The structure and composition of benthic marine habitats are being altered all around the world as groups of foundation species are lost and replaced by groups of lesser ecological complexity. Foundation species create complex physical habitat structure and positively interact with and enhance the resilience of local biological communities (Dayton, 1972; Stachowicz, 2001; Angelini et al., 2011). Marine macrophytes are historically the foundation species found in shallow subtidal temperate environments, large brown seaweeds (i.e. kelps of the orders Laminariales and Fucales) forming forest-like structures on rocky substrates (Teagle et al., 2017) and seagrasses forming vast meadows where sandy substrates dominate. Seaweed and seagrass communities can coexist where rocky and sandy substrates meet and create a unique set of interactions, however, there are relatively few studies that formally test these interactions in temperate regions. Of those that do, the majority present evidence for negative outcomes for either group citing some level of resource competition, with little evidence for facilitation usually (but not exclusively) in favor of the seagrass (reviewed in Richard and Quijón, 2023).

In the Mediterranean Sea, macroalgal species of genera Cystoseira, Ericaria and Gongolaria (order Fucales; hereafter collectively referred to as Cystoseira sensu lato), are considered the dominant canopy forming species for temperate rocky reefs, but in recent decades have dramatically declined (Airoldi and Beck, 2007; Iveša et al., 2016; Mangialajo et al., 2008; Orlando-Bonaca et al., 2021a; Thibaut et al., 2005). The presence of any extant marine forest habitat is considered a positive indicator for ecosystem health (Ballesteros et al., 2007). The importance of conserving these areas is highlighted by research describing the ecological value (Cheminée et al., 2013; Piazzi et al., 2018; Pinna et al., 2020) and by habitat level protection under the Barcelona Convention (United Nations Environment Programme/Mediterranean Action Plan (UNEP/MAP), 2013). Similarly, ecologically valuable seagrass habitats are declining globally (Duarte et al., 2008; Orth et al., 2006; Waycott et al., 2009), including in the Mediterranean Sea (Dunic et al., 2021; Chefaoui et al., 2018). Cymodocea nodosa (Ucria) Ascherson is one of few seagrass species native to the Mediterranean, and forms highly valuable meadows across its distribution along the Mediterranean and Atlantic coasts (Chefaoui et al., 2016) and in the Mediterranean, meadows are considered a protected habitat under the Bern Convention (Council of Europe, 2000). It is projected that a dramatic loss of C. nodosa cover will occur over the coming decades (Chefaoui et al., 2018), despite some potential refuge areas in cooler regions, including the Adriatic Sea, where an increase in cover could occur (Chefaoui et al., 2018), particularly as the species may replace more vulnerable Posidonia oceanica (Linnaeus) Delile meadows under threat (Burgos et al., 2017).

In the northern Adriatic Sea, a unique cold-temperate bioregion of the Mediterranean (Bianchi and Morri, 2000), fucalean forests have declined over the past decade, where only a scattering of remnant populations is documented along the Italian (e.g. Mancuso et al., 2018), Slovenian (e.g. Orlando-Bonaca et al., 2021a) and Croatian (e.g. Iveša et al., 2016) coastlines that make up this region. Among these isolated forests, the species Gongolaria barbata (Stackhouse) Kuntze in particular is commonly observed as a surviving species, an anomaly given a strong regional history of mixed species forests (Iveša and Devescovi, 2014; Iveša et al., 2016). Along the west Istrian Croatian coast, Iveša et al. (2022) have identified a significant remanent population of G. barbata in a costal lagoon where it coexists with the seagrass C. nodosa. Interestingly, the functionally similar fucoid, Cystoseira foeniculacea (Linnaeus) Greville, is also found to persist in a nearby coastal lagoon (Stjuža) in Strunjan Bay, Slovenia where it also coexists amongst a high biomass of seagrasses (C. nodosa and Zostera noltei Hornemann, Battelli and Catra, 2021, 2023). It is therefore perhaps the case that coastal lagoons can represent local refuges for supporting a high biomass of marine macrophytes and/or interactions between seaweed and seagrasses contribute to their simultaneous persistence in these environments.

Šćuza Lagoon is 68.6 ha in area, has a maximum depth of approximately 1.5m and C. nodosa is the dominant habitat forming macrophyte across most of the lagoon as it is dominated by sandy substrate, save for approximately 2 ha along the western edge where a shift to rocky/rubble substrate supports a significant settlement of G. barbata (Iveša et al., 2022). Historically, Šcuza Lagoon was used as an aquaculture basin and thus was heavily fished but the activity was abandoned, and the lagoon is now protected from such activities under the European Union Natura 2000 guidelines. The lagoon is isolated from the sea due to the construction of a pedestrian footbridge and thus there are a range of biotic and abiotic factors that contribute to its unique environment. Importantly, the thermal range of the lagoon is extreme, where the temperature can dip below zero and reach up to 34°C (Iveša et al., 2022). It is therefore quite remarkable that G. barbata, with a previously recorded thermal maxima of 23°C (throughout the Adriatic Sea, Ercegović, 1952), thrives here when it is currently rare elsewhere along the coastline (Iveša et al., 2016). In contrast, C. nodosa is less sensitive to higher temperatures, which was shown experimentally (Olsen et al., 2012) and by its widespread distribution (Chefaoui et al., 2016). The response of both species to extreme cold temperatures is not yet well understood.

It is possible that temporal alternation in the growth cycle of the dominant macrophytes in the lagoon contributes to the coexistence of the two species. G. barbata has a diplontic monophasic lifecycle that is typical of fucalean species (Ercegović, 1952; Rindi et al., 2023 and references therein). Locally, the growth of vegetative branches occurs in the autumn months (Iveša et al., 2022), hypothesized to be triggered by a temperature threshold of ~18°C (Bilajac et al., 2024). During this time, receptacles develop, and fertility is induced in the late winter/early spring when biomass peaks (Iveša et al., 2022; author observations). Annual (branch) material is then lost during the summer months, and the alga enters a “dormant phase” where perennial material remains and prominent, smooth cauloid apexes are evident. During this time, cauloids can become covered in epiphytes (Iveša et al., 2022) and growth is significantly reduced, particularly at high temperatures (Ercegović, 1952; Iveša et al., 2022; Bilajac et al., 2024). C. nodosa however, is a flowering marine angiosperm, capable of both sexual and clonal reproduction (Hemminga and Duarte, 2000), but the strength of either strategy is highly variable and regionally specific (Máñez-Crespo et al., 2020). C. nodosa meadows show a high turnover of biomass (Cunha and Duarte, 2007), can show variations in growth pattern based on depth (Olesen et al., 2002) but generally peak leaf growth is observed in the summer months (Cunha and Duarte, 2007; Olesen et al., 2002; Pérez and Romero, 1994), patterns which have been confirmed for meadows in the northern Adriatic (Bračun et al., 2020; Orlando-Bonaca et al., 2016). Epiphytes recruit heavily on older C. nodosa leaves throughout the year, which are then shed in the months prior to summer (Bračun et al., 2021).

Understanding mechanisms that facilitate the survival of any extant Cystoseira s.l. forest patches across the Mediterranean provides valuable information for formulating effective conservation practices. The identification of this population is important given its documented decline and current rarity along west Istrian coastline (Iveša et al., 2016) and similarly is under threat in wider Mediterranean (e.g. Thibaut et al., 2015; Orlando-Bonaca et al., 2021a; Tamburello et al., 2022). We hypothesize that the population’s unique morphology could provide insight into its persistence here. Falace and Bressan (2006) demonstrated that the morphology of G. barbata can be plastic based on environmental cues. Observationally, Šćuza’s G. barbata population shows obvious morphological distinction from other isolated patches found at other sites outside of the lagoon along the west Istrian coastline, possibly suggesting an evolutionary adaptation to the lagoonal environment (Iveša et al., 2022). This could be partially in response to competitive interactions and/or coexistence with C. nodosa, as we have not observed this interaction amongst remnant patches elsewhere. Within the lagoon itself there are two growth forms, one where a holdfast attaches to the rocky substrate, as is typical for this species and another “detached” form that is free floating where the holdfast is absent or non-functional (Iveša et al., 2022). Despite this free-living lifestyle however, the detached form is spatially restricted to the south-western part of the lagoon where it is primarily found trapped amongst seagrass fronds (Iveša et al., 2022). This observation suggests that there may be an interaction between the two dominant macrophytes, however here we focus on the attached form that is present both amongst seagrass and on exposed boulders for comparison.

In this study, the density and maximum thallus height of G. barbata was measured in-situ using quadrates placed within two distinct settlement habitats (raised submerged rocks outside C. nodosa and pebbles embedded within the C. nodosa seagrass meadow). Based on observation, we expect that G. barbata will grow to a larger maximum height when amongst seagrass compared to when exposed on larger rocky substrates. We aim (1) to test whether the presence of seagrass influences localized settlement and morphology of G. barbata and, (2) to document the presence of a unique marking (characterized by a damaged cauloid apex) that is observed among the G. barbata population largely within Šćuza but occasionally outside of the lagoon as well (author observations). Specifically, this study aims to understand the strength of a seaweed-seagrass interaction in influencing population dynamics of a threatened canopy-forming macroalga.

The population dynamics of the ecologically important macroalgal species G. barbata are influenced by the presence or absence of the seagrass C. nodosa in a coastal lagoon of the northern Adriatic Sea. Šćuza Lagoon hosts what is potentially one of the last significant populations of G. barbata along the western Istrian coastline (Croatia). Understanding the ecological context in which it persists here is valuable for informing targeted conservation and restoration efforts in other locations where it is declining or has been lost. G. barbata reached a greater maximum height during its peak growth phase in the autumn/winter months of 2022/23 when living amongst C. nodosa compared to when it had settled on exposed boulders. This height advantage was maintained over the spring/summer months in 2023 when only perennial parts of the algae were present, suggesting an adaptation of the entire growth form that is not limited to the seasonal generation of adventitious branch material. We found evidence that recruitment of G. barbata was influenced by settlement amongst seagrass compared to on boulders, observing more recruitment to the latter. However, further experimental work would be required to understand the overall similarity in density observed between the two settlement strategies beyond the recruitment stage. Our results support previous work describing the high plasticity in growth forms of G. barbata in reponse to localized environmental conditions. In Šćuza Lagoon, an alternation of peak growth periods likely contributes to the persistence and coexistence of significant populations of ecologically valuable marine macrophytes G. barbata and C. nodosa.

Recruitment of G. barbata is highly influenced by benthic topography (Devescovi, 2015), and within Šćuza Lagoon, it is also possible that boulder substrate provides greater recruitment opportunity comparted to areas dominated by unconsolidated substrate interspersed beneath the seagrass canopy. Amongst seagrass, G. barbata settles on small stones and pebbles and other unconsolidated hard substrate that form a secondary benthic substrate layer. A reduced surface area of appropriate recruitment substrate may explain a reduction in the number of small individuals or recruits observed here for the 2022/23 season. Alternatively, low recruitment amongst seagrass could indicate competitive exclusion during the period of peak cover for C. nodosa where boulder substrates may provide refuge from such competition. However, the difference in recruitment was not maintained when comparing the overall density, implying that opportunistic recruitment amongst seagrass, while less common, is not a barrier to survival. Amongst G. barbata that develop amongst seagrass, it is possible that taller cauloids and heightened branch material observed here is driven by light limitation at the recruit stage enforced by the C. nodosa canopy, but this would require further investigation. Nevertheless, such a fine-scale difference in form is consistent with research conducted elsewhere that shows a high degree of morphological variation for G. barbata across relatively small geographic ranges (Falace and Bressan, 2006; Orlando-Bonaca et al., 2022). Additionally, in Šćuza Lagoon, amongst the seagrass there are “detached” individuals that are not fixed to any substrate (author observations). This ecotype was described by Iveša et al. (2022) and is spatially restricted to the south-west of the lagoon but is often found trapped amongst the seagrass meadow or amongst R. tinctoria complexes that are also commonly observed amongst seagrasses. This high adaptability in form, coupled with the demonstrated physiological adaptation to extreme thermal conditions for this population (Iveša et al., 2022; Bilajac et al., 2024), suggests a resilience worth investigating in the context of its long-term survival and its potential as a donor population for fucalean restoration along the broader coastline.

Damage to perennial cauloid apices is an observed feature among Cystoseira s.l. species along the western Istrian coast which was also observed for G. barbata within Šćuza Lagoon (author observations). This feature was consistently recorded more frequently amongst individuals that had settled on boulders compared to those settled amongst seagrass, even in the summer months when cauloids host a high epiphyte biomass that conceal some such markings. It is possible that this feature is a sign of herbivory, however outside of the lagoon, where urchins are present, cauloid damage of this nature is also frequently observed. The herbivorous fish Sarpa salpa is present within and outside of the lagoon and is a voracious grazer of Cystoseira s.l (Vergés et al., 2009; Gianni et al., 2017; Orlando-Bonaca et al., 2022, 2021). However, S. salpa is not strictly herbivorous throughout its lifecycle (Dobroslavić et al., 2012; Havelange et al., 2008) and can also target C. nodosa and its epiphyte load for feeding (Marco-Méndez et al., 2017) and further investigation would be needed to better understand this interaction. It is also possible that herbivorous invertebrates could also graze on G. barbata in Šćuza Lagoon, and this is an important consideration given that early life stage Cystoseira s.l. species are susceptible to invertebrate grazing (Monserrat et al., 2023). However, this would not necessarily explain damaged cauloid apices observed for adult individuals. The ecological significance of damaged cauloid apices, the ecological relevance of seagrasses providing protection from it here and its prevalence amongst Cystoseira s.l populations elsewhere warrants further investigation.

Understanding the mechanisms that allow the survival of remnant marine forests is of increasing conservation importance because harnessing any observable positive interactions within extant forests may benefit conservation efforts. In Šćuza Lagoon it is likely crucial that there are very few sea urchins observed (Iveša et al., 2022), contributing to the observed abundance of both G. barbata and C. nodosa as both species are susceptible to urchin overgrazing (Agnetta et al., 2015; Cvitković et al., 2024; Fernandez et al., 2012; Giakoumi et al., 2012; Iveša et al., 2016; Nikolaou et al., 2023; Thibaut et al., 2005). In the context of at risk marine algal forests, there is increasing evidence for the role of habitat refugia (such as coastal lagoons, rockpools and microstructures) in providing refuge from herbivory and thus supporting the survival of isolated patches of canopy forming seaweeds (Battelli and Catra, 2021; Zarco-Perello et al., 2021; Monserrat et al., 2023) and long-term exclusion of over-abundant grazers is currently a major roadblock to the success of marine forest restoration in the Mediterranean (Verdura et al., 2023) and worldwide (Eger et al., 2022). Furthermore, water quality and excessive nutrient input has been shown to cause large-scale negative impacts on Cystoseira s.l. populations in the past (Iveša et al., 2016), and for Šćuza Lagoon the impact of nutrient input and fluctuations in water chemistry caused by nearby terrestrial activity is unknown yet makes the survival of the forest here more notable. It is important to study the ecology of persistent populations of at-risk species to understand any adaptive mechanisms that may be important considerations for conservation efforts. In the case of Šćuza Lagoon, here we show that the adaptability of G. barbata, even at the micro-environment level may contribute to its survival in a unique lagoonal environment.