Insights into species diversity of associated crustose coralline algae (Corallinophycidae, Rhodophyta) with Atlantic European maerl beds using DNA barcoding

Cristina Pardo *, Ignacio Bárbara ¹, Rodolfo Barreiro ² & Viviana Peña ³

Grupo BioCost, Dpto. de Bioloxía, Facultade de Ciencias, Univ. da Coruña, 15071 A Coruña, Spain

* Corresponding author: cristina.pardo.carabias@udc.es, http://orcid.org/0000-0002-2534-3961

¹ Ignacio.barbara@udc.es, http://orcid.org/0000-0003-1779-0224

² rodolfo.barreiro@udc.es, http://orcid.org/0000-0002-6049-5215

³ vpena@udc.es, http://orcid.org/0000-0001-7003-3850

INTRODUCTIONTOP

Atlantic European rhodolith beds, also known as maerl beds, are marine coastal habitats of high diversity composed of unattached non-geniculate coralline red algae —maerl/rhodolith— mixed with gravel, shells, and pebbles overgrown by crustose coralline algae (CCA; v.gr., Cabioch 1969, 1970; Bosence 1976; Peña & Bárbara 2009; Hall-Spencer & al. 2010; Adey & al. 2015). The life-history of coralline red algae involves an alternation among three phases: a haploid gametophyte, a diploid carposporophyte —carried by the haploid female gametophyte after fertilization of carpogonium—, and a diploid tetrasporophyte (Irvine & Chamberlain 1994). In two major Atlantic European maerl species, Phymatolithon calcareum (Pallas) W.H.Adey & D.L.McKibbin and Lithothamnion corallioides (P.L.Crouan & H.M.Crouan) P.L.Crouan & H.M.Crouan, sporangial conceptacles have been occasionally reported —usually in free-living growth forms, and rarely in encrusting plants—, while gametangial conceptacles were even more rare (i.e., sexual conceptacles), and were only found as associated CCA in maerl beds from Brittany (Cabioch 1969, 1970; Adey & McKibbin 1970; Maggs 1983; Woelkerling & Irvine 1986; Irvine & Chamberlain 1994; Mendoza & Cabioch 1998; Peña & Bárbara 2004; Peña & al. 2014). Several authors (Lemoine 1910; Cabioch 1969, 1970; Freiwald 1995; Mendoza & Cabioch 1998) have pointed out the contribution of the CCA to the development of further unattached maerl plants: after germination of the carpospores produced in CCA, juvenile tetrasporophytic plants grow also as crusts, and afterwards erect branches formed by these crusts break off, and give rise to unattached maerl plants. In Breton maerl beds, this type of recruitment was reported as the dominant (Cabioch 1969, 1970; Mendoza & Cabioch 1998). Nonetheless, others authors (Bosence 1976; Woelkerling 1988; Peña & al. 2014) have reported that fragmentation of free-living maerl thalli is the main method of propagation in the maerl beds.

In recent years, several molecular studies have been focused on the diversity and systematics of maerl-forming species in Atlantic Europe (Carro & al. 2014; Hernández-Kantún & al. 2014, 2015a, 2015b; Pardo & al. 2014; Peña & al. 2015b). However, the diversity and specific composition of their associated CCA are little-known. In Brittany, two DNA barcoding works found crustose plants of Phymatolithon calcareum and Phymatolithon lamii (Me.Lemoine) Y.M.Chamb., which showed that species composition of associated CCA could share some similarities with unattached maerl plants (Peña & al. 2014, 2015b). Based on previous results obtained for maerl-forming species of two study areas from Brittany and Galicia (Pardo & al. 2014), we identified their associated CCA using a combination of DNA barcoding —COI-5P— and morpho-anatomical features. Apart from the species diversity of associated CCA, we compared the composition of both growth-forms (maerl and CCA) in order to know if the associated CCA could be involved in the recruitment of new unattached maerl plants.

MATERIAL AND METHODSTOP

Sampling collectionTOP

A total of 16 CCA specimens, all of them epilithic over pebbles, were collected in two Atlantic European maerl beds in 2011 (table 1, fig. 1): Molène Archipelago —Brittany, France— and Ons Archipelago —Galicia, Spain—. Both archipelagos are located in two protected marine areas: Parc Naturel Marin d’Iroise —PNMI—, and Parque Nacional Marítimo Terrestre das Illas Atlánticas de Galicia —PNMTIAG—, respectively. Collections were carried out by dredging at < 10 m in Brittany, and SCUBA diving at 13 m in Galicia, as complementary to maerl-forming species surveys (Carro & al. 2014; Pardo & al. 2014; Peña & al. 2014). After collection, material was air-dried, and vouchered in silica. Specimens were observed and photographed under stereomicroscope, and were deposited in the herbarium SANT (Thiers 2016; table 1).

Molecular studiesTOP

For the DNA extraction, a subsample was obtained by grounding of a small portion of each plant selected, choosing free epiphytes areas under stereomicroscope. DNA was extracted and purified using the DNeasy® Blood & Tissue Kit —Qiagen, Valencia, CA— following manufacture’s recommendations. A fragment of 664 bp of the 5’ end of the mitochondrial gene cytochrome oxidase I —COI-5P— was amplified using primers GazF1 and GazR1 (Saunders 2005). Amplifications PCR were performed in a Biometra TProfesional Basic thermocycler following Saunders & McDevit (2012). Amplification success was evaluated by running the reactions in agarose gels. After excess of primers and nucleotides were removed with shrimp alkaline phosphatase and exonuclease I enzymes. PCR products were sequenced using the Sanger method at Macrogen facilities —http://www.macrogen.com—. The sequences obtained were checked, edited and aligned with the program Geneious v. 5.6.6 —Biomatters, New Zealand—, and lodged in BOLD and GenBank (table 1). Haplotypes obtained were searched for matches in BOLD and GenBank databases to their molecular taxonomic identification (table 2). For the molecular analyses, publicly available COI-5P sequences for both maerl-forming and CCA taxa, particularly from Atlantic Europe, were included (Carro & al. 2014; Pardo & al. 2014; Peña & al. 2014, 2015a, 2015b; Hernández-Kantún & al. 2015a), as well as a sequence of Corallina officinalis L. as outgroup (Pardo & al. 2015; table 3). In total, 30 sequences were used to generate a Maximum Likelihood —ML— tree. A bootstrap with 1,000 replicates was applied in Mega v. 6.0 with defaults settings (Tamura & al. 2013), using the General Time-Reversible model, Gamma distributed with Invariant sites —GTR+G+I—, identified previously in ModelTest v. 2.1.3 (Darriba & al. 2012) as the best-fitting substitution model using an Akaike Information Criterion —AIC— and Bayesian Information Criterion —BIC.

Morphological studiesTOP

CCA selected for molecular studies were examined under stereomicroscope, and representative fragments were examined by Scanning Electron Microscope —SEM, model JEOL JSM 6400, Univ. da Coruña—. Vegetative and reproductive features considered diagnostic and anatomical terminology followed Irvine & Chamberlain (1994).

RESULTSTOP

Molecular resultsTOP

COI-5P sequences were obtained for the 16 CCA specimens collected (table 1). They were grouped in 13 haplotypes —CCA_hap-1 to CCA_hap-13—: five from Molène Archipelago and eight from Ons Archipelago (table 1). These 13 haplotypes were delimited in nine CCA species belonging to four genera (fig. 2): Phymatolithon sp. 5 —CCA_hap-2—, Phymatolithon sp. 6 —CCA_hap-3—, Phymatolithon calcareum —CCA_hap-1—, Phymatolithon lamii —CCA_hap-4—, Lithothamnion sp. 2 —CCA_hap-5 and 6, both differing in 1 bp—, Mesophyllum sp. 3 —CCA_hap-7 and 8, both differing in 1 bp—, Mesophyllum sp. 4 —CCA_hap-9 and 10, both differing in 1 bp—, Lithophyllum hibernicum —CCA _hap-12 and 13—, and Lithophyllum sp. 3 —CCA_hap-11— (figs. 3, 4).

Four CCA species were detected in the Breton maerl bed, and were successfully identified by matching with BOLD and GenBank databases (table 2): Phymatolithon calcareum, Phymatolithon lamii, and Lithophyllum hibernicum. Likewise, the species named as Phymatolithon sp. 6 —CCA_hap-3; table 1) scored with Phymatolithon sp. 3AT crust from NW Atlantic —Canada—. By contrast, the Galician CCA haplotypes did not match with any available sequence, but according to the ML-tree performed they were resolved into five different taxa (fig. 2): Phymatolithon sp. 5, Lithothamnion sp. 2, Mesophyllum sp. 3, Mesophyllum sp. 4, and Lithophyllum sp. 3.

Morpho-anatomical examinationTOP

Phymatolithon sp. 5 (fig. 3a) showed domed epithallial cells, and subepithallial cells as short as cells subtending them (fig. 5a); cell fusions between contiguous filaments were common (fig. 5b); multiporate asexual conceptacles showed white pore plates (fig. 3d), surrounded by rim and with roof slightly depressed (fig. 5c). Phymatolithon sp. 6 (figs. 3b, e, g, i, j, 4q) and Phymatolithon calcareum (fig. 3h, k) showed sexual uniporate conceptacles, white coloured in surface view, flushed with thallus surface to sunken in Phymatolithon sp. 6, while raised in Phymatolithon calcareum. Phymatolithon lamii (fig. 3c, f) showed asexual multiporate conceptacles with white pore plates flushed with thallus surface or slightly sunken.

Lithothamnion sp. 2 (fig. 4a, b) showed epithallial cells mainly flared and somewhat flattened, subepithallial cells as long or longer than cells subtending them (fig. 5d, e). Cell fusions between cells of contiguous filaments were common (fig. 5f); asexual multiporate conceptacles with sunken pore plate and white in colour (fig. 4d, e).

Mesophyllum sp. 3 (fig. 4c, g) showed epithallial cells somewhat flared, with subepithallial cells longer than cells subtending them (fig. 6d); cell fusions between contiguous filaments (fig. 6e); multiporate asexual conceptacles with sunken pore plates (figs. 4f, k, 6a); conceptacle chambers were elliptical, 90 µm height by 137 µm in diameter, and conceptacle roof of 11 µm thickness (fig. 6b, c). Mesophyllum sp. 4 (fig. 4h, j) showed epithallial cells somewhat flared, subepithallial cells as long as or longer than cells subtending them, and cell fusions common between contiguous filaments (fig. 6f); uniporate sexual conceptacles were prominent on thallus surface (fig. 4l, n); multiporate asexual conceptacles showed sunken pore plates (fig. 4l, n), and also raised poro plates at the same time in the plant CPVP-769 (fig. 4n); conceptacle chambers were elliptical, 77 µm height by 189 µm in diameter, and conceptacle roof of 40 µm thickness (fig. 6g, h).

Lithophyllum hibernicum (fig. 4o-q), and Lithophyllum sp. 3 (fig. 4i, m), showed uniporate conceptacles slightly flushed with thallus surface.

DISCUSSIONTOP

In our study, nine CCA species were recorded associated with two Atlantic European maerl beds: Molène —Brittany— and Ons —Galicia— Archipelagos. Despite that these CCA collections consisted on epilithic crustose corallines growing partial or entirely over pebbles, we are aware that could be interpreted as “rhodolith” given that this term is extensively applied to unattached nodules with non-algal core (Irvine & Chamberlain 1994; v.gr. see fig. 6 in Basso & al. 2009). While the four CCA species associated with the Breton maerl bed matched with available sequences from northern Atlantic coast —Phymatolithon calcareum, Phymatolithon lamii, Lithophyllum hibernicum, and Phymatolithon sp. 3ATcrust—, the five CCA taxa from the Galician maerl bed did not obtain any record. However, according to the molecular and morphological data obtained, CCA collected in Galicia were identified up to genus level as Phymatolithon sp. 5, Lithothamnion sp. 2, Mesophyllum sp. 3, Mesophyllum sp. 4, and Lithophyllum sp. 3. The absence of any match with public databases containing more than 4,000 COI-5P sequences of coralline red algae pointed out the cryptic diversity still uncovered in coralline algae.

The nine taxa delimited corresponded to the same four maerl-forming genera in Atlantic Europe —Lithothamnion, Phymatolithon, Mesophyllum, and Lithophyllum (Irvine & Chamberlain 1994; Pardo & al. 2014; Hernández-Kantún & al. 2015a)—. Nevertheless, the number of associated CCA species recorded was higher than the number of maerl species cited in both studied areas (Carro & al. 2014; Pardo & al. 2014; Peña & al. 2014, 2015b): four CCA vs. two maerl species in Molène Archipelago —Phymatolithon calcareum and Lithothamnion corallioides—; and five CCA vs. three maerl species in Ons Archipelago — Phymatolithon calcareum, Phymatolithon lusitanicum, and Lithothamnion corallioides—. The similarity in the species composition between the both growth-forms is approximately 25% in the Breton maerl bed, but totally dissimilar —0% shared species— in the Galician bed.

In Molène Archipelago the dominant genus in the CCA studied was Phymatolithon, while in Galicia was Mesophyllum, this latter having a southern distribution along the Atlantic European coasts (Guiry & Guiry 2016). This gradual replacement of both species with latitude has been also observed in the composition of major maerl-forming species in the OSPAR maritime area, including Brittany and Galicia (Carro & al. 2014; Pardo & al. 2014).

According to the literature, Phymatolithon calcareum is considered as a major maerl-forming species in Atlantic Europe (Irvine & Chamberlain 1994; Carro & al. 2014; Pardo & al. 2014). Our record of CCA with uniporate sexual conceptacles in the Breton maerl bed agrees with previous reports of gametophytes —including the carposporophyte— under crustose forms, all of them from this region (Mendoza & Cabioch 1998; Peña & al. 2014). This observation would corroborate that, at least in Brittany, the gametangial stage of Phymatolithon calcareum occurs as associated crusts —i.e. attached—, while unattached maerl plants of this species are sporophytes (Cabioch 1969, 1970; Mendoza & Cabioch 1998). However, in adjacent areas such as the British Isles or Atlantic Iberian Peninsula where maerl beds have been intensively studied, fertile gametophytes —crusts or unattached— have been never recorded yet (Irvine & Chamberlain 1994; Peña & al. 2014, 2015b). Given that all Atlantic European records of gametangial stages of Phymatolithon calcareum are still restricted to Brittany (Mendoza & Cabioch 1998; Peña & al. 2014), we presume that this Atlantic region plays an important role in the life-history of this major maerl-forming species.

Phymatolithon lamii is a crustose species occurring from the intertidal to subtidal rocky shores of the Atlantic coasts from Arctic Norway to South Atlantic Iberian Peninsula (Chamberlain 1991; Irvine & Chamberlain 1994; Peña & al. 2015b). Our collection in the Breton maerl bed as associated CCA corroborated recent records of this species from the same maerl bed —Molène archipelago (Peña & al. 2015b).

A recent molecular study (Hernández-Kantún & al. 2015a) confirmed that Lithophyllum hibernicum is a common intertidal crust that has been repeatedly misidentified in the Atlantic European literature as Lithophyllum incrustans, which also occurs as both encrusting and maerl forms —i.e., Pardo & al. (2014), as Lithophyllum fasciculatum (Lam.) Foslie and Lithophyllum dentatum (Kütz.) Foslie—. In our study, Lithophyllum hibernicum was recorded subtidally in the Breton maerl bed as associated CCA.

A non-common morpho-anatomical observation was detected in the Galician specimen identified as Mesophyllum sp. 4 (CPVP-769): both types of conceptacles —uniporate sexual and multiporate asexual— occurred in the same crust surface. A similar observation was found in the Galician maerl species Mesophyllum sphaericum V.Peña & al. in Galicia (see fig. 2a in Peña & al. 2011).

Regarding the development of new maerl plants in the Atlantic European maerl beds, two main mechanisms have been suggested in the literature: i) vegetative fragmentation of own free-living maerl plants (Bosence 1976; Woelkerling 1988; Peña & al. 2014); ii) involvement of associated CCA in the recruitment of new maerl plants —i.e., detached branches of crusts (Lemoine 1910; Cabioch 1969, 1970; Freiwald 1995; Mendoza & Cabioch 1998)—. Our study pointed out that the latter hypothesis could be applied, but only for Phymatolithon calcareum in the Breton maerl beds, where gametangial —plus carposporangial— and sporangial stages occur under different growth-forms —CCA and maerl, respectively (Cabioch 1969, 1970; Mendoza & Cabioch 1998; Peña & al. 2014)—. However, in the Galician maerl beds the species composition of associated CCA in comparison with the main maerl-forming taxa —Phymatolithon calcareum, Phymatolithon lusitanicum, and Lithothamnion corallioides— appeared to be different. This together with the absence of maerl species records bearing gametangial —and carposporangial— structures in this region —attached or unattached—, and the occasional records of sporangial conceptacles observations (Peña & Bárbara 2004), the fragmentation of own free-living maerl plants should be considered as the main propagation method in the Galician maerl beds.

CONCLUSIONTOP

The present study carried out in two protected Atlantic European maerl beds pointed out that maerl beds harbours a high species diversity of associated CCA, which may not share the same species composition that maerl-forming algae. Also a high cryptic diversity was found in the associated CCA, as occurs in the maerl-forming species (Carro & al. 2014; Pardo & al. 2014) and in geniculate coralline algae (Walker & al. 2009; Pardo & al. 2015). Further extensive studies in other European maerl beds are necessary to understand this unexpected diversity of associated CCA. Nonetheless, according to our results, the crustose growth-forms should be taken into account in management actions of Atlantic European maerl beds.

ACKNOWLEDGEMENTSTOP

This research was supported by Ministerio de Ciencia e Innovacion of Spain (CTM2010–18787, partially co-founded by FEDER, Fondo Europeo de Desarrollo Regional) and Xunta de Galicia (10MMA103003PR). Cristina Pardo acknowledges financial support by Xunta de Galicia (Axudas á etapa predoutoral do Plan I2C, 2011) and Spain’s Ministerio de Educación (Programa FPU, 2010). Viviana Peña acknowledges support by Universidade da Coruña (Contrato programa-Campus Industrial de Ferrol). We acknowledge to Parc Naturel Marin d’Iroise and Parque Nacional Marítimo Terrestre das Illas Atlánticas de Galicia by allowing our surveys. We are also very grateful to Jacques Grall for fieldwork collaboration in Molène Archipelago (Brittany).

REFERENCESTOP