Clone wars: asexual reproduction dominates in the invasive range of Tubastraea spp. (Anthozoa: Scleractinia) in the South-Atlantic Ocean

Introduction

The marine environment is continuously subjected to multiple stressors, many of which are associated with human activities (e.g., over-exploitation of resources, pollution, climate change and invasive species) (Halpern et al., 2014; Gallardo et al., 2016). Among these stressors, invasive species are considered to be a major threat to biodiversity (Molnar et al., 2008) with the potential to quickly trigger changes in native communities and the ecosystem services and functions, which can have wide-ranging negative impacts. There are numerous examples of marine invasions which impact humans or native biota, such as in the Mediterranean Sea with the invasion of the ctenophore Mnemiopsis leidyi, which caused the collapse of the fishing industry (Shiganova, 1998), the algae Womersleyella setacea, that negatively affected sponge reproduction (Caralt & Cebrian, 2013) and the lionfish Pterois spp., responsible for a reduction in the native fish recruitment in the Atlantic (Albins & Hixon, 2008).

Scleractinian corals are known to play a key role in the marine environment by building structurally complex and highly diverse ecosystems (Reaka-Kudla, 1997). As ecosystem engineers that are under threat globally (Hoegh-Guldberg, 1999; Pandolfi et al., 2003), scleractinian corals are rarely seen as an environmental risk. However, three scleractinian species from the genus Tubastraea were introduced and are spreading rapidly throughout the Western Atlantic Ocean (De Paula & Creed, 2004; Fenner, 2001; Fenner & Banks, 2004; Sammarco, Atchison & Boland, 2004; Sammarco, Porter & Cairns, 2010; Capel, 2012; Sampaio et al., 2012; Costa et al., 2014; Silva et al., 2014), threatening native and endemic species (Mantellato et al., 2011; Santos, Ribeiro & Creed, 2013; Creed, 2006) and fouling man-made structures and vessels.

Tubastraea is an azooxanthellate dendrophyllid genus from the Pacific and Indian Oceans that was first reported in the Caribbean in 1943 (Vaughan & Wells, 1943). Since then, three species have been identified in the Western Atlantic Ocean: (1) T. coccinea, now reported along 9,000 km of coastline of the Western Atlantic Ocean from Florida (26°47′N, 80°02′W) (Fenner & Banks, 2004) to Southern Brazil (27°17′S, 48°22′W) (Capel, 2012); (2) T. tagusensis, along the Brazilian coast (De Paula & Creed, 2004); and (3) T. micranthus in the Gulf of Mexico (Sammarco, Porter & Cairns, 2010). All three are considered opportunistic species most likely associated with transport on ships and/or oil platforms in the Caribbean, Gulf of Mexico and Brazilian coast (Cairns, 2000; Castro & Pires, 2001; Sammarco, Porter & Cairns, 2010).

Once established, invasive species can alter the structure of local communities, displacing and outcompeting native species (Vitousek, 1990; Mooney & Cleland, 2001; Lages, Fleury & Menegola, 2011; Cure et al., 2012; Santos, Ribeiro & Creed, 2013; Miranda, Cruz & Barros, 2016). In contrast to the native range, where Tubastraea is largely restricted to shaded or marginal habitats, studies on oil rigs in the Gulf of Mexico have shown that both T. coccinea and T. micranthus are excellent competitors and can overgrow other species (Hennessey & Sammarco, 2014; Sammarco et al., 2015). Similarly, in Brazil, T. coccinea and T. tagusensis can cover up to 100% of the available surface in some areas (Mantellato et al., 2011), killing native and endemic coral species upon direct contact (Creed, 2006; Santos, Ribeiro & Creed, 2013; Mantellato & Creed, 2014; Miranda, Cruz & Barros, 2016).

Fast growth rate, rapid range expansion, early reproductive age, propagule pressure and a wide variety of reproductive and survival strategies are biological characteristics usually associated with invasion success (Sax & Brown, 2000; Sakai et al., 2001; Lockwood, Cassey & Blackburn, 2005; Sax et al., 2007). Tubastraea species possess all of these characteristics (Cairns, 1991; Ayre & Resing, 1986; Glynn et al., 2008; Harrison, 2011; Capel et al., 2014; De Paula, Pires & Creed, 2014), which are enhanced by the fact that within the invaded areas they generally lack natural predators and dominant competitors. In addition, a large number of infested vectors (e.g., oil platforms and monobuoys) have been recorded transporting Tubastraea spp. along the Brazilian coast, leading to rapid range expansion throughout the Southwestern Atlantic Ocean (Creed et al., 2016).

Asexual reproduction improves coral ability to reach high abundance (Ayre & Miller, 2004) and may be an important trait of many invasive species, mainly in the first stage of invasion (Taylor & Hastings, 2005). When associated with early reproductive age and high propagule pressure it can rapidly increase abundance. Asexual production of brooded planulae has been reported in several anthozoans, including actinarians (Ottaway & Kirby, 1975; Black & Johnson, 1979), octocorals (Brazeau & Lasker, 1989) and scleractinians (Stoddart, 1983; Ayre & Resing, 1986). Although T. coccinea and T. diaphana appear to reproduce mainly by asexually produced larvae (Ayre & Resing, 1986), there is no information for their congeners, and the proportion of sexual versus asexual reproduction remains unknown within the genus. Furthermore, Ayre & Resing (1986) were able to score only two allozyme loci to infer asexual production of brooded larvae of Tubastraea spp. and the use of a larger number of more polymorphic loci, such as microsatellites, is desirable to corroborate their findings.

Although Tubastraea species are spreading rapidly and changing local benthic communities throughout the tropical Western Atlantic, information about their genetic diversity and reproductive strategies are still scarce. The study of reproductive strategies of invasive species is fundamental to understanding the invasion process, preventing new invasions, development of effective management strategies, and resolving the ecological and evolutionary processes involved in their invasion success (Sakai et al., 2001; Sax et al., 2007). However, to date there was no molecular marker developed to perform such studies with Tubastraea. Here, we report 12 novel microsatellite loci specifically developed for T. coccinea and cross-amplified in T. tagusensis and investigate the clonal structure and genetic diversity of populations of these alien invasive corals in the Southwestern Atlantic Ocean.

Materials and Methods

Sampling and DNA extraction

Microsatellite development was performed using samples of T. coccinea collected from Búzios Island (23°47′S, 45°08′W, 6 m in depth) and also from a monobuoy (IMODCO 4) at the São Sebastião channel (23°48′S, 45°24′W, 5 m of depth), Brazil. Additional samples of T. coccinea and T. tagusensis, collected from Todos-os-Santos Bay (TSB), northeastern Brazil (12°49′S, 38°46′W), and Ilha Grande Bay (IGB) (23°06′S, 44°15′ W), southeastern Brazil (∼24 colonies/species/locality), were used to test the markers and evaluate their genetic diversity (Fig. 1). Samples were preserved in 96% ethanol or CHAOS buffer (Fukami et al., 2004) prior to extraction. Total DNA was extracted using the Qiagen DNeasy tissue and blood kit following the manufacturer’s instructions or using the Phenol:Chloroform method described by Fukami et al. (2004).

Results

Genetic diversity and population structure

Only unique MLLs were used to assess genetic diversity and population structure in each species. For both species, TSB had higher number of alleles, allelic richness and number of private alleles compared to IGB, with T. coccinea presenting the more accentuated differences (Table 2). There were no significant deficits of heterozygosity; both observed (Ho) and expected (He) heterozygosity were similar when comparing between localities for both T. coccinea (TSB: 0.80 and 0.56; IGB: 0.77 and 0.45) and T. tagusensis (TSB: 0.67 and 0.43; IGB: 0.64 and 0.37). The inbreeding coefficient (F_IS), although not significant, was negative for both localities and in both species, indicating an excess of heterozygotes (Table 2).

F_ST and $G_{\mathrm{ST}}^{″}$ values were 0.06 (p = 0.08) and 0.13 (p = 0.07) for T. coccinea and indistinguishable from zero (p = 0.69 and p = 0.69) for T. tagusensis. The lack of significant population structure among the sampled localities indicates similar allele frequencies for both species across these sites. Although Bayesian analysis recovered two genetic clusters for T. coccinea for both ranges of K tested, these groups are not a function of population structure between localities (Fig. 2), but instead, reflect the presence of population structure within each locality. Furthermore, there is no evidence of interbreeding between the two clusters, and the F_ST values between these sites is likely a result of the strikingly different proportion of these two groups in each site. In contrast, no clustering was observed between or within localities for T. tagusensis, with the most likely K value being one for both ranges of K tested (Fig. 2).

Discussion

The novel microsatellite markers reported herein will enable further studies regarding the genetic diversity and population structure of Tubastraea spp. corals in the Atlantic and native ranges of these invasive populations. Using these microsatellites, this study shows that both invasive coral species (T. coccinea and T. tagusensis) have high proportions of clones at both localities on the Brazilian coast with identical multilocus lineages (MLLs) found in sites separated by more than 1,500 km. The results indicate that asexual reproduction dominates in the invasive range of Tubastraea spp. in the Southwestern Atlantic and despite the large distance between localities, no significant population structure could be found. In contrast, there are clear signs of population structure across this same region in an endemic spawning coral species (Mussismilia hispida, Azevedo, 2015).

Our results support previous work reporting reproduction via asexual larvae in T. coccinea (Ayre & Resing, 1986). Likewise, the high proportion of clones found at both sampled localities for T. tagusensis indicates likely reproduction by asexual larvae for this species also, a reproductive mode previously recorded for only three scleractinian species: Pocillopora damicornis (Stoddart, 1983), Tubastraea diaphana and T. coccinea (Ayre & Resing, 1986). Indeed, a study on the reproductive strategies of T. coccinea and T. tagusensis in the Southwestern Atlantic observed a small number of spermaries and the presence of embryos and planula at different times of the year, concluding that asexual reproduction could be important for both species (De Paula, Pires & Creed, 2014). For most corals, clonality is a result of mechanical fragmentation due to physical disturbances (Foster et al., 2013; Nakajima et al., 2015). T. coccinea and T. tagusensis, however, are not prone to fragmentation, so the high number of clones observed for both species in this study seems more likely to result from asexually produced larvae. Nevertheless, it is desirable to confirm the production of asexual larvae for both T. coccinea and T. tagusensis by performing paternity studies in the future.

For invasive species, asexual reproduction can be crucial in the first stage of invasion, when sexual partners are scarce or absent, because it significantly enhances the chances of survival for the colonists (Taylor & Hastings, 2005). Successful invasions originating from a few clonal genotypes have been previously recorded for plants (Ren, Zhang & Zhand, 2005; Liu et al., 2006) and other cnidarians (Reitzel et al., 2008). Asexual reproduction is dominant in the invasive range and it may have contributed to the invasive success of Tubastraea in the Southwestern Atlantic, where the rocky shores provide a suitable habitat and release from enemies (Enemy Release Hypothesis, Keane & Crawley, 2002). At IGB both studied coral species have high percentage of clones and an extremely low genotypic evenness, indicating that most colonies are clones belonging to the same genet. Sampling more areas surrounding each collection site is needed to thoroughly examine clonal diversity for these regions, but particularly in TSB where samples were more widely spaced, this observation supports the role of asexual reproduction in increasing local abundance. Gregarious settlement has been previously observed for both T. coccinea (Glynn et al., 2008; De Paula, Pires & Creed, 2014), and T. tagusensis (De Paula, Pires & Creed, 2014), although these studies did not determine if the aggregated larvae were sexually or asexually derived. It is noteworthy that T. coccinea has higher numbers of MLLs, clonal richness and genotypic evenness at TSB than at IGB, suggesting increased occurrence of sexual reproduction or a greater number of successful colonists at the former site. Rates of sexual and asexual reproduction can be highly variable among geographic regions in other corals (Baums, Miller & Hellberg, 2006; Noreen, Harrison & Van Oppen, 2009; Combosch & Vollmer, 2011; Gorospe & Karl, 2013), but it remains unknown what governs the difference in the proportion of sexual and asexual reproduction at different localities. Several factors can influence both genotypic and genetic diversity in invasive species, including the number of invasions, the genetic diversity of the source population(s) and a variety of biological factors, such as the main reproductive strategy adopted by the species (Dlugosch & Parker, 2008). Although sexual reproduction might also occur in Tubastraea, the results obtained for T. coccinea might be an effect of the occurrence of recent multiple introductions from different native populations (Roman & Darling, 2007). Another hypothesis would be the presence of cryptic species, which has been found in other scleractinian corals (Pinzón & Weil, 2011; Warner, Van Oppen & Willis, 2015; Nakajima et al., 2017). Morphological analyses combined with molecular data including native populations are necessary to corroborate this hypothesis.

A decrease in genetic diversity as a result of a small founding population has been previously recorded for several invasive populations (Roman & Darling, 2007; Geller et al., 2008; Johnson & Woollacott, 2015; Wrange et al., 2016; but see Gaither et al., 2010; Gaither, Toonen & Bowen, 2012 for counter-examples). Here, we report excess of heterozygosity for both populations of both species and the presence of up to 100% heterozygous individuals at some loci (Table 1). High levels of heterozygosity can result from an isolate-breaking effect, when multiple introductions mix previously separated native populations (Holland, 2000; Hamilton, 2010). However, in this case, there is no evidence of mixing between the two genetic clusters (Fig. 2), indicating that they are not interbreeding. Thus, it seems more likely that TSB and IGB were colonized by different native populations followed by recent transport between localities without sufficient time for them to interbreed, although the possibility of cryptic species that are incapable of interbreeding should also be considered. If the first scenario of introduction by different native populations proves true, the high heterozygosity could be either a result of a founder effect in which the new area was, by chance, colonized by a higher number of heterozygote genotypes, or due to a higher fitness of the heterozygote genotypes, either of which could be propagated by asexual reproduction (De Meeus & Balloux, 2005). Alternatively, Gaither, Toonen & Bowen (2012) showed that introduced fishes in Hawai’i with a known history actually had higher and more even genetic diversity than was observed in the native range, and such an effect could also explain the observed pattern here. In contrast to what is observed with T. coccinea, we recover only a single genetic cluster for T. tagusensis between both populations. This single cluster could result from either invasion of both localities from the same source population, or a secondary invasion along the Brazilian coast from the original locality being spread to another. Unlike T. coccinea, which is now considered cosmopolitan (Cairns, 2000), T. tagusensis has a restricted distribution (Cairns, 1991) and may have naturally low genetic diversity. The distinction between these species is reminiscent of the pattern reported by Gaither, Bowen & Toonen (2013) in which population structure of species in their native range predicts the diversity and rate of spread in the invasive range.

Considering that (i) both T. coccinea and T. tagusensis brood larvae competent for only ∼18 days (in aquaria) that typically display gregarious settlement (Glynn et al., 2008; De Paula, Pires & Creed, 2014) and (ii) the absence of Tubastraea in extensive areas between the two localities, it is highly unlikely that they are connected through larval dispersal. On the other hand, oil platforms are known to be moved between these regions (Sampaio et al., 2012), and are considered the main vector for the introduction of Tubastraea into the southwestern Atlantic (Castro & Pires, 2001; Creed et al., 2016). Thus, our data showing a lack of structure between localities, and the occurrence of shared MLLs for each species among these distant sites, indicate that anthropogenic vectors, such as oil platforms, monobuoys, or other vessels have played an important role in dispersing these alien invasive species, and possibly assisting other species to spread along the coast (Almeida et al., 2015; Creed et al., 2016).

Conclusions

Invasive Tubastraea spp. are spreading quickly throughout the Atlantic, in some areas covering up to 100% of the available surface (Mantellato et al., 2011) and outcompeting native and endemic species (Mantellato et al., 2011; Santos, Ribeiro & Creed, 2013; Creed, 2006). Despite this documented impact and concern, little is known about the genetic diversity and reproductive strategies of Tubastraea species globally. This study provides the first survey of genetic diversity and likely reproductive strategies along the southwestern Atlantic coast, demonstrating that asexual reproduction has an important role in the population dynamics of both T. coccinea and T. tagusensis and is probably a relevant feature leading to their invasive success. Results also indicate that there were likely at least two different populations of T. coccinea introduced into the southwestern Atlantic. A molecular systematic examination of the genus is highly recommended in order to check for the occurrence of cryptic species. Future studies should focus on the identification of potential source populations and the global phylogeograpy of Tubastraea with the goal of tracking and limiting future invasions, as well as the establishment of effective management and prevention strategies.

Supplemental Information