. Concurrently, their chloroplasts were well-developed, spindle-shaped, and arranged peripherally along the cell walls (Fig. 4A and D), a configuration that may optimize light capture. Furthermore, increased starch accumulation alongside a marked reduction in osmiophilic globules within these chloroplasts (Fig. 4C and F) suggests a shift in carbon metabolism and storage under warming. Correlation analysis further corroborates this integrated response. We found that leaf AQY was positively correlated with whole-plant growth traits, such as height and basal diameter, and negatively correlated with leaf structural traits, such as SLA (all P < 0.05) (Figure S2). This indicates that the enhanced photosynthetic capacity under warming is not an isolated physiological change but is functionally coupled with both improved whole-plant growth and a specific leaf structural transformation towards denser tissue.

These structural and growth-level adaptations were paralleled by a strategic reallocation of resources at the physiological level. Despite a lower overall leaf nitrogen content per unit mass, HL seedlings achieved higher photosynthetic rates. This indicates a preferential allocation of nitrogen to the photosynthetic apparatus, thereby enhancing photosynthetic nitrogen-use efficiency, which is a well-documented acclimation strategy to high temperatures [46]. The concurrent increase in G_s and P_n strongly suggests that the primary driver of enhanced photosynthesis was improved biochemical capacity within the mesophyll, rather than a relaxation of stomatal limitation. Together, these coordinated structural and physiological changes enabled T. sinense seedlings to maintain a high carbon gain under simulated warming, despite adopting a denser leaf construction with lower SLA.

Coordinated cellular defense underpins thermal acclimation

Beyond structural and photosynthetic adjustments, successful thermal acclimation in plants fundamentally relies on cellular homeostasis under stress [47]. Our findings indicate that T. sinense seedlings acclimate to warming by deploying a specifically tuned cellular defense program, rather than mounting a broad, non-specific stress response. This acclimation strategy features a selective reconfiguration of the antioxidant system: CAT activity was significantly upregulated, whereas POD activity decreased and SOD remained unchanged in HL seedlings (Fig. 6D-F). This precise modulation points to a strategic shift toward CAT-mediated H₂O₂ detoxification, a pathway associated with efficient oxidative stress management in heat-tolerant plants [47]. The success of this targeted approach is unequivocally demonstrated by the significantly reduced levels of malondialdehyde (MDA) in warmed seedlings (Figure S1), confirming effective containment of membrane lipid peroxidation [48].

Concurrently, seedlings accumulated substantial amounts of key osmoprotectants, namely soluble sugars, proline, and soluble proteins (Figs. 6A-C). These compatible solutes are known to act synergistically to maintain cellular turgor, stabilize proteins and membranes, and thereby safeguard metabolic homeostasis under thermal stress [33, 49]. The integration of this targeted oxidative defense with robust osmotic adjustment constitutes the core physiological machinery that underpins the observed enhancements in seedling growth and photosynthetic performance under warming conditions.

Conclusion and conservation implications

This study reveals that T. sinense seedlings employ an integrated acclimation strategy in response to warming, coordinating morphological, physiological, and biochemical adaptations. However, despite the enhanced performance of seedlings that survive the vulnerable establishment stage, their overall recruitment is severely constrained by extremely high early mortality under warmed conditions. Consequently, future climate warming is likely to restrict natural regeneration by increasing seedling mortality during this critical life stage.

To address this establishment bottleneck, conservation strategies for T. sinense should prioritize interventions at the early seedling stage. Specifically, management efforts should focus on two complementary actions: (1) moderating understory microclimates at natural recruitment sites to buffer extreme temperature and soil moisture fluctuations during the first month post‑germination, thereby directly mitigating the primary cause of seedling mortality; and (2) conserving and enhancing genetic diversity across populations to improve seedling adaptive capacity and survival through the critical first-month establishment phase. By securing the seedling establishment phase, these targeted measures can help translate the species’ inherent physiological resilience into successful recruitment in a warming climate.

Authors’ contributions

Hongyan Han and Xiaohong Gan conceived and designed the experiments. Wenjing He drafted the manuscript and performed the statistical analyses. Hongyan Han revised the manuscript. Rong Wang collected samples and performed the experiments. Luwei Yang and Zili Wan contributed to the statistical analyses. Hamid Aly contributed to the manuscript through comprehensive language polishing and stylistic refinement. All authors read and approved the final versions of the manuscript.

Funding

This study was funded by National Natural Science Foundation of China (No. 32400303; No.32070371), the Natural Science Foundation of Sichuan Province (No.2023NSFSC1272), and the Innovation Team Funds of China West Normal University (KCXTD2022-4). Doctoral Research Start-up Fund (22kE022).

Data availability

All data that support the findings of this study will upload to the designated public database, once the article is considered for acceptance.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.