Karyotypic stasis and its implications for extensive hybridization events in corallivores species of butterflyfishes (Chaetodontidae)

Abstract

The butterflyfishes (Chaetodontidae), emblematic inhabitants of coral reef environments, encompass the majority of known coralivorous species and show one of the highest hybridization rates known among vertebrates, making them an important evolutionary model. The vast knowledge about their life history and phylogenetic relationships contrasts with scarce information on their karyotype evolution. Aiming to expand the cytogenetic data of butterflyfishes and evaluate their karyotype evolution in association with evolutionary aspects, we conducted an extensive cytogenetic analysis in 20 species (Heniochus pleurotaenia and 19 Chaetodon spp.) from the Atlantic and Indo-Pacific regions, comparing the karyotype macrostructure and the arrangement of the 18S and 5S rDNA repetitive DNA classes in their chromosomes. The results demonstrate that butterflyfishes underwent a period of karyotypic stasis, as evidenced by their homoploid and structurally identical basal karyotype, which has 2n = 48 acrocentric chromosomes and is shared by 90% of species. Only C. trifascialis (2n = 48; FN = 50) and C. andamanensis (2n = 48; FN = 52) stood out because they both had karyotypes that diverged due to pericentric inversions. The microstructural arrays of 18S rDNA and 5S rDNA sequences were primarily comprised by single and independent loci on homologous chromosomes, indicating that there was little reshuffling among sets of orthologue chromosomes of species. Geographical comparisons revealed similar karyotypes between individuals of C. striatus from the Greater Caribbean and those of the coast of Brazil, corroborating previous data of gene flow through Amazon/Orinoco plume. The conservative chromosomal patterns in the butterflyfishes, likely overcome the limitations related to segregation and pairing of heterospecific complements and reinforce their contribution to the high degree of hybrid viability and introgression in Chaetodon species.

1. Introduction

Integrative cytogenetic approaches in marine fish groups have indicated that specific biological aspects and historical contingency can act together as causal effects of chromosomal changes [[1], [2], [3]]. Marine fishes display contrasting karyotype evolution patterns, which can vary from high dynamic rates to stasis, a process in which chromosomal change rates are absent or slow, allowing chromosomal patterns to persist for long periods of time [4,5]. This scenario is observed in a variety of groups, including plants [[6], [7], [8]], amphibians [9], birds [10] and marine fishes [4,5,11], including some groups inhabiting coral reefs [12], ecosystems that represent diversity hotspots, with which fish have coexisted for more than 50 million years [13,14]. The strict associations and coevolutionary adaptations with corals produced remarkable effects on the diversity of some fish groups [[14], [15], [16]], however, their influence on their cytogenetic patterns need be more well understood.

Of the more than 5000 fish species associated with coral reefs, 41 species feed directly on them, 61% of which belong to the Chaetodontidae family [17,18]. Known as butterflyfishes, this family comprises 137 species [19], some with ecological relationships that are extremely dependent on corals [[20], [21], [22], [23]]. Their exuberant colors and body shapes [24] attract great commercial interest and contribute to their worrying conservation status [25,26]. Chaetodontidae are divided into two lineages based on their phylogenetic relationships: bannerfishes, and the butterflyfishes [[27], [28], [29]]. Belonging to the latter lineage, the genus Chaetodon Linnaeus 1758 that is the most diverse (88 spp.) [30,31] is subdivided into four main clades (Clades 1–4) according to the patterns of ancestry, distribution and feeding habits [14,27,32,33]. The ecological patterns, geographical distribution and the diversified habits of butterflyfishes have been related to ecomorphological, dietary, and behavioral adaptations [14,18,29,34]. Although the karyotype dataset of these fishes, which encompasses 15% of the species, offers a robust evolutionary database, it has not yet been associated with phylogenetic patterns or other biological aspects, such as a noteworthy hybridization rate, when compared to other reef fish groups [[35], [36], [37], [38]]. Preliminary cytogenetic findings from butterflyfishes suggest a conservative evolution, common to many groups of marine Percomorpha [4,12,39,40], and an evolutionary scenario of karyotypic stasis.

The slow karyotype changes in marine fishes, among other factors is associated with the maintenance of gene flow between large areas of species distribution [4,5,11]. Many butterflyfishes have a wide geographic range and can be found in the Indian and Pacific oceans at the same time or in large portions of the Atlantic Ocean, making them good models for the study of cytogenetic variation in multigeographic environments. Additionally, population cytogenetic comparisons in the Atlantic butterflyfishes can assess the divergent effect of biogeographic barriers, as the Amazon/Orinoco outflow, which has been associated with interpopulation karyotype variations [12,41] and diversification chromosomal in sibling-species between the Brazilian and Caribbean provinces [42].

In Percomorpha groups with essentially conserved karyotypes, the use of chromosomal markers based on repetitive DNAs can be decisive for the identification of karyotypic variations (e.g. Refs. [12,43]). Repeated sequences, such as rDNA, have been linked to mechanisms that promote chromosomal rearrangements [16,44,45], interpopulation variations [1,46], karyotypic diversification [12,47,48], or serve as markers to track karyotypic trends in broader biogeographical contexts [1,3].

Cytogenetic analyses of a wide range of butterflyfish species, considering their phylogenetic and biogeographical contexts, are necessary to clarify obscure aspects of karyotype evolution and its relationship with the intense hybridization processes occurred in this group. In this sense, here we investigated a set of 20 species (Heniochus pleurotaenia and 19 Chaetodon spp. and), geographically distributed in Indo-Pacific and Atlantic regions, regarding their karyotypic patterns and the chromosomal organization of 18S and 5S rDNA sequences, discussing them in a phylogenetic, ecological and distribution perspective. Based on the current findings, we obtained an upfront understanding of the rate of chromosome modifications, clade-specific diversification, and potential connections between biological features and karyotypic changes in the Chaetodon species.

2. Material and methods

2.1. Taxon sampling

We analyzed three males and three females of 20 coral reef butterflyfishes from the Atlantic and Indian oceans (Fig. 1, Table 1). Samples of Chaetodon sedentarius and C. capistratus were obtained in Key Largo, Florida, USA, C. ocellatus from the coast of the State of Rio Grande do Norte, northeast of Brazil and two populations of C. striatus, from the State of Rio Grande do Norte and the Caribbean (Key Largo, Florida). Samples collected in regions of Brazil were authorized by the Brazilian environmental agency ICMBio/SISBIO (License No. 19135-8). The species Heniochus pleurotaenia; C. xanthurus; C. trifasciatus; C. trifascialis; C. octofasciatus; C. andamanensis; C. mellanotus; C. ocellatus; C. collare; C. lunula; C. rafflesii; C. vagabundus; C. auriga; C. decussatus; C. wiebeli; C. lineolatus and C. falcatus were obtained from the Andaman Sea and Gulf of Thailand (Table 1). The handling of collected individuals, in the field, and in the laboratory, followed ethical protocols approved by the Animal Ethics Committee of the Federal University of Rio Grande do Norte, Brazil (Protocol 44/2015) and Khon Kaen University, Thailand (PHD/K0081/2556).

Fig. 1.

Partial world map, highlighting the collection regions of Chaetodontidae species at the Atlantic (A) and Indian (B) oceans. The numbers inside the circles correspond to the species in Table 1.

Table 1.

Species and collection site of the samples herein investigated. Numbers correspond to those in Fig. 1.

	Species	Region	Ocean
1 -	Chaetodon andamanensis	Andaman Sea	Indian
2 -	C. auriga	Andaman Sea	Indian
3 -	C. capistratus	Caribbean	Atlantic
4 -	C. collare	Andaman Sea	Indian
5 -	C. decussatus	Gulf of Thailand	Indian
6 -	C. falcatus	Andaman Sea	Indian
7 -	C. lineolatus	Gulf of Thailand	Indian
8 -	C. lunula	Andaman Sea	Indian
9 -	C. mellanotus	Gulf of Thailand	Indian
10 -	C. ocellatus	Northeast Brazil	Atlantic
11 -	C. octofasciatus	Andaman Sea	Indian
12 -	C. rafflesii	Andaman Sea	Indian
13 -	C. sedentarius	Great Caribbean	Atlantic
14 -	C. striatus	Northeast Brazil	Atlantic
14 -	C. striatus	Great Caribbean	Atlantic
15 -	C. trifascialis	Andaman Sea	Indian
16 -	C. trifasciatus	Gulf of Thailand	Indian
17 -	C. vagabundus	Gulf of Thailand	Indian
18 -	C. wiebeli	Gulf of Thailand	Indian
19 -	C. xanthurus	Andaman Sea	Indian
20 -	Heniochus pleurotaenia	Andaman Sea	Indian

2.2. Chromosomal preparations

The specimens were initially submitted to the mitotic stimulation technique through intramuscular injection of a solution of lysates of bacteria and fungi (7 mg Broncho-Vanxom®️/10 ml of distilled water), in the proportion of 1 ml for each 100 g of weight body [49]. After a period of 24 h, the individuals were sacrificed with an overdose of clove oil, and the mitotic chromosomes were obtained from a suspension of kidney cells, using the technique described by Gold et al. [50]. For chromosomal analyses, a volume of 120 μl of cell suspension was dropped onto a slide covered with a film of distilled water heated to 60 °C, which after drying was stained with 5% Giemsa (pH 6.8) for 10 min. The slides were analyzed under an optical microscope under 1,200X magnification. About twenty metaphases of individuals were analyzed, and the best metaphases were used in the karyotype's preparation. Chromosomes were classified as acrocentric (a), metacentric (m), and submetacentric (sm), according to their arm ratios and the classification suggested for fish karyotypes [35,51].

2.3. Fluorescence in situ hybridization (FISH)

Fluorescence in situ hybridization (FISH) was performed according to the methodology described by Yano et al. [52]. As probes, two tandemly-arranged DNA sequences (named 18S and 5S rDNAs) PCR-obtained from the wolfish Hoplias malabaricus genome were used. The sequences were previously cloned into plasmid vectors and propagated in competent cells of Escherichia coli DH5α (Invitrogen, San Diego, CA, USA), being the first a 1400bp sequence of the 18S rRNA gene and the second a 5S rDNA segment of 120 bp in addition to 200 bp of the non-transcribed spacer [53]. The 18S and 5S rDNA probes were labeled by nick translation with digoxigenin-11-dUTP (Roche) and biotin-14-dATP (Roche), respectively, according to the manufacturer's instructions. Chromosomes were treated with DNAse-free RNAse (20 mg/ml in 2 × SSC) at 37 °C for 1 h, with pepsin (0.005% in 10 mM HCl) at 37 °C for 10 min and fixed with 1% formaldehyde for 10 min and then dehydrated in an alcoholic series at 70%, 85% and 100% for 5 min. Chromosomes were then denatured in 70% formamide/2 × SSC at 72 °C for 5 min. The hybridization solution consisted of 50% formamide, 2 × SSC, 10% dextran sulfate, and the denatured probe (5 ng/μl). After an overnight hybridization at 37 °C, slides were washed in 15% formamide/0.2 × SSC at 42 °C for 20 min, 0.1 × SSC at 60 °C for 15 min, and 0.5% Tween20/4 × SSC at 25 °C. Probe hybridization signals were detected with streptavidin-FITC conjugate (Vector Laboratories, Inc., Burlingame, CA, USA) for the 18S rDNA probe and anti-digoxigenin rhodamine conjugate (Roche) for the 5S rDNA probe. Chromosomes were counterstained with antifading Vectashield/DAPI (1.5 μg/ml) (Vector Laboratories, Inc., Burlingame, USA).

Analyses under fluorescence were performed using an Olympus BX51 epifluorescence photomicroscope, coupled to an Olympus DP73 digital image capture system, using the cellSens software (Olympus Optical Co. Ltd.). The chromosomes were classified according to the ratio between the chromosome arms, in metacentric (m), submetacentric (sm), subtelocentric (st) and acrocentric (a) [51]), following the classification suggested for fish karyotypes [35].

Characteristically, the organization of butterflyfish karyotypes shows very small size differences between chromosome pairs (e.g. Ref. [54]). As a result, mistakes in the identification of chromosomes carrying the rDNA sites might result from size polymorphisms of nucleolar organizer regions (NORs [47]). Therefore, the chromosomes bearing the 18S and 5S rDNA sites were classified according to size, as big (pairs 1–8), medium (pairs 9–18), or small chromosomes (pairs 19–24) (Fig. 5), to establish more accurate phylogenetic comparisons across the species.

Fig. 5.

Karyotypes of species C. rafflesii, C. vagabundus, C. wiebeli, C. auriga, C. lineolatus, and C. decussatus (Clade 4), with conventional staining and in situ hybridization with 18S rDNA (green) and 5S rDNA (red) probes. Scale bar = 5 μm.

3. Results

All butterflyfishes here analyzed have 2n = 48 chromosomes and symmetrical karyotypes, with a small sequential reduction in size from largest to smallest chromosome pairs (Fig. 2, Fig. 3, Fig. 4, Fig. 5). Among the species, 17 Chaetodon species and the bannerfish H. pleurotaenia have karyotypes formed exclusively by acrocentric chromosomes (FN = 48), while C. trifascialis (2 m + 46a; FN = 50) and C. andamanensis (2 m+2st+44a; FN = 52) showed karyotypes with one (m) and two pairs of two-armed chromosomes (m, sm), respectively.

Fig. 2.

Karyotypes of Heniochus pleurotaenia and Chaetodon falcula under conventional staining and in situ hybridization with 18S rDNA (green) and 5S rDNA (red) probes. Scale bar = 5 μm.

Fig. 3.

Karyotypes of C. xanthurus, C. sedentarius (Clade 2); C. trifasciatus; C. octofasciatus; C. andamanensis, and C. trifascialis species (Clade 3), with conventional staining and in situ hybridization with 18S rDNA (green) and 5S rDNA (red) probes. Scale bar = 5 μm.

Fig. 4.

Karyotypes of C. melannotus, C. ocellatus; C. striatus (Great Caribbean and NE coast of Brazil); C. capistratus; C. lunula; and C. collare species (Clade 4), with conventional staining and in situ hybridization with 18S rDNA (green) and 5S rDNA (red) probes. Scale bar = 5 μm.

Mapping of 18S and 5S rDNA sequences in butterflyfish chromosomes showed preferentially unique and independent loci. Accurate identification of chromosomes bearing rDNA sites in Chaetodon is hampered by the small sequential variation in size between chromosomes. In this sense, the categorization by size classes (small, medium, and large) of pairs with 18S and 5S rDNA sites, indicates regular and comparable patterns of phylogenetic conservatism (Fig. 6). The most common pattern of organization of 18S rDNA sites consisted of a single locus on the short arm of a small acrocentric chromosome pair, here tentatively identified as pair 22 (13 spp.), or on the short arm of a medium-sized pair (5 spp.). In C. andamanensis and C. trifascialis, these sites were located respectively in pericentromeric and terminal positions in the short arms of the only metacentric chromosome pair of the karyotype (Fig. 3).

Fig. 6.

a. Karyotype patterns of butterflyfish species from temporal and phylogenetic perspectives (Chaetodon clades redrawn from Fessler and Westneat [27]). Color coding on circle indicates the social solitary behavior [55]. b. Chord diagram of 18S and 5S rDNA -chromosome bearing patterns and number of chromosome arms (fundamental number, FN) among Chaetodontidae clades.

The 5S rDNA sites are generally situated in a pericentromeric region on the medium-sized chromosomal pair, here tentatively designated as pair 17, and appear homologous among the species. C. wiebeli exhibited an extra 5S rDNA site, in the terminal portion of the long arms of pair 5 (Fig. 5). The bannerfish H. pleurotaenia, phylogenetically more basal, and Chaetodon species from the clades analyzed, showed the most frequent pattern of pairs bearing ribosomal sites, in which the 18S rDNA sites occur on the short arms of a small a pair and the 5S rDNA sites in the interstitial position of the long arms of a pair of medium-sized chromosomes (Fig. 2).

The presence of two-armed chromosomes (C. trifascialis, FN = 50; and C. andamanensis, FN = 52) carrying rDNA sites conferred on species from clade 3 the highest amount of karyotypic diversity among the examined phylogenetic clades (i.e., clades 2, 3, and 4) (Fig. 3, Fig. 6a). The 18S rDNA locus was located in the pericentromeric position in the long arms of the metacentric pair in C. andamanensis, whereas they were located in the terminal position of the short arms of the m pair in C. trifascialis (Fig. 3).

Comparing the karyotypes of C. striatus individuals from the Greater Caribbean and the northeastern coast of Brazil (Western Atlantic Ocean) revealed that these individuals shared similar FN and patterns of rDNA sequence organization. The 18S and 5S rDNA sites were located in the pericentromeric position of a small-sized acrocentric pair (pair 22) and in a medium-sized pair (pair 17) in both geographic regions (Fig. 4).

4. Discussion

4.1. Evolutionary karyotype changes in butterflyfishes

The karyotype evolution in Chaetodontidae is markedly brachytelic, with similar cytogenetic patterns being retained and shared from the bannerfish H. pleurotaenia to most Chaetodon species, although these clades have diverged around 32.8 million years ago [14]. This conservative pattern is notable for the extensive phylogenetic sharing of karyotypes with 2n = 48 acrocentric chromosomes, and with rDNA sites organized in simple and independent loci (Fig. 6b), a set of symplesiomorphic traits of Percomorpha groups [12,40]. Indeed, the extensive karyotype conservatism involving Chaetodontidae clades with large divergence periods characterizes a process of karyotypic stasis, whose causes in marine fishes have been attributed to the lower number of geographic barriers, high dispersive potential (larval phase or active migration of adult individuals), large population numbers, and extensive ecological continuity [4,5,11,56].

Given the extensive life history data available for Chaetodon species, it is possible to associate the evolution of karyotypes with their phylogenetic, biological, and biogeographical features. In this sense, although there is a lack of cytogenetic data for clade 1, restricted to the Atlantic (three species), the species of clade 2 (>35 species; 2 species herein analyzed), with food plasticity and composed by corallivores and non-corallivores lineages, together with the species of clade 4 (>30 species; 14 species analyzed), mostly formed by non-coralivorous species [14], both with Indo-Pacific distributions, share basal karyotypic characteristics for the family.

The species of clade 3 (>20 species; 6 here investigated), restricted to the Indo-Pacific, are cytogenetically more diverse, ranging from species with conserved karyotypes (FN = 48) to species with derived karyotypes, by the presence of one or two two-armed chromosome pairs (C. andamanensis, FN = 52; C. trifascialis, FN = 50; and C. plebeius, FN = 50 [57], Table 2). The causes of the breakdown of karyotypic stasis in these species, although cannot be precisely established, may be associated with biological and behavioral features of this clade. In contrast with the others, about 90% of species of clade 3 are obligate corallivores, which promotes strict dependency and territoriality in coral reefs [14]. Furthermore, multiple species exhibit solitary behavior, including the ones with derived karyotypes (C. andamanensis, C. trifascialis, and C. plebeius), in contrast to the pair bonding social behavior observed in butterflyfishes [55,58]. Under specific historical contingencies, these features may facilitate genetic structure and the fixation of chromosomal rearrangements. Indeed, phylogeographic analyses in C. trifascialis revealed a historical demography marked by old population reductions and geographic isolation, characterizing genetic bottlenecks [59], indicating a distinct historical pathway culminating in the breakdown of karyotypic stasis in this species.

Table 2.

Cytogenetic data of Chaetodontidae species grouped into their respective clades (sensu Bellwood et al. [14]; Cowman and Bellwood [33]).

Clade/Species	2n	FN	Karyotype	Ag-NOR	5S/18S rDNA- bearing pair	Geographic Region	References
Bannerfishes
Heniochus acuminatus	48	48	48a	–	–	Japan	[60]
H. pleurotaenia	48	48	48a	–	(S/M)	Thailand	present study
Butterflyfishes
Clade 2
C. sedentarius	48	48	48a	–	–	Brazil/Atlantic	[61]
	48	48	48a	–	(S/M)	Brazil/Atlantic	present study
	48	48	48a	–	(S/M)	EUA/Great Caribbean	present study
C. xanthurus	48	48	48a	–	(S/M)	Thailand	present study
Clade 3
C. trifasciatus	48	48	48a	–	–	Japan/Pacific	[57]
	48	48	48a	–	(M/M)	Thailand	present study
C. triangulum	48	48	48a	2	–	Thailand	[62]
	48	48	48a	2	–	Thailand	[63]
C. trifascialis	48	50	2 m + 46a	–	(M/L)	Thailand	present study
	48	50	2sm + 46a	–	–	Japan/Pacific	[57]
C. octofasciatus	48	48	48a	–	(M/M)	Thailand	present study
C. plebeius	48	50	2 m + 46a	–	–	Japan	[57]
C. andamanensis	48	52	2 m+2st+44a	2	–	Thailand	[64]
	48	52	2 m+2st+44a	–	(L-syntenic)	Thailand	present study
Clade 4
C. mellanotus	48	48	48a	–	(M/M)	Thailand	present study
C. ocellatus	48	48	48a	2	–	Brazil/Atlantic	[65];
	48	48	48a	–	(S/M)	Brazil/Atlantic	present study
C. striatus	48	48	48a	2	–	Brazil/Atlantic	[65,66]
	48	48	48a	–	(S/M)	Brazil/Atlantic	present study
	48	48	48a	–	(S/M)	EUA/Great Caribbean	present study
C. capistratus	48	48	48a	–	(S/M)	EUA/Great Caribbean	present study
C. collare	48	48	48a	–	–	India	[67]
	48	48	48a	2	–	Thailand	[68]
	48	48	48a	–	(S/M)	Thailand	present study
C. lunula	48	48	48a	–	–	Japan/Pacific	[57]
	48	48	48a	2	–	Thailand	[63]
	48	48	48a	–	(S/M)	Thailand	present study
C. rafflesii	48	48	48a	–	(S/M)	Thailand	present study
C. auripes	48	48	48a	–	–	Japan/Pacific	[57,69]
C. vagabundus	48	48	48a	–	–	Japan/Pacific	[57]
	48	48	48a	2	–	Thailand	[70]
	48	48	48a	–	(M/M)	Thailand	present study
C. auriga	48	48	48a	–	–	Japan/Pacific	[57]
	48	48	48a	–	(S/M)	Thailand	present study
C. decussatus	48	48	48a	2	–	Thailand	[54]
	48	48	48a	–	(M/M)	Thailand	present study
C. wiebeli	48	48	48a	2	–	Thailand	[71]
	48	48	48a	–	(S/L-M)	Thailand	present study
C. lineolatus	48	48	48a	2	–	Thailand	[54,63]
	48	48	48a	–	(S/M)	Thailand	present study
C. falcula	48	48	48a	–	(S/M)	Thailand	present study

L = large chromosome; M = medium chromosome; S = small chromosome. References: [60] Arai and Yamamoto, 1981 [61]; Galetti et al. (2006) [57]; Arai and Inoue (1975) [62]; Supiwong et al. (2017a) [64]; Supiwong et al. (2017b) [63]; Na Nongkhai (2014) [65]; Molina et al. (2013) [66]; Affonso et al. (2001) [67]; Nagpure et al. (2006) [68]; Boonsuk (2013) [69]; Ojima and Yamamoto (1990) [70]; Jumrusthanasan and Supiwong (2015) [54]; Supiwong et al. (2015) [71]; Sonsrin and Pinthong, 2018.

Since repetitive DNA constitutes almost 20% of the genome of some Chaetodon species [72], their intrinsic evolutionary high dynamism can be useful to tracking minor changes in their karyotypes. The ribosomal genes, which are included in this diverse genomic fraction, have a strong evolutionary dynamic that is mostly related to duplication events and the temporal accumulation of mutations. Walsh [73], Richard et al. [74], Lopez-Flores and Garrido-Ramos [75], and whose evolutionary tracking have been useful in evolutionary inferences in fishes [76,77]. Despite the high evolutionary dynamism of rDNA in fish karyotypes [47,78], Chaetodontidae species showed a conservative condition with independent arrangements of 18S and 5S rDNA regions in apparently homologous chromosomes. This rDNA arrangements are presents in the bannerfish H. pleurotaenia, as well as in most of the Chaetodon species, from clades 2, 3 and 4, inferring a smaller number of gross changes in the butterflyfish chromosomes. Evolutionarily, this organization of rDNA regions may have older origins, given its occurrence in Pomacanthidae [79], sister group of Chaetodontidae [27] and as a basal condition in Percomorpha (e.g. Refs. [67,56]).

In view of stable set of acrocentric chromosomes bearing rDNA sites, the occurrence of exclusive arrangements of the 18S rDNA and 5S rDNA sites in the two-armed chromosomes of C. andamanensis (pericentromeric 18S rDNA site) and C. trifascialis (18S rDNA/5S rDNA in syntenic array) (Fig. 3), are strong evidence of the participation of 18S rDNA sites as hotspots to pericentric inversions. Although functional 5S rRNA gene copies have already been detected inserted in the intergenic spacers of 18S rDNA in Perciformes lineages [80], as well as in turtles and crocodiles [81], the syntenic arrangement of both rDNA classes is not a typical trait in vertebrate karyotypes [82]. Similarly, chromosomal instabilities involving chromosome pairs carrying rDNA sites have been identified in several other groups of Percomorpha [[83], [84], [85]]. On the other hand, the 5S rDNA sites were conserved in pairs of chromosomes of similar sizes, possibly homeologs, except by the syntenic arrangement with 18S rDNA in C. trifascialis (Fig. 3), and by the occurrence of an additional locus, in C. wiebeli (Fig. 5).

The genomic similarities between Chaetodon species [86], associated with the structural and numerical conservatism of the chromosomes, support a less restrictive role of the karyotype as a postzygotic barrier in species of this genus [65]. However, to determine the extent of homeologous conservation in butterflyfish chromosomes, additional cytogenetic data involving a broader range of chromosomal markers is required.

4.2. The impact of chromosomes in homoploid hybridization in butterflyfishes

Interspecific hybridization events have been reported in representative reef fish groups, such as Labridae and Pomacentridae [36,38,87]. However, this process is more accentuated in Chaetodontidae [[88], [89], [90]], mainly in the Indo-Pacific region, which is home to the greatest diversity and overlapping of species [91,92]. In fact, about half of the butterflyfishes recognize others as reproductive partners and produce viable hybrids [87,90], evidencing very labile pre- and post-zygotic reproductive isolation barriers.

Aspects that promote the breaking of prezygotic barriers in butterflyfishes are well known [87] such as recent divergence; secondary contact of allopatric species; overlapping distributions; diet sharing; rarity of one or both parental species; and random mating behavior [55,93,94]. In contrast, the causes of the low effectiveness of postzygotic barriers in this group and the possible role of cytogenetic patterns in this process are unclear.

Following the breakdown of prezygotic reproductive isolation, meiotic irregularities, or genetic incompatibility, which are unique of each lineage's genetic divergence, may end up resulting in a lower adaptability of the hybrid in comparison to the parents (extrinsic post-zygotic barrier), or its infertility or unfeasibility (intrinsic post-zygotic barrier) [95,96]. Thus, periods that span tens to hundreds of years may be enough to promote genetic incompatibilities [97]; however, intrinsic postzygotic isolations in fishes may require mean divergence times of ∼11.6 Myr [95] up to 28 Myr in centrarchid fishes [98]. Postzygotic effects, according to the Dobzhansky-Muller model [99], come from the accumulation of pleiotropic products obtained from the independent evolution of allopatric populations. Thus, post-zygotic reproductive blocks may be ineffective in cases of recent or incipient evolutionary diversification, as shown in a large number of butterflyfish species [91,100,101], with divergences occurring within 5 Mya or less [14, 27].

Although rare (∼5%), postzygotic blocks may remain ineffective, even among butterflyfishes from different clades [[102], [103], [104]], whose divergences are estimated at around by 14 Mya [27]. Thus, it cannot be ruled out that this long period to acquisition of intrinsic reproductive isolation among butterflyfishes be influenced by the remarkable karyotypic similarities and maintenance of chromosome synteny, which minimize pairing failures and chromosomal segregation [105]. Likewise, high levels of hybridization also occur in Pomacanthidae [36,106], its sister group, whose karyotypic conservatism is also well recognized [12,80].

In a broader view, some cytogenetic characteristics of Chaetodon species can increase the functionality of viable and/or fertile hybrid genomes. The first refers to the extensive diploid stability (2n = 48), which favors chromosome segregation in homoploid hybrids [105]. Secondly, the high macrostructural homogeneity of the karyotypes, formed entirely by acrocentric chromosomes and by the similarities (or even homeology) of the pairs carrying the rDNA sites (18S and 5S rDNA). Indirect evidence from whole chromosome-genome comparisons between Chaetodontidae and Pomacanthidae, two closely related evolutionary families [14], indicates a high level of synteny of their linkage groups [107]. If this is correct, it may be deduced that substantially greater levels of synteny are still preserved among Chaetodon species, despite fast and recent diversification [27]. The dearth of reports on hybridization involving C. trifascialis, C. plebeius, or C. andamanensis, all of which have chromosomes with syntenic configurations altered by pericentric inversions, raises the opportunity of future cytogenomics studies. Last but not least important, homomorphic karyotypes (i.e., absence of differentiated sex chromosomes), present in 95% of the analyzed teleost species [108], constitute an additional cytogenetic condition of butterflyfishes, which favor the occurrence of fertile or viable hybrids. So, the homomorphic complements of hybrids do not follow Haldane's rule (HR) [109], which establishes sterility, rarity, or absence of the heterogametic sex in hybrid offspring [110]. In this sense, the evolutionary retention of karyotypic patterns homoploid, homomorphic and with a high degree of synteny, suggests less unfavorable conditions for the functionality of the chromosomes in the genomes of hybrid butterflyfishes.

4.3. Interpopulation karyotypic similarities in butterflyfishes

The broad geographic distributions of butterflyfish species (e.g. Ref. [111]), sometimes permeated by vast oceanic spaces or biogeographical barriers, provide suitable models for population cytogenetic comparisons in the marine environment. In the Atlantic, the banded butterflyfish C. striatus is distributed from the southeastern USA (Massachusetts) to southern Brazil, with the intermediate position of the Amazon/Orinoco rivers outflow, whose effect on gene flow has produced cytogenetic variations between populations [1,41] or in geminate species [112]. Cytogenetic comparison between individuals from the northeast coast of Brazil and the Great Caribbean, which comprises a large part of its distribution, did not reveal detectable karyotype variations. Similar results occur among individuals of the reef butterflyfish C. sedentarius from the Great Caribbean, analyzed here, and those from the coast of southeastern Brazil [61]. Phylogeographic analyses in populations of C. striatus distributed from the Caribbean to the Western South Atlantic showed high interpopulation connectivity [111], which supports the karyotypic homogeneity of this species between these provinces. Although the population structure of C. sedentarius is not yet known, the data suggest that the long period of larval duration (PLD) of this species, associated with larval swimming ability [113,114] and generalist feeding habits, can act together to maintain gene flow through the Amazon/Orinoco outflow. Finally, comparisons among karyotypes of C. lunula, C. auriga and C. vagabundus from the Pacific and Indian Oceans showed similar patterns (Table 2). However, in butterflyfishes biogeographical barriers may have different effects on closely related species, despite ecological, morphological, life-history similarities, and extensive spatial overlap [115]. Thus, a larger set of cytogenetic information at different points in the distribution of butterflyfish species may clarify intra and interspecific karyotypic stability in the genus Chaetodon.

5. Conclusions

Butterflyfishes have half of their species involved in hybridization events that produce fertile hybrids, demonstrating weak mechanisms of pre- and post-zygotic reproductive isolation. Although the causes related to the low effectiveness of pre-zygotic barriers are widely understood, aspects of the fragility of post-zygotic blocks deserve additional investigation. The combined analysis of butterflyfish cytogenetic patterns, considering macro and microstructural aspects of the chromosomes, reveals very low chromosomal dynamics alongside significant karyotype conservatism. The occurrence of karyotypic stasis in marine Percomorpha has been linked to large populations, high vagility, widespread geographic distribution, and the lack of biogeographical barriers [4,5,11,12]. Indications of this process in the different Chaetodontidae clades are supported by the sharing of the same diploid value (100% spp.), karyotypes composed only of acrocentric chromosomes (88% spp.), symmetrical karyotypes (100% spp.), 18S rDNA sites and 5S rDNA organized into independent (95% spp.) and unique loci (18S rDNA, 100% spp.; 5S rDNA, 95% spp.). Only obligatory coral species with solitary habits have karyotypes with chromosomal changes derived from pericentric inversions, indicating a possible link between biological patterns and karyotype divergence. The two-armed chromosomes bearing 18S rRNA sites in these species have unique arrangements, suggesting that such regions are involved in chromosomal modifications in butterflyfishes. The genus Chaetodon provides a great opportunity to understand the role of karyotypic patterns in the hybridization process due to its huge number of species capable of generating viable hybrids. Although the causes of the low effectiveness of post-zygotic effects remain unknown, evidence of accentuated karyotypic similarity in this genus, highlighted by homoploidy, small variation in repetitive marker DNA (rDNA) sequences, and the absence of sex chromosomes, point to less limiting conditions of chromosomal pairing and segregation favoring hybrid viability and fertility.

6. Funding information

This research was financially supported by the Fundamental Fund of Khon Kaen University, fiscal year 2023, the National Science, Research, and Innovation Fund (NSRF) Thailand, the Thailand science research and innovation fund, the University of Phayao (Grant no. FF66-UoE013), and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) Grant numbers 442626/2019-3 and 301458/2019-7.

Ethics statement

The handling of collected individuals, in the field, and in the laboratory, followed ethical protocols approved by the Animal Ethics Committee of the Federal University of Rio Grande do Norte, Brazil (Protocol 44/2015) and Khon Kaen University, Thailand (PHD/K0081/2556).

Data availability statement

The original datasets presented in the study are included in the article, and further inquiries can be directed to the corresponding authors.

CRediT authorship contribution statement

Wagner Franco Molina: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Sudarat Khensuwan: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Conceptualization. Renata Luiza Rosa de Moraes: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Francisco de Menezes Cavalcante Sassi: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Formal analysis. Gideão Wagner Werneck Félix da Costa: Writing – original draft, Visualization, Methodology, Formal analysis. Davi Zalder Miguel: Writing – original draft, Methodology, Data curation. Weerayuth Supiwong: Supervision, Resources, Methodology, Data curation. Sitthisak Jantarat: Validation, Methodology, Formal analysis, Data curation. Krit Phintong: Methodology, Investigation, Formal analysis, Data curation. Kriengkrai Seetapan: Methodology, Investigation, Formal analysis, Data curation. Sukhonthip Ditcharoen: Methodology, Investigation, Data curation. Alongklod Tanomtong: Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Thomas Liehr: Writing – review & editing, Writing – original draft, Validation, Software, Resources, Methodology, Formal analysis, Data curation. Marcelo de Bello Cioffi: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Methodology, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.