Mediterranean forests dominated by macroalgae of the order Fucales (Ochrophyta) dwell at different depths between the infralittoral and the first meters of the circalittoral (Boudouresque, 1971; Boudouresque, 1972; Ballesteros, 1988; Ballesteros, 1990a; Ballesteros, 1990b; Hereu et al., 2008; Ballesteros et al., 2009). Ericaria crinita (Duby) Molinari and Guiry ( Figure 1A ) and Ericaria mediterranea (Sauvageau) Molinari and Guiry ( Figure 1B ) are two endemic Mediterranean species that can develop highly structured assemblages in the rocky upper infralittoral and sublittoral zones (between 0 and 1 m depth approximately) (Ballesteros, 1992; Ballesteros, 2002). The two species can be found a few meters apart, although they present contrasting life history traits ( Figure 1 ). While E. crinita prefers sheltered or semi-exposed rocky shores (Sales and Ballesteros, 2009; Sales and Ballesteros, 2010; Sales and Ballesteros, 2012), E. mediterranea dwells in wave-exposed or moderately exposed shores (Boudouresque, 1969; Ballesteros, 1988; Gómez-Garreta, 2000). Accordingly, E. crinita populations can live in areas where extreme temperatures are frequent, whereas E. mediterranea develops in open areas where seawater temperatures are buffered by the high hydrodynamism (Verdura et al., 2021). Moreover, while E. crinita is perennial, keeping its branches throughout the year, E. mediterranea is semi-perennial, keeping only the axis throughout the year and losing the branches in autumn, growing back in spring (Ballesteros, 1988; and see Figures 1E , F ). In fact, compared to E. crinita, E. mediterrania is a fast- growing species and has a higher annual biomass production (Ballesteros et al., 1998; Sales and Ballesteros, 2012).

To describe the variability on specific characteristics on thallus growth (perennial vs. semi-perennial) and the contrasting temperature conditions of the habitats where each species can inhabit, in situ temperature and thallus lengths have been monitored for both species in Costa Brava, Palamós (41°86’62.1’’N, 3°17’50.7’’E) in the North-western Mediterranean Sea. Concurrently, populations of each species were monitored monthly throughout 2018 to measure two length variables per individual, from the basal disc to the tip of the apex of the highest main axis and the length from the basal disc to the tip of the longest branch (N=80 per month, and species). Differences among total and man axis length was used as a proxy for each species’ dynamics and productivity. In situ sea temperatures were continuously recorded (hourly records) throughout 2017 and 2018 by positioning high-resolution temperature loggers (HOBO Water Temp Pro v2, ± accurate to 0.21°C) in the specific habitat dominated by each species.

Adult individuals of each species (N=36) were collected at 17^th and 18^th April, 2017 from subtidal rocky shores in the locality of Palamós. In Palamós, the two species (E. mediterranea and E. crinita) inhabit the same area a few meters apart but with different hydrodynamic conditions. Adult specimens were carefully removed from the rock preserving their attaching disc and transported directly to the laboratory. Once in the laboratory, epiphytes and sediment were carefully removed and the specimens were placed into tanks with natural seawater at 18°C and natural irradiance levels for three days for acclimation. Temperatures were raised at a maximum rate of two degrees per day from 18°C until the experimental temperature conditions were reached, marking the start of the experiment.

Fertile branches of E. crinita and E. mediterranea (N= 92 per each) were collected from the same two populations of the adults in late spring and immediately transported to the laboratory. Once in the laboratory, fertile branches were stored at 5°C in dark conditions for 24 hours to stimulate gametes release, before subjecting them to the different treatments.

To analyse the effects of temperature and UV radiation on adults and germlings of E. crinita and E. mediterranea, six different experimental treatments were established by combining three temperature conditions (21°, 24° and 28°C) and two irradiance conditions (visible irradiance -used as a control- and visible irradiance supplemented with ultraviolet irradiance). Temperature conditions were selected to reflect normal summer water temperatures (21°C, set up as the control), maximum summer temperatures (24°C) and extremely high temperatures, experienced during MHWs in the area (28°C) (Verdura et al., 2021).

The two irradiance treatments corresponded to: i) the photosynthetic active radiation (PAR) supplying the natural intensity that these communities receive at 0 m (irradiated with 180-200 µmol m^-2 s^-1 of PAR, Sant, 2003); and ii) the same photosynthetic irradiance adding ultraviolet radiation (PAR+UV) since changes in stratospheric ozone, cloud cover and aerosols continue to affect surface levels of solar UVB radiation (e.g. Bais et al., 2019; Bernhard et al., 2023). The photoperiod, 14:10 (light:dark), was the natural photoperiod for the season.

For adult individuals experiments for each seaweed species were conducted in parallel for 30 days. Six replicate tanks (12L each) were used for each treatment and species, with one adult individual per tank. We therefore set up a total of 36 tanks per species. Each tank had an independent and closed-water system with air continuously pumping and with 2L of natural seawater renovation per day per tank, thus, each tank had a completely natural seawater renewal every 6 days. Each week the natural seawater was collected from the same area where the algae came from and transported and stored to the laboratory in several 60 L drums. The water temperature for each tank was regulated and kept constant (with a maximum deviation of ± 0.4 °C) at the three temperatures using temperature controllers (Teco TK 500). The supply for the absence of ultraviolet radiation treatment tanks came from PAR radiation fluorescent lamps (Master TL-D 36W/386 of Philips), while for the tanks exposed to ultraviolet radiation the same PAR fluorescent lamps were used, plus UVA and UVB radiation fluorescents (Actinic BL TL-K 40W/10-R and Ultraviolet-B 40W/12 RS SLV/25 of Philips, respectively) supplying 180-200 µmol m^-2 s^-1 PAR (12 W m^-2) and 6-11 W m^-2 UVA and 0.3-0.5 W m^-2 UVB. Irradiance in each experimental condition was measured using a Li-1000 SPQA spherical sensor.

For the early life stage experiment, germlings of the two species were obtained in the laboratory facilities. For each species, three replicate culture tanks (0.5L each), each with three microscope slides on the bottom, were used for each treatment (6 treatments * 3 tanks * 3 slides). The 18 culture tanks for each species were filled with sterilised sea water and growth medium (Von Stosch modified by Guiry and Cunningham, 1984) and were maintained in culture chambers to control temperature and radiation conditions. First, to obtain the germlings, six fertile branches (of the corresponding species) were placed in each culture tank until zygote formation was detected (after three days), after which the receptacles were removed. The slides, which could be removed from the tanks for short periods of time, allowed us to observe the germlings and subsequently monitor them under a microscope throughout the experiment. The seawater with the medium was completely renewed twice a week. The temperature conditions (21°C, 24°C and 28°C) were achieved using three different growth chambers (Radiber AGP-360), each of which was subjected to PAR radiation fluorescents (Master TL-D 36W/386 of Philips) and UVA and UVB fluorescents (Actinic BL TL-K 40W/10-R and Ultraviolet-B 40W/12 RS SLV/25 of Philips, respectively) supplying 180-200 µmol m^-2 s^-1 PAR (12 W m^-2) and 6-11 W m^-2 UVA and 0.3-0.5 W m^-2 UVB. Incubator chambers conditions were checked daily, to ensure the proper functioning and that temperature remained constant and correct. For the UV free treatment (PAR treatments), half the tanks in each chamber were covered with polycarbonate filter boxes (transmittance >390 nm) to cut-off UV radiation (Bischof et al., 2002). Irradiance in each experimental condition was measured using a Li-1000 SPQA spherical sensor. The photoperiod was 14:10 throughout the entire experiment, which was carried out at the facilities of the University of Girona.

Effects of temperature and radiation on Ericaria crinita and E. mediterranea individuals were evaluated by measuring several variables. In adult specimens, measures of biomass (wet weight in g) and optimum quantum yield (Fv/Fm) of photosystem II (PSII) were taken once a week. Changes in wet weight (expressed as a percentage) were measured for each specimen at each sampling time. The specimens were dried with absorbent paper before weighing them, following the same procedure for all the samples and sampling times. Optimum quantum yield was used as an indicator of PSII performance to assess photosynthetic efficiency. To do so, macroalgal fronds were dark incubated for 15 minutes, after which, Fv/Fm measurements were estimated by applying a saturation pulse using Pulse Amplitude Modulated Fluorometry (Diving-PAM Underwater Fluorometer, Waltz, Germany). Measuring Fv/Fm following a period of dark adaptation is a common technique for measuring stress in plants (Murchie and Lawson, 2013).

Survival and growth of viable germlings were measured under a microscope once a week. Survival of germlings was calculated counting all the live germlings attached to the microscope slides at each sampling time. Growth was assessed by measuring the maximum length of the thallus of 45 random individuals for each treatment (µm) under a microscope, and using the image analysis software ZEN 2012.

Differences among total length (square root transformation) from natural populations along a year were analysed by Linear Model (LM) with species and months as fixed factors.

The data obtained from both adult and recruit experiments were analysed by mixed-effects models (MM), which allow the inclusion of both fixed and random effects as predictor variables. Different models were fitted to analyse the effect of temperature, radiation and their interaction for each selected response variable. Specifically, wet weight variation in adults was analysed using a linear mixed model (LMM) and the Optimum Quantum Yield with a Generalised Linear Mixed Model (GLMM), with a Poisson error distribution and a logit link function. Survival and growth of germlings were analysed using a GLMM with a binomial error distribution and a logit link function, and a quasi-Poisson error distribution and a log link function, respectively. Additionally, to control for lack of independence between units of observation and to handle repeated measures over time both individual’s identity and the sampling time were considered random factors (Bolker et al., 2009; Bates et al., 2015; Harrison et al., 2018). In all the models, irradiation (2 levels), temperature (3 levels) and species (2 levels) were considered as fixed factors.

Models were fitted using the functions “lm”, “glmer”, “lmer” and “glmmPQL” from lme4 (Bates et al., 2015) and MASS packages (Ripley et al., 2013), and the Wald χ2 test was performed using the “ANOVA” function from the CAR package (Fox and Weisberg, 2019). For all the models, the assumptions of normality and equality of variance were evaluated through graphical analyses of residuals using QQ plot functions (effects, visreg, and car package). For multiple comparisons, we applied Tukey´s post-hoc tests using the “glht” function from the MULTCOMP package (Hothorn et al., 2008). All analyses were performed in the statistical environment R (R Core Team, 2019).

In situ seawater temperatures recorded over one year showed that temperatures in the habitat dominated by E. crinita, characterized by sheltered and shallow conditions, were more extreme (maximums of 31.89 °C and minimums of 8.27°C; Figure 1C ). In contrast, temperatures of the habitat dominated by E. mediterranea, being a more open and exposed environment, remain within more constant ranges (maximums of 28.65 °C and minimums of 11.62 °C; Figure 1D ).

The two species showed different growth dynamics (p < 0.05; Table S1 ), while E. crinita is characterized by constant branch lengths throughout the year (p>0.05; Table S1 ), with branch lengths ranging from 9.25 ± 0.18 cm in November to 12.07 ± 0.15 cm in July ( Figure 1E ), E. mediterranea presented significant variations among months (p<0,05) with the maximum branch lengths in June (26.17 ± 1.15 cm) and a minimum in December (4.87 ± 0.27 cm; Figure 1F ).

UV radiation had no significant effect on the adult individuals of either of the two species ( Figure 2 ; p > 0.005, Table S2 ). However, adults of both species were significantly affected by water temperatures, especially at 28°C, when biomass (especially due to the secondary and tertiary branches losses) and photosynthetic yield dropped drastically after one week for E. mediterranea and after two weeks for E. crinita ( Figure 2 ). Vulnerability to temperature was specie dependent, while at 21°C and 24°C biomass losses (of approx. 26% and 36%, for E. crinita and E. mediterranea, respectively) were comparable among species, at 28°C biomass losses for E. mediterranea were significantly higher than for E. crinita, showing a biomass reduction of ca. 75% and 20% by day 14, respectively. At the end of the experiment (day 30) only E. crinita presented alive thallus, which displayed biomass losses of about 75%.

Similar trends were found for the photosynthetic quantum yield, whose values remained at between 0.6 and 0.8 at 21°C and 24°C for both species (p> 0.05; Table S2 ), in line with usual values found for healthy Cystoseira s.l. species (Abdala-Díaz et al., 2006; de Caralt et al., 2020; Verdura et al., 2021). However, at 28°C the photosynthetic quantum yield dropped by more than half (c.a. 0.258 ± 0.012) as early as day 5 in E. mediterranea (p< 0.001; Table S2 ), while in E. crinita it stayed high and constant under all temperature treatments throughout the experiment (c.a. >0.600; p> 0.05; Table S2 ).

Germlings of both species were significantly affected by temperature and UV radiation, and effects on both variables (survival and growth) were species- and temperature-dependent, although UV radiation always reduced germling survival (p< 0.001, Supplementary Materials Table S3 ; Figure 3 ). High temperatures (28°C) strongly reduced the survival of both species, with less than 10% survival after two weeks in all radiation conditions. UV radiation had a major impact on the E. crinita germlings, leading to zero survival after a few days of UV exposure under all temperature conditions (p < 0.05, Table S3 ; Figure 3 ). At 21 °C and 24°C, survival of the two species under PAR conditions was generally constant, with values ranging from approximately 20% to 50% for both species, while UV conditions lead to the death of almost all E. crinita germlings by days 12 and 5 for the 21° and 24 °C treatments, respectively. However, the survival of E. mediterranea under UV radiation at 21°C and 24°C was similar to the survival under PAR radiation.

E. mediterranea germlings grew faster than the E. crinita germlings. However, temperatures of 28°C negatively impacted the growth of both species (p<0.05, Table S3 ; Figure 3 ). At 28°C, growth was highly reduced in all irradiance treatments for both species, with E. mediterranea individuals under PAR conditions the only survivors at the end of the experiment (ca. 1.700 µm in length). At 21°C and 24°C, the growth of both species under PAR conditions increased steeply, reaching values of ca. 3,000 and 3,500 µm for E. mediterranea at 21°C and 24°C, respectively, and ca 1,300 µm for E. crinita at both temperatures on day 30. However, UV conditions significantly decreased E. mediterranea growth, especially at 24°C when growth was reduced by more than a third (p<0.001, Table S3 ; Figure 3 ).