Methyl jasmonate induces conservative trait shifts that improve drought resilience in clonal grass

Abstract

Recurrent drought events are becoming increasingly frequent under climate change, yet how perennial clonal grasses adjust their functional traits and recovery trajectories across repeated drought–recovery cycles remain poorly understood. Here, we examined whether exogenous application of methyl jasmonate (MeJA) modifies physiological performance, biomass allocation, and recovery dynamics of the clonal grass Festuca rubra under recurrent drought. Plants were exposed to three consecutive drought–recovery cycles in a controlled growth-chamber experiment with factorial drought and MeJA treatments. We quantified physiological traits (chlorophyll concentration, maximum quantum efficiency of PSII), structural traits (specific leaf area, leaf dry matter content), and performance-related traits (ramet number and biomass allocation). Recurrent drought progressively reduced photosynthetic performance and ramet production, with recovery becoming increasingly incomplete across cycles. In contrast, MeJA treatment mitigated drought-induced declines in chlorophyll concentration and PSII efficiency and stabilized recovery responses. MeJA also altered biomass allocation patterns, increasing aboveground biomass and reducing root-to-shoot ratios under drought, in contrast to drought-only plants, which showed enhanced belowground investment. Post-hoc analyses revealed that MeJA + drought plants differed significantly from drought-only plants for aboveground biomass and several functional traits. Overall, our results demonstrate that MeJA modulates drought responses and recovery trajectories in F. rubra, promoting conservative trait expression and improved performance under recurrent drought. These findings highlight the potential role.

Introduction

Drought is among the most pervasive global change drivers, threatening plant productivity, biodiversity, and ecosystem stability^1,2. While the impacts of drought on plants are well-documented^3–5, most existing research on the effects of drought on plants focuses on the effects of a single drought event e.g.,^6–8. In nature, perennial plants, however, do not experience a single drought event, but rather go through repeated drought and recovery cycles³. Studies exploring how plants and ecosystems respond to recurring drought and how these responses change over repeated drought and recovery phases are, however, still quite sparse^9–11.

Repeated stress is known to create stress memory in plants, shaping their future performance^12–14. For example, recurrent droughts can impose cumulative legacies or trigger distinct physiological adjustments compared to a single episode, often reducing recovery capacity and leading to long-term declines in performance^15–17. Consequently, results from single-event studies cannot reliably predict plant and ecosystem responses under repeated stress^18,19. The performance of plants under successive drought-recovery cycles has been studied in annual crops^20–24. Much less is, however, known about how perennial clonal grasses, which dominate temperate grasslands, sustain or adjust their performance across repeated drought–recovery cycles. These species rely heavily on vegetative persistence and clonal spread, making their long-term responses to recurrent drought ecologically critical but poorly understood.

Structural traits (specific leaf area (SLA) and leaf dry matter content (LDMC)) and physiological traits such as leaf water potential, chlorophyll content, and maximum quantum efficiency of Photosystem II (PSII) often return to predrought levels during recovery^15,25,26. However, such apparent resilience may not be sustainable under repeated stress. In perennial grassland dominants, for example, growth-related traits such as ramet production or biomass allocation may remain constrained even when physiological measures recover. This raises the question of which regulatory mechanisms enable plants to sustain or modify recovery trajectories when drought recurs.

Hormonal signalling provides one potential mechanism underlying these dynamics. While abscisic acid (ABA) is recognized as the central regulator of plant drought responses, jasmonates, particularly methyl jasmonate (MeJA), are emerging as additional modulators of drought tolerance and memory^27–29. Unlike ABA, which mainly controls stomatal closure and osmotic adjustment, MeJA influences a broader suite of processes, including antioxidant protection, biomass distribution, and metabolite accumulation that help to maintain photosynthetic efficiency under drought^30–32. Importantly, MeJA has been linked to priming and stress memory in crops or model organisms^33–35, suggesting that it may not only buffer immediate damage but also reprogram recovery trajectories when stress recurs. However, it is unclear how it affects recurrent stress and recovery trajectories in clonal species.

To address this gap, we studied the perennial clonal grass Festuca rubra under factorial combinations of recurrent drought and exogenous MeJA application. We aimed to: (i) Quantify the impact of recurrent drought on plant physiological and fitness-related traits (e.g., SLA, chlorophyll content, ramet number, biomass allocation); (ii) test whether recovery trajectories shift across successive drought–recovery cycles, indicating stress legacy in trait responses; and (iii) evaluate the role of MeJA in modulating stress responses and recovery trajectories.

We hypothesized that: (H1) Recurrent drought will progressively reduce photosynthetic efficiency, structural integrity, and vegetative reproduction, suppressing plant physiological and clonal performance traits across cycles. (H2) Across successive drought–recovery cycles, recovery will become increasingly incomplete, with physiological traits recovering more rapidly than structural or fitness traits, indicating partial stress memory with trait-specific recovery dynamics. (H3) MeJA treatment will mitigate the adverse effects of recurrent drought by promoting more conservative trait syndromes and stabilizing recovery trajectories.

To test these hypotheses, we conducted a controlled growth-chamber experiment combining recurrent drought and MeJA treatments. We measured a suite of physiological and growth traits across multiple drought–recovery cycles.

Materials and methods

Study system

Festuca rubra L. is a dominant perennial grass in temperate and alpine grasslands in the northern hemisphere³⁶. It reproduces sexually and vegetatively, with vegetative spread via intravaginal and extravaginal tillers on rhizomes often dominating over generative reproduction³⁷. We studied the hexaploid cytotype, a widely distributed member of the F. rubra complex, collected from a 300 to 400-year-old meadow in the Krkonoše Mountains, Czech Republic (3.75 km ESE of Pec pod Sněžkou; 50°41′25.165″N, 15°47′41.525″E; 895 m a.s.l.)³⁸. Plant and soil material were collected under permission granted by the Krkonoše National Park Administration (contract no. OSML 38 − 4/2018). A voucher specimen, collected by V. Hadincová, has been deposited at the Herbarium of the National Museum, Prague (PRA-42728). The formal identification of the voucher and experimental material was carried out by Dr. Zuzana Münzbergová (Charles University, Prague), who has extensive experience with this species and has published several studies on its ecology and physiology^37,39–42. The work complies with the IUCN Policy Statement on Research Involving Species at Risk of Extinction and the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES).

Experimental design

The experiment involved a factorial design of drought and MeJA treatments, each with two levels (yes/no). Each treatment had 10 replicates. The soil used was collected from the same meadow where the original plants were obtained. Pots with a volume of 500 ml were filled with this soil, and one F. rubra ramet was planted into each pot. All the ramets were of similar age and size (about 3 weeks old and 3.5 to 4 cm tall). The plants were placed in a growth chamber with long-day conditions, where daytime temperatures ranged from 17 to 20 °C (16 h) and night-time temperatures from 7 to 10 °C (8 h)^37,38. Photosynthetically active radiation (PAR) at canopy level was approximately 350 µmol m⁻² s⁻¹, and relative humidity was maintained at 60–70%, consistent with growth-chamber conditions commonly used for temperate grassland species e.g^37,38. These conditions mimic those at the original site during the peak growing season. The plants were given 21 days to establish before the experiment began. The study included cycles of drought and recovery phases, each lasting 30 days (Drought stage 1 - Recovery stage 1 - Drought stage 2 - Recovery stage 2 - Drought stage 3 - Recovery stage 3).

Control plants were watered daily, maintaining about 1 cm of water above the pot bottom throughout the experiment. Drought-stressed plants received fixed amounts of water (30 ml) at progressively longer intervals: every 7 days during drought phase 1, every 10 days during phase 2, and every 15 days during phase 3. This setup was created to maintain minimal soil moisture for the survival of plants without fully alleviating drought stress. These intervals were chosen to gradually increase drought intensity across successive phases, maintaining stress without causing complete desiccation. During each recovery stage, drought-stressed plants were watered in the same manner as controls. Soil moisture was continuously monitored with TMS dataloggers to verify that drought treatments caused consistent and progressive drying (see Table S1).

During drought phases, MeJA-treated plants received 3 ml of 10 µM MeJA solution applied directly to the soil every 48 h, while drought-only plants received an equal volume (3 ml) of distilled water at the same intervals to equalize liquid input.

To prevent nutrient limitation during prolonged cultivation, minimal fertilization was applied once during recovery phase 2, when initial wilting symptoms were observed in the well-watered control plants. The same fertilizer solution was applied simultaneously and in equal amounts to all plants across all treatments, including control, drought, MeJA, and MeJA + drought treatments. This uniform application ensured that fertilization did not introduce systematic differences among treatments and did not bias the effects of drought or MeJA.

We deliberately chose the soil MeJA application over the more common foliar spraying e.g^43,44, As foliar methods present notable drawbacks, including MeJA’s high volatility leading to environmental loss and uneven droplet deposition, which can result in inconsistent exposure across leaf surfaces. In contrast, soil application offers a more uniform and controlled delivery, promotes root uptake, and substantially reduces volatilization. While partial degradation or microbial transformation of MeJA in soil has been reported⁴⁵, prior studies confirm that root-applied MeJA can be effectively absorbed and trigger physiological and ecological responses of plants^46,47.

Plant performance

We described plant performance by measuring several physiological and fitness-related traits. Plant physiology was characterized by the following traits: chlorophyll concentration, chlorophyll fluorescence (Fv/Fm), specific leaf area (SLA), leaf dry matter content (LDMC), and relative water content in the leaves (RWC). Plant fitness was determined based on the number of ramets and aboveground and belowground biomass. F. rubrareproduces primarily vegetatively and did not flower during the experiment; that is why seed production was not considered a predictor of plant fitness, and the number of ramets and biomass production can be considered as suitable fitness proxies in this case see, e.g³⁷, for a similar approach).

Chlorophyll concentration, chlorophyll fluorescence (Fv/Fm), and the number of ramets were estimated at the end of each stress and recovery phase. We measured SLA, LDMC, RWC, and aboveground and belowground biomass only at the end of the recovery stage 3, as these measurements are destructive and thus could not be taken repeatedly. Biomass is widely used as a fitness indicator in plants⁴⁸, while ramet number is a recognized proxy for fitness in grasses⁴⁹.

Chlorophyll concentration was measured on three random leaves per pot using a CCM 300 chlorophyll meter (Opti-Sciences) and averaged per individual. Maximum PSII efficiency (Fv/Fm) was determined using a FluorPen FP-100 MAX (Photon System Instruments, Czech Republic) after one hour of dark acclimation, with ~ 10 leaves grouped per plant due to the narrowness of F. rubra leaves. Aboveground biomass (g), number of ramets, total plant biomass (g), LDMC, RWC, and specific leaf area (SLA, mm² mg⁻¹), were measured following approaches described by^41–51.

Data analysis

All statistical analyses were performed in R (version 4.2.1). Mixed-effects models were fitted using the lme4 package, with significance assessed via likelihood-ratio χ² tests; figures were produced using ggplot2. We used linear and generalized mixed effect models to analyze the effect of drought, MeJA, and phase, and their interactions on plant performance with pot number as a random factor. These models were used to analyze the variables measured repeatedly. The linear model was used for chlorophyll concentration following a normal distribution, while the generalized model was used for Fv/Fm with a Gamma family distribution, and the number of ramets with a Poisson family distribution.

To compare the relative importance of drought, MeJA, and their interaction over time, we also analyzed the effects of the treatment for each performance measure and phase separately using models assuming the same distributions as above. In each case, we expressed the variance explained by each predictor and compared these values.

To visualize the effects of the different treatments over time (recurrent droughts and recovery), we set the mean value of each performance measure of the well-watered plants without MeJA application separately at each time as a baseline and expressed the performance of all the other plants as a percentage of this baseline.

To study the plant’s recovery after the stress phase, the log response ratio (LRR) was calculated as the natural logarithm (ln) of plant performance during recovery divided by plant performance during the stress phase for each stress phase and performance measure separately.

We used a linear model to analyze the effect of drought, MeJA, and their interaction on the variables measured only once, i.e., SLA, LDMC, RWC, aboveground biomass, and root shoot ratio, as all followed a normal distribution.

All statistical analyses were performed on absolute values. However, for graphical visualization of chlorophyll concentrations, Fv/Fm and number of ramets, normalization to the control (100%) was applied.

The significance of fixed effects in both linear mixed-effects models and generalized linear mixed-effects models was assessed using likelihood-ratio χ² tests based on model comparisons, implemented via the anova() and drop1() functions in the lme4 package. Generalized linear mixed-effects models were fitted with a Poisson distribution for ramet number and a Gamma distribution for Fv/Fm. F-values reported in tables correspond to Wald-type tests obtained via ANOVA.

To compare the relative importance of drought, MeJA, and their interaction across stress and recovery phases, we partitioned explained variation using sums of squares. For each trait and phase, linear or generalized linear models were fitted, including drought, MeJA, and their interaction as predictors, assuming the same distributions as described above. The contribution of each predictor was calculated from the sums of squares and expressed as a percentage of the total model sum of squares. These percentages were used to visualize the relative importance of predictors across phases in the stacked bar plots shown in Figs. 1, 2.

Fig. 1.

Effects of recurrent drought and methyl jasmonate (MeJA) on plant physiological and performance traits across three consecutive drought–recovery cycles. Mean (± 95% confidence intervals) values of (a) chlorophyll concentration, (b) maximum quantum efficiency of PSII (Fv/Fm), and (c) number of ramets measured at the end of each stress (Stress 1–3) and recovery (Recovery 1–3) phase. Treatments include well-watered control, drought-only, MeJA-only, and MeJA + drought. For visualization, values are expressed relative to the mean of the well-watered control at each phase (dotted line = 100%). All statistical analyses were performed on absolute (non-normalized) values. Sample size: n = 10 per treatment.

Fig. 2.

Relative contribution of drought, MeJA, and their interaction to explained variation across stress and recovery phases. Stacked bar plots showing the proportion of explained variation (%) attributable to drought, MeJA, their interaction (Drought × MeJA), and residual variation for (a) chlorophyll concentration, (b) Fv/Fm, and (c) number of ramets during each stress and recovery phase. Values were derived from partitioning sums of squares from models fitted separately for each trait and phase. Higher percentages indicate a greater relative contribution of a given predictor to total explained variation.

Results

Effect of recurrent drought and MeJA on plant traits

Both drought and MeJA significantly affected chlorophyll concentration, chlorophyll fluorescence (Fv/Fm), and ramet number, but the strength and direction of these effects varied across the stress and recovery phases and among traits (Table 1).

Table 1.

The Effects of MeJA, drought, time, and their interactions on plant performance were tested separately for stress and recovery phases. Significant values (p ≤ 0.05) are in bold.

	Chlorophyll concentration				Fv/Fm				Number of ramets
	Stress		Recovery		Stress		Recovery		Stress		Recovery
	F value	p	F value	P	F value	P	F value	p	F value	p	F value	p
Drought	55.63	<0.001	6.77	0.01	77.2	<0.001	2.69	0.103	135.25	<0.001	49.28	<0.001
MeJA	10.31	<0.001	21.81	<0.001	24.53	<0.001	9.79	0.002	17.73	0.025	49.88	<0.001
Time	266.61	<0.001	383.68	<0.001	93.51	<0.001	19.06	<0.001	527.58	< 0.001	1119.89	<0.001
Drought: MeJA	38.45	<0.001	0.46	0.495	18.11	0.008	3.75	0.055	34.35	< 0.001	3.41	0.002
Drought: Time	24.2	<0.001	3.84	0.053	13.29	<0.001	0.12	0.729	8.37	< 0.001	29.15	<0.001
MeJA:Time	46.78	<0.001	0.006	0.936	3.98	0.074	33.73	<0.001	21.33	< 0.001	55.49	<0.001
Drought:MeJA: Time	4.77	0.03	4.78	0.031	13.4	0.001	10.4	<0.001	0.52	0.466	4.77	0.028

During stress phases, drought strongly reduced chlorophyll concentration, Fv/Fm, and ramet number. MeJA mitigated these declines: MeJA and MeJA + drought plants maintained higher chlorophyll levels than drought-only plants and, in stress phase 3, significantly exceeded control values (MeJA: F = 10.31, p < 0.001.CCM; MeJA + drought: F = 38.45, p < 0.001 CCM; Drought F = 55.63 p < 0.001 CCM; n = 10; Fig. 1a). MeJA treatment also buffered declines in Fv/Fm, so that Fv/Fm values in MeJA-treated plants did not differ significantly from those in control plants during stress phases (Fig. 1b). MeJA effects on ramet production became evident only in stress phase 3, when MeJA + drought plants produced significantly more ramets than drought-only plants (Fig. 1c).

During recovery phases, MeJA continued to exert strong effects. Drought alone did not significantly alter Fv/Fm during recovery (F = 2.69, p = 0.103; Table 1), whereas MeJA significantly affected chlorophyll concentration (F = 21.81, p < 0.001), Fv/Fm (F = 9.79, p = 0.002), and ramet number (F = 49.88, p < 0.001). The triple interaction among drought, MeJA, and time was also significant for chlorophyll concentration (F = 4.78, p = 0.031), Fv/Fm (F = 10.40, p < 0.001), and ramet number (F = 4.77, p = 0.028; Table 1). By the end of recovery phase 3, MeJA and MeJA + drought plants had significantly more ramets than controls, whereas drought-only plants had fewer (p < 0.05; Fig. 1c).

Recurrent drought and MeJA significantly affected both the drought and recovery stages. However, the effects differed between the stress and recovery stages and were specific to each studied trait (Table 1). The effects of recurrent drought and MeJA were stronger in the stress phase than in the recovery phase. In the stress phases, all the studied traits (chlorophyll concentration, Fv/Fm, and the number of ramets) were significantly affected by individual and interactive effects of predictors (recurrent drought, MeJA, and time) with only a few exceptions, i.e., no significant effect of the interaction of time (representing 1 st, 2nd and 3rd stress) and MeJA in Fv/Fm. The triple interaction between drought, MeJA, and time had a statistically significant effect on Fv/Fm and chlorophyll concentration but not on the number of ramets (Table 1).

Relative importance of drought, MeJA, and their interactions

MeJA accounted for the largest proportion of explained deviance in chlorophyll concentration across stress phases, while drought contributed the least (Fig. 2a). The contribution of MeJA increased progressively from stress phase one to stress phase three, indicating a strengthening hormonal effect across recurrent drought cycles. In contrast, for Fv/Fm and the number of ramets, drought treatment explained the largest proportion of variation during the stress period, whereas MeJA explained comparatively less variation overall (Fig. 2b and c).

As we progressed from recovery phase one to recovery phase three, the proportion of variation explained by MeJA increased for chlorophyll concentration, Fv/Fm, and ramet number. By recovery phase three, MeJA accounted for the largest share of explained deviance among predictors for all three traits, indicating that hormonal effects became increasingly dominant during later recovery stages (Fig. 2a–c).

Recovery responses after recurring drought

Recovery was quantified using the log response ratio (LRR), which compares trait values during recovery relative to the preceding stress phase within each treatment. Using this metric, drought-only plants showed higher LRR values for chlorophyll concentration and ramet number than MeJA-treated plants, indicating a stronger relative rebound from stress, although this did not necessarily correspond to higher absolute trait values (Fig. 3a and c), based on the LRR analysis, MeJA + drought plants showed significantly higher recovery of Fv/Fm than drought-only plants in recovery stages 1 and 2 (Fig. 3B). There was no significant difference in recovery stage 3 between drought and MeJA+drought plants in terms of the number of ramets and Fv/Fm (Fig. 3b and c).

Fig. 3.

Recovery responses following drought quantified as log response ratios (LRR). Mean (± 95% confidence intervals) LRR values for (a) chlorophyll concentration, (b) Fv/Fm, and (c) number of ramets across the three recovery phases. LRR was calculated as ln (recovery value/preceding stress value) within each treatment; positive values indicate relative improvement during recovery. Treatments include control, drought-only, MeJA-only, and MeJA + drought. Significant differences among treatments were determined using linear models (p < 0.05). Sample size: n = 10 per treatment.

Resource allocation trade-offs under recurrent drought and MeJA

MeJA, drought, and their interaction significantly affected most performance traits measured at the end of the experiment (Table 2). Specifically, SLA was significantly influenced by MeJA (F = 93.54, p < 0.001) and showed a significant MeJA × drought interaction (F = 25.81, p < 0.001), whereas drought alone had no significant effect on SLA (F = 0.11, p = 0.74). In contrast, LDMC was significantly affected by drought (F = 7.53, p = 0.010) and MeJA (F = 36.43, p < 0.001), but their interaction was not significant (F = 0.84, p = 0.365; Fig. 4a). MeJA-treated plants had the highest LDMC values compared to control plants, while drought-stressed plants had lower LDMC values (Fig. 4b). Drought-only plants showed significantly lower relative water content (RWC) than both control and MeJA + drought plants (Drought: F = 35.41, p < 0.001), whereas MeJA-only plants did not differ significantly from controls (p > 0.05; Fig. 4c). MeJA also significantly affected RWC (F = 22.27, p < 0.001), and the MeJA × drought interaction was significant (F = 23.70, p < 0.001; Table 2). There was no significant difference in relative water content between the well-watered control plants and those treated with MeJA only (Fig. 4c).

Table 2.

Effect of MeJA, drought, and their interaction on plant performance measures tested at the end of the experiment. Significant values (p ≤ 0.05) are in bold.

	SLA		LDMC		RWC		Aboveground biomass		Root to Shoot Ratio
	F	p	F	p	F	p	F	p	F	p
Drought	0.11	0.74	7.53	0.010	35.41	< 0.001	66.37	< 0.001	107.3	< 0.001
MeJA	93.54	< 0.001	36.43	< 0.001	22.27	< 0.001	34.89	< 0.001	8.60	0.005
MeJA +Drought	25.81	< 0.001	0.84	0.365	23.70	< 0.001	19.79	< 0.001	38.87	< 0.001

Fig. 4.

Effects of drought and MeJA on structural and water-related traits measured at the end of the experiment. Mean (± SE) values of (a) specific leaf area (SLA), (b) leaf dry matter content (LDMC), and (c) relative water content (RWC) for control, drought-only, MeJA-only, and MeJA + drought treatments at the end of Recovery Phase 3. Different lowercase letters above bars indicate significant differences among treatments based on Tukey-adjusted post hoc comparisons of estimated marginal means (p < 0.05) following linear models. Traits were square-root transformed for statistical analyses; values shown are back-transformed to the original scale. Sample size: n = 10 per treatment.

MeJA + drought plants had significantly higher aboveground biomass than drought-only plants (Table 2; Fig. 5a). The main effects of drought (F = 66.37, p < 0.001), MeJA (F = 34.89, p < 0.001), and their interaction (F = 19.79, p < 0.001) were all significant (Table 2). The root–shoot ratio differed significantly among treatments (Drought: F = 107.3, p < 0.001; MeJA: F = 8.60, p = 0.005; MeJA × drought: F = 38.87, p < 0.001; Fig. 5b). Drought-only plants exhibited the highest root–shoot ratio, indicating increased investment in below-ground biomass. In contrast, both MeJA-only and MeJA + drought plants showed significantly lower root–shoot ratios than drought-only plants and did not differ from each other based on Tukey-adjusted post hoc comparisons (p < 0.05).

Fig. 5.

Effects of drought and MeJA on biomass allocation at the end of the experiment. Mean (± SE) values of (a) aboveground biomass and (b) root-to-shoot ratio measured at the end of Recovery Phase 3 for control, drought-only, MeJA-only, and MeJA + drought treatments. Different lowercase letters indicate significant differences among treatments based on linear models followed by Tukey-adjusted post hoc contrasts (p < 0.05). Sample size: n = 10 per treatment.

Discussion

This study examined how recurrent drought and MeJA interact to shape the stress and recovery dynamics of Festuca rubra, a dominant clonal grass in temperate grasslands. Our findings provide strong support for the first and third hypotheses, while the second hypothesis was only partially consistent with the results. As predicted in (H1), recurrent drought suppressed photosynthesis and growth during stress phases, with physiological traits such as Fv/Fm recovering more rapidly than structural or fitness traits, creating an asymmetry between quick physiological recovery and constrained clonal expansion. We also found clear evidence that MeJA buffered declines in photosynthesis and water status, promoted structurally conservative leaves, and altered biomass allocation in favor of more stable recovery and improved overall performance under recurrent drought. By contrast, rather than showing reduced physiological declines across drought and recovery cycles, drought-only plants accumulated stress costs, with repeated collapses during stress, unstable rebounds during recovery, and an eventual failure to reach control levels.

Effects of recurrent drought on Festuca rubra during stress and recovery

Recurrent drought in F. rubra led to strong declines in chlorophyll concentration, Fv/Fm, and ramet production during stress phases, with impacts becoming progressively more severe across cycles. These cumulative effects align with drought legacy patterns observed in other perennials^52,53, yet our results highlight a distinct response in clonal grasses. Unlike many crops and woody species, where repeated drought can induce acclimation and reduce physiological damage over time^22,24, F. rubra showed no such adjustment. Instead, it increasingly constrained performance, revealing vulnerability in clonal species whose persistence relies heavily on vegetative growth.

A key novelty of our study lies in revealing trait-specific recovery trajectories following recurrent drought. While photosynthetic efficiency (Fv/Fm) consistently rebounded to control levels after each recovery phase, indicating reversible photoinhibition and short-term resilience of the photosynthetic machinery, fitness-related traits, such as ramet production and aboveground biomass, showed incomplete and progressively diminished recovery. This divergence highlights that physiological resilience does not necessarily translate into performance recovery. The limited recovery of fitness traits can be attributed to two key factors: (1) They are integrative outputs of multiple physiological and structural processes, making them inherently slower to rebound. (2) They are structurally and carbon-intensive, requiring sustained resource availability and investment, which recurrent stress likely compromises⁴².

This asymmetry in recovery aligns with earlier reports that structural traits recover more slowly and less completely than physiological traits under climate stress^49,54. Together, these patterns suggest that recurrent drought imposes hidden legacy effects, where apparent resilience in photosynthetic traits masks deeper constraints on clonal performance. Initial bursts of compensatory growth may provide short-term recovery, but ultimately give way to cumulative performance declines, jeopardizing the long-term persistence of clonal grasses under increasing climate variability.

While drought typically reduces SLA by producing smaller, denser, and more conservative leaves^55,56, our drought-only plants showed the opposite pattern, with an increase in SLA, making them more exploitative in resource use. In our case, the increased SLA coincided with a lower relative water content (RWC), indicating a compensatory strategy aimed at maximizing carbon gain with less investment under reduced water availability. However, such a trait combination, thin leaves with low water content, likely increases vulnerability to further stress rather than conferring resilience. This could be an important strategy to avoid drought rather than resilience. The observed increase in root/shoot further implies a strategic shift toward stress avoidance (via greater belowground allocation), yet this comes at the expense of shoot-based productivity and clonal spread, both critical for persistence in competitive grassland environments and highly important for species mainly reproducing through clonal reproduction.

Together, these findings reveal a distinctive form of partial stress memory in F. rubra: physiological traits recover rapidly, yet growth and water relations retain the imprint of past drought. To our knowledge, this asymmetrical memory has not previously been documented in clonal perennial grasses, and it highlights an important, overlooked mechanism that may strongly influence their long-term persistence under increasingly frequent drought events.

MeJA modifies drought–recovery dynamics

Exogenous MeJA fundamentally changed plant responses to drought. During stress phases, MeJA consistently buffered declines in chlorophyll concentration and Fv/Fm, and these protective effects became stronger across cycles. This progressive enhancement is consistent with hormone-induced priming and stress memory^34,57. The novelty here lies in demonstrating such priming in a perennial clonal grass, whereas most previous evidence comes from annuals or model species, such as Arabidopsis^47,58. Our results show that MeJA can establish cumulative resilience across cycles, directly relevant to long-lived species repeatedly exposed to drought.

MeJA also reshaped recovery strategies in ways not observed in drought-only plants. While drought-exposed plants increased SLA, a risky compensatory adjustment that reduced water status, MeJA-treated plants maintained lower SLA, higher LDMC, and RWC comparable to control levels⁵⁹. This indicates that MeJA promotes a conservative, resource-use strategy, reducing vulnerability to subsequent stress. The key novelty lies in demonstrating that MeJA does more than alleviate immediate drought damage: it actively reprograms post-stress recovery, steering plants away from maladaptive compensatory responses toward sustainable structural adjustments. This provides new insight into hormone-mediated regulation of recovery trajectories, particularly in a perennial clonal grass species where such dynamics are largely unexplored.

Biomass allocation patterns reinforced this distinction. Drought-only plants invested heavily into roots, a classic avoidance strategy to enhance water uptake^60,61. MeJA+drought plants, however, prioritized aboveground growth, with higher shoot biomass and lower root-to-shoot ratios. Importantly, ramet production in MeJA-treated plants declined less severely during stress and recovered more slowly but consistently after rewatering. Unlike drought-only plants, which ultimately failed to return to control levels later recovery phase, MeJA+drought plants were able to reach control levels after each cycle. This indicates that MeJA reduced the cost of each drought event and stabilized recovery dynamics, promoting a more sustainable form of resilience compared with the unstable compensatory strategy observed in drought-only plants. Because MeJA was applied through the soil, these effects may also involve altered plant–microbe interactions, a potential mechanism that broadens the scope of jasmonate influence beyond direct plant physiology^46,47.

Together, these findings demonstrate that MeJA produces better outcomes than drought alone by both sustaining photosynthetic efficiency during stress and reprogramming recovery pathways. The novelty here is twofold: first, revealing that MeJA strengthens stress memory in a clonal perennial grass, and second, showing that MeJA fundamentally alters recovery strategies, leading to trait combinations that balance resistance and resilience more effectively than drought-induced responses alone.

Ecological and applied perspectives

Our findings highlight the potential of methyl jasmonate (MeJA) as a powerful modulator of plant responses to recurrent drought. By promoting conservative trait syndromes and stabilizing recovery trajectories, MeJA enhances the resilience of clonal grasses, such as Festuca rubra, which play a crucial role in the structure and function of temperate grasslands.

In the context of climate change, where extreme drought events are becoming more frequent, such hormonal modulation offers a promising avenue for maintaining vegetative persistence, functional stability, and ecosystem services in grassland systems. The success of soil-applied MeJA in our experiment supports its use as a practical and ecologically relevant tool for improving drought tolerance in perennial grasses, with potential applications in restoration ecology, conservation management, and sustainable grassland productivity.

Conclusion

We found that MeJA treatment significantly improved the plant’s ability to withstand and perform better under recurrent drought events. MeJA plays a crucial role in preserving chlorophyll concentration, which is essential for photosynthesis and ultimately leads to enhanced photosynthetic activity. We also observed complex interactions between drought and MeJA in shaping plant performance (ramet numbers, leaf traits, and root-to-shoot ratio) through different phases of recurrent drought. Notably, MeJA-treated plants exposed to drought demonstrated a unique survival strategy distinct from only drought-exposed plants. Our findings highlight the potential of MeJA as a valuable tool in mitigating the adverse effects of recurrent drought on plant performance, offering useful insights for ecosystem management and conservation in the face of changing climate conditions.

Appendix 1

Moisture content

Table S1.

The table contains the pots’ volumetric moisture content from all the treatments determined using TMS loggers.

	Stress Phase 1		Recovery Phase 1		Stress Phase 2		Recovery Phase 2		Stress Phase 3		Recovery Phase 3
	Max	Min	Max	Min	Max	Min	Max	Min	Max	Min	Max	Min
Control	0.551	0.551	0.551	0.551	0.551	0.551	0.551	0.532	0.534	0.526	0.534	0.532
Drought	0.458	0.344	0.551	0.551	0.411	0.192	0.551	0.532	0.238	0.021	0.534	0.532
MeJA	0.551	0.551	0.551	0.551	0.551	0.0551	0.551	0.534	0.534	0.526	0.534	0.534
MeJA+Drought	0.458	0.344	0.551	0.551	0.411	0.192	0.551	0.534	0.238	0.021	0.534	0.534

Author contributions

TB planned and designed the research with ZM’s help, and TB set up the experiment. TB analyzed data and performed all statistical analyses with DT and ZM’s help. TB wrote the paper draft with in-depth suggestions, inputs, and editing from ZM and DT. All the authors approved the final version of the manuscript.

Funding

The study has been supported by the funding given to Zuzana Munzbergova by the Czech Science Foundation (project no. 22–00761 S) and partly by a long-term research development project No. RVO 67985939 of the Czech Academy of Sciences and institutional support for science and research of the Ministry of Education, Youth, and Sports of the Czech Republic.

Data availability

The datasets generated and/or analyzed during the current study will be made available by the corresponding author upon acceptance of the manuscript, through submission to an appropriate public repository such as Zenodo.

Declarations

Competing interests

The authors declare no competing interests.