Diversity of tropical macroalgae in the New Zealand marine aquarium trade

ABSTRACT Exotic species often slip through international borders undetected. Increased global trade has increased the frequency of species introductions. The marine aquarium trade is a significant vector of species introductions, including algal introductions. Molecular barcoding of tropical macroalgae entering the New Zealand aquarium trade was implemented using various molecular markers. Both NCBI BLAST searches and maximum-likelihood phylogenies were used to identify the isolates. A total of 62 species of tropical macroalgae were identified from coral rocks. Some species found are known as invasive elsewhere, for example, Caulerpa cylindracea, C. racemosa, C. sertularioides, Ulva ohnoi and Chaetomorpha vieillardii. All three major groups of algae were well represented with 26 species of red algae, 24 species of green algae and 12 species of brown algae. Temperature tolerance of some of these algae to minimum sea surface temperatures was tested. Results show that one species Chaetomorpha vieillardii can survive at Auckland minimum winter sea surface temperatures. Our findings have important implications for New Zealand biosecurity, as not only are a large diversity of exotic macroalgae entering the New Zealand marine aquarium trade unregulated, but there is also the potential for them to survive in New Zealand waters.


Introduction
Species introductions can have a detrimental effect on the environment (Gozlan et al. 2010). Humans facilitate many species introductions through their movement (Zimmermann et al. 2014). Species introductions can be considered a more significant environmental stressor than eutrophication and toxic chemical release (Chapman 1995). An introduced species is one transported to a non-native habitat by natural pathways or human activity (anthropogenic introductions) (Alpert 2006). The absence of population regulatory forces can lead to the exponential population growth of introduced species to the detriment of native flora and fauna, making an introduced species an invasive species (Briggs 2012).
Increasing global sea surface temperature increases the risk of invasion of non-indigenous marine species from warmer climates (Stachowicz et al. 2002). Global warming causes species range shifts latitudinally and may facilitate invasion, for example, as tropical species range further polewards (Occhipinti-Ambrogi 2007). The largest threat to biotic communities may not be from increasing annual mean temperature but the maximum and minimum temperatures giving invaders a head start over native species (Stachowicz et al. 2002).
The aquarium trade is a significant vector of species introductions (Gertzen et al. 2008). However, the trade has received little attention and risks have not been well studied (Calado and Chapman 2006). One-third of aquatic invasive species that have had a large detrimental impact on native biodiversity are either ornamental or aquarium species (Padilla and Williams 2004). The Ministry of Primary Industries (MPI) takes biosecurity threats seriously (MAF Biosecurity New Zealand 2009a, 2009b. There are also large economic costs in loss of production and for pest eradication or control post-invasion (Goldson et al. 2015). Online trading of flora and fauna has long been a risk to New Zealand's native biota as introductions are inevitable (Derriak and Phillips 2010). Biosecurity risk assessment for the importation of live fish, invertebrates and aquarium plants to New Zealand is primarily based around the introduction of pathogenic microbes (MAF Biosecurity New Zealand 2009a, 2009b.
In New Zealand, livestock importers are allowed to import corals, invertebrates, fish and aquarium plants from all countries under strict protocols set out by MPI (MAF Biosecurity New Zealand 2009a, 2009b, including a quarantine period. Corals, bivalves and the rocks that they are attached to harbour algal spores and juveniles; many too small or slow growing to be visualised during the 4-6 week quarantine period. Macroalgae are frequently introduced worldwide both intentionally for commercial purposes and unintentionally. Vectors of unintentional introduction include hull fouling (Mineur et al. 2006), plus aquaculture equipment fouling (Murphy et al. 2016). A review in 2007 found that 277 species of algae are introduced and some algal families contain many more successful invaders than others. For example, the green algal families Caulerpaceae and Codiaceae, the brown algae of the Fucaceae and Alariaceae, and the red algal families Ceramiaceae and Rhodomelaceae are particularly prevalent (Williams and Smith 2007). These introduced algae can have devastating effects (e.g. Conklin and Smith 2005;Piazzi and Ceccherelli 2002).
Some algal introductions can be directly attributed to the aquarium trade. The most publicised introduction is Caulerpa taxifolia (M.Vahl) C.Agradh released from the Oceanographic Museum, Monaco (Meinesz and Hesse 1991). It is common practice to maintain macroalgae in sumps or refugia as a method of trapping nutrients from aquaria. Species kept include Caulerpa racemosa (Forsskål) J.Agardh, C. serrulata (Forsskål) J.Agardh and C. taxifolia (Zaleski and Murray 2006). The filamentous green alga Chaetomorpha sp. previously thought to be a safe alternative to Caulerpa for growing in aquaria may be of high introduction risk due to its wide range of temperature tolerance, fast growth and ease of fragmentation (Odom and Walters 2014).
Algal taxonomy is an important discipline and taxonomic knowledge is required for research in ecology, physiology, bio-assessment and algal genetics (Manoylov 2014). Many algae species are polymorphic (Verbruggen 2014) making it often difficult to morphologically identify to species level with accuracy. Algal cryptic and pseudo-cryptic species diversity is another issue for morphological taxonomy. Molecular barcoding is a powerful tool for delimiting species (Hebert et al. 2003). Molecular barcoding is the use of individual or multiple gene sequences (often protein-coding) to identify a specimen to genus or species. The technique assists morphological species identification, and increasing certainty in taxonomic conclusions. Molecular species delimitation has been utilised in the study of brown (Buchanan and Zuccarello 2018;Vieira et al. 2014), red (Cianciola et al. 2010;Muangmai et al. 2014) and green algae (Lee and Kim 2015;Verbruggen et al. 2007).
Physiological tolerance limits the distribution of all species. The ability of a species to withstand changes in the environmental conditions makes it more or less adaptable to habitat change during introduction (Zerebecki and Sorte 2011). Invasive species often have broader physiological tolerance limits than native species giving them an advantage in establishment (Higgins and Richardson 2014). Temperature is the largest determinant for survival and reproduction in macroalgae (Graiff et al. 2015), and summer maximum temperature and winter minimum temperature are more important in determining thermal tolerance limits than mean temperature (Graiff et al. 2015). In New Zealand, Undaria pinnatifida sporophytes have temperature tolerances higher than the native kelp Lessonia variegata J.Agardh (Bollen et al. 2016). Aquarium release of Caulerpa taxifolia has occurred in Japan but establishment has failed, this may be due to winter temperatures being lower than its tolerance limit (Komatsu et al. 2003).
In this study, we undertaken to discover the macroalgal diversity of saltwater tropical marine aquaria, collected from retailers and hobbyist, entering New Zealand.

Sample collection
Two pet shops in Wellington New Zealand that sell ornamental tropical marine aquarium fish and invertebrates were visited when new coral orders arrived, either monthly or bimonthly. Many corals were also purchased from five online retailers and importers. Algal recruits were collected from the aragonite rock, bivalve shells and concrete to which the corals were attached. Corals that were purchased (Woodhouse, personal collection) were monitored and over time, algal specimens that grew from the rocks were sampled. A knife was used to break the substrate at the algal holdfast, if one was present. Marine aquarium hobbyists across New Zealand were contacted and some donated algal specimens for identification. Coral species from which samples were collected were recorded (Table S1). A small sample of alga was dried in silica gel for DNA extraction and the reminder as a voucher specimen. In some cases, a live specimen was grown in culture for morphological identification or physiological tolerance testing. There were a few cases where no voucher was prepared, as only enough tissue after DNA extraction was present (Table S1).
Algal specimens collected within the New Zealand marine aquarium trade were inspected under a dissecting microscope to remove detritus and epiphytes. Specimens with excessive overlapping branches had branches removed as required to expose the morphology of the alga and then the specimen was pressed on herbarium paper. Specimens were also preserved in 100% ethanol if too small to press and mounted on microscope slides in 60% Karo (corn syrup)/1% formalin. Aniline blue staining was used to stain the cytoplasm of ethanol preserved specimens. The edges of the coverslip were sealed with nail polish after 2 weeks to prevent over-drying. Vouchers were deposited at Te Papa Tongarewa (WELT; Index Herbariorum: http://sweetgum.nybg.org/science/ ih/) (Table S1). Micrographs were taken using an Olympus BX63 microscope with a DP80 camera (Olympus). Herbarium photographs were taken using an iPhone 7. Figures were annotated in Canvas Draw 5 (Canvas GFX, Plantation, FL, USA).

Molecular identification
Algae were preliminarily identified morphologically, and primers selected based on these identifications. The molecular identification of specimens followed broadly standard procedures as presented elsewhere (Zuccarello and Paul 2019). Silica gel-dried algal samples were extracted with a modified CTAB method (Zuccarello and Lokhorst 2005), or a Chelex extraction protocol for small tissues (Goff and Moon 1993). DNA extracts were stored at −20°C.
The PCR reaction volume was 30 μl and consisted of ∼0.1-0.4 μg genomic DNA, 1.25 nmol of each dNTP, 7.5 pmol of each primer, 1× reaction buffer, 2.5 mM MgCl 2 , 4-5 µl of 0.25% bovine serum albumin and 1 unit of Taq polymerase (Bioline, Meridian Bioscience Inc., USA). The standard amplification conditions were as follows: denaturing at 94°C for 5 min, 36 cycles of denaturing at 94°C for 1 min, annealing at 45-55°C for 30 s and extension at 72°C for 1 min and a final extension at 72°C for 5 min. Primer-specific annealing temperatures are shown in Table S2.
Agarose gel electrophoresis was used to identify successful amplification of the target region. PCR products were prepared for sequencing using Exo Sap-IT following the manufacturer's protocol (USB Corporation). Cleaned products were sequenced commercially (Macrogen Inc., Seoul, Korea). Sequences were edited using the Geneious software package (Geneious Prime, https://www.geneious.com) to remove primers and low-quality sequence.
Sequences were queried against the National Center for Biotechnology Information sequence database (NCBI). GenBank using the Basic Local Alignment Search Tool (BLAST) (Johnson et al. 2008). The BLAST top hit was recorded along with the % identity.
Alignments were built for each gene and genus of alga sampled using the top BLAST hits for each of our sequences and GenBank sequences of named species in the top BLAST hit list. In some cases, examples of closely related genera or other species were included in the alignment. Alignments were made using MAFFT implemented in Geneious. Alignments of protein-coding genes were translated and trimmed to start at the first codon position to allow for codon position partitioning.
Maximum-likelihood (ML) analyses were implemented using IQ-tree 2.0 (Trifinopoulos et al. 2016). IQ-tree was used to select the molecular evolution models (Modelfinder) (Kalyaanamoorthy et al. 2017) and construct (ML) trees with 500 bootstrap replicates. Models selected for each data set were selected using the BIC criterion and are presented in Table S3. Trees were annotated in Figtree v1.4.4 (Rambaut 2009) and Canvas Draw. Sequences were deposited in GenBank, and accession numbers, plus WELT voucher numbers of sequenced specimens presented in Table S1.

Culture experiments
Survival of algal samples was tested at New Zealand minimum sea surface temperatures determined from National Oceanographic and Atmospheric Administration (NOAA) data for Auckland and Wellington (seatemperature.org). Four culture chambers were set to the NZ minimum sea surface temperatures as follows: minimum temperature in Wellington winter (11±1°C), minimum temperature in Auckland winter (14±1°C), and for the other two incubators, 20±1°C a mid-range temperature and a control temperature matching reef aquaria, 25±1°C. HOBO loggers were used to track the temperatures (over 3 days) to insure temperature did not fluctuate more than 1°C. Fluorescent tubes provided photosynthetically available radiation (PAR) of 10-20 µmol −1 m 2−1 s −1 . The light cycle was 16:8 light:dark. Sterile dishes with lids were filled with 200 ml sterile (steam sterilised for 10 min at 100°C) seawater with quarter-strength modified Provasoli's enrichment medium (West 2005), and salinity adjusted to 35 psu. Two drops of 1 mg/ml germanium dioxide were added per dish to inhibit diatom growth (Lewin 1966).
Algal species identified as frequent arrivals or those traded by hobbyists for nutrient capture were selected from aquaculture tanks at 25±1°C (Woodhouse aquaria) for use in the temperature tolerance experiment. The algal samples were cut into 12 approximately equal lengths. Specimens were kept at 25°C for 24 h post-cutting in dishes to allow for recovery. Samples were then randomly allocated to separate sterile dishes, 3 dishes per algal specimen per temperature were placed randomly in the chamber. No acclimatisation procedure was undertaken.
Survival was measured by bleaching of the alga. The alga turning white (loss of pigments) or disintegration was considered as no survival. Specimens were observed weekly for chloroplast constriction or signs of reproduction. Any changes in morphology were recorded. Any remaining pigmentation was considered as 'survival'. Survival/no survival was determined after 2 weeks. Some specimens were kept for up to 4 weeks to confirm survival.

Results
The list of algae collected, analysed and identified is presented in Table S1. In total, we collected 82 specimens from tropical aquaria: 15 brown algae, 33 red algae and 34 green algae. A few examples will be presented in detail.

The genus Caulerpa
Eight samples were identified as Caulerpa (A25, A26, A40, A41, A45, A79, A91, A111) and amplified with partial tufA. BLAST searches were performed and an ML tree was built containing a total of 52 tufA sequences of Caulerpa spp. Caulerpella ambigua Okamura (KM186521) was used as the outgroup. The following ML tree ( Figure 1) and identification for our Caulerpa spp. are presented.
Sample A41 was 2 bp different from A91 in tufA. Both samples had a top BLAST hit and an ML phylogenetic placement (99%) with C. chemnitzia. The clade containing these samples also contains sequences named Caulpera peltata J.V.Lamouroux (KC153510) an earlier name for C. chemnitzia (Belton et al. 2014). We designated samples A41 and A91 as Caulerpa chemnitzia.

Caulerpa cylindracea Sonder
Sample A25 ( Figure 2C) is a siphonous green alga with a thick stolon, 2.5 mm diameter, thick assimilators, 2 mm diameter and 4-5 cm length, with one row of elongate ramuli with rounded tips on opposite sides of the assimilator. Specimen A25 top BLAST hit was C. okamurae Weber Bosse (KX809677). In the ML tree ( Figure 1) of tufA sequence data, A25 grouped in a clade containing C. okamurae, C. racemosa and C. cylindracea with high support bootstrap (93%). A25 is identical to all the sequences in the clade. Belonging to the clade containing A25 is a sample of C. cylindracea from the type locality, designated a reference specimen (JN851143; Belton et al. 2014). The clade includes samples identified as C. racemosa, an often incorrectly named species of Caulerpa (Belton et al. 2014). We designated specimen A25 as Caulerpa cylindracea.

Caulerpa lentillifera J. Agardh
Sample A26 ( Figure 2D) is a siphonous green alga with thin assimilators, 1-1.5 mm diameter, 2-6 cm long, with rounded ramuli, 1-1.5 mm diameter. Assimilators have ramuli along the entire length except for 5-7 mm from attachment to the infrequently branched stolon, many short rhizoids are present along the length of the stolon.
Both BLAST searches and the ML phylogeny indicate that this specimen is C. lentillifera (Figure 1). We designated A26 as Caulerpa lentillifera.

Caulerpa nummularia Harvey ex J. Agardh
Sample A79 ( Figure 2E) is a siphonous green alga with flattened mushroom-shaped ramuli, assimilators are thin, <0.5 mm diameter and <2 cm long, with few ramuli, the  Supplementary Table S1). GenBank Accession numbers and species designation associated with samples given. The outgroup, Caulerpella ambigua (KM186521) was removed for clarity. Scale bar = substitutions per site. stolon is thin, <0.5 mm diameter, and has many short rhizoids, <7 mm length. Sample A111 is very similar morphologically ( Figure 2F).
Sample A79 and A111 group based on tufA sequence data with representative specimen C. nummularia Harvey ex J.Agardh (JN817665; Belton et al. 2014) with high bootstrap support (Figure 1). We designated sample A79 and A111 as Caulerpa nummularia.
Sample A40 is within a clade, based on ML tree (Figure 1) containing C. racemosa f. macrophysa (Sonder ex Kützing) Svedelius and C. racemosa. Caulerpa racemosa is known to be highly variable morphologically with many varieties and forms, some of which have been synonymised into C. racemosa (Belton et al. 2014). Caulerpa racemosa var. macrophysa (Sonder ex Kützing) W.R.Taylor, needs further taxonomic work, the type sequenced, but many also be an environmental variety of C. racemosa. We designated specimen A40 as Caulerpa racemosa.
Caulerpa sertularioides (S.G.Gmelin) M. Howe Sample A78 ( Figure 2H) is a siphonous green alga with feather-like assimilators that are between 1.5 and 2.5 cm in length and 1 mm in diameter. The assimilators are dark green and are frequent along the stolon. The stolon is 1 mm in diameter. Cylindrical pinnules, ending in sharp points are present on opposite sides of the assimilator, giving the alga a feathery appearance. Rhizoids are present on short off branches from the stolon. This sample was not sequenced but we designated A78 as Caulerpa sertularioides based solely on morphology.

Caulerpa serrulata (Forsskål) J.Agardh
Sample A45 ( Figure 2E) is a siphonous green alga with a thick stolon, 2 mm in diameter, short assimilators with flat dichotomous branching blades with serrated margins. The short rhizoids are spaced irregularly along the stolon.
Sample A45 is within a clade containing mostly sequences from samples identified as C. serrulata, but also a sample identified as C. cupressoides (Vahl) C.Agardh (DQ652336) with moderate support bootstrap (84%). We designated sample A45 as Caulerpa serrulata.

The genus Lobophora
Lobophora (Dictyotales) is a well-known tropical reef genus with many cryptic species. While practically all specimens worldwide were designated as L. variegata (J.Lamouroux) Womersely ex E.C.Oliveira, over 40 cryptic or pseudo-cryptic species are now known (Vieira et al. 2014(Vieira et al. , 2017 with some being assigned species names. All Lobophora spp. samples found in this study had similar external morphology with a flattened thallus attached to the substrate with rhizoids, light to dark brown pigmentation with a darker ring of apical cells on the growing margin of the thallus. Most samples were small so were preserved on microscope slides. We collected nine samples of Lobophora, eight of which were sequenced with partial cox3. An ML tree was built using our sequences and samples selected from GenBank (648 bp alignment) for a total of 102 Lobophora spp. samples. Zonaria flabellata (Okamura) Papenfuss (JQ364040) was used as the outgroup (Figure 3). Of the eight samples sequenced ML phylogenetic analysis indicates that they are seven different species.
Both samples A7 ( Figure 4A) and A32 ( Figure 4B) were round flattened thalli with light brown pigmentation. A noticeable pale ring and darker apical cell margin was present. Rhizoidal holdfasts anchored the specimens to aragonite substrate. These samples matched L. asiatica Z.Sun, Ji.Tanaka & H.Kawai. We designated samples A7 and A32 as L. asiatica.

The genus Chaetomorpha
Three samples identified as morphologically Chaetomorpha sp. (Cladophorales) were found during this study. Sample A80 (Figures 5 and 6A) and A112 were formed of large-sized cells, 400 µm in diameter and 600 µm-1 mm in length, filaments are dark green and are tangled in rounded balls. These samples were unattached to the substrate. ML phylogenetic analysis ( Figure 5) of partial LSU (22 taxa, 583 bp) grouped these samples with several Chaetomorpha species including C. vieillardii (Kützing) M.J. Wynne (LT607222). We designate these two samples as C. vieillardii. The taxonomy of many members of the Cladophorales is known to be difficult, due to few characters and morphological convergence, although molecular insights and new nomenclature is aiding in producing monophyletic genera (Boedeker et al. 2016). One of these new genera is Lurbica, a genus segregated from Chaetomorpha. Chaetomorpha vieillardii has been proposed for the relatively large celled tropical specimens of Chaetomorpha (Wynne 2011).
Specimen A82 ( Figure 6B) has thin unbranched upright filaments up to 6 cm long and was attached to the substrate, similar to Chaetomorpha. Samples A82 is in a clade with other GenBank specimens newly transferred to Lurbica, but differs by 13 bp from the samples in this group, possibly indicating a new species. We designated sample A82 as Lurbica sp.

Physiological tolerance
A total of 10 algal samples (7 red, 1 brown and 3 green) were tested (Table 1).
Samples A25, A80 and A62 were maintained for 2-4 weeks in the experimental conditions to determine their survival as they appeared to remain viable (Table 1).
All samples survived at the control temperatures of 20°C and 25°C during the 2-week experimental period. All samples at the Auckland minimum winter SST of 14°C did not survive at 2 weeks, or 4 weeks, except Chaetomorpha vieillardii (A80) which survived for 4 weeks. No samples at the Wellington minimum winter SST of 11°C survived 2 weeks. After 1 week at 11°C, Chaetomorpha vieillardii (A80) ( Figure 6C,E,H) showed  Caulerpa cylibndracea --+ + A62 Dictyota friabilis --+ + chloroplast constriction and did not recover. Chaetomorpha vieillardii (A80) at 14°C treatment remained normally pigmented (Figure 6D,F,G) with no signs of bleaching or chloroplast constriction.

Discussion
Our data reveal a large diversity of tropical macroalgae arriving in New Zealand through the marine aquarium trade. A total of 62 different species were identified, although some samples could only be identified to the genus level. There is a high diversity of tropical macroalgae in the European marine aquarium trade (Vranken et al. 2018), as well as in Brazil (Torrano-Silva et al. 2013). Macroalgal diversity in New Zealand tropical marine aquaria has been understudied (Smith et al. 2010). The presence of non-indigenous algae in aquaria poses the risk of the introduction of potentially invasive species. The impacts of such species introductions can be large, including decrease in native species richness (Casas et al. 2004), biofouling of Figure 5. ML tree of Chaetomorpha spp., and associated genera, using partial LSU sequence data. Only bootstrap support values >70% displayed. Samples collected in bold (designations in Supplementary  Table S1). GenBank Accession numbers and species designation associated with samples. The outgroup was Cladophoropsis membranacea (AM503489) removed for clarity. Scale bar = substitutions per site.
aquaculture equipment (Fitridge et al. 2012;Pochon et al. 2015) and great effort and cost of eradication attempts. The aquarium trade is a known vector of algal introductions including the introduction of the cold tolerant Caulerpa taxifolia to the Mediterranean (Meinesz and Hesse 1991). Other Caulerpa spp. introduced outside their native range includes C. chemnitzia (Sghaier et al. 2016), C. cylindracea (Bernardeau-Esteller et al. 2015) and  H). A, Chaetomorpha vieillardii, A80. Scale bar = 2 cm. B, Lurbica sp., A82. Scale bar = 2 cm. Chaetomorpha vieillardii at weekly intervals during the first 3 weeks of the 4-week temperature tolerance experiment. C, E and G are at 11°C over weeks 1, 2 and 3, respectively. D, F and H are at 14°C over weeks 1, 2 and 3, respectively. Scale bars = 400 µm.
C. racemosa (Nyberg and Wallentinus 2005;Piazzi et al. 2003), our results found these three species in New Zealand tropical aquaria. Some Caulerpa species have caused great detriment to native ecosystems due to shading and outcompeting native species (Meinesz et al. 2001;Piazzi et al. 2003Piazzi et al. , 2016. A previously study has recorded the presence of a non-invasive strain of C. taxifolia and C. racemosa in the New Zealand Aquarium trade (Smith et al. 2010).
Propagule pressure is an important determinant of introduction success (Duggan et al. 2006). There are two aspects of propagule pressure; the initial number of individuals introduced and the frequency of subsequent immigration (Drake and Lodge 2004). The more individuals introduced (larger propagule numbers), the higher the probability of establishment. An example of propagule pressure in a natural system is the large populations of the introduced alga Undaria pinnatifida in marinas, facilitating its spillover to nearby rocky reef ecosystems (Epstein and Smale 2018). One species of red alga Hypnea, H. wynnei, was recorded four times. Two of the three Valonia species were also found multiple times, with two occurrences of V. ventricosa and three occurrences of V. macrophysa. Species in the Peyssonneliaceae were found multiple times. This indicates, if this reflects the importation incidence, that propagule pressure could increase the risks of release and establishment in New Zealand. The importation of corals into New Zealand occurs on a regular basis. We cannot conclude whether the multiple occurrences of certain species of macroalgae that we found in our study are due to multiple introductions to New Zealand or the reproduction of macroalgae in the importers' aquaria and their subsequent recruitment to different coral rocks. Our results do indicate that some species are prevalent in aquaria and so propagule pressure, and chance release, from personal aquaria and retailers is still a concern.
Our results also show the presence of at least five tropical Cladophorales (Chlorophyta) genera in New Zealand marine aquaria. The green algal genus Cladophora is known to contain some introduced species, including C. ruchingeri (C.Agardh) Kützing, possibly recently introduced to the Marlborough Sounds in New Zealand (Pochon et al. 2015).
Chaetomorpha vieillardii is the only species tested that survives Auckland minimum sea surface temperatures (14°C). A bloom-forming strain of C. viellardii is present in Guam (University of Guam https://cnas-re.uog.edu/wp-content/uploads/2018/09/ IInvasive_speciesGuam.pdf). Chaetomorpha vieillardii has been found for sale in the European aquarium trade along with another unidentified species (Vranken et al. 2018). With knowledge on the potential for invasiveness in Chaetomorpha spp., methods of safe disposal and implications of unsafe disposal have been discussed (Odom and Walters 2014). It has been found that Chaetomorpha sp. can survive extended periods ∼6 days in freshwater, this means that algal tissue or fragments in aquarium water may survive and reach the ocean if disposed of into storm water drains (Odom and Walters 2014).
The ability of an algae to survive and reproduce in aquaria could have implications for its potential spread. Many algae have the capability to regenerate from small fragments or holdfasts making eradication from an aquarium difficult. Hypnea wynnei had cystocarps our samples, showing the potential for sexual reproduction. Sexual reproduction allows for increased genetic variants in populations. The fluctuation in water parameters that occurs in aquaria may provide selection pressures, selecting for weedy strains and developing strains more tolerant to changes in environmental conditions and potentially being more invasive.
Polymorphic species have always been an issue for taxonomists. Many algal genera contain polymorphic species; examples include Sargassum (Mattio and Payri 2011), Halimeda (Verbruggen et al. 2005) and Caulerpa (Gacia et al. 1996;Sauvage et al. 2013). Caulerpa, for instance, can be highly polymorphic, making species assignment problematic (Belton et al. 2014). One of our samples Caulerpa nummularia (A111) had flattened mushroom-shaped ramuli when grown under low intensity lighting in a refugium, once relocated to higher light the morphology of new ramuli changed to flattened with serrations. Molecular barcoding therefore is an effective way of distinguishing species that have overlap in morphological boundaries (Verbruggen et al. 2005(Verbruggen et al. , 2007.
Cryptic diversity is another issue that makes morphological identification difficult due to little or no morphological differences between some species (Cianciola et al. 2010). Two brown algal genus well known for cryptic diversity are Lobophora (Vieira et al. 2014). A recent study identified many genetically distinct entities of Lobophora spp. that can be considered species (Vieira et al. 2014). Our results show that at least seven different species of Lobophora have been found in the New Zealand aquarium trade. One species L. asiatica was found twice. There were a few samples that also may be new species of Lobophora. Sample A60 designated by us as Lobophora sp1 are different from any other Lobophora species sequenced so far. Red algal genera containing cryptic diversity have also been found by molecular barcoding (e.g. West 2002, 2003;. This shows the importance of molecular identification in accurately identifying algae. Cryptic diversity needs to be taken into account to correctly determine the diversity of algae entering the New Zealand marine aquarium trade as more species may be entering than recognised by morphological characters. Cryptic species can have different physiological tolerance characteristics, as shown by Muangmai et al. (2015) and in cryptic species of Dictyota sp. at the Canary Islands (Tronholm et al. 2010). The variation in temperature tolerance between cryptic species suggests that some cryptic species of algae will be better suited to New Zealand winter minimum SST than others and therefore pose more of a biosecurity risk than others.
We used BLAST searching along with ML phylogenies to determine the identity of our samples. From our results, the top BLAST hit and sequences of high identity to the samples were often in the ML clade containing our samples, and in other cases, the top BLAST hit was sister to our sample.
Some taxonomic groups are understudied therefore have inadequate sequences available in GenBank for effective species-level identification. This is apparent in the ITS1 gene in fungi (Nilsson et al. 2006) and the rbcL gene in diatoms ). Some of our samples, that could only be identified to genus level, had large pairwise distances to the closest GenBank sequences. Sample A2 (Accession TBA), a branching red alga, is 13.4% different in cox1 to its closest GenBank sequence, Lomentaria divaricata from Canada. Sample A4 (Accession TBA) had 9.2% pairwise difference in cox1 to Halopeltis willisii from North Carolina USA. Samples A2 and A4 are either of described species that have not been sequenced using cox1, or may be undescribed species.
The names of sequences in GenBank must be taken with some caution. For example, within our Caulerpa serrulata clade, there is a sequence named C. cupressoides (DQ652336) from Indonesia; this may have been a misidentification as it has only 0.5% pairwise difference in tufA to our sample A45 and samples of C. serrulata. Another limitation of GenBank sequences is that some sequences have not been updated when species or generic names change. An example, sequences named Caulerpa peltata, an entity that has been synonymised with Caulerpa chemnitzia (Belton et al. 2014). Our Cladophorales tree also has species present that have been formally renamed but not updated on GenBank by the submitter, for example, Cladophora montagneana has been renamed Willeella brachyclados (Wynne 2016). Misidentification and incorrect naming of sequences submitted to GenBank is a common occurrence, 65% of over 600 Ganoderma spp. fungal sequences surveyed were misidentified or ambiguously labelled (Jargalmaa et al. 2017).
This study found a large diversity of tropical macroalgae entering the New Zealand Aquarium trade on imported corals. These findings have important implications for New Zealand biosecurity, as there are currently no measures in place to prevent invasive algal species from passing through quarantine undetected as spores or germlings. Strict monitoring of marine aquarium livestock during quarantine and a longer quarantine period would be needed to detect the presence of exotic macroalgal hitchhikers. The outlawing of the sale of certain algal genera has been suggested (Smith et al. 2010). A ban of the sale of nine Caulerpa spp. is in place in California USA, with a genus wide ban in San Diego (Diaz et al. 2012). Banning sale and the propagation of potentially invasive algae will need enforcement and education for it to succeed.
Many aquarists are unaware of the impacts of releasing aquarium specimens or wastewater into natural environments (Duggan et al. 2006;Martin and Coetzee 2011). The finding of Chaetomorpha vieillardii with cold tolerance to Auckland minimum winter temperature is alarming as northern New Zealand has a sea surface temperature that rarely declines below 14°C. As this strain of Chaetomorpha vieillardii can tolerate low temperature, grows into very large unattached balls and can reproduce from single cell fragments (Odom and Walters 2014), it is a biosecurity threat in northern New Zealand. The general public needs to have access to information on the proper disposal of unwanted aquarium livestock and wastewater to reduce the risk of an unintentional release of an algae.
Having access to more marine aquarium livestock importers and a larger sample size of coral rocks would be useful. However, we did find a very large diversity of species. Not knowing the collection location (anecdotally mostly from Indonesia, also tropical Australia) of many of the marine invertebrates that our samples came from was a limitation of this study. Collection location information would have been useful for determining habitat compatibility to New Zealand, as some of the corals kept in New Zealand aquaria are in fact subtropical, possibly associated algae will also have low temperature tolerance.
Testing a broader range of physiological tolerance parameters, for example, salinity might be helpful in determining the introduction risk of tropical algae within New Zealand. Our study used the survival/death of tropical macroalgal samples at different temperatures. Survival/death was adequate for this study but doesn't take into account whether or not the surviving algae are stressed. Further work in this area is needed to determine the risks.
A survey, along with dried sample collection for DNA extraction, for algae kept in aquarium refugia or reported by marine aquarium hobbyists would be useful to better gauge the prevalence of certain algal genera in New Zealand aquaria. Sampling from coral rocks that have been freshly imported into New Zealand and are in quarantine would be useful as we will achieve a larger sample size and collect algae that may otherwise detach, be removed or die back while in quarantine. We will also be able to observe if there are signs that the algae are reproducing during quarantine.