Epiphytism is a widespread strategy in marine benthic communities on rocky bottoms where the competition for space is high (Rindi and Guiry 2004, Belegratis et al. 1999). Macroalgae provide an ideal primary substratum for the growth of epiphytic algae (Lutz et al. 2010), which significantly contribute to primary productivity (Belegratis et al. 1999) and, in turn, provide habitat for other organisms (Karez et al. 2000). Epiphytism entails both benefits and disadvantages for the host plant (Wahl 1989). In general, the detrimental effects of epiphytes are indirect (Fricke et al. 2011). They increase drag on hosts, increasing their risk of breakage and dislodgement, particularly in hydrodynamically stressful environments (Anderson and Martone 2014). Furthermore, growth rates and reproductive output of heavily epiphytized hosts may decrease as a result of the reduction in irradiance reaching the frond (Sand-Jensen 1977) and the partial nutrient depletion of water after its passage through the epiphytic layer (Muñoz et al. 1991). Favourable consequences of epiphytism have also been postulated, including herbivore avoidance and protection against desiccation for intertidal hosts (Anderson and Martone 2014).

Gelidium corneum (Hudson) J.V. Lamouroux (formerly G. sesquipedale) is a very important benthic primary producer in the Cantabrian Sea (Rico and Fredriksen 1996, Gorostiaga et al. 1998, Díez et al. 2003) and forms extensive stands at depths down to 9-15 m in the subtidal zone off the Basque coast (Gorostiaga 1995, Borja et al. 2004). This canopy-forming species plays an important ecological role by providing habitat and shelter for many other organisms (Borja et al. 2004, Bustamante et al. 2014). The algae Dictyota dichotoma, Plocamium cartilagineum and Asparagopsis armata (Santos 1994) and the bryozoans Electra pilosa and Scrupocellaria spp. (Bustamante et al. 2014) have been reported as its main epibionts. The seasonal pattern of G. corneum growth substantially influences the development of the epiphytic load. This perennial species with apical growth (about 7-8 cm yr^–1) reaches its maximum frond elongation rate in late spring and summer (Gorostiaga 1994). Under the high light availability of this period, G. corneum forms a dense canopy as a result of the fast process of frond branching, which favours the settlement and growth of epiphytes (Gorostiaga 1994, Santos 1994). In autumn and winter a considerable loss of branches occurs due to the effect of storms, resulting in a decrease in the standing stock of G. corneum (Gorostiaga 1994) and in its epiphytic load.

G. corneum used to be the main raw material for agar extraction along the Atlantic shores of Spain, Portugal and Morocco, providing these countries with almost 50% of the world harvest in the 1980s (McHugh 1991). In the particular case of the Basque coast, G. corneum was only exploited by scuba diving in the 1970s and between 1992 and 1999 (Borja et al. 2013). However, since the turn of the 21^st century many G. corneum beds have suffered a significant decline on several coastal stretches of the Basque coast (Díez et al. 2012, Borja et al. 2013), where this macrophyte exhibits morphological (Díez et al. 2012) and biochemical (Quintano et al. 2013) symptoms of stress. It has been suggested that climate-driven factors such as increased seawater temperature in combination with changes in local factors that include light and nutrients availability and wave energy are involved in the declines of G. corneum beds (Díez et al. 2012, Borja et al. 2013). Since epiphytism potentially affects the growth and survival of G. corneum and may interact with other stressors, information about factors affecting epiphyte development is valuable. Likewise, epiphyte composition and abundance on G. corneum are liable to change according to environmental conditions, and thus this particular flora can play a greater role as a bioindicator (i.e. stress symptoms), as has been reported for other macrophytes (Russell et al. 2005, Giovannetti et al. 2010).

Wave exposure and irradiance have long been recognized as major factors in regulating the distribution of algae (Lüning 1990). In this regard, epiphyte growth and colonization are generally negatively affected by increased wave energy, whereas increased light favours epiphyte development (Michael et al. 2008). In this study we examine two indirect factors, coastal orientation (north and northwest; N and NW) and depth (3 and 7 m), as proxies for wave exposure and light availability, respectively. We test the effects of these two factors and their interaction on the total epiphytic load, alpha diversity (species richness, Shannon, Simpson and evenness measures) and multivariate structure of the epiphytic flora growing on the red alga G. corneum in subtidal waters off the Basque coast. We expected the abundance and diversity of epiphytes to be negatively correlated with wave exposure (higher loads in north facing coastal stretches) and positively correlated with light intensity (higher loads on shallow bottoms).

A total of 12 macroalgae taxa were recorded growing over G. corneum, most of them being morphologically simple forms belonging to the Ceramiaceae, Callithamniaceae and Bonnemaisoniaceae families (Table 2). The total epiphytes/G. corneum biomass ratio was higher at N-facing locations than at NW-facing locations (Fig. 2). However, differences in relation to depth were only found at N-facing locations, with the total epiphytes/G. corneum biomass ratio being higher at 3 m depth (Fig. 2). In this regard, univariate PERMANOVA analyses detected a significant interaction between coastal orientation and depth (Pseudo-F_1,4=42.31, p=0.004). With respect to species richness (Fig. 3A), the number of taxa in N-facing locations was significantly higher than in NW-facing locations (Pseudo-F_1,4=18.89, p=0.013). Diversity also differed from one coastal orientation to the other. Both Shannon (Fig. 3B) and Simpson (Fig. 3C) diversities were significantly higher at N-facing locations (Pseudo-F_1,4=87.43, p=0.001 and Pseudo-F_1,4=96.09, p=0.001, respectively). Finally, Pielou J’ evenness (Fig. 3D) was also higher at N-facing locations (Pseudo-F_1,4=13.79, p=0.027). In all cases depth showed no effect on community measures.

Principal coordinates analysis (PCO) relates the pattern of distribution of samples to the macroalgae taxa recorded as G. corneum epiphytes (Fig. 4). PCO axis 1, which explained 72.2% of the total variation, mainly reflected differences related to the interaction between coastal orientation and depth. The points for the samples collected at NW-facing locations, at depths of both 3 and 7 m, are located towards the left of the plot, whereas the samples collected at N-facing locations are placed towards the right, where samples collected at 3 and 7 m are separated. Indeed, PERMANOVA analyses on multivariate data detected a significant interaction between coastal orientation and depth (Table 3). Pairwise comparisons indicated that epiphytic abundance and composition at 3 and 7 m was significantly different within N-facing locations, whereas no differences were found between epiphytes from the two depths at NW-facing locations. Therefore, three different groups were distinguished: Group 1 (N; 3 m), Group 2 (N; 7 m) and Group 3 (NW). The two-dimensional projection is highly representative of the full data cloud since it explains 80.5% of the total variation. Spearman correlations show that P. cartilagineum, D. dichotoma, Acrosorium ciliolatum and Falkenbergia rufolanosa are positively correlated with PCO axis 1, whereas Callithamnion tetragonum has a negative relationship (Fig. 4).

The SIMPER routine identified Group 3 (NW) as the most distinctive group, with average dissimilarities of 31.11% and 24.35% with respect to Group 1 (N; 3 m) and Group 2 (N; 7 m), respectively (Table 4). This procedure identified P. cartilagineum, D. dichotoma, A. ciliolatum and F. rufolanosa as the species which contributed most to the dissimilarities between Group 3 and the other two groups (Table 4). In the case of the dissimilarity between Group 1 and Group 2, the contribution of A. tenuissimum and A. tripinnatum was also relevant. Regarding the abundance of different taxa, P. cartilagineum, D. dichotoma, and A. ciliolatum showed the highest biomass ratios in all the groups whereas other taxa only appeared in one group: Trailliella intricata (Group 1), Antithamnionella ternifolia (Group 2), C. tetragonum and Antithamnion nipponicum (Group 3) (Table 5). Group 1 registered the highest biomass ratios for most taxa, which decreased gradually towards Groups 2 and 3.

We expected the abundance and diversity of epiphytes to be negatively correlated with wave exposure (coastal orientation) and positively correlated with light intensity (negatively with depth). Our results partially support this hypothesis. Significant interactive effect of these two factors was observed for species composition and abundance of epiphytic flora. G. corneum populations living at the most exposed locations had the lowest epiphyte load and the least diverse community. However, depth only had a significant effect on the distribution of epiphytes below a certain threshold of wave action, with the epiphytic load being 30-40% greater on shallow bottoms. This pattern suggests that in shallow waters hydrodynamic pressure on epiphytic flora could be a more determinant factor in regulating epiphytes than light availability.

The composition of epiphytes on G. corneum is found to be dominated by P. cartilagineum, D. dichotoma and A. ciliolatum. The first two of these taxa have been described as common annual epiphytes accompanying G. corneum (Santos 1994, Gorostiaga 1995). Furthermore, hydrodynamism appeared to have a significant influence on the distribution of some morphologically simple taxa, whereas F. rufolanosa, A. tenuissimum, A. tripinnatum, A. ternifolia, and G. flaccida showed a preference for relatively lower-energy conditions, C. tetragonum and the alien species A. nipponicum were more abundant at the most exposed locations. It is also noteworthy that the epiphytic community on G. corneum found in this study was extraordinary low, especially compared with that found in northwestern Spain (Galicia) for other canopy-forming species thriving in semi-exposed to exposed areas (Otero-Schmitt and Pérez-Cirera 1996). The latter authors found 30 to 70 different species of epiphytic flora, mainly rhodophytes, growing in four different species from the genus Cystoseira. In this sense, the frondose morphology and cartilaginous texture of G. corneum contrasts with the tree-like morphology and leathery texture of the species of the genus Cystoseira, which might enhance the notable differences found in the diversity of epiphytic flora between the two macrophytes. Differences in epiphytic loads regarding morphological features and longevity of the host have been reported in the case of seagrass species (Lavery and Vanderklift 2002).

With respect to the epiphytic loads in the present study, very exposed (NW) areas showed significantly lower epiphytic abundances than exposed (N) areas. This pattern is consistent with other studies from northwestern Spain (Galicia) (Peteiro and Freire 2013) and northern Europe (Pedersen et al. 2012, Baer and Stengel 2014), which found the epiphytic loads on canopy-forming species to be lower on more exposed coastal stretches. Moreover, very exposed locations (NW) in our study showed lower species richness and diversity, coinciding with that found for epiphytic flora (Kersen et al. 2011) and fauna growing on canopy-forming macrophytes (Norderhaug et al. 2012, 2014), according to whom lower species diversities are found under frequent or even rare ecological disturbances (e.g. wave action). Indeed, wave exposure is one of the major factors directly or indirectly influencing the structure of algal assemblages (Rindi and Guiry 2004, Nishihara and Terada 2010, Borja et al. 2013). Although increasing wave action is a key factor in supplying nutrients for macroalgae (Martins et al. 2013), hydrodynamic pressure can also remove epibionts and prevent them from settling (Kersen et al. 2011). The mechanical stress produced by higher wave-action (Kraufvelin 2007) could favour the removal of epiphytes at NW oriented locations in our study, whereas at N oriented locations lower hydrodynamic levels could favour their development. Similarly, the lower species richness and diversity and the smaller quantities of morphologically simple forms found at NW-facing locations in this study might be explained by the higher sensitivity of some taxa to the effects of wave action (Nishihara and Terada 2010). However, there are some exceptions, such as C. tetragonum and A. nipponicum, which showed affinity to more exposed coastal stretches, as also reported by Secilla (2009).

Light is a prime factor regulating the distribution of algae since the survival and productivity of photosynthetic organisms are closely linked to light intensity and quality (Prado et al. 2007, Michael et al. 2008). Increasing irradiance levels usually favour algal development and growth (Altamirano et al. 2000, Colombo-Pallotta et al. 2006). Therefore, due to the exponential decay of light intensity with increasing depth (Gross et al. 2003, Martins et al. 2013), shallower waters may be expected to contain higher epiphyte loads (Muñoz and Fotedar 2010, Tsirika et al. 2007). Our findings partially support this hypothesis. At the N-facing locations (with less wave energy) the epiphytic load was significantly greater in shallow waters than in deeper waters. However, our results do not show this pattern in the NW-facing locations, where the potentially greater epiphytic growth under high light conditions may be mitigated by the higher wave action (Liria et al. 2009, Galparsoro et al. 2010). Nevertheless, it should be noted that factors other than light, such as water movement, temperature, sedimentation and nutrient availability, may vary with depth. Therefore, experimental field and laboratory studies should be conducted in order to separate the effect of light availability on the distributional patterns of epiphytes from other depth-related factors.

In conclusion, our findings suggest that the interactive effects of wave action and light availability seem to be involved in the diversity, abundance and composition patterns of the epiphytes growing on G. corneum. Since epiphytes may cause some damage to their hosts, future field and laboratory experiments should be carried out to assess the effects of these assemblages on G. corneum populations. Furthermore, considering the significant decline undergone by G. corneum in pristine coastal stretches of the Basque coast (Díez et al. 2012, Borja et al. 2013), the combined effects of epiphytism and potential environmental stressors on this canopy-forming species would need to be separated to generate better predictive models of the distribution of G. corneum under hypothetical climate change scenarios.