Notes on the marine algae of the Bermudas. 15. Dichotomaria huismanii (Galaxauraceae, Rhodophyta), a new species in the D. marginata complex from the western Atlantic

Standard methods

Offshore collections from Bermuda and St. Croix, Virgin Is. were made using SCUBA, with site locations marked with a Garmin™ eTrex H (Olathe, KS, USA). Small portions of each specimen were dried on silica gel for DNA extraction and other samples were preserved in 4–5% formalin in seawater for sectioning, the remainder being pressed onto herbarium paper as a permanent voucher. Liquid preserved samples were decalcified in 8% HCl, sectioned with a model 880 American Optical freezing microtome (San Diego, CA, USA), with sections mounted in 30% corn syrup with acidified 1% aniline blue in a ratio of 20:1 with a few drops of formalin as a medium preservative. Live specimens chosen for DNA analysis were photographed using a Canon Powershot s90 digital camera (Canon Inc., Tokyo, Japan) and dried herbarium specimens were scanned on an HP 309a Photosmart Premium scanner (Hewlett-Packard Company, Palo Alto, CA, USA). Photomicrographs were taken using Zeiss Axioskop 40 microscope (Carl Zeiss, Oberkochen, Germany) equipped with a model 11.2 Spot InSight 2 digital camera (Diagnostic Instruments, Sterling Heights, MI, USA). The digital images were composed in Adobe Photoshop™CS6 v. 13.0.1 (Adobe Systems, San Jose, CA, USA). Voucher specimens of some numbers are deposited in MICH, NY, the Bermuda Natural History Museum and CWS’s personal herbarium, the holotype in the first. Herbarium abbreviations follow the online Index Herbariorum (http://sweetgum.nybg.org/ih/) and standard author initials were taken from Brummitt and Powell (1992).

Molecular methods

Silica dried samples for DNA analysis were ground in liquid nitrogen and stored at -20°C. DNA was extracted from 0.1–0.5 μl ground material using the Sigma-Aldrich (St. Louis, MO, USA) GenElute Plant Genomic Miniprep Kit according to the manufacturer’s protocol, with 500 μl of modified lysis solution (50 μl 10% TWEEN 20 and 5 μl of 20 mg ml^-1 ProK), as well as 1 h of incubation at 23°C followed by 20 min on ice (Saunders and Druehl 1993).

DNA was amplified via polymerase chain reaction (PCR) with the Takara Ex-Taq DNA polymerase kit (PanVera, Madison, WI, USA) in an Eppendorf AG Mastercycler epGradient thermal cycler (Eppendorf, Hamburg, Germany). To assign all specimens to species groups, two oligonucleotide primers were used for both sequencing and amplification of the COI-5P mitochondrial marker, GWSFn (Le Gall and Saunders 2010) and GWSRx (Saunders and McDevit 2012). A denaturation cycle of 94°C for 4 min was followed by 38–42 cycles of 94°C for 1 min, 45°C for 1 min, 72°C for 1 min, and a final extension of 72°C for 7 min. Specimens were likewise sequenced for the plastid-encoded rbcL operon. Amplification and/or sequencing reactions for rbcL were conducted using the primers (F43-R753; F615-RrbcS start) and the thermal profile referenced in Wiriyadamrikul et al. (2014). All amplified DNA was treated with the Qiagen (Redwood City, CA, USA) QIAquick PCR Purification Kit following the manufacturer’s protocol. The purified PCR product was sequenced at the Rhode Island Genomics and Sequencing Center using the Applied Biosystems Inc. 3130xl Genetic Analyzer (Life Technologies, Grand Island, NY, USA).

COI-5P barcode sequences from representatives in the Galaxauraceae, including some available through GenBank and those newly determined here, were aligned using the MUSCLE (multiple sequence comparison by log-expectation) alignment program in Geneious (v. 6.1.8 available from http://www.geneious.com). To visually characterize genetic variability among specimens, the UPGMA clustering algorithm was applied to the COI-5P alignment (42 specimens, 538 sites) with Tamura-nei-corrected distances (default setting). The resulting tree (Figure 1) was used to demarcate species groupings. Based on these genetically related groups (=species), and with comparative data available from GenBank, one specimen from each species and/or geographic location was selected for phylogenetic analysis using rbcL sequences (Figure 2). The best models of evolution for the individual gene regions rbcL (62 taxa, 1249 sites) were determined in jModelTest 2 (volume 2.1.5; Darriba et al. 2012). The selected phylogenetic model (GTR+I+G in both instances) was used to complete both maximum likelihood (ML) and Bayesian analyses for each gene. The ML phylogeny was estimated using the RAxML graphical user interface (Silvestro and Michalak 2011) with branch support calculated using 1000 bootstrap replicates. Bayesian analysis of rbcL was conducted in MrBayes v.3.2.2 (Ronquist and Huelsenbeck 2003) and run with four parallel chains (three heated+one cold) with branch lengths optimized during the run for one million generations. Stationarity was attained after the first 425,000 generations (burnin=4250 trees) and posterior probabilities were estimated based on the remaining trees. The rbcL phylogenetic tree (Figure 2) includes Liagora and Scinaia as outgroups, and was manipulated for presentation using FigTree software (http://tree.bio.ed.ac.uk/software/figtree/). Specimens used in our molecular analyses are recorded in Table 1.

Figure 1:

Species groups in the Galaxauraceae determined with UPGMA clustering of the COI-5P genetic barcode, with Scinaia and Liagora as outgroups. Sequences generated for this study are in bold text.

Figure 2:

Phylogeny of the Galaxauraceae based on maximum-likelihood analyses of rbcL sequences. Branch values are bootstrap supports for ML (1000 replicates) followed by Bayesian posterior probabilities expressed as a percentage of support out of 1. An asterisk (*) indicates 100% support for both robustness metrics while a dash (-) indicates support values <60%. Sequences generated for this study are in bold text.

Basionym

Corallina marginata J. Ellis et Solander 1786, p. 115, pl. 22 figure 6 (Figure 3).

Figures 3–6:

Dichotomaria marginata.

(3) Lectotype of Corallina marginata (=Dichotomaria marginata): Ellis and Solander illustration (1786, pl. 22 figure 6). (4) Sample from St. Croix (STX032), scale bar=2 cm. (5) Section through outer cortex showing subcortical cell bearing a pair of assimilatory outer cortical cells, scale bar=25 μm. (6) Surface view of outer cortical assimilatory cells (STX054), scale bar=25 μm.

Description

Rosy-brown plants to 10 cm tall and 13 cm across, composed of dichotomously branched flattened axes with thickened margins, these showing faint annulations; axes mottled in appearance when dried; internodes 4.7–13.7 mm long and 1.6–2.6 mm wide; axes branching dichotomously at angles of 72–89°; cortex of sporophytes composed of two layers of cells, outer cortical assimilatory cells subspherical to ovoid, 20–40 μm long and (21–) 33–49 μm diameter, and an inner layer of flared stalk cells each bearing 1–2 assimilatory cells, 15–36 μm long and 5–10 μm diameter (at narrowest point); stalk cells borne typically in pairs though occasionally singly on the outer surface of the inner cortex; stalk cells most commonly unicellular, rarely two-celled; inner cortex consisting of typically one, occasionally two, layers of transversely ovoid to rectangular cells, 50–115 μm diameter and 25–68 μm high; subcortical cells subspherical at margins; medullary filaments extending from subcortical cells, 7–13 μm diameter; gametangia and tetrasporangia not seen in our sequenced specimens.

Type locality

“… on the shore of one of the Bahama islands”, West Indies, western Atlantic Ocean (Ellis and Solander 1786).

Distribution

Formerly reported as pantropical, genetically verified specimens are known from Guadeloupe, Puerto Rico and St. Croix, USVI (West Indies).

Selected collections

St. Croix, Virgin Is. – T.R. Popolizio (TRP)/C.E. Lane (CEL)/E.D. Salomaki (EDS) 13-24-1 [STX032] (Figure 4), TRP/CEL/EDS 13-25-9 [STX054] (see Table 1 for collection details).

Remarks

As discussed above, two St. Croix collections genetically matched other sequenced specimens from Guadeloupe and Puerto Rico, as well as the description of Dichotomaria marginata provided by Papenfuss et al. (1982). This has allowed us to identify STX032 and STX054 (Figures 1 and 2) as true representatives of D. marginata in the West Indies, a species with a type locality in the Bahamas (Silva et al. 1996). All of the genetic collections made in St. Croix identified as D. marginata have the anatomy of sporophytes defined for the species, unfortunately none bearing tetrasporangia. Thus, we cannot provide tetrasporangial size and shape nor the anatomy or reproduction in gametophytes of these genetically significant D. marginata vouchers. Although there are reports about these features in earlier literature (e.g. Børgesen 1916, Howe 1918b, Taylor 1960, Papenfuss et al. 1982), without a molecular tie to this species where crypsis is now reported, we are hesitant to link that information here.

Similar to what Huisman et al. (2004) found for the species, tetrasporic St. Croix representatives of D. marginata have “the presence of a distinctly thickened margin (particularly noticeable when dried)” and a “cortex of the tetrasporophyte bearing paired subspherical cells (at least some with a small terminal spine)” (Figure 5). As was pointed out in the Ellis and Solander (1786) protologue, the flattened axes of D. marginata are more pronounced when removed from the water, the first collections being washed up and dried on a Bahamian beach. Our specimens have three, or occasionally four, layers of cells including stalk and outer assimilatory cells. The assimilatory cells of St. Croix D. marginata are subspherical to ovoid in cross-section (Figure 5), and appear circular to ellipsoidal in surface view (Figure 6). Some, but not a majority, of the assimilatory cells are apiculate, these cells usually paired on stalk cells that are mostly paired on outer subcortical cells (Figure 5). The subcortical cells are transversely ovoid to rectangular, and occasionally subspherical at the margins (Figure 5).

Howe (1918b) was the first to consider Galaxaura occidentalis Børgesen (type locality=Virgin Islands) as the gametophyte generation of G. marginata, and Børgesen (1920) and subsequent workers (Taylor 1960, Papenfuss et al. 1982) followed his supposition. Howe (1918a) mentioned gametophytic specimens in his account of G. marginata in the Bermuda flora, but it remains unclear whether he was referring to actual Bermuda gametophytes or rather to the gametophytes he was working on at the same time from the West Indies (Howe 1918b). We have been able to observe the gametophytic holotype of G. occidentalis [in C], most reasonably the alternate generation of the sporangial D. marginata, but this requires genetic confirmation. It would be helpful to find sporophytes and gametophytes of D. marginata in a single Bahamian or West Indian population to confirm this supposition, both generations matching genetic sequences already linked to this species from the West Indies.

A second genetic species, Dichotomaria sp. 1CRB, was collected in St. Croix and St. Kitts (Figures 1 and 2, Tables 1 and 2), and the St. Croix isolate represents the gametophytic stage of an undescribed species in the D. marginata complex. The St. Croix specimen has anatomical similarities with the type of G. occidentalis; unfortunately, it is genetically distinct from local and regional D. marginata. We only have two representatives of this species from the West Indies, thus until additional specimens are obtained and sequenced, Dichotomaria sp. 1CRB will remain undescribed. Whether its gametophytic stage represents what was described as G. occidentalis requires further investigation. Nevertheless, its sequence demonstrates that there are at least two species in the D. marginata complex in the West Indies.

Description

Rosy-brown plants to 8 cm tall and 12 cm across, composed of dichotomously branched, flattened axes with thickened margins above, the axes subterete below; axes mottled in appearance when dried, annulations, when present, faint; internodes 3–9 mm long and 1.2–2.0 mm wide; axes branching dichotomously at angles of (32–) 41–54 (–68°); cortex of sporophytes composed of two layers of cells, outer cortical assimilatory cells subspherical to ovoid, 28–50 μm length and 22–48 μm diameter, and an inner layer of flared stalk cells, each bearing one or two assimilatory cells, 23–45 μm length and 5–13 μm diameter at narrowest point; stalk cells borne typically in pairs though occasionally alone on the outer layer of the inner cortex; stalk cells unicellular; inner cortex consisting of two layers of transversely ovoid to rectangular cells, 24–44 (–70) μm long and (38–) 58–89 μm diameter; subcortical cells subspherical at margins; medullary filaments extending from subcortical cells, closely adherent and parallel, running longitudinally the length of the axis, 13–20 μm diameter; tetrasporangia and gametophytes unknown.

Etymology

The epithet “huismanii” honors Dr. John M. Huisman, the foremost expert on the Nemaliales today, for, among other things, his prolific systematic work on the order over the past three decades, culminating in his recognition of the D. marginata complex using genetic techniques (Huisman et al. 2004, Kurihara and Huisman 2006).

Holotype

T.R. Popolizio 12-117-5 [BDA1492], 20 Sept. 2012, Hog Breaker, north shore Bermuda I., 32°27′47.7″N, 64°49′48.9″W, Bermuda, western Atlantic Ocean, depth 12 m [MICH] (Figure 8); isotype Herb. CWS.

Paratypes

Bermuda – A.B. Hervey, P.B.-A. no. 1930 [Collins et al. 1913, as Brachycladia marginata], 22 Jan. 1913, near low water mark south shore, Bermuda I. [Herb. CWS] (Figure 7); CWS/CEL 10-5-13 [BDA0024], 19 August 2010, reef off Frick’s Beach, 32°19′56.0″N, 64°40′20.7″W, Tucker’s Town, Bermuda I., depth 10–12 m (Figure 9); TRP/CWS 12-151-11 [BDA1669], TRP/CWS 12-158-3 [BDA1709], TRP/CWS 12-170-9 [BDA1805] (see Table 1 for collection details).

Distribution

Endemic to Bermuda as presently known.

Remarks

The new species is presently very rare in Bermuda and, when found, there are few individuals in the population. When Collins and Hervey (1917) first reported Galaxaura marginata from Bermuda, they cited only their shallow water collection from Gravelly Bay (P.B.-A. no. 1930; Figure 7), but clearly there were enough specimens available at the time to make herbarium specimens for each of the 80 distributed sets of the exsiccata they were ultimately placed in (Fahey and Doty 1955). Collins and Hervey (1917) did not distinguish their collections in any feature from the standard characteristics of G. marginata taken from earlier sources available to them at the time. Interestingly, all of the Dichotomaria huismanii specimens we have collected show many truncated branch tips that have been grazed by herbivores (Figures 8 and 9). Often on these plants, closely adherent medullary filaments that run longitudinally through the axes emerge from the truncated tips, and occasionally from branch nodes, as caespitose tufts of loose, uniseriate filaments, appearing much like a dense epiphytic growth of an acrochaetioid alga (Figures 8 and 9; we sequenced these tufts and they were a 100% match to the macroscopic plant on which they form). Taylor (1960) may have observed the same for D. marginata when he mentioned “branches occasionally tipped with a brush of deciduous hairs”, but he made no mention of herbivory. It is worth noting that all of our recent collections from Bermuda are much smaller (to 4 cm tall, 7 cm across) and less robust than those distributed in 1913 fascicles of P.B.-A (8 cm tall, 12 cm across; Figure 7). The reduced habit appears to be a trend seen in many seaweeds collected in Bermuda during the latter part of the 20th and first part of the 21st centuries, these being smaller than herbarium collections made in the early years of the 20th century. We believe increased fish herbivory in Bermuda may be responsible for the more diminutive plants of today, as was shown for two small Botryocladia spp. growing on the reefs compared to large, fully formed, non-grazed plants in display tanks of the Bermuda Aquarium (Schneider and Lane 2008). Increased herbivory could also account for the fewer individuals found today in Bermuda populations as compared with the distant past, as well as a lack of specimens from shallow waters, the only habitat reported for this species by Collins and Hervey (1917).

Thus far, the new species is known only from Bermuda and represents the only flattened member of the genus in the local flora. Dichotomaria huismanii is genetically distinct from D. marginata in the West Indies (4.8% bp differences in COI-5P sequences, 2.4% in rbcL), the islands of Bermuda being isolated from the closest populations of the generitype by a distance >1350 km to the northeast of the Bahamas. St. Kitts/St. Croix isolates (Dichotomaria sp. 1CRB, Table 1) are sister to the new species in the COI-5P tree (Figure 1), but show 4.1% bp differences with it (22 nucleotide differences out of 538). The rbcL clade where D. huismanii is resolved (Figure 2) also includes a specimen identified as “D. marginata” from the Gulf of California (GenBank AY688022). The rbcL sequences from Bermuda and Mexico are 0.6% different from each other (there is no COI-5P data for the Mexican isolate). The Bermuda and Mexico species are each 0.6% different in rbcL bp from the two Caribbean sequences of Dichotomaria sp. 1CRB (Table 1).

Aside from the geographic separation and genetic difference, a few morphological characteristics present themselves as means for differentiating D. huismanii from its cryptic counterpart in the western Atlantic, D. marginata. In gross morphology, D. huismanii is difficult to distinguish from D. marginata, but subtle differences can be found anatomically. The dimensions of cortical cells, thallus height and width, and the length and width of internodes all demonstrate a great amount of overlap between the two species (Table 2). But unlike D. marginata, none of the assimilatory cells on D. huismanii are apiculate (Figures 11–13), a character that may be of use in western Atlantic populations of Dichotomaria after a more thorough examination of isolates is made from the region of the Bahamas, throughout the Caribbean and south to Brazil. The new species has four cortical cell layers including stalk and outer assimilatory cells (Figure 12) and in general has greater diameter medullary filaments (13–20 μm) than the generitype D. marginata (7–13 μm), which mostly shows three cortical cell layers (Figure 5). The dichotomous axes of D. huismanii branch at markedly narrower angles than St. Croix specimens of D. marginata, 32–68° vs. 72–89°. Whether any of these anatomical character differences will hold up as more isolates from the Caribbean are sequenced remains to be seen.

Using the chloroplast rbcL gene, along with the western Atlantic species mentioned above, D. huismanii clusters with two Indo-Pacific species, D. tenera and D. intermedia (Figure 2). The only detailed description of D. tenera since its genetic analysis and resurrection by Huisman et al. (2004) is found in De Clerck et al. (2005) based on specimens from Kwazulu-Natal, South Africa, the geographic area from which Huisman et al. (2004) used isolate sequences to segregate it from Caribbean D. marginata. These Indian Ocean specimens have somewhat broader flattened branches (1.5–3.0 mm) than D. huismanii (1.2–2.2 mm), and unlike the new species have axes that are hirsute above the holdfasts and produce outer cortical cells that are apiculate at the margins (De Clerck et al. 2005). Specimens of D. tenera from Mauritius are reported to have outer cortical cells that are 38–42 μm tall and 27–30 μm diameter (Børgesen 1942, as Galaxaura tenera Kjellman), well within the range of these cell sizes in D. huismanii. Their branch angles (measured from Børgesen 1942, figure 24) are from 50–75°, a slightly wider angle range than that of the new species reported here (Table 2). A specimen attributed to D. tenera from South Africa and pictured by Kylin (1938), as G. tenera), has much narrower branch angles, 40–55°, than the Mauritian sample.

The second species that genetically clusters with Dichotomaria huismanii (Figure 2), D. intermedia (R.C.Y. Chou) Wiriyadamrikul, M.J. Wynne et S.M. Boo (type locality=Galapagos Is.), is considerably larger (to 23 cm) than the new species (to 8 cm). Despite its overall large size, D. intermedia has similar cell dimensions for outer cortical cells on tetrasporophytes, 30–50 μm tall and 25–35 μm diameter (Wiriyadamrikul et al. 2014), as compared to those for D. huismanii, 28–50 μm tall and 22–48 μm diameter (Table 2), but unlike the new species, these cells are apiculate.

Morphological comparisons of Galaxaura and Dichotomaria species from locations other than their type localities with species from different parts of the world create similar problems to those we have already alluded to for genetic data from distant locations. For example, prior to the availability of molecular sequencing and comparison of species, Chou (1945) used Pacific Costa Rican samples of G. stupocaulon Kjellman (type locality=Brazil) to characterize this species, and make comparisons with other members that would be included in the D. marginata complex today (Huisman et al. 2004, Wang et al. 2005, Kurihara and Huisman 2006, Wiriyadamrikul et al. 2014). Oliveira (1977) noted the “great similarity” of Galaxaura angustifrons Kjellman (type locality=Brazil) from Brazil to G. veprecula Kjellman (type locality=Madagascar) from the Philippines and Ecuador based on Chou’s (1947) observations and measurements. G. veprecula was subsequently placed in synonymy with G. tenera by Papenfuss and Chiang (1969), species later subsumed in G. marginata (Papenfuss et al. 1982). Finally, G. tenera was resurrected as Dichotomaria tenera by Huisman et al. (2004) based on LSU sequence data (and the morphology of terminal cortical cells) of specimens from South Africa. It seems obvious from these convoluted pathways that historical taxonomic decisions involving Galaxaura and Dichotomaria species need to be reassessed from the fresh perspective of DNA sequencing, as has been done in selected recent studies (Huisman et al. 2004, Wang et al. 2005, Kurihara and Huisman 2006, Wiriyadamrikul et al. 2014). Accordingly, what has been combined in the past under D. marginata, has already begun fragmenting back to a greater number of species.

For their Indian Ocean and Australian specimens of Dichotomaria “marginata”, Huisman et al. (2004) were able to resurrect species names for flattened taxa from those areas with specimens that did not genetically match the sequence of true D. marginata from Puerto Rico. The resurrected D. australis and D. tenera had been among a large number of species that, over the years and culminating with the Papenfuss et al. (1982) circumscription, were placed in synonymy with D. marginata. In the western Atlantic, there are early names presently considered junior synonyms of D. marginata: G. angustifrons, G. frutescens Kjellman (type locality=Brazil), G. occidentalis and G. stupocaulon. None of these taxa were ever reported from Bermuda (Taylor 1960). The above three species presently in the D. marginata complex with type localities in Brazil do share some resemblance to both D. marginata and D. huismanii, and are presently considered synonyms of the former. Observations for these species on type specimens is critical to linking anatomical measurements/characteristics to recently collected genetic species from near their type localities in this highly cryptic complex of species. As none of the historical species were described locally, we choose to describe our Bermuda specimens as a new species from the northern limit of distribution of the genus in the western Atlantic. Genetic information from other areas to the south may prove this species to have a more widespread distribution than presently delimited.

The separation and description of a new cryptic species from Bermuda for records that in the past were considered a part of a pantropical species’ range with a western Atlantic type is similar to other nemalialean examples already genetically distinguished in the islands. What was identified in Bermuda as Helminthocladia calvadosii (J.V. Lamouroux ex Duby) Setchell for over a century was shown to be H. kempii Popolizio, Chengsupanimit et C.W. Schneider (Popolizio et al. 2013), and what was known as Liagora ceranoides J.V. Lamouroux since the early 1900s in the islands was recently described as L. nesophila Popolizio, C.W. Schneider et C.E. Lane (Popolizio et al. 2015).