Conspecificity of two morphologically distinct calcified red algae from the northwest Pacific Ocean: Galaxaura pacifica and G. filamentosa (Galaxauraceae, Rhodophyta)

Species of the calcified red algal family Galaxauraceae are distributed widely throughout the shallow marine waters of the warm temperate, subtropical, and tropical regions (e.g., Littler and Littler, 1997;Abbott 1999;Huisman, 2006). The unique ability of these red algae to incorporate calcium carbonate into their thalli makes them critical elements in the carbon budget, biomineralization, reef building processes, and coastal marine ecosystems. Research related to their biodiversity is necessary to understand their contribution to the carbon cycle of tropical marine ecosystems because calcium carbonate deposition in algae is affected by environmental conditions and differs from one species to another (Stanley et al., 2010). Seven genera were historically recognized in the family Galaxauraceae, including Actinotrichia Decaisne, 1842, Galaxaura Lamouroux, 1812, Tricleocarpa Huisman and Borowitzka 1990, Gloiophloea J. Agardh, 1872, Scinaia,1822, Nothogenia Montagne, 1843, and Whidbeyella Setchell and Gardner 1903 (reviewed in Huisman, J.M 2006). Huisman et al. (2004a) used molecular analyses to resurrect the genus Dichotomaria Lamarck 1816 from synonymy under Galaxaura, and split the family Galaxauraceae into two different families, Galaxauraceae and Scinaiaceae. The former currently comprises four calcified genera, Actinotrichia, Dichotomaria, Galaxaura, and Tricleocarpa; whereas the latter consists of four non-calcified genera, Gloiophloea, Scinaia, Nothogenia, and Whidbeyella.

The genus Galaxaura was first erected by Lamouroux (1812). Kjellman (1900) recognized 62 species and described 47 of these species as new, suggesting that the genus Galaxaura is extremely diverse. However, later studies challenged this taxonomic system. Howe (19171918) suggested that the species diversity of Galaxaura could be inflated because of the different external morphologies in the sexual (i.e., gametophytic) and sporophytic (i.e., tetrasporophytic) stages, referred to as the dimorphic life history (e.g., Wang et al. 2005). Guided by this concept, two fundamental studies that investigated the calcified red algal family Galaxauraceae in the Indian Ocean and Australian region reduced a significant number of different species to a few pantropical species such as Galaxaura rugosa, G. marginata (= Dichotomaria marginata), and G. obtusata (= D. obtusata ) (Papenfuss et al. 1982;Huisman and Borowitzka 1990). The elucidation of a possible relationship between two morphologically different (e.g., gametophytic vs. tetrasporophytic) species of the calcified red algal genus Galaxaura remains largely unexplored, although the concept of the dimorphic life history was only recently proven and the synonymy of some Galaxaura species has been demonstrated (Huisman et al. 2004b;Kurihara et al. 2005;Wang et al. 2005). Wang et al. (2005) suggested that the genus Galaxaura should have more diverse species in the tropical oceans and that the synonymy of numerous species based solely on morphological observations should be reassessed carefully after showing that the monophyly of a pan-tropical species, G. rugosa, was not supported by their molecular analyses.

To examine this issue, we collected two specimens of the genus Galaxaura from Taiwan and the Philippines. Based on their morphological characters, the specimens were initially identified as two different species. They were identified according to the taxonomic keys in the literature as Galaxaura pacifica Tanaka (Tanaka, 1935) and G. filamentosa Chou ex Taylor (Chou, 1945). The thallus of G. pacifica comprises a lower villous portion where branches are very hairy (i.e., villous) and composed of numerous assimilatory filaments and an upper smooth portion where branches are very smooth (i.e., glabrous) and lack assimilatory filaments (Tanaka, 19351936). In contrast, the branches of G. filamentosa are extremely villous throughout the thallus (Chou, 1945). Thus, G. pacifica was considered to be in the sexual phase (i.e., gametophytic plants) (Tanaka 1935), whereas G. filamentosa was speculated to be in the asexual phase (i.e., tetrasporophytic plants) (Chou, 1945), but in separate species because of their distinct morphologies. These two species have been previously documented to occur in the coral reef areas of Taiwan by Tanaka (1935) and Itono (1977). This study examined these specimens morphologically and compared their rbc L sequences, and additionally examined morphological variation in several samples from different locations in southern Taiwan. The results of these studies are presented herein.

The relationship of the four Galaxauraceae genera Actinotrichia, Dichotomaria, Galaxaura, and Tricleocarpa in our phylogenetic analyses is consistent with those of Huisman et al. (2004a), Wang et al. (2005), and Liu and Wang (2009). In the genus Galaxaura, it is noteworthy that many ambiguous taxa within the G. rugosa assemblage require further morphological examinations, particularly considering that only one species can be identified with certainty as G. cuculligera based on morphological characters (Kurihara et al., 2005).

This study provides the first evidence that two morphologically distinct species, G. pacifica and G. filamentosa, should be treated as a single species, G. pacifica. Galaxaura pacifica is characterized by having multi-axial thalli, glabrous, slightly rugose terminal branches, and densely villous branches in the lower part of the plant (Tanaka, 1935); the latter is a unique morphological trait (see the illustration of Figure 11 of Tanaka, 1936) that can be used to separate G. pacifica from the other species of the genus Galaxaura. In contrast, G. filamentosa is characterized by villous branches occurring throughout, stretching from the lower portion to the terminal portion of the thallus with only long assimilatory filaments shown in the cross section of its branches (Chou, 1945). Based on these characteristics, Chou (1945) considered it a new species. The evidence from our molecular analyses however revealed that these two species should be treated as the same species since their rbc L sequences were unexpectedly identical, consistent with the assumption made by Huisman and Borowitzka (1990). We propose to combine these two species as a single species, G. pacifica, with this name having taxonomic priority over G. filamentosa. From our morphological observations, G. pacifica is now characterized as (1) a glabrous or villous multi-axial plant with either glabrous or villous branches and a discoid holdfast; (2) glabrous branches with a three-celled layer of cortex; (3) villous branches with numerous long assimilatory filaments arising from the swollen basal cells; (4) three-celled carpogonial branches (consisting of a carpogonium, hypogynous cell, and basal cell) that are produced from the replacement of one of the dichotomous vegetative cortical filaments (referred to as type I), between the dichotomous vegetative cortical filament (referred to as type II), or the mixture of both type I and type II growth patterns (referred to as type III); (5) gonimoblast initial produced from the fertilized carpogonium; (6) involucral filaments derived from the basal cell of the carpogonial branch remaining at the base throughout cystocarp development; and (7) a distinct fusion cell derived from the fusion of 7-10 inner gonimoblast cells without the involvement of the hypogynous cell, the sterile branches from the hypogynous cell and the basal cell.

Two possible scenarios could explain the different morphological features in G. pacifica. First, the external morphology of Galaxaura species can be dimorphic (i.e., gametophyte and tetrasporophyte) during their life history (e.g., Huisman and Borowitzka, 1990;Huisman et al., 2004b;Kurihara et al., 2005;Wang et al., 2005). This dimorphic life history in Galaxaura was first observed by Howe (1917;1918), who noted that different life history stages (i.e., gametophyte vs. tetrasporophyte) in Galaxaura (at the time including Dichotomaria) can have two different external morphologies with different cellular structures (in Galaxaura) or same external morphology with two different cortical structures (in Dichotomaria) - a phenomenon referred to as dimorphism. Recently, dimorphism was supported by evidence from molecular analyses (Huisman et al., 2004b;Kurihara et al., 2005;Wang et al., 2005). Second, the villous plant might merely be an immature form of the gametophyte (Huisman, pers. comm.), although some studies have considered them as possible tetrasporophytic forms of G. rugosa (e.g., Chou, 1945;Papenfuss et al., 1982). Huisman and Borowitzka (1990) indicated some precautions with directly treating a villous thallus as a tetrasporophytic form of Galaxaura, particularly for specimens without any evidence of tetrasporangia. Indeed, no tetrasporangial structures have been found in G. “filamentosa” (= G. pacifica) from Taiwan (AY688006; Figure 1) and the Philippines (AY688004; Figure 1), consistent with the observations of Chou (1945) and Huisman and Borowitzka (1990). Svedelius (1953) and Huisman and Borowitzka (1990) indicated that the reproductive structures of G. filamentosa have never been found and its cortical structure is vegetatively identical to the hirsute (i.e., villous) basal portion of G. rugosa, suggesting that further work may show that G. pacifica might prove to be a synonym of G. rugosa. Thus, for the first time, this study provides evidence to further support these earlier proposals that G. filamento sa should be considered as the same species as G. pacifica. It should be noted that the species G. rugosa described in Huisman and Borowitzka (1990) may comprise several different species (including the G. rugosa assemblage and the G. pacifica assemblage) as shown by Wang et al. (2005) and this study. For example, the specimen in Figure 10 of Huisman and Borowitzka (1990) is externally similar with our materials in the G. rugosa assemblage, and that in Figure 11 of the same paper is externally similar to our materials comprising the G. pacifica assemblage. Thus, it is likely that the specimens with obvious villous basal portions cited by Huisman and Borowitzka (1990) (e.g., the specimen in Figure 11 of their paper) should be considered phylogenetically close to the G. pacifica assemblage. To verify the possibility whether the villous plant is a tetrasporophyte or not, a further study to investigate their ploidy levels should be required to clarify this issue, among others. According to the materials collected from Xiao-Liu-Qiu Island, Taiwan, the villous plants might be merely the remnants of the basal villous part of the thallus after the distal glabrous portions of the thallus die off in the summer. As also evidenced by G. “filamentosa” from Sorsogon, Philippines, this plant was collected from the warmer regions of the Pacific Ocean where water temperature is comparable to that in higher latitudes of the Pacific Ocean in the summer. Considering that we never saw the seasonal co-occurrence of G. “filamentosa” and G. pacifica, our observations therefore support the hypothesis that certain environmental cues (e.g., temperature) might stimulate the development of G. filamentosa-type plants from G. pacifica-type plants.

Previously, Papenfuss et al. (1982) proposedd that G. pacifica should be considered synonymous with G. rugosa because their morphological characters were not distinguishable. In contrast, Kurihara et al. (2005) found that G. pacifica could be separated from G. rugosa based on molecular evidence. Our results are consistent with their observations and further support that G. pacifica should be recognized as an independent species different from G. rugosa. Contrary to the observations by Papenfuss et al. (1982), we can morphologically separate G. pacifica from the other species of Galaxaura based on three morphological grounds: 1) obvious villous branches in the lower part of gametophytic plants, 2) long assimilatory filaments without swollen basal cells in the villous branches, and 3) three different growth patterns of carpogonial branches (explained in more details below).

Compared to other species of Galaxaura, we found that carpogonial branches of G. pacifica have an unusual growth pattern. Some carpogonial branches arose to replace one of the dichotomous terminal cortical filaments (type I) while some of them arose between the dichotomous terminal cortical filaments (type II). Occasionally, two carpogonial branches, one of which originated from type I and the other originated from type II, grew on the same terminal cortical filament, suggesting a mixture of type 1 and type 2 growth pattern of carpogonial branches (referred to as type III). After fertilization, only one of them could successfully develop into a carposporophyte. Once one of them was fertilized, the others seemed to stop their development and senesce subsequently. To quantify the frequency of these three different types of carpogonial branch growth patterns, we randomly observed 30 different carpogonial branches and found that the type I (25 out of 30) is the most common type of carpogonial branch growth pattern and the type II (3 out of 30) and the type III (2 out of 30) are less frequent. Previously, only type I carpogonial branch was observed in the genera Dichotomaria, Galaxaura, and Tricleocarpa (Huisman and Borowitzka, 1990;Wang et al., 2005b), whereas the type II carpogonial branch was only observed in the genus Actinotrichia (Wang and Chiang, 2001;Liu and Wang, 2009). Therefore, the ontogeny of carpogonial branches has been suggested as a useful character for separating different genera among the members of the family Galaxauraceae (e.g., Wang and Chiang, 2001;Liu and Wang, 2009). Considering that both types of carpogonial branches can be found in G. pacifica, the value of this character in separating genera is questionable. Instead, this mixture of growth pattern types of carpogonial branches might be a useful taxonomic feature at the species level to separate G. pacifica from other species in the genus Galaxaura.

Two caveats in this study should be addressed. The first is that the rbc L sequences of our specimens from Xiao-Liu-Qiu Island, Taiwan and Sorsogon, the Philippines genetically differ from the specimen obtained from the type locality of G. pacifica. In total, 27 nucleotides of rbc L sequence (approx. 2%; uncorrected P- distance) were found to be different between these two populations. Such a sequence divergence exceeds the species-level boundary (< 0.5%), which was delineated based on previous rbc L sequence analyses (e.g., Kurihara et al., 2005). Compared to morphological descriptions of gametophytic plants of G. pacifica in Tanaka (1935), we could not however recognize any morphological difference between these two entities. Furthermore, we failed to recognize any morphological differences between the two genetically different specimens under study, which were initially identified as G. “filamentosa” (AY688006 and AY688004 in Figure 1). Both specimens show numerous long assimilatory filaments without a swollen basal cell, and their external morphological features cannot be distinguished (unpublished data). Rbc L sequences between these two specimens differed by about 2.6%. Overall, our results implied that morphological traits within the G. pacifica assemblage vary in a subtle manner despite their genetic heterogeneity. The species definition could be challenging in the light of taxonomic studies, particularly when the available evidence is limited. Without examining if a cross between any two populations of G. pacifica can produce viable hybrids or if there is any gene flow among them, the species boundary in G. pacifica will remain unresolved and subject to continuous disputes. Therefore, based on our limited information, the typological species concept, which is considered a conservative approach (e.g., Kurihara and Huisman, 2006), was applied to our case since no morphological difference can be found among the G. pacifica assemblage to convincingly argue for its distinctness. The second caveat is that G. filamentosa may be still a valid species since as yet no one has obtained rbc L sequences from materials collected from the type locality in the Revillagigedo Islands, Mexico. Further work may provide additional evidence supporting its synonymy with G. pacifica.

Clearly, the morphological and reproductive structures in Galaxaura show many similarities and can be sources of confusion. Collective evidence from a combination of approaches including molecular analyses, population genetic analyses, and detailed morphometric comparisons will be required to unequivocally delineate the species-level boundary of Galaxaura in the future. This attempt to use molecular tools for the elucidation of relationship between two morphologically different (e.g., gametophytic vs. tetrasporophytic) species of the calcified red algal family Galaxauraceae is only beginning to scratch the surface of this taxonomic puzzle. So far, only two cases have been argued. The first example is the pair of G. rugosa and G. subverticillata (= G. rugosa) on the basis of large subunit rDNA sequence analysis (Huisman et al., 2004b). The second example is the pair of D. apiculata and D. hystrix (= D. apiculata ) based on rbc L and ITS sequence analyses (Kurihara et al., 2005). Thus, our study provided another example in attempting to match two morphologically different species (from the northeast Pacific Ocean), G. pacifica and G. filamentosa, and to recognize them as a single species, G. pacifica. Since numerous species have been historically established in the genus Galaxaura and Dichotomaria based largely on their cortical structures (e.g., Kjellman, 1900), the biodiversity of the genus Galaxaura, as well as the genus Dichotomaria, was definitely overestimated in previous studies (i.e., Kjellman, 1900). Our study provides another example of documented dimorphism in different stages of the life history within a single species of Galaxaura (first proposed by Howe, 19171918), and sheds light on the biodiversity and unique biology of the calcified red algal genus Galaxaura.