Introduction

Globally, coastal ecosystems are experiencing ecological, functional, and structural changes due to increasing and interacting anthropogenic pressures (Hennige 2019; Halpern et al. 2015). Global pressures, such as ocean warming and increased intensity of storm events, combined with local pressures, such as habitat fragmentation and loss, and trophic imbalance due to overfishing, seriously threaten marine biodiversity (O’Hara et al. 2021). Coastal regions are particularly threatened, as this is where most human activities are concentrated today and will be even more so in the future (Halpern et al. 2015; Merkens et al. 2016). In these areas we find the rocky shore seaweeds. They are important foundation species that occur in the intertidal and upper subtidal worldwide, structuring diverse communities of associated flora and fauna (Spruzen et al. 2008; Christie et al. 2009; Clayden et al. 2014). Seaweed growth and morphology may contribute to ecosystem function as they are directly related to the complexity of the habitat (providing niches for numerous species, impacting the seaweeds function as shelter etc.) and food availability for associated species. Seaweed habitats provide a range of ecosystem services, including habitat provision for fish and invertebrates (Christie et al. 2009), nutrient cycling and regulation of water quality (Cotas et al. 2023, and references therein), provision of biomolecules and food when harvested (Mac Monagail et al. 2017), and carbon storage, export, and sequestration (Krause-Jensen and Duarte 2016).

Intertidal seaweed communities are, by nature, adapted to high variations in environmental conditions. They experience heat and desiccation stress at low tide (Davison and Pearson 1996, and references therein), which may influence seaweed physiology and community structure (Kautsky and Kautsky 1989; Nygård and Dring 2008). They also experience varying levels of wave exposure, which impacts seaweed growth and morphology (e.g., Bonsdorff and Nelson 1996; Ruuskanen and Bäck 1999; Long et al. 2013), clean the seaweed of periphyton, including epiphytic turf algae, and expose a larger surface area of fronds to light through movement (Koehl and Daniel 2022). The seaweeds themselves may modify waterflow and reduce heat and desiccation stress for the associated species (Bulleri et al. 2018). On top of the harsh environmental conditions that may act on intertidal seaweed ecosystems, they also experience pressures from human activities. This includes climate change related stressors, such as altered levels and patterns of heat waves and wave exposure (increasing wind speed and frequencies of storm events, Weitzman et al. 2021) and trophic downgrading of ecosystems, causing overgrazing (Steneck et al. 2002; Estes et al. 2011). Such interacting pressures can lead to the degradation and fragmentation of coastal habitats, reducing patch size and connectivity, and have severe effects on seaweed structure (e.g., size, shape, biomass) and ecosystem functioning, hampering their ability to provide important ecosystem services (Walker and Kendrick 1998; Mineur et al. 2015).

The effects of habitat fragmentation on intertidal seaweed communities have received little attention. Increasing the knowledge on the effects of habitat fragmentation on growth and morphology of these foundation species is important, as it will contribute to our understanding of drivers of change in structure and ecosystem functioning. In intertidal seaweeds, habitat fragmentation could possibly alter the small-scaled wave exposure (i.e., hydrodynamics) conditions, as larger seaweed patches likely attenuate water flow more than smaller patches. This, in turn, could have important repercussions on seaweed growth and morphology. If this is the case, the effects of habitat fragmentation may vary depending on wave exposure. Additionally, larger patches likely shade the seaweed fronds more than smaller and more fragmented patches, likely impacting the photosynthesis and growth of the seaweeds. These effects may also vary depending on the vertical position on the shore – for example, if smaller patches are less effective at retaining moisture when exposed to air at low tide, then intertidal seaweeds may be more negatively affected by habitat fragmentation than subtidal seaweeds.

The combined impact of habitat fragmentation and wave exposure may increase the susceptibility of ecosystems to other stressors, as multiple stressors often interact (Crain et al. 2008; Wahl et al. 2011). Herbivory, or grazing, is well known to have important top-down and bottom-up effects in intertidal seaweed systems (Lubchenco 1983; Bracken et al. 2014). The grazing pressure in coastal areas could possibly also be linked to habitat fragmentation and wave exposure. Habitat fragmentation produces more edges, which may alter the distribution of grazers through edge-effects. While this has not been studied in intertidal rockweeds, Reeves et al. (2022) showed that kelp forests were more susceptible to overgrazing when more fragmented. Wave exposure may affect grazing patterns directly or indirectly. For example, strong wave action can increase the probability of grazers (such as snails or sea urchins) becoming dislodged and increase the toughness of seaweed (Long et al. 2013). Increased toughness might reduce seaweed palatability, though Molis et al. (2015) have found that snail traits are phenotypically plastic, adjusting to the toughness of their food. Additionally, predation pressure on grazers may vary depending on wave exposure level (Menge 1978; Boulding et al. 1999). Therefore, habitat fragmentation, wave exposure, grazing and seaweed growth and morphology may be related in various ways. This study investigated these relationships in a mesocosm-based experiment by answering the following questions: (1) how do habitat fragmentation and wave exposure interact and impact seaweed growth (measured as tip elongation and branching)? (2) how is grazer abundance (i.e., coverage of gastropods) in seaweed beds impacted by habitat fragmentation and wave exposure? and (3) does gastropod coverage impact seaweed growth? We focused our study on the seaweed species Fucus vesiculosus, which is considered a foundation species in Nordic marine ecosystems (Graiff et al. 2015) and which dominates the rocky shore intertidal along the Norwegian coast. Habitat fragmentation and wave exposure was controlled as part of the experimental design, but the effect of gastropod coverage was included after observing rapid increases in numbers of gastropods in the basins.

Methods

The mesocosm basins and experimental setup

This experiment was conducted in 12 hard bottom (concrete) mesocosm basins at NIVA’s Marine Research Station Solbergstrand (59.62° N, 10.65° E), located in the outer Oslofjord, southeast of Norway. Each basin is 4.75 m long and 3.65 m wide, with a volume of 12 m3 and a water depth of 1.3 m at high tide (Bokn et al. 2001). The basins are individually supplied with surface water (1 m depth) from just outside of the station. The water has a residence time in the basins of 2–3 h, resulting in water temperature and salinity mimicking that of the surrounding Oslofjord surface water (Kraufvelin et al. 2020). Each basin has four steps descending to the basin floor, each step being 0.4 m wide. Tides are simulated according to a semi-diurnal tidal pattern with a tidal amplitude of 36 cm, so that all steps are covered with water at high tide, but steps 1 and 2 are exposed to the air at low tide (i.e., representing the intertidal zone). The basins have a wave machine generating constant wave action, with 18 strokes per minute. The wave machine may produce waves at two different levels, corresponded roughly to a wind strength of 5 m/s (gentle breeze at the Beaufort scale) in the high wave treatment and 2.5 m/s (light breeze) in the low wave treatment (Kraufvelin et al. 2010, 2020).

In March 2022, rocks and boulders with Fucus vesiculosus attached were collected from two donor sites in the intertidal and upper subtidal (down to approx. 1.5 m depth). The two sites were approx. 1 km from each other and 6 km from the research station, both having similar environmental conditions. The rocks and boulders with seaweeds were placed in the 12 basins at two different fragmentation levels (patch sizes). There were three replicate basins for each combination of habitat fragmentation and wave exposure level. Six of the basins had F. vesiculosus placed out as lots of tiny patches (LT, each patch with a size of approx. 0.25 m × 0.25 cm) and six basins had one single large patch (SL, approx. 1.3 m × 1.6 m) (Fig. 1). The seaweed covered the four steps in the basins, both the intertidal (step 1 and 2) and the upper subtidal (step 3 and 4). The total coverage of the rock with seaweeds were approximately the same in all basins, regardless of fragmentation pattern, and the number of patches in the LT basins were the same for all basins. Between the LT patches, there were seaweed free corridors varying between 0.5 and 0.8 m in width. The SL patches were centered in the middle of the steps so that they had approx. 1 m of seaweed free rock outside of the patch.

Fig. 1
figure 1

Drone image (Photo: NIVA) of the 12 mesocosm basins. Six of the basins (organized as 3 pairs) experienced high wave exposure, and six had low wave exposure (marked in the figure). For each pair of basins, there was one with a single large (SL) patch of seaweed, and one with lots of tiny (LT) patches

Full size image

Seaweed growth

To monitor seaweed growth, we measured tip elongation, i.e., increases in tip length (excluding heavily grazed tips, see chapter 2.4), and the degree of branching that the individual F. vesiculosus experienced during the time of experiment. Apical tip length is a parameter commonly used to measure growth (Strömgren 1994; Bäck et al. 1992; Bonsdorff and Nelson 1996) and branching is one of the many seaweed properties representing morphological variation that responds to environmental factors (Bäck 1992; Ruuskanen and Bäck 1999).

In total, 72 tips were marked and selected for monitoring, three individuals in the intertidal (step 1) and three in the upper subtidal (step 3) in each of the 12 basins. The three individuals on each step were distributed evenly within the “single large” (SL) patch or amongst the available “lots of tiny” (LT) patches. The tips were selected at the beginning of the experiment amongst the healthy seaweed individuals. Most of the selected tips had multiple tip endings, which will henceforth be referred to as apices. The length of these apices was measured from just above the youngest bladder (under which a cable tie was attached for identification), providing a dataset consisting of several replicate length measures representing one tip. Branching was measured as the change in number of apices recorded per seaweed individual during the experiment. The properties were measured weekly, and a time-series was collected in two periods of 6-weeks duration, in early and late summer (i.e., 24.06–29.07 and 17.08–21.09 2022).

Gastropod coverage

From the first (June-July) to the second (August–September) period, we observed an increase in the number of gastropods within the genus Littorina sp., with abundances varying between basins. To quantify the abundance of these grazers, we measured Littorina sp. coverage on step 2 (the lower intertidal) in each basin. As we had high densities of gastropods and too little time to record them, we visually identified the coverage as presence or absence in each of 16 5 × 5 cm large cells within a 20 × 20 cm frame. Three frames were randomly placed within seaweed patches and three in the seaweed-free corridors outside of the patches. Coverage was estimated as the percent of all 16 cells in the frames having Littorina sp. presence. We include only the data from August and September, as this was when grazing became an observable issue in the basins.

The green crab Carcinus maenas, an important predator in the Oslofjord, was frequently observed in the mesocosm basins during the experiment. During monthly visits, the crabs were captured and distributed evenly in the basins to assure comparable levels of predation on gastropod abundances.

Calculations and criteria for data exclusion

As a measure of tip growth, we decided to use the rate of increase in length, i.e., the regression slope extracted from linear models using the lengths from all apices within each basin. This approach was preferred over calculating the average length increase from the beginning to the end of the experiment, as the latter approach would have implied a loss of temporal resolution in the data. Linear regression analyses were conducted for length of apices against time for each of the 72 replicate tips.

Before the analyses, some data were excluded based on three criteria. First, tips that were lost after two weeks were excluded, as this period was considered too short to accurately assess growth. Second, negative tips elongation or branching data were excluded, as this was assumed to be due to heavy grazing or loss of seaweed parts. Third, tips where severe grazing was observed were excluded from the analysis. This resulted in a total of 15 out of the 72 replicate tips being excluded, leaving a total of 57 seaweed tips for the analyses. In no case were all replicates lost for a given step in a specific basin. All data were uploaded to Zenodo (Hayes et al. 2024a, 2024b).

Statistical analyses

All statistical analyses were run using the packages lme4 (Bates et al. 2015), car (Fox and Weisberg 2018), and emmeans (Lenth 2022) in R (R Core Team 2021).

Analyses of seaweed tip elongation and branching

Tip elongation was analyzed using linear mixed models (LMMs; Jiang 2007), with wave exposure, fragmentation level and their interaction included as fixed factors. Basin was included as a random factor in order to take the dependency of data in the same basin into account. The analyses were done separately for the intertidal (step 1) and the upper subtidal (step 3). Linear mixed models were checked for normality of error distributions using the Shapiro–Wilk test (Shapiro and Wilk 1965) and diagnostic error distribution plots were examined. Significance of fixed factors and their interaction was tested in two-way ANOVAs of the LMMs, using type III sum of squares. In the case that there were no significant effects, one-way tests were also run for each of the fixed factors separately. Branching, being count data, were analyzed in generalized linear mixed models (GLMMs; Jiang 2007), with wave exposure and fragmentation level as fixed factors and basin as a random factor, separately for step 1 and step 3. These GLMMs were run using a negative binomial distribution and a logarithmic link function. The GLMMs were all checked for overdispersion. Significance of fixed factors and their interaction in all GLMMs was tested with the Wald Chi-squared test.

Analyses of gastropod coverage

Gastropod coverage was analyzed against wave exposure and habitat fragmentation using LMMs, including habitat type (seaweed patch or seaweed free corridors) and basin as random factors. The analyses were done separately for August and September. Model assumptions were checked using the Shapiro–Wilk normality test and by examining diagnostic error distribution plots. Significance of fixed factors was tested in two-way ANOVAs of the LMMs, using type III sum of squares. Additionally, a one-way ANOVA was run to analyze the difference in gastropod coverage between the seaweed patches and the seaweed free areas in between and around the patches.

To learn about the possible impact of gastropod coverage on growth, we analyzed tip elongation and branching using LMMs and GLMMs respectively, with the predictors wave exposure, habitat fragmentation and their interaction as fixed factors, and gastropod coverage (at basin level) as a continuous predictor. Analyses were done for August and September separately. All LMM error distributions were checked for normality with the Shapiro–Wilk test, visual examination of error histograms and qq-plots, and GLMMs were checked for overdispersion.

Results

Effects of habitat fragmentation and wave exposure on seaweed growth

Tip elongation was not significantly related to wave exposure, fragmentation level or the interaction between them (Table 1, Fig. 2). Supplementary Table 1 shows the data on branching and tip elongation. Branching (i.e., the change in number of apices recorded per seaweed individual during the experiment), was significantly impacted by wave exposure in the intertidal (Table 2) and was higher in basins with low than high wave exposure (Fig. 3). The relationship was not significant for the upper subtidal. There was no significant relationship between branching and habitat fragmentation or the interaction between wave exposure and habitat fragmentation. A total of 18 apices were lost during the experiment, due to grazing, mechanical damage, drying or other unknown reasons.

Table 1 Two-Way ANOVA results for the effect of the predictor variables habitat fragmentation level (FL, single large versus lots of tiny patches), wave exposure (WE, high versus low) and their interaction on tip elongation in the intertidal and upper subtidal
Full size table
Fig. 2
figure 2

Tip elongation (i.e., linear regression slope of tip-lengths over time) of Fucus vesiculosus for high and low wave exposure (left panel) and for both levels of habitat fragmentation (right panel; LT lots of tiny patches, SL single large patch) for the intertidal and upper subtidal. Horizontal lines: median value, box: the 25–75 quartile, whiskers: data outside of the middle 50%, dots: outliers (i.e., 1.5 times the interquartile range)

Full size image
Table 2 Wald Chi-Squared test results for the effect of the predictor variables habitat fragmentation level (FL, single large versus lots of tiny patches), wave exposure (WE, high versus low) and their interaction on branching in the intertidal and upper subtidal
Full size table
Fig. 3
figure 3

Branching of Fucus vesiculosus (i.e., changes in number of apices over time) for high and low wave exposure (left panel) and for both levels of habitat fragmentation (right panel; LT lots of tiny patches, SL single large patch) for the intertidal and upper subtidal. Horizontal lines: median value, box: the 25–75 quartile, whiskers: data outside of the middle 50%, dots: outliers (i.e., 1.5 times the interquartile range). The asterix marks the significant relationship

Full size image

Relationships between habitat fragmentation, wave exposure gastropod coverage, and seaweed growth

In August, gastropod coverage varied from 0 to 50% and in September it varied between 0 to 31%. Gastropod coverage was higher inside seaweed patches than in the seaweed free areas (August: one-way ANOVA, F = 33.05, P < 0.001; September: one way ANOVA, F = 24.46, P < 0.001). In August, there was a significant impact of wave exposure in interaction with fragmentation (Table 3), i.e., the effect of wave exposure on gastropod coverage dependent on habitat fragmentation level. In basins with lots of tiny (LT) patches, gastropod coverage was higher at low than at high wave exposure, whereas the opposite was the case for basins with single large (SL) patches (Fig. 4). In September, there was no such relationship. Supplementary Table 2 shows the data on gastropod coverage. There was no significant relationship between tip elongation or branching and gastropod coverage, neither in August nor September.

Table 3 Two-Way ANOVA results for the effects of the predictor variables habitat fragmentation level (FL, single large versus lots of tiny patches), wave exposure (WE, high versus low) and their interaction on gastropod coverage in August and September
Full size table
Fig. 4
figure 4

Gastropod coverage (%) at high and low wave exposure for lots of tiny (i.e., high level of habitat fragmentation) and single large (, i.e., low level of habitat fragmentation) patches, measured in August and September 2022. Horizontal line: median value, box: the 25–75 quartile, whiskers: data outside of the middle 50%, dots: outliers (i.e., 1.5 times the interquartile range)

Full size image

Discussion

This study provides the results from a mesocosm study on the relationships between wave exposure, habitat fragmentation, grazing and Fucus vesiculosus growth. We found that F. vesiculosus in the intertidal exhibited a significantly higher degree of growth through branching in basins with low compared to high wave exposure. Several other morphological traits of F. vesiculosus have been shown to be related to wave exposure (Bäck 1993; Long et al. 2013; Barboza et al. 2019), and our results support field observations showing higher branching in F. vesiculosus living on less-exposed shores (Kalvas and Kautsky 1993). Little is known about the physiological mechanisms responsible for branching on healthy F. vesiculosus tips, although irregular branching has been observed to result from the inhibition of apical dominance (Moss 1964). Intertidal species experience severe heat and desiccation stress for extended periods of time (Davison and Pearson 1996, and references therein), which may be more intense at low than high wave exposure due to a lack of water splash. This stress may result in a greater level of tip tearing or breakage, and thereby increased branching (due to inhibition of apical dominance). An explanation for the lack of effect of wave exposure on branching in the subtidal might be that the heat and desiccation stress is lower here. The lower level of branching at high wave exposure might also be an adaption to wave sweeping, as branched fronds experience higher drag stress than less branched fronds (Starko et al. 2015; Martone et al. 2012), which will be a bigger challenge in the intertidal than in the subtidal. Also, more branching may potentially help seaweeds to maintain moisture during low tide, but knowledge on this is lacking. We found no effects of habitat fragmentation on seaweed branching. Further studies into the growth physiology of this species will elucidate which other environmental factors are responsible for branching.

We found no significant effect of habitat fragmentation, wave exposure, or their interaction, on F. vesiculosus growth measured as tip elongation. While the effects of habitat fragmentation on intertidal seaweed growth have not been widely studied, these results contrast with several other studies showing effects of wave exposure on F. vesiculosus growth. Total thallus length has been generally shown to be reduced by increased wave exposure (Kalvas and Kautsky 1993; Bäck 1993; Barboza et al. 2019). However, this relationship has not been well established with respect to tip elongation. Bonsdorff and Nelson (1996) did find significantly higher tip lengths in F. vesiculosus in sheltered than wave exposed areas, however they found higher variation within sites of the same wave exposure than between wave exposure levels. Therefore, while wave exposure likely has important effects on total thallus growth, it is possible that this effect is not observable when considering growth of a specific tip, as has been done here.

In the second half of our experiment, high abundances of gastropods and heavy grazing on the seaweeds became progressively observable. Snails of the genus Littorina, along with certain large-bodied isopods (such as those of the genus Idotea) and amphipods, are important grazers in intertidal seaweed habitats in the Oslofjord (Kraufvelin et al. 2020). In August and September, gastropod coverage was significantly higher within seaweed patches than in the seaweed free areas between and outside of the seaweed patches. Seaweeds likely offer snails more food and more protection from predators (for instance crabs), hydrodynamic forces compared to “naked” substrate. Various gastropod species were considered together, so more information on different behaviors by different species could arise if species were analyzed separately, but this was not done due to time constraints.

Our results show an effect of wave exposure on gastropod abundance (measured as coverage) that depended on the level of habitat fragmentation. In basins with lots of tiny seaweed patches, increased wave exposure reduced gastropod abundance, while increased wave exposure was associated with increasing coverage of gastropods in basins with a single large patch of seaweed. A possible explanation for this finding might be related to differences in how wave exposure is dampened by the seaweeds. High wave exposure might dislodge gastropods (Denny 1985) and the waves are most likely attenuated more effectively in the basins with single large patches than in those with lots of tiny patches. Therefore, we propose that in basins with lots of tiny patches, high wave exposure reduces gastropod abundance because it increases their susceptibility to dislodgement. Further studies are needed to test these specific mechanisms. It is worth noting that the interaction between wave exposure and fragmentation was only significant in August. This most likely means that the difference disappears as the summer turns to autumn and the communities associated with seaweeds change.

Another possible explanation for the combined effect of wave exposure and fragmentation on the gastropod coverage might also be related to differences in predation patterns. The green crab Carcinus maenas is an important predator of gastropods (Buschbaum et al. 2007) and is the most common crab species in the Norwegian intertidal. Despite the number of craps being kept (to the best of our abilities) similar in all basins, the effect of their presence might have varied. Predation on intertidal rocky shores is found to be more intense in sheltered than in exposed areas due to the difficulties foraging in the heavy surf (Menge 1978; Boulding et al. 1999). Furthermore, dense intertidal algal canopies may favor predation more than sparse canopies (Menge 1978). Therefore, in general, the gastropods in our study may have been more susceptible to predation by crabs in basins with low wave exposure and larger, denser patches. We did not study gastropod size and morphology, but studies have shown that wave sheltered areas select for thicker shells, as the predator pressure is higher in these areas (Boulding et al. 1999). Further studies are needed to fully understand the possible variation in gastropod ecotypes and their relation to wave exposure.

The lack of a clear relationship between gastropod abundance (coverage) and tip elongation indicates that taxa other than gastropods could have been involved in the F. vesiculosus grazing. Large isopods and amphipods were frequently observed in the basins. Isopods of the genus Idotea have been shown to structure F. vesiculosus populations in the Baltic Sea (Engkvist et al. 2000). The abundance of isopods and amphipods was not monitored throughout this experiment, due to the difficulties associated with in-situ measurement of these highly mobile and small animals. Future replication studies, including grazing control (such as exclusion cages), will shed more light on the mechanisms behind these processes and may also uncover effects of habitat fragmentation and wave exposure on seaweed growth that were not studied here.

There are many seaweeds properties representing morphological variation that could have been used to assess seaweed growth. These includes thallus and stipe length and width, number of and distance between dichotomies, midrib width and number of tips (Ruuskanen and Bäck 1999). We focused our study on change in tip length and branching, as these are parameter commonly used to measure growth (Bäck et al. 1992; Strömgren 1994; Bonsdorff and Nelson 1996; Ruuskanen and Bäck 1999). However, more properties should be analyzed in order to fully understand how growth is impacted by environmental conditions and pressures.

Conclusion and future work

Our study showed that wave exposure impacts Fucus vesiculosus morphology (branching) in the upper intertidal. While habitat fragmentation did not have any direct effect on the seaweed growth parameters in this study (i.e., branching and tip elongation), it impacted gastropod (grazer) coverage in an interaction with wave exposure. The results presented here thereby contribute to increased knowledge on the dynamics between seaweed growth and grazer abundance, and the relationships to habitat fragmentation and wave exposure. However, more studies on seaweed growth physiology are needed in order to disentangle the effects of different environmental factors. More controlled studies on the impact of environmental variables on grazer abundance and ecotype are also needed, and studies using exclusion chambers should be applied in order to prevent uncontrolled grazing impacting the studies and the ability to draw conclusions from them.

Despite the lack of a fit-for-purpose study design on the effect of grazing, our study still emphasizes the important of discussing the potential dangers that overgrazing events can pose to marine habitats, also in the intertidal (as have been studied and discussed for many years in e.g., the kelp forests, Ling et al. 2015). As increased winds, more frequent storm events and more habitat fragmentation is expected in the future, our findings on the effect of these interacting pressures on grazer abundance is relevant for the discussion on the importance of safeguarding the connectivity of intertidal habitats. Further studies, both in experimental setups and in the field, are needed to confirm and challenge our results, and to continue unravelling the relationship between the variables under study.