First evidence for backcrossing of F1 hybrids in Acropora corals under sperm competition.

Acropora is a species-rich genus of reef-building corals with highly diverse morphologies. Hybridization among intercrossing species potentially influences species diversity within Acropora. However, the mechanisms that allow hybridization/backcrossing remain unknown. Although we tested a limited number of species, we hypothesized that Acropora gametes in the Indo-Pacific may preferentially fertilize conspecific gametes despite their compatibility with heterospecific gametes, leading to infrequent hybridization between potentially intercrossing species. In this study, F1 hybrids of Acropora florida and A. intermedia showed specific fertilization trends. For example, sperm had the ability to backcross with the parental species even in the presence of sperm from the parental species. Also, eggs of the hybrids produced from A. florida eggs and A. intermedia sperm ("FLOint") exhibited self-fertilization. Since a low ratio of hybridization between A. florida and A. intermedia is predicted, the population size of hybrids should be small. Therefore, self-fertilization would facilitate reproduction of the hybrid in nature, while remaining sperm could outcompete parental species sperm to backcross with eggs. Although we succeeded in breeding two colonies of hybrids, it is reasonable to speculate that hybrids show a high tendency to choose the most efficient sexual reproduction tactics.

Introduction

Hybridization is considered a mechanism for evolutionary innovation. Introgressive hybridization is caused by the repeated backcrossing of hybrids to the parental species[1]. Introgressive hybridization results in new gene combinations, leading to transgressive phenotypes[2,3]. Extensive hybridization is associated with the rapid diversification of species[4]. Moreover, hybrids can potentially occupy new habitats, differentiating them from the parental species[5]. Implications of hybridization for adaptation and hybrid fitness have been suggested[6-8], but the ways in which introgression occurs in nature are still unknown.

In the Indo-Pacific, the reef-building coral Acropora spp. is species rich (> 110 species)[9], and there is a potential relationship between hybridization and high species diversity[8]. For example, tabular species such as A. hyacinthus can form species complexes[9,10], and gene flow among such species complexes occurs in a complex manner[11]. In addition, intermediate morphologies among intercrossing species imply that admixture events are associated with morphological diversity and similarity[10,12,13]. Morphological similarity is associated with hybridization/introgression in Caribbean[14-17] and Indo-Pacific Acropora[8,12]. In both the Indo-Pacific and Caribbean, co-occurrence of spawning times/dates and gamete compatibility is related to introgression[18-20]. For repeated hybridization events among intercrossing species, the backcrossing of F1 hybrids to the parental species must occur, but the reproduction of F1 hybrids has not yet been fully investigated in the Southern Japanese Indo-Pacific, which is a hybrid hotspot area with high species richness of the coral Acropora[8,9].

For introgression between two species, F1 hybrids from the two species must backcross with the parental species. However, such reproductive strategies, including the fertilization mechanisms of F1 hybrids in the Indo-Pacific, have not been fully elucidated, because Indo-Pacific hybrids other than A. florida and A. intermedia[21] have not been successfully raised to their spawning age of approximately 7 years[21]. In nature, repeated hybridization between parental species arising from the backcrossing of F1 hybrids has been demonstrated[18] in A. prolifera, the hybrid of two species inhabiting the Caribbean[14]. Therefore, although the importance of hybridization in the species-rich Indo-Pacific reef-building coral Acropora has been posited, there is no evidence showing how hybrids reproduce at the gametic level, or how they backcross and/or mate with each other.

Although introgression in the coral Acropora has been shown[7,8,11,22], how F1 hybrids backcross with the parental species remains unknown. Our previous study showed that sperm of the F1 hybrids named “FLOint” and “INTflo”, bred from A. florida and A. intermedia, are compatible with eggs of the parental species[21]. In addition, eggs of the hybrid FLOint, raised from A. florida eggs and A. intermedia sperm, showed high selfing[21]. Ecologically, mating opportunities within F1 hybrids are more limited than are backcrossing opportunities with the parental species due to smaller numbers of F1 hybrid colonies. For hybrids to backcross with the parental species, the eggs of the parental species must accept hybrid sperm but tentatively prefer to mate with conspecific sperm in the presence of both conspecific and heterospecific sperm[23]. To clarify how gametes of the F1 hybrids mate, we examined whether hybrid sperm can outcompete parental species sperm, and whether hybrid FLOint eggs always show selfing. Using the results of these analyses, we show fertilization trends of the gametes of F1 hybrids (A. florida and A. intermedia) in the Indo-Pacific that lead to backcrossed and F2 generations of hybrids.

Materials and methods

Coral collection

In 2016, Acropora florida fragments from six colonies were collected from Majanohama, Akajima (Aka Island), Japan (26° 120 N, 127° 170 E), and two F1 hybrid colonies (INTflo: A. intermedia eggs × A. florida sperm; FLOint: A. florida eggs × A. intermedia sperm) were kept at Aka Island port. In 2016, we detected no A. intermedia with mature eggs, and thus A. intermedia spawning was supposed to have occurred during the previous full moon. In 2017, the hybrids (FLOint and INTflo) were used in experiments[12,21], and two F1 hybrids, measuring approximately 30–50 cm, were transferred to Sesoko Station from Aka Island and maintained in an aquarium tank. In 2017, fragments from seven colonies of A. florida and 12 colonies of A. intermedia were collected from Sesoko Island (26° 37 N, 127° 51 E). All colonies and fragments were kept in a running seawater tank at the Akajima Marine Science Laboratory in 2016 or Sesoko Station at the University of the Ryukyus in 2017 until 1–5 days before their predicted spawning date.

Spawning observation and gamete collection

Corals were observed at 20:30 from 5 days before their predicted spawning date on a full moon. When bundles were observed at the mouth of each polyp, the corals were transferred to a tank filled with seawater, and the time of spawning was recorded. The bundles were collected from the colony using plastic pipettes (Table 1), and gametes were separated into sperm and eggs using 100-µm plankton mesh, following Morita, et al.[24]. The sperm concentration of the isolated spermatozoa was determined using a hemocytometer. The final sperm concentrations were adjusted to approximately 104, 105, or 106 sperm/mL for use in the subsequent crossing experiments.

Table 1

Spawning time of the coral Acropora florida, A. intermedia, and hybrid (FLOint, INTflo) in 2016 at Aka Island and 2017 at Sesoko Island.

Date	Time	Species	Colony names for fertilization experiments
2016 Aka
June 19th	21:40	A. florida	*f74	Full moon:June 20th
	21:50	A. florida	*f1,f15
	22:00	A. florida, FLOint	*f11, *f13
2017 Sesoko
June 6th	22:08	A. intermedia	*i5, *i72	Full moon:June 9th
	22:09	INTflo
	22:12	A. intermedia	*i2
	22:13	A. intermedia	*i75
June 16th	21:29	FLOint
	21:35	INTflo
July 7th	22:10	A. florida	f77	Full moon:July 9th
July 8th	21:56	A. florida	f77, f79
July 11th	22:01	A. florida	f72

*Indicate colonies used for crossing experiments.

Crossing experiments and paternity tests

In the crossing experiments, approximately 200 eggs were transferred to 5 mL filtered seawater and 5 mL sperm suspension according to previously described methods[23,25]. The experiments were performed using A. florida and FLOint hybrid gametes on 19 June 2016 at Aka Island, and A. intermedia and INTflo hybrid gametes on 6 June 2017 at Sesoko Island. The spawning dates of A. florida and the two hybrids (INTflo and FLOint) did not coincide in 2017; INTflo spawned on 6 and 16 June and FLOint on 16 June, but A. florida spawned on 7, 8, and 11 July. Therefore, crossing experiments using FLOint and A. florida were not conducted.

For the crossing experiments, sperm concentrations were 104, 105, and 106 sperm/mL for the sperm non-choice and sperm choice tests (fertilization trials in the presence of both parental species and hybrid sperm). In the crossing experiments, only colonies that spawned on the same day were used. A. florida and the FLOint hybrid were used in 2016, and INTflo and A. intermedia were used in 2017 (Table 1, Supplementary Table 1). The fertilization ratio was recorded based on whether or not eggs showed cell division within 2–4 h after mating at a temperature of 29 °C. Larvae from 3 to 4 days after fertilization were preserved in 99.5% ethanol and used for paternity tests. Paternity tests were performed to confirm which sperm fertilized the egg in the sperm choice test. DNA was extracted from the larvae, and microsatellite analysis was performed using the extracted DNA as a template with the markers 11745m3 and 11401m4[26], according to Kitanobo et al.[23]. For each marker, fewer than two alleles were detected in the hybrids, and the same alleles were consistently detected from sperm and tissues, suggesting that the hybrids were not chimera (Supplementary data 1). Some analyses were conducted using acrylamide gel electrophoresis and others using fragment analysis with the ABI 3130xl or 3730xl DNA Sequencer (Applied Biosystems, Waltham, MA, USA). Microsatellite Analysis v1.0 (Applied Biosystems) software (https://www.thermofisher.com/order/catalog/product/4381867) was used to score the sizes.

Statistical analyses

We conducted Tukey’s honestly significant difference (HSD) tests to evaluate differences in multiple comparisons. Welch’s two-sample t-tests were used to confirm differences in the fertilization ratio when using heterospecific sperm. All statistical analyses were performed using R version 4.01[27].

Ethical approval

All applicable international, national, and/or institutional guidelines for sampling, care, and experimental use of organisms for the study have been followed, and all necessary approvals have been obtained (No. 31-30).

Results

F1 hybrid spawning times and dates

Hybrids (FLOint; one colony) and A. florida (five colonies) released gametes in June 19–21, 2016 on Aka Island. A. intermedia did not release gametes synchronously with the hybrids. A. intermedia with mature eggs were not found, and it is likely that most A. intermedia had already spawned around the full moon at the end of May (Table 1). In 2017, spawning of the hybrid INTflo and A. intermedia coincided, but A. intermedia, A. florida, and the hybrid FLOint did not spawn on the same date (Table 1). Most colonies of A. florida spawned in July rather than June.

Percentage of eggs fertilized in crosses among conspecifics

Crossing experiments using hybrids and parental species were conducted using A. florida and FLOint hybrid gametes on 19 June 2016, and A. intermedia and INTflo hybrid gametes on 6 June 2017. Fertilization ratios between the parental species were low in A. florida (< 11%) (Table S1). In contrast, fertilization ratios among A. intermedia were high under both low and adequate sperm concentrations, but the colony (I3_i2) had very low fertility ratios (Table S1).

Inherent selfing and backcrossing of F1 hybrid eggs

Self-fertilization of the INTflo and FLOint gametes was examined. INTflo gametes did not show self-fertilization in the presence of low (104 sperm/mL) to high (106 sperm/mL) sperm concentrations (Table S1). By contrast, as in our previous study[21], the eggs of FLOint showed high selfing ratios of 76%, 89%, and 76% under low, moderate and high sperm concentrations, respectively (104, 105, and 106 sperm/mL) (Table S1).

To investigate the inherent ability of hybrid eggs to backcross with the sperm of the mother species, A. florida or A. intermedia, gametic compatibility was examined. Hybrid eggs showed a high ratio of fertilization to the sperm of the mother species; INTflo crossed with the sperm of A. intermedia, and FLOint showed a high ratio of fertilization to A. florida sperm (Fig. S1). Moreover, there was no significant difference in fertilization ratio among the different sperm concentrations (104, 105, and 106 sperm/mL) for each hybrid (FLOint eggs × A. florida sperm: Tukey HSD P > 0.05, INTflo eggs × A. intermedia sperm: Tukey HSD P > 0.05).

Since high self-fertilization was observed for the eggs of FLOint in this and our previous study[21], we performed paternity testing to determine whether FLOint eggs can be fertilized by A. florida sperm in sperm choice experiments (fertilization trials in the presence of both parental species and hybrid sperm). The results showed that most eggs were self-fertilized in the presence of both A. florida sperm and FLOint sperm (Fig. 1). On the other hand, most INTflo hybrid eggs backcrossed with A. intermedia sperm in the presence of INTflo sperm (Fig. 1).

Figure 1

Backcrossing or self-fertilization of hybrid eggs. Sperm choice experiments using hybrid eggs are shown. INTflo and FLOint eggs showed different features: INTflo eggs backcrossed with the sperm of the mother species, whereas FLOint eggs self-fertilized as reported in Isomura et al.[21]. Light grey bars indicate the fertilization ratio of hybrid sperm (self-fertilization), and the dark grey bars indicate the sperm of the parental species. Microsatellites were checked using electrophoresis or Microsatellite Analysis v1.0 (Applied Biosystems) software (https://www.thermofisher.com/order/catalog/product/4381867).

High compatibility of F1 hybrid sperm and eggs of the parental species

INTflo and FLOint sperm both backcrossed with eggs of the maternal parent of each hybrid (Fig. S2). For A. florida eggs, the ratio of fertilization to conspecific sperm was much lower (Table S1) than in previous trials, such as that conducted in 2015[21], although the reason for this is unclear. For A. intermedia eggs, the fertilization ratio did not change with sperm concentration (A. intermedia eggs × INTflo sperm: Tukey HSD P > 0.05). As in A. intermedia, the fertilization ratio of A. florida eggs also did not significantly differ with sperm concentration (A. florida eggs × FLOint sperm: Tukey HSD P > 0.05).

Can F1 hybrid sperm compete with conspecific sperm to backcross with eggs of the maternal species?

To examine backcrossing and self-fertilization ratios in the presence of parental species and hybrid sperm (A. intermedia and INTflo sperm or A. florida and FLOint sperm, respectively), paternity tests were performed. Fertilization ratios in the sperm choice experiments were not significantly different among combinations (Fig. S3; INTflo eggs × INTflo sperm and A. intermedia sperm: Tukey HSD P > 0.05, A. intermedia eggs × INTflo sperm and A. intermedia sperm: Tukey HSD P > 0.05).

Eggs of the parental species (A. florida and A. intermedia) were fertilized by hybrid sperm in the presence of conspecific sperm. In case of INTflo sperm backcrossing, the hybrid sperm fertilized A. intermedia eggs independent of the sperm concentration of the parent species (Fig. 2, Tukey HSD P > 0.05) or hybrid sperm concentration (Tukey HSD P > 0.05). FLOint sperm also backcrossed with parent species A. florida eggs independent of hybrid sperm concentration (Fig. 2, A. florida egg × A. florida sperm and FLOint sperm; T-test, mt = − 0.1, df = 1.78, P > 0.05).

Figure 2

Backcrossing of hybrid sperm with parental species. Sperm choice experiments using eggs of the parental species are shown. Light grey bars indicate the fertilization ratio of parental species sperm, and white bars indicate the sperm of the hybrid species. Microsatellites were checked using electrophoresis or Microsatellite Analysis v1.0 (Applied Biosystems) software (https://www.thermofisher.com/order/catalog/product/4381867).

Discussion

In our study, the F1 hybrid FLOint showed adequate fertilization patterns for reproduction, with high rates of backcrossing to the parental species and most of the eggs showing selfing. Our results showed that hybridization may arise when colony numbers decline due to heavy bleaching events. For example, A. florida eggs could hybridize with A. intermedia sperm at low sperm concentrations (FLOint)[23]. Conversely, A. intermedia eggs were preferentially fertilized by conspecific sperm when exposed to both A. intermedia and A. florida sperm[23]. Since the recovery of reefs after heavy bleaching often takes more than 10 years, and reef species composition changes frequently[28,29], the post-bleaching F2 generation is predicted that they need to reproduce within lower number of colonies. In contrast to the Caribbean hybrid A. prolifera (< 25% selfing)[14], FLOint showed more than 95% selfing (Fig. 1). Six to fourteen eggs and 106 sperm are packed into Acropora gamete bundles[30,31]; thus sperm not involved in the self-fertilization of bundled eggs can mate with other unfertilized eggs. However, for backcrossing, spawning synchronicities of the hybrids FLOint and A. intermedia and INTflo and A. florida were not observed. Therefore, relationships between spawning synchronism and backcrossing with parental species still need to be clarified.

Sperm from the hybrids FLOint and INTflo potentially mate with parental species eggs when hybrid and parental species sperm compete. Sperm choice experiments in this study showed that the competencies of FLOint and INTflo sperm were high enough to outcompete parental sperm even when the number of hybrid sperm was far lower than that of the parental species sperm (Fig. 2). To support this result, repeated backcrossing is suggested in Caribbean Acropora[18]. In addition, there were fewer than two alleles of microsatellites, indicating polyploidy of the hybrids involving two chromosomes, unlike that reported in a previous study[32]. Therefore, fusion of the parental species is predicted to occur, but as the two parental species A. florida and A. intermedia are morphologically distinct, lineage fusion does not occur extensively at present.

From the present study, interspecific hybridization and introgression are suggested, but molecular based analyses for examining admixture events between the parental species are needed. From our preliminary SNP-based analyses, gene flow occurred among intercrossing species showing spawning synchronicity and high gamete compatibility (Kitanobo et al., unpublished data). However, the detection of hybridization is influenced by methodological differences[10], and hybrid lineages are rare (five species)[8]. Moreover, integrative approaches from breeding trials and morphological and molecular based analyses indicate that morphologically distinguishable species can be reproductively isolated and can evolve independently[10]. Contrary to a previous study[33], tabular species do not cross with other morphospecies[10], but A. florida and A. intermedia gametes show high rates of intercrossing. In addition, gametes showed specific fertilization patterns according to sperm concentration, and the patterns of the F1 hybrids also indicate that backcrossing is highly probable. However, it seems difficult to distinguish between hybridization and incomplete lineage sorting if an admixture event occurred in the early speciated Acropora (< 6 Ma)[34].

From our study, a slight admixture event between A. florida and A. intermedia may be ongoing, but detailed comprehensive studies involving other intercrossing species are needed. In this study, we focused on only two intercrossing species, A. florida and A. intermedia, but A. intermedia shows high rates of crossing with other sympatric and synchronous spawning species[35]. As discussed above, the morphologically distinct species of A. florida and A. intermedia show slight differences in spawning times and dates (Table 1); thus, these two species are tentatively at lower risk of hybridization. A. intermedia and other intercrossing species such as A. gemmifera are more likely to hybridize due to their overlapping spawning times[36].

Our study also shows that delimitation of species is suspected in the morphologically distinct intercrossing species A. florida and A. intermedia. Although our study used limited numbers of hybrids, the sperm competency of F1 hybrids was sufficient to provide opportunities for mating with parental species (backcrossing) and selfing in FLOint. This would be beneficial to the production of the F2 generation in cases of solitary spawning due to reduced spawning synchronicity or a reduced number of colonies of the parental species (Fig. 3). These features are congruent with high rates of introgression events under past climate changes[22,37], and hybrid hotspots are located at biogeographic borders including Southern Japan[8]. Although the unique fertilization patterns of F1 hybrids are potentially not involved in ongoing hybridization, they may be a footprint of the past hybridization of ancestral species, and these results can be used to understand the complex history of the coral Acropora.

Figure 3

Schema of F1 hybrid reproduction. Flow from A. intermedia and A. florida, F1 hybrids, to reproduction of F1 hybrids. According to our studies, hybridization can occur when the sperm concentration is low, and an F1 hybrid (FLOint) can backcross or reproduce asexually. Bold lines indicate probable pathways, which are supported by this and our previous studies; faint grey lines indicate pathways that may not occur because of the low probability of hybridization.

Supplementary Data.

Supplementary Figure 1.

Supplementary Figure 2.

Supplementary Figure 3.

Supplementary Table.