Interspecific diversity in PSII photoinhibition and photoprotection of benthic diatoms during consecutive light and heat stress

Diatoms inhabiting estuarine intertidal flats are regularly exposed to episodes of high irradiance and elevated temperature during daytime low tides, conditions prone to inducing strong photoinhibition of photosystem II (PSII). However, the extent to which different species cope with repeated exposure to these combined stressors, and how photoacclimation modulates these responses, remains poorly known. This study examined the interspecific diversity in PSII photoinactivation, photorepair and photoprotection of three estuarine pennate diatom species exposed to three consecutive daily periods of combined high light and heat: two benthic species, Navicula phyllepta and Entomoneis paludosa, and one tychoplanktonic species, Cylindrotheca closterium. Unialgal cultures of each species were also acclimated to contrasting low- (LL) and high-light (HL) regimes to further evaluate the influence of photoacclimation on the responses to light and heat stress. In the two benthic species, PSII photoinactivation was largely counterbalanced by efficient photorepair, leading to near-complete recovery after each stress event. In contrast, the tychoplanktonic C. closterium showed markedly higher photoinactivation rates and limited photorepair capacity, resulting in an overall poor recovery. The tested species also differed strongly in the operation of the xanthophyll cycle (XC). Although XC-mediated de-epoxidation of the pigment diadinoxanthin into diatoxanthin increased under stress in all LL-acclimated cultures, it was strongly reduced in the HL-acclimated benthic species. Despite pronounced short-term effects of each stress exposure, no significant accumulation of irreversible PSII damage occurred over the 3 days, likely due to extended recovery intervals that favored an effective PSII photorepair. Overall, the results show that species-specific physiological traits and photoacclimation state play a dominant role in determining the capacity of diatoms to cope with the extreme conditions found in the estuarine intertidal environments.

A distinctive feature of the environment experienced by microalgae inhabiting intertidal sedimentary flats is the high and systematic variability of key abiotic factors across multiple and overlapping time scales (Admiraal, 1984; Serôdio and Catarino, 1999; Serôdio et al., 2008; Laviale et al., 2015). The superimposition of the day-night and tidal cycles drives short-term (hours to days) fluctuations in abiotic conditions which can reach extreme levels, particularly when diurnal low tide periods coincide with midday. The microalgal communities inhabiting these intertidal sedimentary habitats, commonly dominated by diatoms of the group of the pennates and collectively known as microphytobenthos (MPB), are regularly exposed to the combined stresses of high irradiance, extremes temperature and salinity, and resulting osmotic stress (Brotas et al., 2003; Migné et al., 2007; Chevalier et al., 2010; Savelli et al., 2018), as well as localized carbon limitation resulting from high photosynthetic rates (Vieira et al., 2016).

These extreme environmental conditions can negatively impact on the photosynthetic activity and productivity of MPB. Exposure to high solar irradiance, in particular, has the potential to cause photoinhibition, the light-induced reduction in photosynthetic efficiency and productivity. Photoinhibition results from the imbalance between direct photoinactivation of photosystem II (PSII) and the counteracting action of photorepair processes, leading to a net loss of functional PSII and the consequent capacity to capture light, generate ATP and NADPH, and ultimately fix carbon (Tyystjärvi, 2013; Murata and Nishiyama, 2018; Campbell and Serôdio, 2020). The deleterious effects of high light are exacerbated when combined with extreme levels of other abiotic factors (MacIntyre et al., 2002; Nishiyama and Murata, 2014; Zavafer et al., 2015). These extreme conditions can occur on consecutive days in the intertidal habitat, potentially inducing cumulative stress in MPB communities if photodamage suffered during one diurnal emersion period is not fully repaired before the next.

Like other photosynthetic organisms, pennate diatoms mitigate light stress by activating physiological photoprotective mechanisms. The most relevant of these is non-photochemical quenching and its associated xanthophyll cycle (XC), which dissipates excess absorbed light energy as heat, reducing the excitation pressure delivered to PSII and preventing its photoinactivation (Lavaud and Goss, 2014). Concurrently, cells activate the repair of photoinactivated PSII units, through de novo synthesis of the D1 protein (Campbell and Serôdio, 2020).

The photoinhibitory impact of abiotic stress on MPB, and the photoprotective responses activated to deal with it, can be expected to vary with two key factors: (i) species composition: studies carried out on individual diatom species grown in culture have shown that different species may show markedly different behaviors when facing abiotic stress (Barnett et al., 2015; Lavaud et al., 2016; Blommaert et al., 2017), making it difficult to predict the behavior of the natural, multi-species communities; (ii) photoacclimation state: comparative studies on low- and high-light acclimated cultures (Blot et al., 2011) or on natural MPB communities across different seasons (Bártolo et al., 2023) have shown that the phenotypical adjustment of the photosynthetic apparatus to growth light conditions, namely, through changes in the cellular pigment content, can substantially affect the susceptibility to PSII photoinactivation and repair capacity.

Previous studies on the effects of abiotic stress on MPB have primarily focused on single stress events, conducted under laboratory conditions (Laviale et al., 2015; Blommaert et al., 2017) or in situ (Brotas et al., 2003; Serôdio et al., 2008). Furthermore, most experiments carried out in controlled conditions have applied individual abiotic stressors (Savelli et al., 2018), most commonly high light (Serôdio et al., 2012), failing to assess the consequences of repeated exposure to combined stress factors. Additionally, these studies have typically examined single species (Blommaert et al., 2020; Giossi et al., 2025) or whole MPB assemblages (Bártolo et al., 2023), overlooking the likely high interspecific diversity within these communities. These shortcomings call for studies that simultaneously assess the effects of repeated and combined abiotic stressors on multiple, ecologically distinct MPB species, in order to better resolve the interspecific variability of photoinhibitory and photoprotective responses.

This work examines the interspecific diversity in photoinhibition and photoprotection of MPB diatoms subjected to combined, repeated high light and heat stress during consecutive days. Specifically, this study aims to: (i) evaluate the interspecific diversity in the response to combined high light and heat, by comparing the stress responses of three diatom species, differing in preferred habitat: two benthic species, Navicula phyllepta and Entomoneis paludosa, and one tychoplanktonic species, Cylindrotheca closterium; this enables assessing the extent of variability between species in physiological traits and stress responses, and to test whether this variability is linked to habitat preference; (ii) evaluate the cumulative effects of consecutive stress events, by repeatedly subjecting the same cultures to three daily periods of combined high light and heat, simulating successive diurnal low tides, and testing whether consecutive stress induces cumulative net photoinhibition or whether recovery periods, simulating night or submersion, are sufficient to fully recover; (iii) assess the role of photoacclimation in abiotic stress responses, by comparing cultures acclimated to low- and high-light regimes, and testing whether photoacclimation state modulates photoprotection capacity, susceptibility to PSII photoinactivation, and the efficiency of PSII photorepair. The effects of light and heat stress were quantified by measuring the rate constants of PSII photoinactivation and photorepair, and the capacity of photoprotection, through the operation of the xanthophyll cycle. This aimed to quantify the effects of abiotic stressors on direct PSII photoinactivation and on inhibiting photorepair, the relative contribution of photoprotection and photorepair in mitigating PSII photoinactivation.

Diatom species

Experiments were carried out on three species of estuarine pennate diatoms representative of benthic and tychoplanktonic forms: N. phyllepta Kützing and E. paludosa (W. Smith) Reimer, and C. closterium (Ehrenberg) Reimann & Lewin. The three species were isolated from intertidal mudflats in the Ria de Aveiro, Portugal. N. phyllepta and E. paludosa were collected in the Canal de Ílhavo (Vista Alegre; 40°37′12″N, 08°44′54″W) and C. closterium was collected in the Esteiro de São Pedro (40°37′40″N, 8°39′32″W). These species were selected based on their known different sensitivity to light stress and habitat preference (Lavaud and Lepetit, 2013; Barnett et al., 2015). The diatom species were identified based on morphological traits and, for E. paludosa and C. closterium, confirmed by molecular analysis (see below).

The three diatoms were grown in semi-continuous batch cultures in an incubator (Versatile Environmental Test Chamber MLR-352, Panasonic, Japan) at 18 °C under a 12 h light:12 h dark cycle. Cultures were grown under 120 µmol photons m -2 s -1 (high light, HL) and 10 µmol photons m -2 s -1 (low light, LL) for several weeks to induce HL- and LL-acclimated contrasting photophenotypes, respectively. Incident light intensity was measured using a spectroradiometer and quantum light meter (SpectraPen Mini, Photon Systems Instruments, Czech Republic). Cultures were grown in 250 mL Erlenmeyer flasks with 150 mL natural, autoclaved and GF/F 1.2 µm (Whatman International, UK) filtered seawater with silica-enriched f/2 medium. Replicated cultures were grown in separate flaks prior to the start of the experiments.

Photoacclimation state

The photoacclimation state of the LL- and HL-grown cultures was characterized by measuring light-response curves (LC) of the relative electron transport rate, rETR, given by Genty et al. (1989): where E is the incident irradiance, F m ʹ and F are the maximum and steady-state fluorescence levels measured under actinic light, using a PAM fluorometer (see below) (Table 1). LCs were generated by sequentially applying eight increasing levels of actinic light, each lasting 120 s (14, 69, 162, 326, 468, 786, 1385, 2191 µmol photons m -2 s -1). Samples were dark acclimated for 15 min before the start of the LCs. LCs were characterized by fitting the model of Eilers and Peeters (1988) to individual curves using a procedure written in Microsoft Visual Basic and based on Microsoft Excel Solver (Frankenbach et al., 2018), by estimating the initial slope, α, the maximum rETR level, rETR m, and the photoacclimation parameter, E k. The photoinhibition parameter of the model was not used due to the difficulty inherent to its interpretation and because direct information on photoinhibition was measured by k PI and k REC (see below). Three replicated LCs were measured for each culture.

Combined light and heat stress

The combined effects of light and heat stress were evaluated on the LL- and HL-acclimated photophenotypes of each species. The response of samples exposed to periods of high light intensity and elevated temperature (650 µmol photons m -2 s -1 and 35 °C, for 90 min), during three consecutive days, was compared to replicated culture flasks kept under growth light intensity and temperature conditions (Figure 1). These conditions are within the range of conditions occurring during diurnal low tides in the intertidal mudflats from which the diatoms used in this study were isolated from Barnett et al. (2015). Control samples were kept in the incubator under growth light and temperature conditions during the total duration of the experiment. Samples to be subjected to stress were removed from the incubator only during the period of exposure to stress conditions and were returned immediately afterwards. These samples were transferred to the hotter conditions, and only after 10 min necessary to reach the desired temperature of 35 °C the high light was turned on. Thirty minutes before the start of stress exposure, a subsample (1.5 mL) was taken from each culture flask (control and stress) and transferred to a spectrophotometer cuvette (6030 Optical Glass, Hellma, Germany). For the cultures to be exposed to stress, a second subsample was transferred to a similar cuvette to be treated with the PSII photorepair inhibitor lincomycin. Lincomycin is widely used for inhibiting the de novo synthesis of the chloroplast-encoded PSII protein D1 and thus photorepair of photoinactivated PSII. For relatively short incubation periods and moderate concentrations, as the ones used in this study, lincomycin does not have relevant side effects and, due to the uniquely high turnover rate of the D1 protein, it can be assumed to only affect PSII repair (Campbell and Tyystjärvi, 2012; Bachmann et al., 2004). Cuvettes were used to facilitate the direct measurement of photophysiological parameters in the PAM fluorometer and to allow treatment with the PSII photorepair inhibitor lincomycin without affecting the whole culture in each flask (see below). Samples in cuvettes were exposed to the same irradiance and temperature conditions as the culture flasks they were transferred from. This was repeated on each of the three consecutive days of the experiment, thus allowing the samples in the culture flasks to be cumulatively exposed to the high light and heat conditions (Figure 1). All experiments were carried out using cultures in the beginning of the exponential growth phase.

The maximum quantum yield of PSII, F v /F m, was measured immediately before the start (following 15 min dark acclimation; initial time, T i) and every 3 min during the 15 min after the end of the stress period, in the dark (recovery time, T r), in all samples. For calculation of PSII inactivation and photorepair rates, the values measured after 15 min in the dark were considered (see below). Additional subsamples (10 mL) were collected at the start (T i) and end of the stress period (final time, T f), from both control and stress (not treated with lincomycin) samples, for quantification of photosynthetic pigments and cell concentration (see below). The applied light intensity and duration of stress exposure were determined from preliminary tests so that in lincomycin-treated samples F v /F m > 0, to allow the reliable determination of rate constants of PSII photoinactivation and photorepair (see below).

Experimental setup

The combined light and heat stress was applied by immersing the culture flasks and the cuvettes in a shallow water bath inside an acrylic recipient (32 × 24.5 × 8.5 cm) with a transparent bottom (Figure 1). The water temperature was maintained by circulating it through an external thermostatic bath (Clifton, Nickel-Electro Ltd, UK) and periodically monitored. The flasks were kept from floating by using vinyl-coated lead rings (SP Bel-Art, USA) and the cuvettes were secured in place using a custom-designed and 3D-printed holder. Culture flasks and cuvettes were illuminated from below, by a flexible rectangular RGB LED panel comprising 8 × 32 5050SMD LEDs (BTF Lighting, China) placed immediately below the bottom of the recipient and facing upwards. The LED panel covered an area long and wide enough to provide homogenous illumination throughout the area where the flasks and cuvettes were placed. The LED panel was slightly curved longitudinally (fixed in position using a custom-designed 3D-printed holder), to focus light on the area of the water bath where the samples were positioned, thus maximizing incident light intensity. The LEDs were powered by 5V-30A-150W power source and controlled by an Arduino Uno R3 microcontroller. Heating from the LEDs was effectively prevented by using three 5 V fans. The LEDs providing a mixture of red (630 nm), green (518 nm) and blue (460 nm) light, corresponding to an incident PAR of 650 µmol photons m -2 s -1, as measured with a calibrated submersible spherical US-SQS/L micro-quantum sensor (Heinz Walz GmbH, Germany), were positioned at mid-height of a cuvette filled with growth medium. The combined emission spectrum of the three types of LEDs was measured using a SpectraPen Mini spectroradiometer and found to be identical to the spectrum presented in (Serôdio et al., 2018).

PSII photoinactivation and photorepair

The susceptibility to PSII photoinactivation was evaluated essentially as described in (Bártolo et al., 2023). The rate constant of PSII inactivation, k PI (s -1) was determined by measuring the relative decrease between post- and pre-stress F v /F m values (at T r and T r, respectively): in lincomycin-treated samples, following (Campbell and Tyystjärvi, 2012):

The PSII photorepair capacity was evaluated by determining the rate constant of PSII photorepair k REC (s -1) following (Ragni et al., 2008), based on the light-induced relative decrease of F v /F m in untreated samples:

Lincomycin (lincomycin hydrochloride; Alfa Aesar, Germany) was added at a final concentration of 2 mM, in the dark, 30 min before the start of the stress period. It was prepared from a stock solution with pH of 6.8 ± 0.2 using a 0.14 M NaOH solution (Sigma-Aldrich, USA). This concentration was determined by preliminary tests, showing that it enabled attaining maximum inhibition. All samples were supplemented with NaHCO 3 (4 mM final concentration) 30 min before the start of the light stress exposure, to avoid carbon limitation of photosynthesis during the light stress experiments.

Chlorophyll fluorescence

All in vivo chlorophyll (Chl) fluorescence parameters were measured using a Multi-Color PAM fluorometer (Walz) with an MCP-D detector unit fitted with an RG 665 long pass filter (>650 nm, 3 mm RG665, Schott AG, Germany) and controlled by the PamWin V3.12w software (Walz). White light was used for actinic light and saturating light pulses (800 m), and blue light (440 nm) was used for measuring light. Measurements were carried out in the cuvettes where samples were exposed to control and stress conditions, using the ED-101US/MD optical unit, coupled to a magnetic stirrer (PHYTO-MS Miniature Magnetic Stirrer, Walz).

Pigment extraction and quantification

10 mL of homogenized culture from each culture flask were harvested by vacuum filtration through a 1.2 µm pore size, 25 mm diameter polycarbonate filter (Isopore 0.2 μm PC Membrane, Merck Millipore KGaA, Germany). The filters were immediately flash-frozen in liquid nitrogen and transferred to 2 mL microtubes for storage at −80 °C until pigment extraction. Samples were extracted in cold-buffered 95% methanol (2% ammonium acetate) containing 1 mg L -1 trans-β-Apo-8′-carotenal (Sigma-Aldrich) as an internal standard. The samples were vortexed for 30 s and incubated at −20 °C for 15 min. Following incubation, the samples were filtered into 2 mL HPLC dark glass vials through a 0.22 µm pore cellulose acetate syringe filter (Whatman International). Throughout the entire extraction process, both the samples and the solvent were kept in the dark and on ice to prevent pigment degradation. Extracted samples were stored at −80 °C for no longer than 24 h prior to HPLC analysis.

Samples were analyzed by HPLC (Shimadzu, Columbia, MD, USA) equipped with a photodiode array detector. Pigment signals were detected at 430 nm (for Chl a) and at 450 nm (for diadinoxanthin and diatoxanthin). The solvent gradient was prepared with an injection volume of 100 μL at a flow rate of 1 mL min −1. The stationary phase was a C18 Kromasil 100 column (particle size of 3.5 μm; diameter, 4.6 mm; length, 250 mm; Kromasil, Sweden) coupled to a C18 guard column Kromasil 100 (particle size of 3.5 μm; 4.6 mm diameter; Kromasil). The column temperature was set to 20 °C. HPLC solvents were 80% methanol:20% ammonium acetate, 1 mol L -1 (v:v) (A), and 60% methanol:40% acetone (v:v) (B). All solvents were HPLC grade and were filtered through a 0.22 µm pore PTFE membrane. Separation was achieved with a linear gradient: 0 min: 20% B; 3 min: 20% B; 23 min: 100% B; 43 min: 100% B; 58 min: 20% B; 65 min: 20% B. The HPLC was calibrated with crystalline Chl a (0.0175–0.175 µg), DT (0.00429–0.06435 µg) and DD (0.00429–006,435 µg) standards from DHI (Hørsohn, Denmark).

The de-epoxidation state was calculated as: where DD and DT are the molar concentrations of the XC pigments diadinoxanthin and diatoxanthin, respectively. When there was newly synthesized DT after the light stress exposure, the newly formed DES was calculated as: where ΔDT is the increase of the DT concentration between before (‘constitutive’ DT) and after the light stress exposure (Croteau et al., 2021).

Cell concentration

Cultures were homogenized by repeated pipetting prior to sampling for counting. Subsamples (1 mL) were fixed in Lugol solution (final concentration of 0.1% Iodine; AppliChem GmbH, Germany) and stored at 4 °C. Cell counts were carried out using the improved Neubauer counting chamber (Marienfeld GmbH and Co.KG, Germany).

Species identification

Species identification was based on morphological observations using a light microscope (BS.1153-PLPHi bScope, Euromex, Netherlands) on permanent microscopy slides prepared as described by Bártolo et al. (2023). Identification was based on the morphological features of the valves, including its shape, length and width but also raphe presence, length and shape, patterns within the striae, and ornamentation features of the frustule (Round et al., 1990; Ribeiro, 2010; Coste and Rosebery, 2011).

Species were also identified through sequencing of the rbcL gene. Cells were harvested from actively growing culture by centrifugation (10 min, 4000 rpm, Universal 16A, Hettich Zentrifugen, Germany) and genomic DNA was extracted using the DNA extraction kit NucleoSpin® Plant II (Macherey-Nagel, Germany). To amplify the rbcL gene two primer pairs mixes were used (Abarca et al., 2014) (Supplementary Table S1). Polymerase chain reactions were performed in a final volume of 25 μL reaction mixtures containing 15.75 μL of PCR grade water (NZYTech Lda, Portugal), 6.25 μL of NZYTaq II 2x Green Master Mix (NZYTech), 1 μL of each primer (0.1 μM) and 1 μL of the DNA template. Negative control was made using PCR grade water instead of DNA template. PCRs were performed in a thermocycler (C1000 TouchTM Thermal Cycler, Bio-Rad, USA) with the following conditions: an initial denaturation step for 5 min at 95 °C, followed by 33 amplification cycles of 1 min at 95 °C, 1 min at 54 °C and 1 min at 72 °C, and a final elongation for 5 min at 72 °C. PCR products were analyzed by 1.5% agarose gel electrophoresis, purified using the NZYGelpure kit (NZYTech) and sequenced by StabVida (Portugal) using Sanger sequencing. The nucleotide sequences were analyzed with Finch TV 1.4.0 (Gepspiza Inc). Forward and Reverse sequences were aligned using Clustal Omega (EMBL-EBI, UK; Sievers et al., 2011). The resulting consensus sequences were compared against the core nucleotide BLAST database (core_nt) (Altschul et al., 1990; Ye et al., 2006) to identify the closest matching sequences. Sequences were deposited in the Nucleotide database (NCBI), with accession numbers: PZ221692, PZ221693 and PZ221694.

Statistical analysis

Measurements for comparing photoacclimation states (LC parameters and Chl a, DD and DT cellular content) and for comparing untreated and lincomycin-treated samples were compared by applying the Student’s t-test. Measurements made on different species, photoacclimation state, and days were compared by applying three-way Analysis of Variance (ANOVA). The effect size of experimental factors (species, photoacclimation state and day of experiment) was estimated by partial Eta squared (η 2) associated to the main effects of a 3-way ANOVA. Assumptions of normality and homoscedasticity were verified prior to analysis using the Shapiro–Wilk test and Levene’s test, respectively. All statistical analyses were carried out using SPSS Statistics (IBM, USA).

Photoacclimation state

The applied growth light conditions were sufficiently distinct to induce marked and contrasting LL- and HL-acclimation states for the three species, as confirmed by their light-response curves of rETR (Equation 1 α, rETR m and E k), Chl a content per cell, and relative content of XC pigments (DD and DT) per cell (Figure 2). For all species, α was lower under HL than LL growth conditions, whereas rETR m and E k were higher (Figures 2A–C). However, in the case of E. paludosa, the differences in LC parameters between LL- and HL-acclimated cultures were not significant due to the large variability between replicates (t -test, p > 0.05). In all other cases, differences were significant (t -test, p < 0.05). Chl a per cell was higher in LL-than in HL-grown cultures of the benthic species N. phyllepta and E. paludosa (t -test, p < 0.001 in both cases) but not for C. closterium (t -test, p = 0.317) (Figure 2D).

In all species, both DD and DT content per Chl a was higher in HL-than in LL-grown cultures (t -test, p < 0.02 in all cases). Again, smaller differences between LL- and HL-acclimated cultures were observed in the case of C. closterium. Regarding the DT pool, the variation between LL- and HL-grown cultures was also markedly larger in the case of N. phyllepta and E. paludosa, showing increases of 11.9 and 3.0 times, respectively, compared to C. closterium, for which DT increased by only 1.6-fold.

Effects of combined light and heat stress

During the first day of exposure, the combined light and heat stress event induced clear species-specific and photoacclimation-dependent changes in F v /F m, with lincomycin treatment exacerbating the decline in PSII quantum yield (Figure 3). In all cases, the cultures started in good physiological condition, as denoted by F v /F m values above 0.65. No substantial differences were observed between lincomycin-treated and untreated samples prior to stress exposure. Following stress exposure, the untreated N. phyllepta and E. paludosa recovered rapidly to pre-stress levels, reaching F v /F m values also similar to the ones measured in the controls (samples not exposed to stress), both in LL- and HL-acclimated cultures (Figures 3A,B,D,E). By contrast, C. closterium failed to recover from the applied stress, its F v /F m remaining at levels substantially below those of the controls (Figures 3C,F). The treatment with lincomycin resulted in large negative effects on the capacity to recover from stress exposure, with F v /F m values remaining below control levels, from ca. 0.1 (N. phyllepta HL) to 0.5 (E. paludosa LL). Cylindrotheca closterium showed the poorest results, showing F v /F m remaining at or below and below 0.1 without recovery.

Considering the data expressed as changes relative to the pre-stress state through the 3-day experiment (%F v /F m; Equation 2; Figure 4), the most pronounced aspect was again the marked difference between the two benthic pennate species, N. phyllepta and E. paludosa, and the tychoplanktonic species C. closterium. In the case of the untreated samples, the benthic species showed a strong recovery capacity, reaching a minimum of 89.6% and 99.5% of the controls for LL- and HL-acclimated samples, respectively (Figures 4A,B,D,E). The HL-acclimated cultures recovered marginally better, and in the case of N. phyllepta, samples exposed to stress even exceeded slightly the controls (Figure 4A). In contrast, C. closterium showed a much weaker recovery, %F v /F m remaining below 27.2% of the controls in all but one case (HL, day 2) (Figures 4C,F).

As for N. phyllepta and E. paludosa, HL-acclimated samples of C. closterium outperformed their LL-acclimated counterparts, the former ranging between 24.2% and 52.7%, and the latter averaged only 17.3% over the 3 days (Figures 4A,B,B,E). Considering the lincomycin-treated samples, much lower recovery rates were measured, typically below 35% of the controls. Entomoneis paludosa was an exception, showing a post-stress %F v /F m averaging 66.8% (Figures 4B,E). Nevertheless, highly significant differences between lincomycin-treated and untreated samples were found in all cases (t -test, p < 0.001). Importantly, the photoacclimation state also affected the response to combined light and heat stress, with lincomycin-treated HL-acclimated samples showing lower recoveries than the LL-acclimated ones, for all tested species and days of the experiment (with the exception of C. closterium on day 1). Although the %F v /F m of the controls and untreated samples were relatively unchanged over the 3 days of the experiment (ANOVA, p > 0.052 in all cases) (exception: E. paludosa (p = 0.003) and C. closterium (p = 0.019) stress, untreated; Figures 4B,C), the values measured in lincomycin-treated samples often showed significant variations (ANOVA, p < 0.05 in all cases except for N. phyllepta (LL), for which p = 0.092). In the case of N. phyllepta (HL) and E. paludosa (LL), %F v /F m decreased consistently over time (Figures 4A,E).

Photoprotection vs. photorepair

The relative contribution of enhancing PSII photoprotection (in the sense of mitigating photoinactivation) and activating photorepair and photoacclimation states is shown in Figure 5. The contribution of photoprotection and photorepair was inferred from the %F v /F m measured in lincomycin-treated samples (reflecting the extent to which photoinactivation was prevented) and from the difference between the %F v /F m measured in untreated and lincomycin-treated samples, respectively. The results revealed substantial interspecific variation in the strategies used to cope with the light and heat stress, resulting from distinct balances between photoprotection and photorepair mechanisms. Navicula phyllepta achieved a high overall recovery mainly through efficient photorepair of damaged PSII, rather than by preventing photoinactivation (Figures 5A,D). The low %F v /F m values observed in lincomycin-treated samples are indicative of a high susceptibility to PSII photoinactivation, associated with a weak photoprotective capacity. On the other hand, the large difference between the %F v /F m measured in untreated and lincomycin-treated samples is indicative of a high photorepair capacity. This pattern was observed in both LL- and HL-acclimated samples throughout the 3 days of the experiment, but the role of photorepair was more pronounced in HL-acclimated cultures. In the case of HL-acclimated cultures, the relative importance of photorepair increased gradually with time.

In contrast, E. paludosa achieved near-complete recovery based both on a more pronounced photoprotection combined with a good repair capacity (Figures 5B,E). This is supported by the higher %F v /F m observed in lincomycin-treated samples, indicating a lower photoinactivation capacity. The relative contribution of photoprotection and repair varied markedly between LL- and HL-acclimated cultures. In the latter, photorepair remained as the dominant process, with moderate susceptibility to photoinactivation. In contrast, LL-acclimated samples exhibited low photoinactivation, suggesting that photoprotection is the primary mechanism in operation. However, this protective effect appeared to decline over time.

In the case of C. closterium, the balance between photoprotection and photorepair is less clear, varying with growth light conditions and photoacclimation state (Figures 5C,F). In LL-acclimated cultures, photoprotection dominated, associated with low photoinactivation and limited repair activity. In HL-acclimated cultures, the results were more variable (with an anomalous value observed on day 2) but overall showing that the contribution of photoprotection was not as pronounced as in LL cultures, and that photorepair remained relatively weak.

Photoinactivation vs. photorepair

Navicula phyllepta and E. paludosa showed very similar values of the rate constants of photoinactivation and photorepair under LL and HL growth light conditions, although much higher values were measured for the former than the latter species (Figures 6A,B,D,E). In the case of the tychoplanktonic C. closterium, the rates of photoinactivation were always higher than the corresponding rates of photorepair, a result consistent with the overall more limited capacity to recover from the applied light and heat stress: with the exception of the HL samples on day 2, k PI exceeded k REC by more than 2.7 and 4.0 times for HL- and LL-acclimated cultures (Figures 6C,F). In all three species, both k PI and k REC were consistently higher in HL-than in LL-acclimated cultures. Large differences were found for N. phyllepta and E. paludosa, for which k PI and k REC of HL-acclimated samples surpassed by 2.2–3.2 times the values of LL-acclimated ones. In most cases, k PI (Equation 3) and k REC Equation 4 changed significantly over the days of the experiment (ANOVA, p < 0.03 in all cases). Only in the case of N. phyllepta (HL) (Figure 6A) and C. closterium (LL) (Figure 6F) a consistent and parallel variation of the two rates was observed. In the first case, revealing a gradual increase in photoinactivation due to cumulative abiotic stress, but compensated by increased photorepair capacity; in the second, showing that the successive exposure to stress caused an improvement in photoprotection that reduced photoinactivation, but also a decrease in photorepair.

Xanthophyll cycle and photoprotection

Also when considering the results of XC pigment de-epoxidation (DES; Equation 5) a very distinct response was observed between the benthic (N. phyllepta and E. paludosa), and the tychoplanktonic species (C. closterium). In LL-acclimate cultures, exposure to light and heat resulted in a substantial increase in DES denoting the activation of the XC and the conversion of pigment DD into DT (t -test, p < 0.001 in all cases; Figures 7D–F). DES reached moderate values for N. phyllepta and E. paludosa, with averages of 33.9% and 23.7% (Figures 7D,E), respectively. Much higher values were observed in the case of C. closterium, where DES reached above average values 63.4% (Figure 7F). However, in the case of HL-acclimated cultures, markedly different responses were found for the benthic species N. phyllepta and E. paludosa. While C. closterium continued to show a large light-induced DD de-epoxidation, although not as pronounced as in the LL-acclimated samples (DES averaging 51.3%; t -test, p < 0.001; Figure 7C), both N. phyllepta and E. paludosa did not show an increase in DES (t -test, p > 0.109; Figures 7A,B). In all cases for which the XC was activated during the light and heat stress (all LL-acclimated cultures and C. closterium HL), DES decreased gradually and significantly over the 3 days of the experiment (ANOVA, p ≤ 0.001 in all cases). The decrease between day 1 and 3 ranged from 43.3% in E. paludosa (LL) to 47.1% in C. closterium (LL). Of notice is the fact that in all cases, but especially in HL-acclimated N. phyllepta cultures, a significant amount of DT was present already before the exposure to high light (constitutive DT) (Table 2).

Further analysis of the dataset, for N. phyllepta and E. paludosa, revealed an inverse relationship between the high light-activated operation of the XC, as measured by ΔDES (Equation 6), and the suffered photoinactivation, as measured by k PI (Figures 8A,B). Samples, either LL- or HL-acclimated, in which the DD de-epoxidation was limited under stress conditions showed substantially higher photoinactivation rates. The inverse situation was found in samples showing high ΔDES, for which k PI remained at low levels, resulting in a highly significant correlation between the two indices when data from both LL- and HL-acclimated cultures are considered (p = 0.009 and p < 0.001 for N. phyllepta and E. paludosa, respectively). A similar relationship was not detected in C. closterium, for which both DES and k PI reached high values (Figure 8C).

Relative effects of species, photoacclimation and cumulative stress

The quantification of effect sizes (η 2) for diatom species, photoacclimation state (HL vs. LL) and days of experiment on k PI, k REC, and DES (at T f) revealed moderate to strong effects in all cases, with a clear hierarchy among factors: diatom species was associated to the largest variability in k PI and DES (η 2 = 0.902 and 0.981, respectively), followed by photoacclimation state (η 2 = 0.798 and 0.916) and by duration of stress exposure (η 2 = 0.422 and 0.854). This pattern was only slightly different in the case of k REC, where photoacclimation state appeared as the most impactful factor, closely followed by species, while experimental duration remained the least important factor.

Interspecific diversity in the response to high irradiance and heat

The primary finding of this study is the substantial variation observed in response to light and heat stress between the tested diatom species. Overall, the responses of the two benthic species were quite similar to each other and markedly different from those of the tychoplanktonic species. These interspecific differences were, however, modulated by a complex interplay between photoacclimation state and day of experiment.

The two benthic species, N. phyllepta and E. paludosa, shared a high capacity to recover from the applied light and heat stress. This large recovery capacity is likely associated with their shared trait of having equivalent constant rates of PSII photoinactivation (k PI) and photorepair (k REC), indicating that photorepair capacity was, for each species, sufficient to offset the suffered photoinactivation. Interestingly, the balance of these two processes differed between the two species: N. phyllepta relied more on highly efficient photorepair, while E. paludosa appeared to depend on a combination of photoprotection and photorepair. However, E. paludosa can increase its photosynthetic efficiency under high irradiance (higher rETR m and E k) more than N. phyllepta, which may contribute to decrease excitation pressure and D1 photoinactivation. In contrast, the tychoplanktonic species C. closterium showed a limited recovery capacity, consistent with its high photoinactivation rates relative to its counteracting photorepair rates.

The susceptibility to photoinactivation determined for the studied species is consistent with values reported in the literature on MPB assemblages from the source sites of the cultures used in this work. Considering the quantum yield of PSII photoinactivation, a parameter that is independent from the applied stress irradiance, and thus facilitating cross-study comparisons (Φ PI = k PI /E, where E is the stress irradiance (Campbell and Tyystjärvi, 2012)), the values measured for the first day of experiments ranged, on average across species, from 4.64 to 5.22 × 10 −7 m 2 μmol photons -1 (LL and HL-acclimated, respectively), while published values averaged 2.10 (Bártolo et al., 2023) and 3.25 × 10 −7 m 2 μmol photons -1 (Frankenbach et al., 2018). The measured values also align with those compiled for diatoms and microalgal suspensions: 3.72 and 4.06 × 10 −7 m 2 μmol photons -1, respectively (Serôdio and Campbell, 2021).

Another marked difference between the benthic and the tychoplanktonic species concerns the light-induced activation of the xanthophyll cycle, as quantified by the increase in the de-epoxidation state (DES). In all cases, C. closterium was capable of achieving higher conversion rates of DD to DT than N. phyllepta and E. paludosa. Similar results were reported by Barnett et al. (2015), who showed that C. closterium responded to high light stress with a much more pronounced DES increase compared to epipelic forms, including the two species used in this study. Nonetheless, this enhanced pigment-based photoprotection capacity was insufficient to prevent proportionally higher net photoinhibition or a deficient capacity to recover from the applied stress. These results support that the operation of the XC needs to be complemented by efficient PSII photorepair to successfully mitigate the photoinhibitory effects of abiotic stress.

The markedly different strategies of the benthic and tychoplanktonic species might be associated with their preferred habitats: the sedimentary environment and the alternation between the water column and the sediment, respectively. In the case of the benthic forms, cells remain in the upper layers of the sediment for prolonged, continuous periods during low tides, where they are regularly exposed to high solar irradiance and elevated temperatures (Serôdio et al., 2008; Chevalier et al., 2010). It thus appears highly advantageous for these species to possess more efficient photoprotection and repair mechanisms to cope with these particularly stressful environmental conditions. In contrast, tychoplanktonic cells frequently alternate between the extreme conditions of emersed sediments and periods in the typically turbid estuarine water column (Lavaud and Lepetit, 2013; Barnett et al., 2015), where light is attenuated and temperature does not reach the extreme levels of the exposed sediments. Tychoplanktonic species would thus not benefit as much from having highly efficient photoprotection and repair mechanisms, as it may be more favorable to allocate resources to cope with low light conditions rather than dealing with extreme high light stress.

Role of photoacclimation

Despite being well marked, the differences found between species were strongly modulated by the photoacclimation state of the cultures. The applied growth light regimes induced the formation of distinct photophenotypes, as deduced from LC parameters, XC pigment and Chl a cellular content. HL-grown cultures used higher irradiance levels more efficiently, showing higher light-saturated photosynthetic activity (rETR m and E k) and larger cellular DD and DT pools; LL-grown cultures optimized light capture capacity, as denoted by higher α and higher Chl a content per cell. As before, C. closterium differed from the other species in not showing pronounced differences in cellular Chl a content between LL- and HL-grown cultures. This may indicate a lower phenotypical plasticity in this species, related to the less extreme and variable abiotic conditions typically experienced by tychoplanktonic cells.

Changes in photoacclimation state comprise adjustments to the pigment composition, including the XC pigments associated with photoprotective energy dissipation, and the ultrastructure of the photosynthetic apparatus to prevailing growth light conditions (Falkowski and LaRoche, 1991). These phenotypic adjustments are known to modify the susceptibility to PSII photoinhibition, either by affecting direct PSII photoinactivation and/or by influencing PSII photorepair mechanisms (Han, 2002; MacIntyre et al., 2002; Blot et al., 2011). The results of this study show that the photoacclimation state affected the response to light and heat stress in multiple ways. First, it modified the relative contribution of photoprotection and photorepair of PSII. HL-acclimated cultures relied more on enhanced photorepair than on reducing photoinactivation. This might be associated with the high energy cost of the PSII D1 protein repair cycle, which is facilitated by the higher ATP production possible under high-light growth conditions (Murata and Nishiyama, 2018). Also, HL-acclimated cells are likely already investing more in D1 repair due to the higher photodamage they systematically experience under constant high light growth conditions. HL-acclimation also resulted in a marked reduction in Chl a content per cell, a typical acclimatory response in algae and plants (Han, 2002; MacIntyre et al., 2002; Lepetit et al., 2012). This reduction lowers the capture of incident light and the delivery of absorbed excitation energy to PSII, thereby contributing to minimizing direct PSII inactivation (Lavaud et al., 2002).

These mechanisms likely explain the overall higher %F v /F m recovery observed for HL-acclimated cultures, in comparison to LL-acclimated ones, as reported for plants and cyanobacteria (Blot et al., 2011; Keren and Krieger-Liszkay, 2011). In the present study, %F v /F m reached values exceeding one in some cases, an effect that can be attributed to non-photochemical quenching formed in the dark, prior to high light exposure; this phenomenon is not observed in plants or green algae but frequently reported for benthic pennate diatoms (Serôdio et al., 2005; Bailleul et al., 2015; Blommaert et al., 2021).

Another way in which photoacclimation affected the response to light and heat stress was by increasing both k PI and k REC in HL-acclimated cultures. Similar results have been reported by Ragni et al. (2008) who showed comparable values for both rate constants and a parallel increase following transfer from low to high light growth conditions. Several other studies have also shown that HL-acclimated cultures became more susceptible to direct PSII photoinactivation than LL-acclimated ones, which can be offset by an upregulation of counteracting photorepair mechanisms (Baroli and Melis, 1996; Jeans et al., 2013; Serôdio et al., 2017).

A more pronounced difference between the two photophenotypes referred to the cellular pool of XC pigments DD and DT, and their interconversion upon stress exposure. HL-acclimated cultures presented a much larger cellular DD and DT content relative to their LL-acclimated counterparts, as commonly observed in diatoms and other photoautotrophs (Schumann et al., 2007; Lepetit et al., 2012; Barnett et al., 2015; Lacour et al., 2018; 2020; Pniewski and Piasecka-Jędrzejak, 2020). In particular, the substantial amount of constitutive DT is not only common in HL-acclimated diatoms grown in culture but is observed in natural MPB communities acclimated to moderate light intensities (Jesus et al., 2009; Chevalier et al., 2010), explaining at least partially the large values of DES that were measured.

The combined exposure to high irradiance and elevated temperature is well known to promote a marked increase in DES in diatoms (Laviale et al., 2015; Giossi et al., 2025). However, this type of response was not observed in the case of the HL-acclimated cultures of the benthic forms N. phyllepta and E. paludosa, that showed no significant ΔDES. Similar results have been observed previously. A study of natural benthic diatom-dominated samples (Pniewski and Sylwestrzak, 2018), likely representing communities acclimated to high ambient light, reported a DES increase under light stress, but also observed a significant reduction in this increase when high light was combined with heat exposure. Furthermore, Phaeodactylum tricornutum cultures exhibited only marginal increases in DT accumulation and DES upon exposure to high irradiance and high temperature, interpreted as resulting from a limitation of enzymatic de-epoxidation kinetics (Giossi et al., 2025).

Several factors may cumulatively explain the limited ΔDES observed for HL-acclimated N. phyllepta and E. paludosa. First, HL-grown cells already possessed large constitutive pools of DD, DT, and total XC pigments, as well as comparatively high basal DES values before high light exposure. The amounts of pre-existing DT may have been sufficient to cope with the applied stress, reducing the physiological need to further convert DD into DT. Also, the applied irradiance (650 µmol photons m -2 s -1), may not have been stressful enough for HL-acclimated cultures, which had already adjusted their photosynthetic apparatus to high light during growth. As stated before, this irradiance level was set based on preliminary tests so that k PI and k REC could be reliably quantified. Second, not all DD available in HL-acclimated cells might be accessible for de-epoxidation. Acclimation to high-light conditions may alter the portion of the DD pool that is structurally constrained, not readily accessible to the DD de-epoxidase enzyme (Schumann et al., 2007; Kuczynska et al., 2020). Third, high temperatures accelerate electron transport rates and Calvin cycle enzyme kinetics, contributing to lower the ΔpH across the thylakoid membrane and suppressing the activation of the DD de-epoxidase, thereby reducing DES. Moreover, elevated temperatures increase thylakoid membrane fluidity and permeability, promoting proton leakage and further diminishing the ΔpH-dependent activation of the DD de-epoxidase leakage (Goss and Latowski, 2020). Finally, changes in lipid composition of the thylakoid membrane induced by HL-acclimation might have also played a role. HL-grown diatoms typically increase the proportion of negatively charged lipid sulfoquinovosyldiacylglycerol (SQDG) at the expense of monogalactosyldiacylglycerol (MGDG) (Lepetit et al., 2012). Because the latter lipid stimulates the activity of the DD de-epoxidase while SQDG inhibits it (Goss et al., 2009), the shift toward SQDG induced by HL-acclimation may constrain the DD de-epoxidation capacity. These combined effects might thus explain the observed reduced DT formation under combined high light and high temperature in HL-acclimated cultures.

The limited XC activity in the benthic species N. phyllepta and E. paludosa provided an opportunity to evaluate the photoprotective role of the XC: cells unable to produce the photoprotective pigment DT suffered the highest rates of PSII photoinactivation, supporting the conclusion that the XC effectively protects PSII from abiotic stress. In the case of C. closterium, the efficient DD de-epoxidation in both LL- and HL-acclimated cultures prevented the large variation in photoprotection conferred by the XC recorded for the other two species. Both DES and k PI reached high values, indicating an intrinsic high susceptibility to photoinactivation, against which the operation of the XC does not seem to be effective. An explanation for this result is that, in this species, heat stress may act primarily by affecting the repair of photoinactivated PSII rather than by directly increasing photoinactivation. In this case, the efficient DD de-epoxidation, while promoting the reduction of excitation pressure, would but not be effective if avoiding net photoinhibition. Similar results were reported by Bártolo et al. (2023), who found a negative linear relationship between photoprotection capacity (as assessed by nonphotochemical quenching of Chl fluorescence) and PSII photoinactivation in MPB samples exposed to light and heat stress.

Cumulative effects of consecutive stress events

Unlike most studies, which address abiotic stress responses using a single exposure period, this work tested the cumulative effects of acute light and heat stress events. In general, the results revealed only minor or inconsistent variations over the course of the 3-day experiment, indicating that interspecific variability and photoacclimation state were more influential in determining the responses to stress than the number of consecutive stress days. The most widespread and consistent response to applied stress observed in this study was the decline in high light-induced DD de-epoxidation, as measured by the increase in DES, over the 3 days. This decline was progressively attenuated over the days for all species and photoacclimation states with a functional XC (all except HL-acclimated N. phyllepta and E. paludosa cultures). This trend may be indicative of cumulative negative impacts on XC function or can result from ongoing photo- and thermoacclimation to the intermittent stress. Either way, it appears to be a species-specific behavior.

An immediate explanation for the absence of significant cumulative effects in some of the measured physiological parameters is that the applied stress conditions, although causing substantial deleterious effects, were not severe enough to induce irreversible damage over the 3-day experiment. To this result also contributed the long periods between consecutive stress events (22.5 h), which included several hours under growth irradiance (10.5 h) and darkness (12.0 h), at growth temperature. PSII repair requires the de novo synthesis of the D1 protein, a process favored by conditions applied between stress periods: low to moderate light intensities support the ATP production required for the D1 protein turnover while keeping new PSII photoinactivation rates below the counteracting photorepair ones (Campbell and Tyystjärvi, 2012; Murata and Nishiyama, 2018; Campbell and Serôdio, 2020). The observed strong recovery from 1 day to the next could also be attributable to PSII photorepair carried out during the night periods following stress exposure. The capacity to turnover PSII proteins in darkness has been shown in centric diatoms (Li et al., 2016), making it plausible that a similar mechanism might operate in benthic species.

The observed interspecific differences in the response to heat stress might have implications for MPB species composition under scenarios of increasing heatwave frequency. Species with higher photorepair capacity and more efficient PSII repair, such as the benthic forms here studied, are likely to be favored under more frequent extreme events, which can possibly drive changes in composition and productivity of intertidal MPB communities. A decline in tychoplanktonic forms such as C. closterium, disadvantaged by their limited photorepair capacity and poor recovery from cumulative stress, could alter biofilm structure, nutrient cycling, and benthic-pelagic coupling, as these species can be periodically resuspended and contribute to pelagic primary production. Changes in MPB community composition would likely cascade into higher trophic levels, given its central role a food source for meiofauna, benthic invertebrates, and shorebirds (Hope et al., 2019; Saint-Béat et al., 2013).

Constraints

Despite providing novel results on the interspecific variability of photophysiological responses in MPB diatoms, this study has two main constraints worth acknowledging, that also point to directions for future research. The first concerns the experimental conditions used to simulate light and heat stress. In intertidal environments, MPB communities can be exposed to more extreme conditions than those applied here, including higher irradiances, temperatures frequently exceeding 35 °C during summer low tides, and UV-B radiation, all of which vary substantially on short time scales (Serôdio and Catarino, 1999; Serôdio et al., 2008; Laviale et al., 2015). The conditions used (constant 650 µmol photons m -2 s -1, no UV-B, and 35 °C) may thus not be fully representative of the most extreme conditions occurring in situ. Nonetheless, their controlled and uniform application across species enabled a valid comparison of interspecific differences in photophysiological responses. The second aspect relates to the possibility that the responses documented here may differ from those expressed under in situ conditions. In the sediment, motile diatoms can use vertical migration to adjust their position within the sediment photic zone, regulating the experienced light stress, reducing the need for physiological photoprotection and repair to be fully activated. However, the suspension-based approach here used allowed the characterization of the inherent photophysiological capacities of each species, without the confounding influence of migratory behavior, revealing significant intrinsic differences between co-occurring MPB species.

This study showed that the tested diatom species differed markedly in their responses to repeated light and heat stress, with species type and photoacclimation state appearing as more determining of the final responses than the number of consecutive stress events. The two benthic species N. phyllepta and E. paludosa maintained a high capacity to recover from each stress period, based on an efficient PSII photorepair, while the tychoplanktonic species C. closterium showed consistently lower recovery capacity, associated to higher PSII photoinactivation rates. A consistent cumulative effect across species was the gradual reduction in DD de-epoxidation on successive days, indicating that this type of photophysiological response was especially sensitive to repeated stress. Despite the strong effects observed following each stress exposure, irreversible PSII photodamage did not accumulate, probably due to the long intervals between consecutive stress periods that allowed an effective PSII photorepair.