INTRODUCTION
Marine macroalgae serve as foundational components of coastal ecosystems and function as primary producers and structural habitat providers for a wide variety of associated organisms (Miller et al. 2011, Paine et al. 2021, Saha et al. 2024). Terminology describing organisms inhabiting the surfaces of other living hosts has been applied inconsistently leading to ambiguity. The term epibionts (or epibiota) broadly refers to organisms that reside on the external surfaces of other organisms, typically without causing direct harm to their hosts (Wahl 2008). Epiphytes are more narrowly defined as sessile organisms—such as bacteria, fungi, algae, protists, and small animals—that colonize the surfaces of marine macrophytes (Burfeid-Castellanos et al. 2021, Bjorbækmo et al. 2023). Some taxa blur these distinctions; for example, certain oomycete fungi in the genus Olpidiopsis are obligate parasites of algal hosts (Zuccarello et al. 2024). Distinguishing between epiphytic and parasitic relationships can therefore be challenging, as host–symbiont interactions may range from benign coexistence to pathogenicity (Bjorbækmo et al. 2023). These epiphytic communities significantly contribute to overall marine biodiversity, facilitate nutrient cycling, and support complex trophic webs (Chen et al. 2021, El-Khaled et al. 2022). Given their ecological significance, understanding the factors that shape the assembly, composition, and diversity of epiphytic communities is essential for predicting changes in coastal biodiversity and ecosystem functioning in response to environmental variability and anthropogenic impacts.
The composition of epiphytic communities is influenced by various biotic and abiotic factors, including water quality, nutrient availability, predation pressure, competition, and the chemical and physical properties of the host algal surface (Saarinen et al. 2018, Malik et al. 2020, Koehl and Daniel 2022). Host macroalgal morphology characteristics directly influence microhabitat complexity, attachment stability, resource availability, and colonization dynamics (Saarinen et al. 2018). For instance, structural features such as thallus thickness, surface texture, branching patterns, and overall complexity of macroalgal hosts can create diverse microenvironments that influence epiphyte settlement, survival, and growth (Chemello and Milazzo 2002, Wahl et al. 2012, Bringloe 2023, Lenzo et al. 2023). However, previous studies have predominantly relied on broad functional-form categories—such as coarsely branched (CB), sheet-like (S), or thick leathery (TL)—to investigate epiphytic diversity patterns (Littler and Littler 1984, Ryznar et al. 2021). While useful, these coarse classifications might overlook subtle but ecologically important morphological distinctions that significantly influence community assembly.
The cortical cell layer thickness of macroalgae has rarely been examined as a specific factor influencing epiphyte community structure. Cortical cells form the outermost layers of algal thalli, thus directly mediating interactions with epiphytic organisms. Variations in cortical thickness may affect substrate rigidity, surface stability (Koehl and Daniel 2022), microtopography (Malik et al. 2020), nutrient exchange (Machado and Oliveira 2024), and chemical exudates (Lachnit et al. 2011), potentially shaping both microbial biofilm formation and colonization by larger epiphytic taxa. Despite the plausible ecological significance of this trait, empirical studies linking cortical cell layer thickness explicitly to epiphyte community composition are scarce, leaving substantial gaps in our understanding of host-epiphyte interactions.
Although epiphyte research has advanced considerably at the global scale, studies explicitly addressing host trait-driven community assembly are notably underrepresented in Korean coastal ecosystems. The intertidal regions of Korea, characterized by diverse and abundant macroalgal communities, can function as important biodiversity hotspots by providing habitat for a variety of associated organisms (Keith et al. 2014, Hwang et al. 2020, Manca et al. 2024). This underscores the importance of investigating the ecological mechanisms underpinning host-epiphyte relationships. To address the gaps, this study integrates modern metagenomic techniques with morphological trait analyses in these habitats.
Recent methodological advances in small organelle-enriched metagenomics (SoEM) have provided unprecedented opportunities to comprehensively characterize eukaryotic epiphyte communities. Unlike traditional microscopy or targeted amplicon sequencing, SoEM enables broader and deeper organism detection through shotgun sequencing without targeting specific DNA markers. This provides robust taxonomic resolution across diverse taxa—from protists and diatoms to macroinvertebrates (Kim et al. 2019, Jin et al. 2023, Min and Kim 2023). This advanced genomic approach substantially enhances our understanding of community assembly processes and their ecological determinants.
In this study, we used SoEM methodology to elucidate the compositional structure of eukaryotic assemblages associated with predominant macroalgal taxa (Grateloupia elliptica, Ishige okamurae, Chondrus ocellatus, Grateloupia cornea, Gelidium elegans, Callophyllis sp., Sargassum thunbergii, and Ahnfeltiopsis flabelliformis) collected from an intertidal tidal pool at Jeongdo-ri, on the southwest coast of Korea. Our research had three specific objectives: (1) to characterize the diversity and taxonomic composition of epiphytic assemblages across host taxa, (2) to determine whether morphological traits—particularly cortical cell layer thickness and structural complexity—serve as reliable predictors of community composition and differentiation, and (3) to explore the phylogenetic and ecological assembly processes underlying community differentiation.
This investigation elucidates the mechanistic relationship between host morphological traits and epiphytic eukaryotic assemblage structure, transcending conventional functional-form paradigms. The fine-scale morphological approach used herein facilitates empirical examination of theoretical ecological processes including niche differentiation, environmental filtering, and habitat specialization. These findings contribute to fundamental understanding of marine biodiversity maintenance mechanisms and community assembly theory, while simultaneously informing applied coastal ecosystem management, conservation prioritization strategies, and predictions of community responses to environmental change.
MATERIALS AND METHODS
Study area and sampling
Epiphyte community samples were collected from an intertidal tidal pool located at Jeongdo-ri, on the southwest coast of Korea (34°17′45.3″ N, 126°42′04.0″ E). To minimize ecological disturbance, sampling was limited to the collection of eight dominant macroalgal species: Grateloupia elliptica, Ishige okamurae, Chondrus ocellatus, Grateloupia cornea, Gelidium elegans, Callophyllis sp., Sargassum thunbergii, and Ahnfeltiopsis flabelliformis. Entire thalli of each species were hand-collected during low tide (July 2023) into sterile bags with ambient seawater and transported on ice for immediate processing. In addition, 2 L of seawater from the same pool was collected to account for non-epiphytic organisms.
Data acquisition for host morphological traits
To avoid additional collection and ensure data reliability, morphological trait data for the eight host species were obtained from published literature, verified databases, and photographic archives. Traits included functional form, thallus thickness, cortical cell layer thickness, surface texture, and structural complexity. Where possible, we prioritized measurements from identical species collected in geographically proximate Korean populations (e.g., Kim et al. 2023; complete references in Supplementary Table S1). Trait accuracy was cross-verified using high-quality specimen images from credible academic sources (Supplementary Table S2).
Trait definitions were as follows:
- Functional form: classified as CB, S, or TL following Littler and Littler (1984).
- Thallus thickness: measured in micrometers (μm)from peer-reviewed morphological descriptions.
- Cortical cell layer thickness: determined as the number of cortical cell layers in published transverse-section micrographs.
- Surface texture: categorized qualitatively based on morphological surveys.
- Structural complexity: calculated as the perimeter-to-area ratio from ImageJ (Schneider et al. 2012) analyses of at least 10 representative specimen images per each species.
This approach allowed for comprehensive trait characterization while minimizing ecological impact.
Epiphyte isolation and sequencing
Epiphytes were removed from each macroalgal thallus by gently brushing in sterile seawater, and the resulting suspension was filtered through 0.45 μm cellulose acetate membranes (Advantec, Tokyo, Japan) under gentle vacuum to retain cells. DNA extraction followed the SoEM protocol (Jin et al. 2023): pellets were homogenized in buffer (250 mM sucrose, 30 mM Tris-HCl, 10 mM EDTA, and pH 7.5) with 0.2 and 2 mm zirconia beads using FastPrep-24 (MP Biomedicals, Irvine, CA, USA). Small organelles were concentrated by modified differential centrifugation (Jo et al. 2019), and DNA was purified using the Monarch Genomic DNA Purification Kit (NEB, Ipswich, MA, USA).
Sequencing libraries were prepared with the TruSeq Nano DNA Kit (Illumina, San Diego, CA, USA), fragmenting ~100 ng DNA to ~350 bp inserts using a LE220 Focused-ultrasonicator (Covaris, Woburn, MA, USA). Libraries were checked on a Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA) and sequenced on an Illumina MiSeq (2 × 301 bp). Raw reads were deposited in the National Center for Biotechnology Information (NCBI) SRA (BioProject PRJNA1082063; SRR31896593–SRR31896601).
Bioinformatics
Raw reads were quality-filtered using Trimmomatic v0.39 (Bolger et al. 2014), followed by merging with FLASH v1.2 (Magoč and Salzberg 2011), retaining only sequences longer than 300 bp. Taxonomic identification of operational taxonomic units (OTUs) used BLASTn against the NCBI nucleotide (nt) database (accessed October 2024) with E-value <1e-10, retaining the top bit-score hit. OTUs with fewer than five reads and non-eukaryotic (i.e., Bacteria, Archaea, Viruses) or non-marine lineages were removed. The taxonomic classification and habitat metadata were annotated using the World Register of Marine Species (WoRMS) database with the worrms package v0.4.3 (Chamberlain and Vanhoorne 2023).
Statistical analyses and community structure
Community analyses were performed on presence-absence OTU matrices. Hierarchical clustering used dissimilarity with group significance assessed by the similarity profile (SIMPROF) testing using the clustsig v1.1 (Whitaker and Christman 2014). To further assess compositional differences in relation to host morphological traits (thick vs. thin cortical layer hosts), we conducted non-metric multidimensional scaling (NMDS) and analysis of similarity (ANOSIM) using the vegan v2.6-10 (Oksanen et al. 2025). Similarity percentage analysis (SIMPER) quantified contributions to within-group similarity and between-group dissimilarity. All analyses and visualization were performed using R software v4.4.1 (R Core Team 2024).
Phylogenetic community analyses
Faith’s phylogenetic diversity (PD) and the net relatedness index (NRI) were calculated using the picante v1.8.2 (Kembel et al. 2010) from phylogenetic tree for OTUs constructed via phylostratr v0.2.1 (Arendsee et al. 2019) with branch lengths assigned by the Grafen method in ape v5.8 (Paradis and Schliep 2019). Null communities were generated through randomization, with the expected NRI value standardized to zero, and these null simulations were used to infer phylogenetic clustering or overdispersion.
RESULTS
Host morphological traits
The eight macroalgal species exhibited substantial interspecific variation in morphological traits potentially influencing epiphyte colonization (Table 1). The functional forms consisted predominantly of CB species (Ishige okamurae, Chondrus ocellatus, Grateloupia cornea, Gelidium elegans, Sargassum thunbergii, and Ahnfeltiopsis flabelliformis), with one S species (Callophyllis sp.) and one TL species (Grateloupia elliptica). Thallus thickness varied significantly, ranging from a minimum of 100 μm in A. flabelliformis to a maximum of 1,000 μm in G. elliptica. The cortical cell layer thickness was particularly variable, ranging from very thin (1–3 cells in Callophyllis sp.) to very thick (14–25 cells in I. okamurae). Surface textures were categorized qualitatively and showed diverse patterns: leathery textures were common in thicker-layered species, while cartilaginous and fleshy textures characterized many thin-layered hosts. Structural complexity scores ranged widely (from a minimum of 8 cm−1 in G. elliptica to a maximum of 138 cm−1 in A. flabelliformis), indicating broad variation in structural complexity across host species.
Epiphyte richness and specificity
Sequencing of epiphytic communities yielded robust datasets with 3.5–4.7 million high-quality merged reads per sample (Supplementary Table S3). Rarefaction analyses confirmed sufficient sequencing depth, with all curves reaching asymptotes (Supplementary Fig. S1), indicating comprehensive capture of epiphytic species richness.
Species overlap analyses revealed marked differences in epiphyte specificity relative to the ambient seawater community (Supplementary Fig. S2). Each host showed substantial proportions of unique species (66–84%), highlighting strong host-specific selection rather than passive colonization from the environment. Callophyllis sp. and I. okamurae supported particularly high numbers of unique taxa, further confirming morphological filtering effects. Conversely, hosts such as G. elegans exhibited a greater proportion of seawater-derived taxa, suggesting reduced filtering or selective pressure. Following the removal of OTUs shared with ambient water, OTU richness varied considerably among host species (Fig. 1, Supplementary Fig. S2): Callophyllis sp. exhibited the highest count (172 OTUs), while A. flabelliformis hosted the fewest (36 OTUs), demonstrating significant variability in epiphyte richness associated with host morphological traits.
Morphological influences on community structure
Hierarchical clustering based on Jaccard dissimilarity revealed clear differentiation in epiphytic community composition among macroalgal hosts (Fig. 1). Two statistically significant clusters emerged (SIMPROF test, p < 0.05). The first cluster grouped hosts characterized by thick cortical layers (G. elliptica, I. okamurae, C. ocellatus, and G. cornea). These hosts generally supported higher species richness and were dominated by macroinvertebrate taxa, particularly Arthropoda (Malacostraca, Hexapoda), Annelida, and Mollusca. The second cluster included thin-layered hosts (G. elegans, Callophyllis sp., S. thunbergii, and A. flabelliformis) characterized by lower richness and dominance of microalgal taxa, especially Bacillariophyta (diatoms), suggesting morphological control over epiphyte colonization.
NMDS clearly separated epiphytic communities according to host cortical layer thickness, independently of host taxonomic affiliation (Rhodophyta vs. Ochrophyta) (Fig. 2). ANOSIM analysis confirmed statistically significant compositional differences between thin and thick cortical layer groups (R = 0.885, p = 0.024), emphasizing cortical cell layer thickness as a primary determinant of community structure.
Host morphology-driven epiphyte communities
Venn analyses (Fig. 3A) further demonstrated compositional divergence driven by host morphological characteristics. Thick cortical layer hosts collectively supported a large proportion of unique epiphyte taxa (48%), characterized predominantly by macrofaunal groups (Fig. 3B). In contrast, thin-layered hosts exhibited a higher proportion of unique taxa dominated by Bacillariophyta and other microalgae, emphasizing strong morphological control over epiphyte community assembly processes.
SIMPER analysis quantified within-group similarity and between-group dissimilarity, identifying taxa driving compositional differences between morphological groups (Table 2). Thin-layered hosts showed moderate within-group similarity (42.7%), predominantly driven by Bacillariophyceae (diatoms, 58.2% contribution). Thick-layered hosts exhibited higher within-group similarity (51.3%), with Arthropoda, specifically Malacostraca (24.7%) and Hexapoda (17.4%), dominating contributions. The high average dissimilarity between morphological groups (72.0%) highlighted substantial compositional divergence, primarily driven by the contrasting dominance of microalgal and macrofaunal taxa.
PD and community assembly patterns
Phylogenetic analyses revealed contrasting community assembly patterns among macroalgal hosts (Fig. 4). Faith’s PD varied substantially among host species, notably illustrated by contrasting values in two structurally complex hosts: Callophyllis sp. exhibited the highest PD, whereas A. flabelliformis showed one of the lowest. NRI indicated that most thick-layered hosts, such as I. okamurae and C. ocellatus, had strong phylogenetic clustering (positive NRI), indicating selective environmental or trait-based filtering. Only one species, G. cornea, exhibited a negative NRI, though its small magnitude suggests merely a weak tendency toward phylogenetic overdispersion (negative NRI), likely falling within stochastic variation. Thin-layered hosts mostly had NRI values near-zero, indicating community assembly close to random expectations.
DISCUSSION
Host morphological traits as determinants of epiphyte community structure
Our results emphasize the pivotal role of macroalgal morphological characteristics—particularly cortical cell layer thickness, surface complexity, and texture—in structuring epiphytic eukaryotic communities (Table 1). While traditional functional-form classifications (such as sheet-like, coarsely branched, or leathery forms) have often been used to predict colonization patterns in macroalgal ecology (Littler and Littler 1984, Fong et al. 2023, Vranken et al. 2023), we demonstrate here that more fine-scale morphological features offer greater predictive power. Specifically, cortical cell layer thickness emerged as a primary morphological determinant of epiphyte composition and phylogenetic structure, with thick layers favoring macrofauna-rich, clustered assemblages and thin layers favoring diatom-rich, near-random assemblages (Figs 2 & 3).
Previous research has identified host morphological complexity as a significant factor influencing epiphyte recruitment by providing diverse physical habitats (Chemello and Milazzo 2002, Gallardo et al. 2021, Koehl and Daniel 2022, Bringloe 2023, Gibbons and Quijón 2023). Our study extends this knowledge by pinpointing cortical cell layer thickness as a particularly influential morphological attribute, linking host microstructural traits directly to epiphyte community assembly processes (Fig. 4). These findings suggest potential implications for ecological theories concerning habitat complexity, niche differentiation, and the maintenance of biodiversity in marine ecosystems.
Underlying mechanisms linking host traits and epiphyte assembly
Distinct epiphyte communities observed among hosts with thick versus thin cortical layers suggest specific host-associated mechanisms driving community assembly. Thick cortical layers likely offer structurally stable, complex surfaces and deeper microhabitats suitable for attachment, refuge, and resource partitioning among macroinvertebrates such as Malacostraca and Hexapoda (Chemello and Milazzo 2002, Gallardo et al. 2021, Koehl and Daniel 2022). In contrast, thin-layered hosts, such as Callophyllis sp. and Gelidium elegans, were predominantly colonized by photoautotrophic microalgae, especially Bacillariophyceae, likely due to more flexible surfaces and fewer physical or chemical barriers to colonization. Diatoms, being rapid colonizers, exploit thin and structurally simpler surfaces, possibly gaining advantage in early successional or disturbance-prone habitats (Bjorbækmo et al. 2023, Lenzo et al. 2023, Smith et al. 2025).
Moreover, host-specific surface texture and associated chemical properties might also interact with structural traits to modulate epiphyte establishment. For instance, leathery or cartilaginous textures, common in thicker-layered hosts, may chemically or physically deter settlement by certain microbial or algal species, thereby selectively favoring robust macrofauna or tolerant taxa (Chemello and Milazzo 2002). Future work should investigate how chemical traits such as surface-associated metabolites influence these observed structural relationships, further illuminating mechanisms behind morphological filtering.
Phylogenetic perspectives and assembly processes
Phylogenetic analyses provided deeper insight into community assembly processes, showing variable patterns of clustering and overdispersion among hosts (Logares et al. 2020, Milke et al. 2022). Positive NRI values, indicating phylogenetic clustering, suggest that host traits impose selective pressures leading to colonization by closely related taxa, likely sharing similar ecological preferences (e.g., Ishige okamurae in Fig. 4). Such selective filtering is consistent with strong habitat specialization, wherein host traits serve as strong ecological barriers to colonization by less-adapted taxa (Menaa et al. 2020, Pearman et al. 2023, Xie et al. 2023).
Cortical cell layer thickness was an important predictor of phylogenetic structure, but its effects varied. While thick-layered hosts often showed clustering, G. cornea was the only species with a negative NRI (Fig. 4B). However, the magnitude of this overdispersion signal was small, suggesting only a weak tendency toward coexistence of distantly related taxa—possibly within the range of stochastic variation—rather than strong evidence for competitive exclusion or facilitation (Liu et al. 2019, Cappelatti et al. 2020, Vass et al. 2020). This nuance highlights that even in morphologically “filtering” environments, niche differentiation may be subtle and secondary to other processes.
Near-zero NRI values in several thin-layered hosts (e.g., G. elegans, S. thunbergii, A. flabelliformis) further suggest that random settlement from the local species pool can dominate when morphological filtering is weak. Overall, the phylogenetic structure of these epiphytic assemblages appears to result from a combination of habitat filtering, stochastic colonization, and—in some cases—minor niche differentiation, underscoring the complexity of assembly mechanisms in macroalgal-associated communities.
Host specificity and environmental connectivity of epiphyte communities
Venn analyses revealed significant host specificity in epiphyte communities compared to ambient seawater, with most taxa unique to particular macroalgal hosts (Supplementary Fig. S2). This finding indicates that, while seawater may serve as a regional pool supplying epiphytes, selective filtering mediated by host morphological traits strongly shapes the final community composition (Florez et al. 2017).
Distinct morphological traits of host macroalgae appear to underlie substantial variation in the uniqueness of their associated epiphytic communities. Callophyllis sp. exhibited the highest proportion of unique epiphytic taxa (50.1%), while Ahnfeltiopsis flabelliformis harbored the lowest (17.4%), as shown in Supplementary Fig. S2. A closer examination of morphological features in Table 1 reveals potential explanations for this contrast. Callophyllis sp. possesses a sheet-like thallus structure with a minimal cortical cell layer thickness (1–3 cells) and a soft, fleshy surface texture—characteristics that may create a broad, accessible microhabitat favorable to diverse and host-specific colonization. In contrast, A. flabelliformis, while also having a thin cortical layer (3–4 cells), features a more coarsely branched form and a firmer, cartilaginous surface, which could limit attachment niches or favor more generalist colonizers already abundant in surrounding waters. This structural distinction likely results in fewer unique epiphytes being retained exclusively by A. flabelliformis. These findings support the broader hypothesis that specific macroalgal traits, beyond just cortical layer thickness, such as thallus architecture and surface properties, can significantly shape the assembly of epiphytic communities by modulating habitat complexity and colonization filtering mechanisms.
Functional implications of epiphyte composition variation
The clear distinction in taxonomic composition between cortical morphological groups revealed by SIMPER analysis further illustrates potential functional differences between host-associated communities (Table 2). Thick-layered hosts supported arthropod-dominated communities likely contributing to higher trophic complexity, facilitating ecological interactions such as predation, grazing, and nutrient recycling (Wahl 2008, Chemello and Milazzo 2002). Thin-layered hosts with microalgae-dominated communities, particularly diatoms, may influence primary productivity, surface biofilm stability, and initial stages of ecological succession (Littler and Littler 1984, Lenzo et al. 2023, Xie et al. 2023).
Thus, host-driven community composition differences may extend beyond mere species diversity to functional divergence in ecosystem services provided by epiphytes. Future research should explicitly quantify these functional differences, such as rates of nutrient cycling, productivity, and resilience to environmental stressors, to fully understand ecological consequences of epiphyte community shifts driven by host traits.
Future perspectives
This study demonstrates that cortical cell layer thickness is a key morphological trait structuring epiphyte community on macroalgae, influencing species composition, richness, and PD. By applying SoEM, we reveal fine-scale community differences linked to host traits beyond traditional functional-form classifications, advancing a trait-based framework for host-epiphyte ecology. Our findings indicate that habitat filtering, stochastic colonization, and niche differentiation act together to shape community assembly, with their relative influence modulated by host morphology. Limitations such as limited sample replication per species and reliance on literature-based trait measurements highlight the need for replication and direct trait quantification.
Future research should expand sampling across seasons and regions, incorporate direct measurements of morphological and chemical traits, and integrate functional trait analyses of epiphytes. Coupling SoEM with experimental manipulations and long-term monitoring will help disentangle the interplay of filtering, competition, facilitation, and stochasticity. Such integrative approaches will refine our understanding of host-epiphyte assembly processes and guide conservation and restoration strategies for sustaining biodiversity and ecosystem function in coastal environments.
PDF Links
PubReader
ePub Link
Full text via DOI
Download Citation 
Supplement
Print 
