Seasonal ecophysiology of Fucus vesiculosus (Phaeophyceae) in the Northern Baltic Sea

ABSTRACT The brown macroalga Fucus vesiculosus is a foundation species in temperate rocky shores, subjected to seasonally fluctuating environmental conditions. To obtain a more complete picture of the seasonality of F. vesiculosus ecophysiology in the northern Baltic Sea, in situ photochemistry, elemental ratios and chlorophyll a and c content of the alga were investigated in field campaigns conducted in different months throughout the year during 2017. Carbon, nitrogen, carbon to nitrogen ratio and chlorophyll a and c content of the alga varied substantially throughout the year, with highest carbon content observed in summer, and highest nitrogen content in winter. C:N ratio in F. vesiculosus apical tissue ranged from 8.6 in February to 48.3 in July. Chlorophyll a and c content followed inversely the seasonal patterns of ambient irradiance. High chlorophyll a and c content in winter was associated with higher maximum photosynthetic efficiency of energy conversion (Fv/Fm), but not with efficiency of photosynthetic energy conversion under light limitation (α). Electron transport rate correlated strongly with seawater temperature, and the highest electron transport rates were observed in summer and correlated with highest internal carbon content of the alga. Redundancy analysis conducted on measured environmental variables against physiological responses identified day of year, temperature and macronutrients in seawater as the most important variables driving the observed seasonal patterns in F. vesiculosus ecophysiology. The results suggest elevated temperatures may increase Fucus growth and photosynthesis rates in the study area. Highlights Two distinct physiological states of Fucus vesiculosus were identified. The most important variables driving responses were season, temperature and macronutrients. Fucus ecophysiology shows substantial seasonality in the northern Baltic Sea.


Introduction
The brown macroalga Fucus vesiculosus L. is a foundation species in the rocky temperate ecosystems of the Northern Atlantic and the Baltic Sea, extending its distribution into the Arctic (Kalvas & Kautsky, 1998;Berger, 2009) In the Northern Atlantic, F. vesiculosus (hereafter Fucus) forms characteristic ecological communities with other fucoids, especially Ascophyllum nodosum and Fucus serratus (Jenkins et al., 1999;Høgslund et al., 2014;Burel et al., 2022), while in the Baltic Sea, large-scale absence of other large perennial macroalgae, attributed to low salinity (Russell, 1988) emphasizes the ecological importance of Fucus (Kautsky et al., 1992).
Fucus habitats harbour a rich assemblage of associated species such as macro-and microalgae, invertebrates and fish Wikström & Kautsky, 2007;Schagerström et al., 2014) and also act as an important carbon sink in coastal ecosystems (Attard et al., 2019a(Attard et al., , 2019bDuarte et al., 2022), potentially sheltering co-occurring species against environmental changes such as ocean acidification (Wahl et al., 2018). In contrast to annual, fastgrowing filamentous macroalgae, Fucus beds provide biomass and habitat complexity throughout the year (Kraufvelin & Salovius, 2004).
Through most of its geographic range, Fucus grows in the intertidal zone, being subjected to substantial environmental fluctuations, especially when exposed from water (Zaneveld, 1937). Temperature tolerance of Fucus is wide, and shows seasonality related to ambient temperature (Parker, 1960).
In the non-tidal Baltic Sea, Fucus grows permanently submerged, and has evolved to non-tidal conditions over several thousands of years (Russell, 1985), resulting in reduced tolerance to desiccation (Rothäusler et al., 2016), increased tolerance against low salinities and lower temperature tolerance than the Atlantic populations (Russell, 1985(Russell, , 1987Bäck et al., 1992;Nygård & Dring, 2008). In the northern Baltic, Fucus individuals are subjected to seasonal fluctuations in temperature, macronutrients, irradiance and pH (Raven & Samuelsson, 1988;Lindström, 2000;Leppäranta & Myrberg, 2009;Saderne et al., 2013). The photosynthesis rates of Baltic Fucus are low compared with Atlantic populations (Raven & Samuelsson, 1988), while the Baltic populations have lower optimum temperature for photosynthesis (Nygård & Dring, 2008). Highest precipitation in the study area occurs in autumn (Omstedt et al., 1997), and seasonal salinity fluctuations, driven by riverine inflow, are rather minor (Alenius et al., 1998). The Baltic Fucus show distinctive annual patterns in growth, metabolism and reproduction (Lehvo et al., 2001;Al-Janabi et al., 2016a;Attard et al., 2019aAttard et al., , 2019b. In the central and southern Baltic, Fucus has two distinctive reproductive periods, one in May-June and a second in September-November (Berger et al., 2001), while in the area where this study was conducted, only a single reproductive period has been observed, in June (Lehvo et al., 2001). In the study area, the growth of reproductive branches initiates in winter, peaking in April (Lehvo et al., 2001). Highest precipitation rates in the study area occur in late summer, in July/August and are lowest in spring (Irannezhad et al., 2014), outside the reproductive period of Fucus.
Fucus maintains positive growth rates through the year in the northern Baltic, despite low irradiance levels during winter, by utilizing internal mannitol pools as an energy reserve (Lehvo et al., 2001). Highest growth rates occur during summer, and winter growth rates are one-third of summer growth rates (Lehvo et al., 2001).
Laboratory and field experiments have identified several abiotic factors which affect Fucus ecophysiology. Low salinity affects the size and morphology of Baltic Fucus (Ruuskanen & Bäck, 1999, with individuals growing smaller and more branched towards lower salinity. This may be caused by even short exposures to low salinity during the growing season (Ruuskanen & Kiirikki, 2000). Also, Fucus tissue chlorophyll content decreases with declining salinity, but C:N ratio increases, which is associated with declining palatability of algal tissue (Barboza et al., 2019). In contrast, increases in salinity, macronutrients and carbon availability promote photosynthesis and growth in Fucus (Nygård & Dring, 2008).
Previous studies have found internal nitrogen content of F. vesiculosus to vary substantially through the year, with the highest nitrogen content occurring during winter, when seawater nitrogen availability is high (Lehvo et al., 2001). In the western Baltic Sea and in the Atlantic, the highest carbon content in Fucus has been observed in summer (Bordeyne et al., 2015;Graiff et al., 2015a). Chlorophyll content of Fucus thalli varies through the year, with highest chlorophyll concentrations observed in winter (Leskinen et al., 1992).
Macroalgal carbon to nitrogen (C:N) ratios may affect herbivory, as grazers feeding on algal material with suboptimal elemental ratios may have to increase the rates of grazing to obtain sufficient amounts of limiting element (Hemmi & Jormalainen, 2002Urabe et al., 2002). Because of this, the dynamics of Fucus elemental composition may have cascading effects also on higher trophic levels and may affect the seasonal patterns of the intensity of herbivory.
Given its ecological importance, Fucus has been intensively studied for several decades (as reviewed by Takolander et al., 2017a). However, a multivariate statistical investigation into the most important environmental factors affecting Fucus ecophysiology has been missing. Quantifying the seasonal fluctuations of different ecophysiological parameters is important, as Fucus is frequently the subject of experimental studies investigating the effects of environmental changes such as climate change (Takolander et al., 2017b(Takolander et al., , 2019Rothäusler et al., 2018;Rugiu et al., 2018;Milec et al., 2022) and ocean acidification (Graiff et al., 2015a;Al-Janabi et al., 2016a or their combination (Werner et al., 2016a(Werner et al., , 2016bGraiff et al., 2020).
The aim of this observational study is to investigate the seasonal changes in Fucus ecophysiology (in situ photobiology, chlorophyll content and elemental ratios) in the northern Baltic Sea, and to identify which of the most important environmental variables (temperature, irradiance, salinity, macronutrients, pH and carbon availability) are associated with seasonal changes in Fucus ecophysiology.

Fieldwork
Fucus vesiculosus ecophysiology was investigated with in situ underwater chlorophyll fluorescence measurements and laboratory analyses (chlorophyll a and c, carbon and nitrogen content) conducted on sampled specimens. Fucus samples were collected with diving in five fieldwork campaigns in 2017. In each site, five randomly selected Fucus individuals were sampled from the Fucus bed. A thallus tip (~10 cm from the tip) was sampled from each specimen. The tip is the metabolically most active tissue in Fucus, whereas photosynthesis of basal parts is considerably lower (Leskinen et al., 1992). In addition, the basal parts of the alga, having lower phenolic content (Tuomi et al., 1989), are frequently almost totally consumed by the main herbivore, Idotea balthica.
Fieldwork was conducted in the northern Baltic Sea, near Tvärminne Zoological Station (TZS), south-west Finland in 2017. Field trips were conducted in February, May, July, September and November in two sites: Ångbåtsbryggan and Spikarna. In February only Ångbåtsbryggan was visited as Spikarna was unaccessible due to sea ice. In addition, in July 2017 three additional sites in the inner/middle archipelago were included: Ekö, Björnholmen and Danskog (Fig. 1), to obtain more comprehensive data on Fucus ecophysiology along the archipelago gradient during the main growing season. For qualitative estimation of density of Fucus individuals on each site, approximate Fucus cover was visually evaluated by the diver.
Ångbåtsbryggan is located in the immediate vicinity of TZS, in the middle archipelago zone, and is a shallow (4 m) strait with substantial water flow. The deepest bottoms are covered by sand, with a meadow of Zostera marina. Here Fucus grows attached to hard bottoms around small skerries that reach the surface, forming a dense belt where suitable substrate is available. The growing depth of Fucus is approximately from 0.5-2 m. The approximate cover of Fucus on this site ranges from 40-100%.
Spikarna is a shallow skerry in the outmost archipelago zone. It has a stony lagoon 2-4 m deep, which is covered with dense macroalgal vegetation comprising mostly of Fucus (Attard et al., 2019a). The Fucus cover on this site is 100%, as the vegetation covers the lagoon bottom completely. Thallus tips were sampled from the fronds closest to the surface, to avoid sampling fronds that would be subjected to self-shading. The Spikarna lagoon has a narrow passage which opens south into the outer sea.
Danskog is a sheltered and shallow site in the inner archipelago. Fucus grows attached to rocks close (1 m) to the surface. In addition to attached specimens, the site has free-living Fucus which grow unattached on a sandy bottom surrounding the boulders in 1-1.5 m depth. Fucus individuals on this site were sampled from the attached specimens. The substrate cover of attached Fucus individuals was 90%.
Ekö is a semi-sheltered shore in the middle archipelago zone. In this site Fucus grows attached to boulders in 1-1.5 m depth. Fucus individuals occur sparsely, attached on suitable substrate, when available. Approximate cover of Fucus on this site was 10%.
Björnholmen is also a semi-sheltered shore in the middle archipelago zone but located more inland compared with Ekö. The salinity and nutrient gradients existing in the study area, caused by freshwater inflow from the Mustionjoki River into the Pohjanpitäjänlahti Bay, are strongest at this site, as it is located closest to the river. In this site the densest belt of Fucus lies within 1-1.5 m depth. The approximate cover of Fucus individuals on this site, at the depth range of densest belt, is 20-40%.

Environmental parameters
In each site, salinity and temperature were measured at the surface water with a YSI EC 300 conductivity meter at the time of sampling. Irradiance was monitored with a LICOR-LI 1500 underwater light sensor, which was lowered to the bottom with a rope from the boat. Measurements were performed between 10:00 am and 15:00 pm, depending on when each site was visited. In average, 11 replicate irradiance measurements were taken from the surface and then with 1 m intervals down to the bottom. Irradiance at Fucus sampling depth was estimated with linear regression. Due to technical problems, irradiance measurements could not be conducted in the July sampling campaign at Ekö. Temperature, salinity and pH logger (YSI EXO 2 Multiparameter Sonde) data for Ångbåtsbryggan were provided by Tvärminne Zoological Station, covering the entire year 2017 (Fig. 1). Logger data were used only for visualizing seasonal trends in salinity, temperature and pH in Fig. 1, while the on-site instantaneous measurements of environmental data were used in the statistical analysis.
Water samples for determining seawater nitrite and nitrate (NO 2 -+ NO 3 -) and phosphate (PO 4 3-) concentration were collected from the surface water into acid-washed airtight polyethylene bottles. Samples were placed in the cool and dark, and transported to the laboratory, where they were frozen. Nutrient concentrations were determined from the samples by TZS laboratory according to the method of Koroleff (Koroleff, 1983a(Koroleff, , 1983b. To determine seawater pH, total alkalinity, dissolved inorganic carbon (DIC) and pCO 2 , five replicate samples from surface water were collected to machine-washed glass bottles (V = 100 ml) with an airtight seal. Samples were kept in the dark and transported into laboratory. They were stored in the dark to equilibriate to room temperature, after which the determination of alkalinity was performed by titrating 50 ml of sample with 20 or 100 mmol HCl to pH 4.5 (Finnish Standards Association, 1981) which corresponds to ISO standard NS 4754: ISO 9963-1:1994 with 2 µmol-l accuracy. pH of the samples was measured with a tabletop pH meter (Jenway 3510), calibrated with commercial NIST-traceable buffers (Merck Certipur®) using 3-point (pH 4, 7 and 10) calibration. DIC and CO 2 partial pressure in seawater (pCO 2 ) were calculated from temperature, salinity, pH and alkalinity data using R package 'seacarb' (Gattuso et al., 2021). The rela-tionships between different carbon system components of the Baltic Sea have been evaluated by Omstedt et al. (2010).

Ecophysiological parameters measured
To quantify the photosynthetic status of the algae, several chlorophyll fluorescence parameters were measured using a DIVING-PAM pulse-amplitude modulated fluorometer (Walz GmbH, Germany) First, a rapid light curve (RLC) measurement (Ralph & Gademann, 2005), with 20s actinic light illumination time for each light increment, followed by saturating pulse to measure fluorescence (Maxwell & Johnson, 2000) was performed in situ by a diver on a vegetative Fucus thallus tip free of epiphytes.
As the absorbance of the thallus tips was not measured, maximum electron transport rates are given as relative maximum electron transport rate (rETR max ) (Ralph & Gademann, 2005). The rETR vs irradiance curves were fitted to a light curve model (Platt et al., 1981) using R package 'phytotools' (Silsbe & Malkin, 2015). The rETR max and alpha were calculated from the fitted model. Alpha represents the linear slope of the rETR vs. irradiance curve in the light-limited region of the curve and represents the efficiency of photosynthetic energy conversion in conditions where photosynthesis is light-limited. Relative maximum electron transport rate is an approximation of the maximum rate of electrons pumped through the electron transport chain (Beer et al., 2000(Beer et al., , 2001Hawes et al., 2003;Ralph & Gademann, 2005).
After the RLC measurement, the thallus tip was harvested by the diver and placed into a mesh bag. After the dive, the specimens were kept submerged in a seawater container, and each thallus was darkadapted for 15 min with a dark leaf clip, followed by measurement of maximum potential quantum yield (Fv/Fm). After this, the algal samples were divided into Eppendorf tubes and frozen (−80°C) for elemental content and chlorophyll analysis. Five algal thalli from randomly selected separate algae individuals were measured and collected from the sampled sites during each sampling trip.
Carbon and nitrogen content was analysed from the frozen algal samples using LECO TruSpec Micro CHNS elemental analyser. Prior to measurements, samples were dried (24 h at 60°C) and ground with mortar and pestle. Chlorophyll content was analysed from frozen samples with a Shimadzu UV-2550 Spectrophotometer (1 nm precision). A small piece of frozen algae was ground in the dark with a pestle and mortar in liquid nitrogen. Chlorophyll was extracted for 4 h in 99% ethanol solvent. Chlorophyll a and c content were calculated from the absorption spectra following the equations provided by Ritchie (2006).

Statistical analyses
The effects of environmental variables on the ecophysiological parameters measured were analysed with redundancy analysis (Ter Braak, 1987, 1994. More specifically, the explanatory variables used were salinity, temperature, sampling site, dissolved inorganic carbon (DIC), day of the year at the time of sampling, nutrients (NO 2 -+ NO 3 and PO 4 3-) and irradiance. The response variables analysed were alpha, rETRmax,  ), and with dissolved inorganic carbon (DIC) and temperature. Because of this, a composite 'nutrients' variable was created from NO 2 -+ NO 3 and PO 4 3by first standardizing the two variables to zero mean and unit variance and then calculating mean pooled nutrient concentration for each observation. To reduce collinearity of explanatory variables, DIC was omitted from the statistical analysis, which resulted in variance inflation factors between all explanatory variables to remain below 10.
Due to technical issues, irradiance data from site Ekö and (Fv/Fm) values from Björnholmen, Danskog and Spikarna in July could not be obtained. Because of this, all data from Ekö site were omitted from statistical analyses, resulting in 55 observations in total in the analysis. The missing July (Fv/Fm) values for Björnholmen, Danskog and Spikarna sites were imputed from other ecophysiological variables measured together with environmental variables using R package 'Amelia' (Honaker et al., 2011).
The distribution of the response variables was first analysed with detrended correspondence analysis (Hill & Gauch, 1980). As the length of the axis was below 0.6 standard deviations, the responses of ecophysiological parameters to explanatory variables were assumed to be linear. Linearity was further ensured by plotting all response variables against all explanatory variables. As many of the response variables showed non-linear patterns when plotted against day of year, this explanatory variable was first normalized to zero mean and unit variance and then cosine-transformed to ensure linear relationship in relation to the response variables. After these steps, a redundancy analysis (RDA) (Ter Braak, 1987, 1994 was conducted to assess the effects of explanatory variables to the response variables. RDA was conducted with a forward stepwise model selection, i.e. adding explanatory variables into the model one at a time. The significance of the explanatory variables added to the model were assessed with a Monte Carlo permutation test with 10 000 permutations (Legendre & Legendre, 1998). To account for the effects of performing multiple statistical tests, significance levels of the p-values (originally p = 0.05) were adjusted with Bonferroni correction by dividing the original significance level by the number of tests performed, resulting in a significance threshold of 0.007 for rejecting the null hypothesis (Armstrong, 2014). Total adjusted R 2 of the full RDA model (model containing all explanatory variables) and the Bonferroni adjusted significance threshold were used as stopping criteria for the addition of variables into the model (Blanchet et al., 2008).
The importance of explanatory variables (explained variance) was investigated with marginal effects (the ratio of eigenvalue using only one explanatory variable to the eigenvalue when using all variables) and conditional effects (increase in eigenvalue when adding a new variable in forward model selection) (Zuur et al., 2007), and with the Monte Carlo permutation test (Legendre & Legendre, 1998).

Seasonality of environmental parameters
Environmental parameters measured showed substantial seasonal variation throughout the year (Tables 1 and 2). Seawater temperature, measured at the surface at sampling sites, ranged from 0.5-15.7°C. Seasonal changes in salinity dynamics and the carbonate system parameters (CO 2 , CO 3 2-, HCO 3-) were more conservative, whereas irradiance and nutrients in seawater, especially nitrogen (NO 2 -+ NO 3 -) fluctuated substantially, peaking in February at 73 µg l -1 and subsequently dropping to 3 µg l -1 for the remainder of the year (Tables 1 and 2).
Although turbidity or Secchi depth were not measured, the irradiance conditions varied substantially between the sampling sites, which was caused by differences in water transparency, somewhat obscuring the seasonal variation in irradiance. In particular, the Danskog sampling site, located in the inner archipelago ( Fig. 1) had very high turbidity, and consequently the lowest irradiance values of all sites (Table 1) although sampled in July, when the ambient irradiance is highest (Lindström, 2000).
The minimum pH, 7.87, was measured in February at Ångbåtsbryggan, and the maximum, 8.34, at Ekö in July (Table 2).

Statistical relationships between environmental and ecophysiological parameters
The results of RDA permutation test showed three environmental parameters had a statistically significant effect on Fucus ecophysiology (see Fig. 2 for the parameters included in 'ecophysiology'): day of the year (F = 35.23, p < 0.001), temperature (F = 16.96, p < 0.001) and nutrients (F = 3.77, p = 0.002) ( Table 3). These were also the variables with the highest marginal and conditional effects (Table 3).
The RDA ordination built with these three explanatory variables explained 57.8% of variation in the response variables. The first two RDA ordination axes were significant, and explained 45.9% and 10.3% of variation, summing to 56.2% of explained variance. More specifically, rETR max was highly correlated with seawater temperature, whereas in situ measured irradiance was highly correlated with carbon content of the alga (Fig. 2). Two specific physiological states could be identified: a summer state with high electron transport rate, high carbon content and low chlorophyll and N content, and a winter state characterized by high chlorophyll and nitrogen content, both of which co-occurred with higher levels of maximum photochemical yield, (Fv/Fm), reflecting an adaptation of photosynthetic complex to more efficient light capture in winter under low light conditions.

Seasonality of ecophysiological parameters
All ecophysiological parameters measured showed strong seasonality, with somewhat complementary patterns (Fig. 3). Chlorophyll a and c content reached their minimum in July, and maximum in February, the month with lowest ambient irradiance. Nitrogen content of the alga followed this pattern. The maximum photosynthetic efficiency (Fv/Fm) reached its peak in February and November, when ambient irradiance was low, and declined somewhat in July. Light-limited photosynthetic efficiency (α) did not show as clear a pattern.
In contrast, photosynthetic activity, measured as maximum electron transport rate, was highest in summer, when temperature and irradiance peaked (Figs 1, 3). This was also reflected in carbon content of the alga, which had highest values in July and then declined towards the end of the year. C:N ratio of the alga followed closely that of rETR max , peaking in July and declining towards winter (Fig. 3).

The effects of temperature on Fucus ecophysiology
Temperature had a strong effect on the ecophysiological parameters measured. More specifically, chlorophyll content declined substantially with increasing temperature. Chlorophyll a content declined more rapidly compared with chlorophyll c, which caused also declines in chlorophyll a/c ratio. Nitrogen content of the alga and (Fv/Fm) declined in high temperatures (Fig. 4), but this correlation is most likely caused by seasonal nitrogen dynamics, where nitrogen content is low in summer when temperature and irradiance are high (Fig. 3). In contrast, rETR max increased with increasing temperature. Carbon content remained unaffected but declines in nitrogen content caused increases in C:N ratio in higher temperatures (Fig. 4).

Discussion
All ecophysiological parameters measured showed substantial seasonal fluctuations. These were especially pronounced in chlorophyll, carbon and nitrogen content of the alga, and maximum electron transport rate. The ordination results of redundancy analysis identified two distinct physiological states, which correlated mostly with RDA ordination axis 1 (Fig. 2). During summer, Fucus maintained high photosynthesis rates under high irradiance and temperature. Summer is also the period of most rapid growth (Lehvo et al., 2001). High photosynthesis (measured as rETR max ) rates correlated with high internal carbon content, potentially stored as mannitol, which is the main storage compound in Fucus (Bidwell & Ghosh, 1962), and shows similar seasonal fluctuations as carbon content in this study (Lehvo et al., 2001). During summer, both the internal nitrogen reserves of the alga and chlorophyll content were low. As the summer samples were taken in July after the reproductive period, it is possible that some of the internal nitrogen reserves were allocated and lost to reproduction.
In winter, low temperature and irradiance suppressed electron transport rates, but chlorophyll and nitrogen content of the alga were high. This was accompanied by higher (Fv/Fm) values, which suggests that the algae invested nitrogen into building up photosynthetic efficiency under low irradiance in winter conditions (Leskinen et al., 1992). Presumably increased photosynthetic energy transfer efficiency, together with energy from internal storage carbon compounds, would help Fucus to maintain positive growth rates over the course of the year (Lehvo et al., 2001). In this study, the observed carbon content of the alga in winter was low as internal energy reserves were used to support vegetative growth (Lehvo et al., 2001;Meichssner et al., 2021). The initiation of productive structures occurs in October (Lehvo et al., 2001), and maturation continues through the winter in low light conditions and is likely also sustained by internal energy reserves. However, the low rETR max values in winter suggest that growth rate during winter was most likely very slow, as ETR max has been shown to be Day of year C/N ratio a good proxy for Fucus growth rates (Nygård & Dring, 2008). In contrast, internal nitrogen content was highest in winter and spring (Fig. 3). Fucus stores inorganic nitrogen internally to be utilized in summer when higher temperature and irradiance allow for more rapid growth rates but nutrient concentration in seawater is low (Lehvo et al., 2001;Young et al., 2009). Brown macroalgae have generally lower internal protein content compared with red or green algae, and consequently higher C:N ratios, with mean C:N ratios around 20 (Fiset et al., 2019). In this study, carbon to nitrogen ratios showed substantial seasonal variation, ranging from 8.6 in February to 48.3 in July. This was mainly driven by seasonal variations in internal nitrogen content and C:N ratio showed similar seasonal patterns with maximum electron transport rate (Fig. 3). Increasing nutrient availability has been shown to lower the total carbon content and increase the amount of insoluble sugars in Fucus, increasing grazing by herbivores (Hemmi & Jormalainen, 2002). In contrast, high carbon:nitrogen ratios may reduce herbivory through stoichiometric constraints (Urabe et al., 2002). The results of this study suggests that the quality of Fucus as a food resource for herbivores varies substantially by season. The highest C:N ratios, which were observed in July, imply that Fucus is a suboptimal food resource for grazers in summer. This interpretation is supported by experimental results showing amphipod grazers having high preference for filamentous algae (presumably with lower C:N ratio) over F. vesiculosus in September (Goecker & Kåll, 2003), as well as field observations in the study area showing macroinvertebrate grazers aggregating in the filamentous Cladophora glomerata zone instead of the Fucus zone in summer (Salemaa, 1979;Kraufvelin & Salovius, 2004). In the northern Baltic, Fucus is subjected to the most intensive grazing in late summer/autumn, when a new Idotea baltica generation migrates to the Fucus zone, causing grazing of up to 80% of annual algal tissue growth (Korpinen et al., 2010). Low tissue nitrogen content in Fucus, which was observed to occur in summer, has been associated with higher concentrations of phenolic compounds such as phlorotannins (Yates & Peckol, 1993;. Phlorotannins reduce the food assimilation efficiency of the main herbivore, Idotea baltica (Jormalainen et al., 2005), potentially increasing feeding rates. This suggests that the herbivory impacts of juvenile Idotea in late summer may be further intensified by suboptimal chemical composition of the algal tissue grazed, contributing to high grazing rates in late summer. Temperature had a substantial effect on photosynthetic activity (measured as maximum electron transport rate), which declined rapidly under low temperatures (Fig. 4), indicating that photosynthesis rates (and consequently growth rates) in winter and spring were low. This was expected, as many of the molecular components of the electron transport chain are temperature-dependent (Venkataramanaiah et al., 2003;Allakhverdiev et al., 2008). On the other hand, maximum electron transport rate measurements might not fully capture the in situ metabolic patterns in a Fucus canopy. ETR has been shown to be a good proxy for photosynthesis rates under moderate irradiances, but the relationship breaks down under high irradiance (Beer & Axelsson, 2004). Recent aquatic eddy covariance measurements have shown Fucus beds to be net autotrophic almost throughout the year, with high net photosynthesis rates even under low temperature and irradiance in winter (Attard et al., 2019a(Attard et al., , 2019b. Due to collinearity, DIC was omitted from the redundancy analysis. DIC was negatively correlated with temperature (R 2 = −0.83) and positively correlated with nutrients in seawater (R 2 = 0.54). Seawater DIC concentration has been shown to have a positive effect on Fucus growth and electron transport rates (Nygård & Dring, 2008), but despite this, electron transport rates during winter months remained low, which suggests that compared with the effects of temperature, increases in seawater DIC concentration most likely have only minor effects on Fucus ecophysiology, as identified by other studies (Al-Janabi et al., 2016b;Takolander et al., 2019).
The results of the redundancy analysis indicated that salinity, pH, irradiance and sampling site were not statistically associated with the measured ecophysiological parameters. Salinity is the main determinant of the geographic range of Fucus in the Baltic (Serrão et al., 1999;Bäck & Ruuskanen, 2000), and also has a substantial effect on Fucus physiology (Bäck et al., 1992;Nygård & Ekelund, 2006;Nygård & Dring, 2008). The range of salinity values measured in this study (from 5.4-6.1) lies well within the salinity tolerance range of Baltic Fucus (Takolander et al., 2017a), which explains why no statistical associations between Fucus ecophysiology and salinity were found. However, the temporal density of field sampling in this study is unlikely to capture the range of short-term salinity dynamics which may affect Fucus morphology (Ruuskanen & Kiirikki, 2000).
Macrophyte photosynthesis strongly affects seawater pH in shallow coastal zone (Middelboe & Hansen, 2007;Wahl et al., 2018), as high photosynthesis rates cause pH to rise, but high pH also has an effect on photosynthesis (Middelboe & Hansen, 2007). In this study, pH did not have a statistically significant effect on Fucus. This may be explained by traits related to carbon physiology. Previous studies have found relatively minor or no effects of declining pH on Fucus (Pajusalu et al., 2013;Al-Janabi et al., 2016a;Werner et al., 2016a;Takolander et al., 2019). Declining pH could potentially have beneficial effects on macroalgae, if this allows the alga to utilize the seawater CO 2 fraction as a carbon source in photosynthesis (Koch et al., 2013). Potentially, this would allow the downregulation of the activity of cellular carbon concentrating mechanisms (CCMs), which are energetically costly, thus allowing the alga to benefit from declining pH through improved energetics of carbon accumulation (Raven et al., 2011), potentially resulting in higher tissue carbon content under high CO 2 (Takolander et al., 2019). Fucus has a potent CCM (Middelboe & Hansen, 2007), and is able to maintain high photosynthesis rates in high pH conditions, e.g. when aquatic photosynthesis depletes the seawater inorganic carbon pool, causing pH to rise (Middelboe & Hansen, 2007). The range of pH values sampled in this study ranged from 7.87 (Ångbåtsbryggan, February) to 8.34 (Ekö, July). As Fucus has been shown to be able to photosynthesize efficiently even at pH 9.3 (Middelboe & Hansen, 2007), the highest pH values observed in this study are not likely to inhibit photosynthesis rates through carbon limitation. On the other hand, the effects of low pH/CO 2 fertilization on Fucus have been shown to be rather small (Al-Janabi et al., 2016a;Takolander et al., 2019). The lowest pH levels observed in this study occurred in winter, when photosynthesis rates are low due to low temperature and irradiance, and thus the carbon requirements for photosynthesis are also low, which in part explains why no statistically significant pH effects were found.
Irradiance has a substantial effect on macroalgal photobiology (Falkowski & Raven, 2007). However, the analysis conducted in this study did not detect a statistically significant effect of irradiance. This was somewhat unexpected but may partly be caused by variability in ambient irradiance or turbidity in sites at the time of the sampling. As ambient irradiance is highly variable and it was measured only at a single moment in time when each site was visited, the irradiance values measured may not reflect the real amount of irradiance the Fucus individuals are subjected to, but rather a snapshot during the time of sampling. The Bonferroni correction applied in the RDA analysis increases the risk of type II error (Armstrong, 2014), and the p-value for irradiance (p = 0.008) was rather close to the acceptance threshold (p = 0.007). It is also possible that other correlated variables (e.g. day of year) captured partly the irradiance effect in the RDA analysis. Irradiance values were negatively correlated with day of year (R 2 = −0.53) and positively correlated with nutrients (R 2 = 0.54), both of which had a statistically significant effect on Fucus. Many of the responses observed in this study, especially temporal patterns in chlorophyll content and photosynthetic efficiency, are most plausibly interpreted as adaptations to seasonal variations in light availability (Leskinen et al., 1992;Rohde et al., 2008). In this study, irradiance was positively correlated with internal carbon content of the alga, and negatively with chlorophyll content and (Fv/Fm), which together with biological reasoning suggests that seasonal changes in ambient irradiance were also contributing to the observed responses.
This study identified the most important environmental covariates driving the seasonality of Fucus ecophysiology in the northern Baltic to be temperature and macronutrient availability, although some of the observed responses, especially variations in chlorophyll content, may be interpreted as responses to seasonal changes in light availability. However, measurements conducted in this study only offer a snapshot of measurements conducted under different seasons, which may be potentially interpreted as responses to seasonal changes in environmental parameters.
The seasonal patterns observed in Fucus ecophysiology in this study mostly align with earlier studies which have identified substantial seasonality on growth and physiology in Baltic Fucus (Lehvo et al., 2001;Al-Janabi et al., 2016a;Takolander et al., 2019;Graiff et al., 2021;Meichssner et al., 2021). Notable geographic differences were observed, however. Chlorophyll a and c values observed in Kiel Fjord, southern Baltic Sea (Graiff et al., 2021), were approximately three times higher than those observed in this study, whereas carbon and nitrogen contents observed here fall well within the variability observed within Fucus in the Baltic and North Seas in summer. rETR max values were similar to those from Bothnian Fucus (Ekelund et al., 2007). However, in contrast to experiments conducted in the western Baltic (Graiff et al., 2021), in this study the rETR max values in winter remained low, highlighting the inhibitory effects of low temperature and low light availability on photosynthesis, and consequently growth rates, in the northern part of the geographic range of Baltic Fucus.
The Baltic Sea is one of the most rapidly warming large marine areas in the world (Belkin, 2009), with substantial warming projected also to the 21st century (Meier et al., 2022), potentially coupled with increasing nutrient emissions (Meier et al., 2012). In the context of the results of this study, such changes may potentially elevate Fucus photosynthesis and growth rates, given that temperature tolerances are not exceeded in the summer (Graiff et al., 2015b;Takolander et al., 2017b). Indeed, increased seawater temperatures have been observed to increase Fucus receptacle growth rates in the area (Kraufvelin et al., 2012), suggesting that elevated temperatures may result in phenological changes. Due to the ability of Fucus to store energy and nutrients, elevated photosynthesis in the summer could potentially sustain higher growth rates in winter. However, the realized effects will depend on ecosystem-scale responses, as changing abiotic conditions may potentially favour or inhibit co-occurring species such as filamentous algae (Takolander et al., 2017a) or grazers (Werner et al., 2016b).