Recent introduction of a freshwater red alga Chantransia macrospora (Batrachospermales, Rhodophyta) to Okinawa, Japan

The freshwater red algal form taxon, Chantransia macrospora Wood, was recently collected at an artificial reservoir pond in Okinawa, Japan. This species is chiefly characterized by a bluish-green thallus and erect filaments irregularly branched with cylindrical to tapered cells and monosporangia on lateral branchlets. Our culture study confirmed that C. macrospora from Okinawa was the ‘chantransia’ sporophyte of Batrachospermum species. The rbcL sequence of C. macrospora from Okinawa was identical to isolate FG15 of this species from French Guiana in South America. The sporophyte Chantransia macrospora as well as the gametophyte B. macrospora Montagne are distributed in North and South America. However, there have been no records of either C. macrospora or B. macrosporum in Japan or in eastern Asia. Therefore, C. macrospora specimens originally from South America may have been introduced to Okinawa, Japan, through the vector such as ornamental macrophytes.


Introduction
Many introduced species of marine macroalgae have been reported worldwide, for example, the green algae Caulerpa taxifolia (Vahl) C. Agardh and Codium fragile ssp. tomentosoides (van Goor) P. C. Silva, the brown algae Undaria pinnatifida (Harvey) Suringar and Sargassum muticum (Yendo) Fensholt, the red algae Grateloupia turuturu Yamada and Hypnea musciformis (Wulfen) J.V. Lamouroux (Inderjit et al. 2006;Defeo et al. 2009). In addition, several introduced species of freshwater macroalgae, Bangia Lyngbye, Compsopogon Montagne, Hildenbrandia Nardo, have also been found, and the possibility of their introduction was usually reinforced by molecular phylogenetic analyses (Müller et al. 1998;Rintoul et al. 1999;Sherwood and Sheath 1999). Boudouresque and Verlaque (2002) distinguished between introduced or invasive species: an introduced species shows no discernible negative impact, whereas an invasive species has an identified negative impact. Inderjit et al. (2006) suggest that it is better to use the term "introduced" or "non-indigenous" and state whether or not the introduced species has demonstrated a harmful effect. However, they pointed out the difficulties with the exact definition of an invasive species in the absence of a clearly demonstrated negative effect. We followed the perspective of Boudouresque and Verlaque (2002) and regarded a species without discernible negative impact as an introduced species.
Chantransia macrospora Wood, a freshwater filamentous red alga, usually grows in streams or rivers in North and South America [Necchi et al. 1993;Necchi and Zucchi 1995;Kumano 2002, as Audouinella macrospora (Wood) Sheath & Burkholder]. This species is mainly characterized by a bluish-green thallus and erect filaments irregularly branched with cylindrical to tapered cells. Pit plug ultrastructure of C. macro-spora indicated that this species belongs to the Batrachospermales Pueschel & Cole (Pueschel et al. 2000), not the Acrochaetiales Feldmann including Audouinella Bory which once contained this species. Recent culture and molecular studies indicate that C. macrospora is the sporophyte, 'chantransia stage', of Batrachospermum macrospora Montagne (Pueschel et al. 2000;Necchi and Zuccchi 1997;Chiasson et al. 2005). Especially, the rbcL has been utilized in the past to link Chantransia sporophytes with Batrachospermum species identified based on gametophyte characteristics 2007;Vis et al. 2005;2008). Both the sporophyte C. macrospora and gametophyte B. macrosporum are found in North and South America (Kumano 2002). However, there have been no records of either species from other continents and regions, including Japan.
In this study, we identified algal specimens growing on dead grasses and floating plastic trash in an artificial reservoir pond, Senbaru-ike, in Okinawajima Island, Japan, based on morphology as C. macrospora. To further investigate this occurrence, the specimens were cultured to determine if gametophytes would be produced and rbcL gene sequences were compared with the literature to confirm the taxonomic identification and find a potential origin.

Specimens and culturing
Chantransia macrospora specimens were collected in 2006 from an artificial reservoir pond, Senbaru-ike, on the premises of the University of the Ryukyus, Okinawa, Japan (26°14'66"N, 127°45'55"E). Voucher herbarium specimens were deposited in the Herbarium of the Faculty of Science, University of the Ryukyus, Okinawa (RYU-A). The phenology of C. macrospora was surveyed every one or two weeks from November in 2007 to October in 2008. Unialgal cultures were established by removing attached zooplankton and diatoms in agar plates. Culture was maintained in freshwater enriched PES medium (Provasoli 1968), C medium (Ichimura 1971) and IMK medium (Nippon Pharm. Co., Tokyo). The temperatures and photoperiods were regulated at 20 or 25°C, 14:10 h L:D and lighting was 25-35 µmol photons m -2 s -1 . The established culture was deposited in National Institute for Environmental Studies, Japan (NIES). For light microscopy, living field and culture materials were used.

Electron microscopy
Material for thin sections was prepared as follows: equal volumes of fixative [5% glutaraldehyde, 0.5 M sucrose, in 0.1 M cacodylate buffer (pH 7.2)] with or without a drop of osmium tetroxide and culture medium containing growing cells were mixed together at room temperature for 1 h. A fixative based on the culture medium instead of cacodylate buffer was also used. After rinsing twice with 0.05 M cacodylate buffer, cells were postfixed in 2% osmium tetroxide for 1 h, and rinsed once with the same buffer. Thalli were embedded in Spurr's resin (Spurr 1969) after dehydration in a graded ethanol series. Sections were cut with a diamond knife using a Leica EM UC6 Ultramicrotome (Leica Microsystems K.K., Tokyo, Japan) and collected on slot grids coated by Formvar. Sections were double stained with 2% uranyl acetate and lead citrate (Reynolds 1963). Observations were carried out with a JEOL JEM 1011 (JEOL, Tokyo, Japan) transmission electron microscope.

Genetic analyses
Total DNA was extracted from a fresh culture thallus using Nucleon PhytoPure Genomic DNA Extraction Kits (GE Healthcare, Buckinghamshire, UK) according to the manufacturer's protocol. The following gene fragments were amplified and sequenced with the oligonucleotide primers stated below: 1) the nuclear SSU rDNA -three primer pairs SR1 and SR5, SR4 and SR9, SR8 and SR12 (Nakayama et al. 1996); and 2) the chloroplast rbcL -a primer pair F160 and rbcL R and two internal primers R472.2 and R 897 . The PCR products directly sequenced using a CEQ8800 Genetic Analysis System (Beckman Coulter, Fullerton, CA, USA) with the dye-terminator method. Sequences of SSU rDNA and rbcL were deposited in DDBJ (AB503215 and AB503216). A BLAST (Basic Local Alignment Search Tool) search (http://blast.ncbi.nlm.nih.gov/Blast.cgi) was performed to determine closely related sequences.
The SSU rDNA alignment, including a sequence determined here and previously published sequences of 27 species of the Batrachospermales and related orders Pueschel et al. 2000), was aligned using Clustal W 1.83 (Thompson et al. 1997). Rhodogorgon carriebowensis J. Norris et Bucher (AF006089) served as an outgroup. The rbcL alignment, including a sequence determined here and 10 previously published sequences of Chantransia macrospora and 19 sequences of Batrachospermum macrosporum Vis et al. 2008), was aligned manually as there were no gaps among rbcL sequences. Identical sequences were represented by a single sequence in the alignment. Thirteen closely related Batrachospermum species were used as outgroups following Vis et al. (2008).
The phylogenetic relationships among the sequences of the SSU rDNA and rbcL were inferred using maximum parsimony (MP), minimum evolution (ME) and maximum likelihood (ML), which was performed using PAUP*4.0b10 (Swofford 2002). MP and ME analyses were performed using the heuristic search option (with 100 random sequence additions for MP) and the tree-bisectionreconnection (TBR) branch swapping algorithm. All sites were treated as equally weighted, and gaps were treated as missing. To evaluate bootstrap support for individual clades (Felsenstein 1985), the MP and ME bootstrap analyses consisted of 1000 replications of the full heuristic search option with 10 random additions and TBR branch swapping (for MP). The maximum likelihood (ML) analysis was carried out using PhyML 2.4.4 (Guindon and Gascuel 2003;Guindon et al. 2005). The evolutionary models for ME and ML that best fit the data set were determined using a hierarchical likelihood ratio test in Modeltest 3.7 (Posada and Crandall 1998). The selected model for the rbcL data set was the GTR model (Rodriguez et al. 1990) with invariable sites (0.5249) and a gamma distribution parameter of 0.9279. The ML bootstrap analyses were conducted with 500 replications with subtree pruning-regrafting branch swapping.

Morphology
Chantransia macrospora from the artificial reservoir pond Senbaru-ike was bluish green in colour ( Figure 1) and growing on dead grasses, driftwood or floating plastic with water transparency that was approximately 50 cm throughout the year. There was a seasonal pattern and C. macrospora was not observed in late July to September. During this period, the water temperature was almost always more than 30°C.
This species formed dense tufts and was 2.0-2.5 mm in height. The thallus was composed of a uniseriate erect filament branching with narrow angles ( Figure 1A) and well-developed rhizoids. The erect filaments were generally cylindrical to tapered along their height. Axial cells of the erect filaments were 14-22 µm in diameter and 41-72 µm in height. Chloroplasts were discoid to elongate and with three to five pyrenoids ( Figure  1B). Lateral branches irregularly arose from the axial cell and formed monosporangia (19-23 µm in diameter and 31-39 µm in height) on apices of the ultimate branches. Germination of monospores occurred by unipolar elongation to form a rhizoidal filament leaving the spore empty of protoplast ( Figure 1C). The dimensions of vegetative and reproductive structures of a culture isolate overlapped with the ranges of field specimens. After 6-10 months in culture from small segments (< 5mm) of C. macrospora, Batrachospermum gametophytes developed ( Figure 1D-F). The filaments of juvenile gametophytes were clearly distinct from those of C. macrospora in shape and size, as gametophyte cells were disk shaped and significantly shorter and wider than the cylindrical vegetative cells of C. macrospora ( Figure 1D). The gametophytes in later stages formed young whorled filaments and were branched ( Figure  1E, F). The Batrachospermum gametophytes did not form reproductive structures and it was not possible to identify the species. We started culturing small segments (< 5mm) of C. macrospora and observed the formation of Batrachospermum gametophytes twice. However, the particular conditions of the temperature, light intensity and culture media that induced formation of gametophytes were not identified.
The ultrastructure of the pit plugs from the Senbaru-ike culture isolate are shown in Figure  2. Pit plug cores were 0.6-1.0 µm in diameter at the narrow waist, 0.9-1.2 µm in diameter at the ends, and 1.2-1.5 µm in height. The plug core often had projections near the waist (Figure 2A). The outer layer closer to the cytoplasm was unevenly thick and plate-like ( Figure 2A) and sometimes appeared to be assembled from spheroidal units ( Figure 2B), 0.1-0.2 µm thick. Some outer caps were discontinuous ( Figure 2C) and therefore, it was possible to confirm that no membrane was present.

Molecular analyses
The BLAST search indicated that our determined sequence of SSU rDNA was closely related to two sequences (AF199505 and AF199506) of three isolates of Chantransia macrospora (as Audouinella macrospora, Pueschel et al. 2000), which were the only available in Genbank (searched 15 May 2009). The phylogenetic analysis using SSU rDNA sequences of the Batrachospermales and related orders also supported this.
The rbcL sequence of the isolate from Senbaru-ike was identical to isolate FG15 of C. macrospora from French Guiana (AY423414) in South America. The analyzed rbcL data set consisted of 38 sequences of Chantransia macrospora including the Senbaru-ike isolate and Batrachospermum macrosporum. Of the 1227-bp portion analyzed, 797 bp were variable, and 52 bp were phylogenetically informative. The MP analyses produced four MP trees, one of which is shown in Figure 3 (Length = 1156 steps, CI = 0.528, RI = 0.806). The ML (-ln L = 7270.695734) and ME trees are not shown. The topologies of the MP, ML and ME trees were identical to each other except for two arrangements that were not supported by any analytical methods. The Senbaru-ike isolates and FG15 formed a fully-supported clade with other isolates of C. macrospora and B. macrosporum gametophytes from South America. This South American clade was related as a sister clade to a fully-supported clade comprising of isolates of C. macrospora and B. macrosporum from North America (Figure 3).

Discussion
The form genus Chantransia including three species was proposed to distinguish these specimens from taxa of the genera Audouinella and Ptilothamnion Thuret (Vis et al. 2006). Chantransia species represent an alternate life history phase (sporophyte stage) of species of the Batrachospermales and Thoreales Müller, Sherwood, Pueschel, Gutell & Sheath. The three Chantransia species are distinguished from each other by the thallus color in field, the development of erect filaments and cell dimensions. Our Chantransia specimens from Senbaru-ike was identical to C. macrospora in having a bluishgreen thallus and erect filaments irregularly branched with cylindrical to tapered cells and the size of vegetative cells and monosporangia (Glazer et al. 1997;Kumano 2002;2007;Vis et al. 2006). Pit plugs of the Senbaru-ike isolate were characterized by a thick plate-like outer cap layer and the absence of a cap membrane, which are consistent with the C. macrospora (as A. macrospora) in Pueschel et al. (2000). Moreover, the Senbaru-ike isolate produced juvenile gametophytes of Batrachospermum species, indicating that the Senbaru-ike isolate was morphologically identical to other specimens of C. macrospora in previous studies (Glazer et al. 1997;Necchi and Zucchi 1997). The rbcL sequence of C. macrospora from the Senbaru-ike was identical to an isolate from French Guiana which was closely related to other isolates of this species . Therefore, we identified our Chantransia specimens as C. macrospora. This is the first report of C. macrospora from the northwest Pacific Ocean region.
Chantransia macrospora is distributed in North and South America (Kumano 2002, as A. macrospora). This species has been found to be the chantransia stage of B. macrosporum by culture and molecular studies (Necchi and Zucchi 1997;Pueschel et al. 2000;. The geographic distribution of B. macrosporum is also North and South America (Kumano 2002). In molecular phylogenetic analyses, C. macrospora and B. macrosporum were distantly related to the members of the Batrachospermales known from eastern Asia, for example, B. arcuatum Kylin and B. atrum (Hudson) Harvey (Vis et al. 2006;Chiasson et al. 2007). Kumano (2002) reported eight species of bluish-green Audouinella including C. macrospora and C. pygmaea (as A. macrospora and A. pygmaea), and these have not been reported from Japan to date. Geographical distribution of five species includes China, and one of these species, Audouinella chalybaea (Roth) Bory has been showed to produce Batrachospermum gametophytes (Israelson 1942). However, A. chalybaea is clearly distinguished from C. macrospora by much smaller monosporangia and from C. pygmaea by narrower erect filaments (Kumano 2002).
The rbcL sequence differences of C. macrospora and B. macrosporum ranged between 1-2.5% even in South America , and 5.5-6.5% between North and South America , Vis et al. 2008). These results indicate that C. macrospora has relatively high genetic intraspecific divergence. Therefore, considering the geographical distribution and the low genetic variations between the Japan specimens and French Guiana, the most likely explanation of the presence of C. macrospora in Okinawa is via introduction.
Chantransia macrospora from this study may have been secondarily transported from South America, directly or indirectly, to Okinawa, Japan. The freshwater red alga Compsopogon has been suspected to be introduced in the same way that the freshwater green alga Hydrodictyon Roth was introduced by the aquarium or ornamental fish and macrophyte trade (Rintoul et al. 1999). Chantransia macrospora has also been reported to grow on the aquatic fern Ceratopteris in fish tanks (Glazer et al. 1997). Moreover, introduced fishes and water plants have been reported in Okinawa, Japan (Tachihara et al. 2002;Kadono 2002;Yokota 2003). For example, an introduced freshwater fish, Liposarcus disjunctivus Weber, which is native to Amazon, South America, was recently found spreading in Okinawajima Island, including at our collection site, Senbaru-ike, the artificial reservoir pond of University of the Ryukyus (Takeshima and Yoshino 1996). According to Takeshima and Yoshino (1996), after this pond was built in 1977, it was drained completely for over 1 week to construct irrigation facilities in 1983. At that time, L. disjunctivus was not found there. However, L. disjunctivus was found for the first time in the Makiminato River system, which originates from Senbaru-ike, in 1989; subsequently more than 40 specimens have been collected at Senbaru-ike. These results suggest that L. disjunctivus was introduced into Senbaru-ike after 1984. Considering this evidence, it is possible that C. macrospora was also introduced in recent decades by a vector, such as ornamental macrophytes.
Chantransia macrospora can grow in the very eutrophic pond Senbaru-ike, although Chantransia and Batrachospermum species are usually reported in less polluted streams (Necchi et al. 1993;Necchi and Zucchi 1995). This result may indicate that C. macrospora has a high tolerance for polluted conditions. Moreover, the fact that tiny freshwater algae, such as C. macrospora, have been found may indicate that unknown introduced creatures on which C. macrospora grows potentially exist. Rivers on Okinawajima Island, Japan, are small-scale and their environments are unstable (Nishijima 1984). Some reports have previously dealt with freshwater fishes introduced to Okinawajima Island, and have demonstrated ecosystem disturbances by these introductions (Senou 1985;Kouchi 1995). A survey of introduced aquatic plants and algae is urgently needed in Okinawajima Island and also on the other Ryukyu Islands in Japan.