Summer Heatwave Impacts on the European Kelp Saccharina latissima Across Its Latitudinal Distribution Gradient

Kelps are important foundation species in coastal ecosystems currently experiencing pronounced shifts in their distribution patterns caused by ocean warming. While some populations found at species’ warm distribution edges have been recently observed to decline, expansions of some species have been recorded at their cold distribution edges. Reduced population resilience can contribute to kelp habitat loss, hence, understanding intraspecific variations in physiological responses across a species’ latitudinal distribution is crucial for its conservation. To investigate potential local responses of the broadly distributed kelp Saccharina latissima to marine heatwaves in summer, we collected sporophytes from five locations in Europe (Spitsbergen, Bodø, Bergen, Helgoland, Locmariaquer), including populations exposed to the coldest and warmest local temperature regimes. Meristematic tissue from sporophytes was subjected to increasing temperatures of Δ+2, Δ+4 and Δ+6°C above the respective mean summer temperatures (control, Δ±0°C) characteristic for each site. Survival and corresponding physiological and biochemical traits were analyzed. Vitality (optimum quantum yield, Fv/Fm) and growth were monitored over time and biochemical responses were measured at the end of the experiment. Growth was highest in northern and lowest in southern populations. Overall, northern populations from Spitsbergen, Bodø and Bergen were largely unaffected by increasing summer temperatures up to Δ+6°C. Conversely, sporophytes from Helgoland and Locmariaquer were markedly stressed at Δ+6°C: occurrence of tissue necrosis, reduced Fv/Fm, and a significantly elevated de-epoxidation state of the xanthophyll cycle (DPS). The variations in phlorotannins, mannitol and tissue C and N contents were independent of temperature treatments and latitudinal distribution pattern. Pronounced site-specific variability in response to increasing temperatures implies that exceeding a threshold above the mean summer temperature exclusively affect rear-edge (southernmost) populations.


INTRODUCTION
Temperature is a major factor controlling global biogeographic patterns of many marine benthic organisms, such as macroalgae (Lüning, 1990;Adey and Steneck, 2001;Wiencke and Bischof, 2012). Hence, global warming will likely provoke shifts in the latitudinal distribution of algal species (King et al., 2018;Smale, 2020). Changes in temperature regimes, such as an increase in frequency and amplitude of marine heatwaves (MHWs), often impair primary production or even survival of macroalgae, with consequences for standing stock biomass or even loss of important foundation communities as well as the shift into a novel ecosystem status (Frölicher et al., 2018;Harris et al., 2018;Smale et al., 2019;and references therein), such as shifts to turf algae communities (Wernberg et al., 2013(Wernberg et al., , 2016a. Sea surface temperature (SST) has increased by 0.63 • C globally in the period 1986-2005compared to 1850-1900(IPCC, 2019a. However, on a regional scale SSTs rose even higher, for example, on Helgoland (North Sea) > 1.6 • C between 1962 and 2010 (Wiltshire et al., 2010). Under different climate change scenarios, the global mean SST is likely to increase by 1.6 • C up to more than 4 • C until the end of this century and additionally MHWs are expected to become more severe (Frölicher et al., 2018;IPCC, 2019a). MHWs are defined as periods of SST increases of 3-5 • C above the longterm mean, which can last from a few days to months (Meehl and Tebaldi, 2004;Hobday et al., 2016). Worldwide, the annual duration of MHWs has already increased by 54% between 1925 and 2016 (Oliver et al., 2018).
Considering temperature tolerance limits, species with a wide latitudinal distribution range are generally more tolerant than narrow-ranging species (Wiencke et al., 1994;Kelly et al., 2012;Sunday et al., 2015). Among trees, species with a temperate to Arctic distribution will respond neutrally or even positively to future warming near their cold-range edge (Reich et al., 2015). This response might also apply to broadly distributed macroalgal species, such as kelps. In previous studies, different macroalgal populations were often treated as a single homogenous physiological unit, irrespective of their distribution (Reed et al., 2011). However, local adaptation and phenotypic plasticity can result in intraspecific differences in thermal tolerance, and population loss might occur not only among thermal rear edges but also in other regions (King et al., 2018). Recently, studies on the kelp Laminaria digitata, revealed that despite considerable plasticity at intermediate temperatures, plasticity at the upper survival limit was low (Liesner et al., 2020a,b;Franke et al., 2021). Thus, models assuming a uniform climatic envelope for wide-ranging species, particularly the local temperature threshold for survival, may underestimate local adaptation and extinction (Kelly et al., 2012;Filbee-Dexter et al., 2020).
Kelps are large canopy-forming brown algae of the order Laminariales . They are important primary producers and foundation species in coastal ecosystems providing habitat, nurseries and food for many associated organisms (Dayton, 1985;Bartsch et al., 2008). Alterations in distribution patterns of kelp forests in response to global climate change, including expansions and declines, were observed worldwide (Krumhansl et al., 2016). In Europe, species distribution models project a northward expansion of different kelp species (Müller et al., 2009;Assis et al., 2018). Newly exposed hard substrates in Arctic regions are already colonized by kelps (Krause-Jensen et al., 2020), whereas drastic loss of kelps is occurring at the southern distribution boundary (e.g., Voerman et al., 2013;Filbee-Dexter et al., 2020).
One widely distributed kelp along the European Atlantic is the boreal-temperate species Saccharina latissima, which is found from Arctic regions to northern Portugal (Araújo et al., 2016). The species exhibits a high degree of polymorphism and physiological plasticity, and is regarded to exhibit an opportunistic growth strategy (reviewed by Bartsch et al., 2008). Optimum growth of sporophytes is reported between 10 and 15 • C (Bolton and Lüning, 1982). However, the sporophytes can briefly survive high temperatures of 23 • C up to 1 week, but increasing mortality rates were already observed above 20 • C (Fortes and Lüning, 1980;Bolton and Lüning, 1982;Lüning, 1984). In fact, the published information on upper survival temperatures derived from experiments without a preacclimation phase and may not reveal the actual resilience among wild populations where kelps are normally subjected to slowly increasing temperature. Nevertheless, Casado-Amezúa et al. (2019) report a decline in the S. latissima population along the north-western Iberian Coast during the last three decades, and presently, individuals are only rarely found in the North of Portugal (F. Arenas, pers. comm.). Massive alterations in population density and distribution pattern of European S. latissima populations were not only reported at their southern distribution range, but also in their intermediate northern locations, e.g., in the North Sea areas of Helgoland and Skagerrak, and along the southern and southwestern coasts of Norway (Pehlke and Bartsch, 2008;Bekkby and Moy, 2011;Moy and Christie, 2012).
Photosynthesis and growth of S. latissima is suboptimal at low temperatures (Bolton and Lüning, 1982;Karsten, 2007) while exposure to high and sublethal temperatures can result to biomass loss when, e.g., photosynthetic apparatus and other cellular structures are damaged (Andersen et al., 2013;Simonson et al., 2015b) leading to reduced growth. Aside from physiological acclimation, kelps and other seaweeds also use different biochemical protective mechanisms to adjust to temperature variations, such as adjustments in the xanthophyll cycle (de-epoxidation, DPS; Goss and Jakob, 2010).
The general impacts of MHWs on kelps and other foundation seaweeds have been intensively studied (e.g., Bennett et al., 2015;Burdett et al., 2019;Nepper-Davidsen et al., 2019;Straub et al., 2019;Saha et al., 2020), while there are only few studies on thermal plasticity of species across their latitudinal distribution (Winters et al., 2011;Jueterbock et al., 2014;Pereira et al., 2015;Wernberg et al., 2016b;Liesner et al., 2020a). Besides local variations in temperature, variable local thermal susceptibility may contribute to the mortality of European S. latissima, as suggested for other kelps (Wernberg et al., 2010;King et al., 2018). For instance, Wernberg et al. (2016b) showed in a common garden experiment that independent of their latitudinal distribution, the optimum photosynthetic temperature remained the same in different seaweeds in Australia, while Q 10 -values for photosynthesis and respiration and also chlorophyll a concentrations decreased from cooler to warmer locations. Bennett et al. (2019) and Filbee-Dexter et al. (2020) emphasized the relevance of understanding intraspecific variations in ecology, and until now, most experimental studies on temperature stress across latitudes did not consider the different respective local temperatures.
Contrary to the usual common garden experiments, our experimental design aims to reveal the responses of S. latissima to local MHWs. Therefore, we conducted a short-term experiment using wild sporophytes collected from five locations along the European Atlantic coast (i.e., Spitsbergen, Bodø, Bergen, Helgoland, Locmariaquer). We investigated whether different amplitudes of summer MHWs across latitudes will have differential effects on distinct S. latissima populations or not. In our mechanistic experimental set-up we increased the respective local mean summer temperature (control, ±0 • C) by +2, +4 and +6 • C to investigate the relative responses of different kelp populations to MHWs over a large geographical scale.
We hypothesized that the tolerance of European S. latissima to summer MHWs, ranging from +2 to +6 • C, is related to the mean summer temperature experienced in situ. Since the maximum temperature during summer in northern populations is lower than their physiological limit (Bolton and Lüning, 1982;Lüning, 1984), we expect these populations at their expanding edge to benefit or be unaffected (neutral response) by the temperature-amplitude treatments. In contrast, among the rear-edge populations at the southern limit, we expect them to suffer physiological stress from the temperature increase, i.e., when experiencing temperatures higher than the lethal limit determined for S. latissima (Bolton and Lüning, 1982;Lüning, 1984).
Meristematic discs (diameter 22-24 mm) from 20 to 25 nonfertile sporophytes were excised within the 2-10 cm length of the basal lamina. Samples from Bodø, Bergen, Helgoland and Locmariaquer were transported moist, cool (< 15 • C) and dark ( Table 1). All the experiments, except for the Spitsbergen material, were conducted at the Alfred Wegener Institute for Polar and Marine Research in Bremerhaven, Germany. The Spitsbergen samples were processed locally at Ny-Ålesund and no transportation of the samples was necessary. For recovery and wound healing (pre-acclimation phase, Figure 2), the samples were maintained in the control seawater temperature ( Table 1).
During the pre-acclimation phase, four pools of discs were prepared in separate flasks to represent the four replicates. Each pool consisted of 10 meristematic discs excised from five distinct sporophytes, this means: twenty distinct sporophytes were used to ensure the independence of each replicate when representative discs from each pool were harvested during subsequent sampling to measure respective response variables.
In the acclimation phase (Figure 2), the discs were successively acclimated to different temperature amplitudes ( +2, +4, +6 • C) with an increase of 2 • C every second day. Afterwards, the final temperature-amplitude treatments were applied for another week. Respectively, the highest temperature-amplitude treatment ( +6 • C) run for 9 days, while +4 • C run for 11 days and +2 • C for 13 days, according to the treatment scheme (Figure 2). We based the different absolute temperatures on the averaged mean summer sea surface temperatures (SST) at each sampling site (Table 1), using the respective mean summer SST as control ( ±0 • C). Figure 1B displays the satellite-obtained mean summer (June-August) SST between 2016 and 2020 with a resolution of 4 × 4 km at each sampling site (Giovanni Satellite: Acker and Leptoukh, 2007).
The integrity of all replicate discs was monitored to measure survival, vitality (F v /F m ) and growth during the temperature exposure treatments. After the acclimation phase, first sampling was conducted (day 1, data not shown). Five discs were kept for the temperature-amplitude treatments which then run for 1 week (day 1-8, Figure 2). For biochemical analyses samples were shock frozen in liquid N 2 , stored at −80 • C, and freeze-dried (Alpha 1-4 LO plus, Martin Christ Gefriertrocknungsanlagen GmbH, Osterode am Harz, Germany). Statistical evaluation (data not shown) did not reveal any relevant significant differences in the biochemical response parameters between day 1 and day 8. Therefore, this study exclusively presents biochemical results after the temperature-amplitude treatments (day 8).

Physiological Response Variables
At the end of the experiment, all discs were qualified according to their physical state as healthy or dead. A disc was categorized as "dead" when tissue necrosis was observed (Supplementary Figure 1). Disc survival (%) at day 8 was computed as: number of discs day 8 number of discs day 1 · 100 (1) For growth analysis, the discs were photographed (Figure 2: initial, day 1, 2, 4, 6, 8) together with a known 9 cm 2 size object a reference. The algal disc size was then analyzed using the ImageJ (Version 1.52a, Java 1.8.0_112, Wayne Rasband, National Institute of Health, United States) software. As the samples grew with different rates prior to the temperature-amplitude treatments, the initial size of the replicates was adjusted to 100% and the increase in size calculated as % of initial for a better comparison of algal growth. Decomposing discs were excluded from the growth analyses.
As proxy for photosynthetic performance and, hence, of algal vitality, the in vivo chlorophyll-fluorescence of photosystem II (optimum quantum yield, F v /F m ) was measured every day (day 1-8) after 5 min of dark acclimation using a pulseamplitude-modulated fluorometer (Imaging-PAM, Walz GmbH Mess-und Regeltechnik, Effeltrich, Germany), which was set up to determine the amplitude of the initial fluorescence signal (F t ) between 0.15 and 0.2 (SP intensity = 8, SP duration = 3 s), as recommended in the manual (Imaging-PAM M-Series Chlorophyll Fluorometer, Heinz Walz GmbH, Effeltrich, Germany). F v /F m of each disc was monitored in all stages of visible stress, until the tissue was completely decomposed (see Supplementary Figure 1).

Biochemical Response Variables
Pigments, total carbon (C), total nitrogen (N), C:N ratio, mannitol and phlorotannins were analyzed following the protocol in Diehl and Bischof (2021).
To determine differentiations in the photosynthetic apparatus, the ratios of chlorophyll a to accessory pigment pool (Chla:Acc) and of chlorophyll a to xanthophyll pigments pool (Chla:VAZ) were calculated. The de-epoxidation state of the xanthophyll cycle (DPS) was calculated after Colombo-Pallotta et al. (2006):

Statistics
Outliers (Bonferroni, p < 0.05) in the biochemical parameters were excluded from the statistical analyses. All datasets were tested for normal distribution (Shapiro-Wilk test, p > 0.05) and homogeneity of variance (Levene's test, p > 0.05). F-statistic was reported to be robust against a moderate violation of normal distributions at small sample sizes in terms of Type I errors (Blanca et al., 2017), thus parametric tests were performed. The increase of size (% of initial) monitored over several time points (normally distributed after log 10 -transformation) and F v /F m , which was monitored every day (non-normally distributed) was analyzed using a repeated measures two-way ANOVA [population (P) × temperature-amplitude (TA)] followed by a post hoc Tukey's test. All response variables measured on a single time point at the end of the experiment were examined using twoway ANOVAs (P × TA) followed by post hoc Tukey's tests. The level of significance was set to p < 0.05 for normally distributed data. A more conservative level of significance was applied for non-normally distributed data (p < 0.01). All statistical analyzes were conducted with RStudio (Version 1.3.1073, 2020, Boston, MA, United States).

RESULTS
Significant effects of the temperature-amplitude treatments (TA) within each population are marked in Figures 3-5 and Tables 2,3 a Data from 12 m depth, provided by the COSYNA system (Baschek et al., 2017), since they were not measured during sampling. b Giovanni Satellite (Acker and Leptoukh, 2007).
FIGURE 2 | Set-up of the temperature-amplitude experiment with Saccharina latissima. Pre-acclimation phase varied between the sampling sites due to logistic constraints. During the acclimation phase, the control temperature ( ±0 • C) was increased by 2 • C every second day to reach the temperature-amplitude treatments ( +2-+6 • C). Sampling for biochemical analyses was conducted on day 8 of the temperature-amplitude treatments.
using different letters. The statistical results of overall population (P) and TA effects, and their interaction are presented in Table 4. Survival among different populations from the species northern to southern distribution range ( Table 2) varied significantly in response to increasing TA (p < 0.001) and significant interactive effects of P and TA were found (p < 0.001). The mortality of the samples strongly increased at absolute temperatures (AT) > 23 • C ( Table 2 and Supplementary  Figure 2). Samples from Spitsbergen showed survival of 100% (AT 6-12 • C) and samples from Bodø (AT 12-18 • C) and Bergen (AT 16-22 • C) of 95-100%, while samples collected at the locations with highest mean summer temperatures-Helgoland and Locmariaquer-died and decomposed at +6 • C (Helgoland: AT 24 • C, 70% survival; Locmariaquer: AT 25 • C, 65% survival).
For all five populations, size of the discs increased over time (Figure 3; p < 0.001). There were significant differences in growth (as reflected in rising disc diameter in % of initial) between the populations (p < 0.001), independently of the temperature treatments: While Spitsbergen samples grew considerably reaching 160-200% of their initial size after the preacclimation phase, samples from all other populations exhibited reduced growth activity (<130%), being lowest in Helgoland and Locmariaquer. Differences between samples from Bodø and Bergen were non-significant. Even though statistical significances were also detected for the temperature-amplitude treatments (TA) (p < 0.001) and P × TA interaction (p < 0.001), these could not be assigned to temperature increase per se (Supplementary  Figure 3). For instance, growth in Spitsbergen at +2 • C (AT 8 • C) was significantly higher than at +6 • C (AT 12 • C), whereas FIGURE 3 | Size (% of initial) of meristematic discs of Saccharina latissima across latitudes in Europe during the temperature-amplitude experiment. Experimental set-up: mean summer SST (control, ±0 • C) and temperature stress treatments: +2, +4, +6 • C. The gray shading represents the temperature acclimation phase (initial day [i] -day 1). Values are means ± SD (n = 4). Significant differences are marked by different letters.
at ±0 • C (AT 10 • C) and +6 • C (AT 12 • C) growth did not differ significantly and at +6 • C samples from Spitsbergen were considerably larger than samples from Bodø at ±0 • C, even though both treatments were subjected to the same absolute temperature of 12 • C.
Survival and growth responses are supported by the optimum quantum yield (F v /F m ) measurements (Figure 4). We determined significant differences between the five populations (p < 0.001) and significant impact of TA (p < 0.001) and P × TA interaction (p < 0.001) on F v /F m . All samples from Spitsbergen, Bodø and Bergen revealed F v /F m values > 0.6 throughout the experiment, whereas F v /F m values of samples from Helgoland and Locmariaquer markedly decreased with FIGURE 4 | Optimum quantum yield of photosystem II (F v /F m ) of Saccharina latissima across latitudes in Europe during the temperature-amplitude experiment. Experimental set-up: mean summer SST (control, ±0 • C) and temperature stress treatments: +2, +4, +6 • C. Values are means ± SD (n = 4). The dotted line indicates the threshold for a good physiological status. Significant differences are marked by different letters.
increasing temperatures of +6 • C (AT > 23 • C). Clear trends to decreasing F v /F m values from north to south and from lower to higher temperatures were determined. Plotting the F v /F m values of day 8 against the absolute temperatures of the treatments (Supplementary Figure 4) reveals a strong impact of absolute temperatures on quantum yield of S. latissima. A strong diminution was solely detected at absolute temperatures above 23 • C.
The absolute pigment contents of chlorophyll a (Chl a), the pool of the accessory pigments (Acc), the pool of the xanthophyll cycle pigments (VAZ) and the ratios of Chla:VAZ and Chla:Acc sampled on day 8 did not reveal any trends assigning to the latitudinal distribution of S. latissima, even though significant differences were found between the populations (Tables 2, 4). There was also no significant impact of TA and no P × TA interaction on Chl a, Acc and VAZ (Table 4). For the ratio between Chla:VAZ, significant impact of TA (p < 0.01) but not of P × TA were determined (Table 4). Still, trends to decreasing Chl a, increasing VAZ and decreasing Chla:VAZ were detected in the +6 • C treatments from Helgoland and Locmariaquer (Table 2). Even though a significant effect of P × TA on Chla:Acc was found (p < 0.001), no pattern regarding temperature were detected. The de-epoxidation state of the xanthophyll cycle (DPS; Figure 5) was significantly affected by P, TA and P × TA (p < 0.001; Table 4), revealing increasing DPS at the high temperature treatments in the southern populations (AT ≥ 24 • C). DPS of the samples from Helgoland and Locmariaquer both increased significantly at +6 • C (p < 0.001). None of the other biochemical response variables (Table 3)mannitol, C:N ratio, phlorotannins-were significantly affected by the TA or P × TA, but significant differences were present between samples from the five populations in all three parameters (p < 0.001; Table 4), however, without correlation to the sampling latitude. Regarding phlorotannins, the lowest concentrations were found in the samples from Spitsbergen and highest in the samples from Bergen and Locmariaquer. Overall, mannitol concentrations were similar in all populations, though the samples from Locmariaquer exhibited the lowest concentrations. Saccharina latissima from Helgoland exceeded C:N ratios above 20, while samples from all other populations had values of about 13-19, independent of the TA. These differences were mainly based on significantly lower nitrogen (N) concentrations in the Helgoland samples and not on variations in total carbon (C) (

Impacts of Short-Term Temperature Increase on Saccharina latissima Across Latitudes
Results of our study showed that S. latissima sampled across the species' latitudinal distribution reveals differences regarding their physiology and biochemical status in response to temperature increases of +2, +4 or +6 • C above the respective local mean summer temperature. We detected adverse effects of rising temperatures exclusively in the samples of southernmost S. latissima, for which the temperature-amplitude treatments exceeded absolute temperatures (AT) of 20 • C. Northern isolates overall were not impaired. In addition, we found strong sitespecific and temperature-independent differences, which were especially prominent in growth response but also determined in C:N and phlorotannins.
Prior studies reported an increased mortality of S. latissima sporophytes for temperatures above 20 • C (Fortes and Lüning, 1980;Bolton and Lüning, 1982;Lüning, 1984) and blade tissue of S. latissima from Nova Scotia was severely damaged and decomposed after 1-week exposure to high temperatures of 18 and 21 • C (Simonson et al., 2015b). Contrary to these studies, samples from Helgoland and Locmariaquer in our experiment survived a temperature increase by +2 and +4 • C (AT up to 23 • C) and revealed high mortality only in the +6 • C treatment (AT ≥ 24 • C) after 8 days of treatment. Thus, S. latissima from these two locations showed extraordinary resilience to MHWs in summer. The application of a stepwise increase in temperature, despite done within a short timescale, very likely allowed the specimens from Helgoland and Locmariaquer to acclimate and survive up to +4 • C (Terblanche et al., 2007). Thus, we can conclude that S. latissima at their warm-edge distribution can withstand short periods of summer MHWs up to AT 25 • C, although mortality 2 | Survival, chlorophyll a (Chl a), pool-size of accessory pigments (Acc) and pool-size of xanthophylls (VAZ), Chla:Acc ratio and Chla:VAZ ratio of S. latissima across latitudes in Europe after the temperature-amplitude experiment.  already occurred and increases at AT > 23 • C after 1 week. As it has been shown that increased nitrogen supply supports heat tolerance in kelp (Gerard, 1997;Fernández et al., 2020), the application of 1 /2 PES (274 µmol NO 3 − l −1 ; Sarker et al., 2013) in our study could have ameliorated the negative impacts of enhanced temperatures. Moreover, we monitored    survival of excised meristematic discs, although kelp thalli under heat stress usually start degenerating from the tip (Franke et al., 2021). Furthermore, it has to be considered that adult meristematic tissue collected from the field might be more resilient than laboratory cultures or young sporophytes (Hanelt et al., 1997;Heinrich et al., 2016).
Along the entire latitudinal range, growth of the surviving discs did not differ based on the temperature-amplitude treatments but differed between the populations. In contrast to previous studies on Arctic S. latissima working with comparable absolute temperatures [Olischläger et al., 2017 (AT 4 and 10 • C), Li et al., 2020 (AT 0, 8, and 15 • C), Diehl and Bischof, 2021 (AT 4, 6, 8 and 10 • C)], we did not detect enhanced growth with increasing temperatures in the samples from Spitsbergen. Strong site-specific differences in growth became evident in our study. An overall decreased growth from north to south was determined but was independent of absolute temperatures in general. For instance, at AT 12 • C samples from Spitsbergen ( +6 • C) revealed markedly higher growth activity than samples from Bodø ( ±0 • C). Growth in samples from Spitsbergen was highest by far in all treatments, even though the control temperature ( ±0 • C, AT 6 • C) and the +2 • C treatment (AT 8 • C) were below the reported optimum growth temperature range of 10-15 • C (Bolton and Lüning, 1982). Growth was significantly lower in samples from Helgoland (AT 18-24 • C) and Locmariaquer (AT 19-25 • C), which were both exposed to temperatures exceeding the optimum growth temperature of the species. Even at ambient temperature ( ±0 • C) and lower amplitude of temperature increase, samples from Locmariaquer recorded the lowest growth. This suggests that S. latissima in this region is already exposed to the proximate upper temperature tolerance limit and occurrence of MHW can be detrimental to the population. As we collected all meristems in June to early July, according to Lüning (1979) meristems should still have been in some kind of growth mode. As growth activity differs with day length (Fortes and Lüning, 1980), and our latitudinal collection gradient is characterized by a steep day length difference, it might have been that the physiological growth activity state was different besides locations despite sampling at a uniform time period. However, our treatment conditions always applied the same day length condition and thereby it can be excluded that the experimental light setup had an influence on growth. Iken (2011, 2012) suggest that S. latissima exhibits growth plasticity within a genetically fixed seasonal growth window that might be an advantage in locations with high environmental variability, such as the Arctic. The particular irradiance conditions in the Arctic (Polar Day) prior the experiment may have promoted the extraordinary growth activity of the Spitsbergen population, whereas shorter day lengths may have decelerated growth of S. latissima from lower latitudes (Fortes and Lüning, 1980). Consequently, observed growth activity might have been more dependent on internal growth patterns than on their exposure temperatures during the experiment, resulting in S. latissima from the Arctic growing most and from Locmariaquer least in summer. The causes for the observed differences in growth patterns of European S. latissima requires further investigation.
Temperature treatment corresponding to the absolute temperature rather than the respective temperature amplitude treatments affected the optimum quantum yield of S. latissima. For example, samples from Spitsbergen were slightly stimulated at the highest temperature amplitude treatment of +6 • C (AT 12 • C), but samples from warmer locations decreased under +6 • C treatments (AT > 20 • C). Several studies already reported an increase in optimum quantum yield of wild Arctic S. latissima sporophytes exposed to increasing absolute temperatures up to 10 • C (Iñiguez et al., 2016;Diehl and Bischof, 2021). On the other hand, a decrease in F v /F m was reported among populations from southern Norway and Denmark (Andersen et al., 2013;Nepper-Davidsen et al., 2019). Our study showed that regardless of origin, all S. latissima exhibited relatively high F v /F m values (> 0.6) suggesting a healthy physiological status under MHW conditions of absolute temperatures of up to 20 • C. This suggests that seaweeds can withstand short-term increase in SST during summer across latitudes. It is striking that S. latissima from Locmariaquer recorded higher photosynthetic quantum yields compared to Helgoland, even though both populations were exposed to comparable high temperatures. Locmariaquer samples maintained a very good physiological status at AT 23 • C (> 0.65), while F v /F m values for Helgoland samples already declined at AT 22 • C (0.605). This response is similar to that observed for disc survival. If these responses are based on local adaptation or different phenotypic plasticity between the populations, requires further examination.
Saccharina latissima and other kelps exhibit different biochemical mechanisms to acclimate to temperature variations. In general, the effects of temperature, e.g., in the environment or during laboratory experiments, on pigment content and composition were reported to be very complex (e.g., Machalek et al., 1996;Fernandes et al., 2016). However, we did not detect a clear link between variation in pigment concentrations and the species latitudinal distribution range or exposure to absolute temperature treatments. The tested temperature amplitudes only revealed significant impact of the +6 • C treatments. In particular, when surpassing the sublethal thresholds (Bolton and Lüning, 1982;Lüning, 1984), +6 • C imposed considerable stress in S. latissima from Helgoland and Locmariaquer (AT: > 24 • C). During temperature stress, the interconversion of violaxanthin to zeaxanthin (de-epoxidation state of the xanthophyll cycle, DPS) acts as precursory protective mechanism for the dissipation of excessive energy, thereby preventing the generation of reactive oxygen species (ROS) (Goss and Jakob, 2010). Comparable to Nepper-Davidsen et al. (2019) we measured a strong increase of DPS at high temperatures in samples from Helgoland and Locmariaquer, indicating the exertion to counteract negative impacts of ROS. The increasing trend in the pool size of the xanthophyll cycle pigments (VAZ) in the same treatments reflects the importance of DPS as a protective mechanism at high temperatures. Changes in the pigment composition within the photosynthetic apparatus, for instance a decrease in chlorophyll a (Chl a) in response to suboptimal temperatures, is associated to a consequential reduced photosynthetic activity (Wernberg et al., 2016b;Nepper-Davidsen et al., 2019). Interestingly, we did not observe a significant decrease in Chl a under different temperatures in our study. Still, the combination of slightly increasing VAZ and decreasing Chl a resulted in decreases of the Chla:VAZ ratio at +6 • C with potential negative implications for the functioning of the photosynthetic apparatus of Helgoland and Locmariaquer populations.
Phlorotannins, which act as antioxidants and protection against ROS (Amsler, 2008), were not affected by the temperature-amplitude treatments. In accordance with the study by Simonson et al. (2015a), they did not increase significantly at high temperatures as was observed in other brown algae, e.g., Agarum clathratum, but strong differentiations across populations were found. Significantly fewer phlorotannins were measured in the samples from Spitsbergen and the concentrations were comparable to those of Arctic S. latissima in a previous study (Diehl and Bischof, 2021).
The storage carbohydrate mannitol is the main photoassimilatory product in brown algae (Reed et al., 1985). We did not find any temperature effects or differences in mannitol content across latitudes. Diehl and Bischof (2021) proposed that Arctic S. latissima may constantly store high mannitol concentrations as a mechanism to acclimate to frequent environmental variation, which we cannot verify in the present study. In their experiment, Diehl and Bischof (2021) mimicked Arctic summer conditions and maintained the samples under constant 24 h of light exposure. For our current study, we kept all samples at a 16:8-h-light:dark rhythm to eliminate potential light effects at the different latitudes, which might have influenced the synthesis of mannitol.
Nutrient uptake in seaweeds is usually affected by temperature (Roleda and Hurd, 2019). In our experiment, the C:N ratio did not vary in the temperature-amplitude treatments and no pattern of intracellular nitrogen (N) was detected across latitudes. Generally, local variations in C:N were very small and mainly based on changes in total N rather than total carbon (C). N is frequently limiting in coastal systems over summer. Still, only samples from Helgoland revealed C:N ratios of above 20, indicating N limitation (Atkinson and Smith, 1983) despite high supply with N ( 1 /2 PES) during the experiment.

Implications for the Population Biology of Saccharina latissima in Europe
Even though S. latissima tolerates wide amplitudes of different physico-chemical drivers (reviewed by Bartsch et al., 2008), strong declines and changes in biomass were observed from several regions in Europe (Pehlke and Bartsch, 2008;Moy and Christie, 2012;Casado-Amezúa et al., 2019;Filbee-Dexter et al., 2020). Heat tolerance and the capacity for acclimation may differ between genetic strains, thus, populations (Clark et al., 2013). Our study shows that local MHWs in summer only pose stress to S. latissima in regions with higher mean summer sea surface temperatures. Overall, the thermal responses of all five populations sampled along the latitudinal gradient were dependent on the experienced absolute temperatures during the experiments. However, population genetics revealed low connectivity among S. latissima populations in Europe within short distances (Guzinski et al., 2020). Thus, analyses of only one population per site may not be fully representative of the latitudinal range.
Our study is short-term and worked on excised discs of sporophytes incubated in small-volume vessels in the laboratory; therefore, has some limitations. For instance, transportation and different pre-acclimation periods could have affected the sensitivity toward temperature of the samples. Long-term studies on whole sporophytes in outdoor mesocosms may measure different responses and provide a more realistic prognosis for the future. Moreover, our experiments were run in different years (2018, 2019, and 2020) but consistently during summer months (June and July); hence, inter-annual variation in the quality of sporophytes may be imminent but cannot be addressed. Rearing temperature of sporophytes is known to affect the photosynthetic performance of S. latissima (Davison, 1987;Davison and Davison, 1987;Davison et al., 1991), and the importance of temperature history on thermal plasticity of kelps was recently shown by Liesner et al. (2020b). They provided evidence for nongenetic carry-over and cross-generational effects and reported a stimulating effect of cold temperatures during early life stages on several physiological and biochemical responses to high temperatures. Rapid and intense ocean warming does not only occur in summer but has also been detected in winter and is predicted to further increase in the near future (Maturilli et al., 2013;IPCC, 2019b). This might additionally affect the thermal responses and stress resilience of S. latissima to MHWs in summer and should be addressed in future studies. Furthermore, it has to be considered that also at suboptimal temperatures, specimens might be less resilient to other stressors or less competitive against other algae (Wernberg et al., 2010;Moy and Christie, 2012;Andersen et al., 2013;Filbee-Dexter and Wernberg, 2018).
Summarizing, European S. latissima exposed to realistic local summer MHW events reveals that stress responses are rather independent from temperature amplitudes, but mostly dependent on the absolute temperatures experienced across latitudes, especially when surpassing sub-lethal and lethal limits for a period of 8 days. Thus, the increase in frequency and amplitude of MHWs will considerably stress southern S. latissima, causing further decline of rear-edge populations while northern populations stay safe for the time being.

DATA AVAILABILITY STATEMENT
The datasets analyzed for this study can be found in the PANGAEA Database (https://doi.org/10.1594/PANGAEA. 931637).

AUTHOR CONTRIBUTIONS
ND and KB planned the experiments. MR helped collecting in Bodø and helped designing the experimental set-up. ND conducted the experiments and analyzed the data. UK and IB provided the facilities for experiments and biochemical analyses and helped with interpretation. IB supported the experimental procedure. KB supervised the project. ND wrote the manuscript, which was reviewed, revised, and approved by all co-authors. All authors contributed to the data interpretation and discussion.