Introduction

Stingless bees are important pollinators of native plants in tropical and subtropical regions of the world (Heard 1999). The number of stingless bees found in the Neotropics is high; 418 species have been described to date (Camargo and Pedro 2013; Pedro 2014). These species are grouped into 34 genera (including one that is extinct) (Camargo and Pedro 2013; Melo 2016). There are also 89 valid species grouped into 15 genera in the Indo-Malayan/Australasian region (Rasmussen 2008) and 19 species belonging to six genera in Africa (Eardley 2004). In Brazil, 244 species are found, distributed among 29 genera with valid names (Camargo and Pedro 2013; Pedro 2014).

Cytogenetic analysis can contribute greatly to the understanding of the types of chromosome changes that might have occurred during the differentiation of these species and how chromosome evolution occurred in the different taxa of this group. Cytogenetics also helps to solve questions related to systematics and the existence of morphologically cryptic species and species complexes in several groups of organisms (Crossa et al. 2002; Van Daele et al. 2004; Ravaoarimanama et al. 2004; Vicari et al. 2005; Rossi et al. 2005; Tinni et al. 2007; Maurício et al. 2012).

The first cytogenetic study on Meliponini was published by Kerr (1948), and since then, more than 100 species have been karyotyped. The first analyses were restricted to the description of chromosome numbers, but with the emergence of new techniques, it was possible to describe and compare the patterns of heterochromatin constitution (Rocha et al. 2002), distribution and localization of genes or regions (e.g., 18S rRNA, 5S rRNA) (Fernandes et al. 2011; Martins et al. 2013; Lopes et al. 2014) and whole chromosomes (Martins et al. 2013).

Some studies have pointed to some possible mechanisms generating karyotype changes within Meliponini. Kerr and Silveira (1972) proposed that polyploidy would be the main mechanism responsible for the karyotypic evolution of Meliponini, followed by subsequent changes in the chromosome number, such as Robertsonian rearrangements. In contrast, data published by Pompolo (1992), Rocha et al. (2003a, b) and Brito et al. (2005) suggested that the karyotype evolution of Meliponini follows the same cycle of changes proposed by the Minimum Interaction Theory (Imai et al. 1988; Imai 1991; Hoshiba and Imai 1993). According to this theory, karyotype evolution involved an increase in chromosome number by centric fission, resulting in a reduction of chromosome size and an increase of heterochromatic regions to stabilize the telomeres. This process would reduce the occurrence of deleterious chromosomal translocations (for details, see Imai 1991; Imai et al. 2001). In some species of the genus Melipona, however, there seems to have been an increase in heterochromatin content without fission having occurred (Rocha and Pompolo 1998; Rocha et al. 2007). This shows that despite the Minimum Interaction Theory being widely used to explain karyotype evolution in Meliponini, different mechanisms are probably involved in this process, which requires a more detailed analysis of the information available.

The last review on cytogenetic data in Meliponini was performed by Rocha et al. (2003a), and this has been used as reference in several works. However, since its publication, several species have been analysed cytogenetically, and studies on phylogenetic relationships in Meliponini in general (Rasmussen and Cameron 2010) and some genera in particular (Ramirez et al. 2010; Rasmussen and Cameron 2007) have been published, allowing a new evaluation and the development of hypotheses about the evolution of the karyotype of Meliponini. Thus, we see the need to re-evaluate the available cytogenetic data, focusing mainly on the chromosome number variation and its distribution in this tribe.

Materials and methods

Information on the chromosome number of different species of stingless bees presented in this review was obtained by consulting the papers published in specialized journals, theses and/or dissertations and abstracts presented at scientific events (Table 1). Thus, all described karyotypes, even those presenting different chromosome numbers were included. It is noteworthy, however, that when the same data were presented at any event and subsequently published in specialized magazines, sometimes we chose to include only the published work, especially when the summary did not carry information about the geographical location of the samples analysed.

Table 1 Overview of the karyotypic data described for Meliponini species
Full size table

The nomenclature used for the species was based on the proposals of Camargo and Pedro (2013) for the Brazilian species, Rasmussen (2008) for the Indo-Malayan/Australasian species and Eardley (2004) for the African species. Thus, we present a synonymy table (Table S1) to allow comparisons with the original work and make information more clear and accurate.

It is noteworthy that many of the data described in Table 1 were obtained from studies conducted in the 1950s, 1960s and 1970s, using the squashing cytogenetic technique available at the time. This may explain the discrepancies observed when comparing such results with the most current data, when some of these species were reanalysed using more refined techniques for the preparation of slides (Rocha et al. 2003a; Oliveira et al. 2013). In such cases, in the present study discussion (when there was no evidence that the differences were due to the study of different populations, the occurrence of B chromosomes or chromosome polymorphisms), we considered the chromosome number obtained in the most recent work due to better definition of the number and morphology of chromosomes provided by the techniques used. Additionally, when the published work only provided the species haploid or diploid number, double or half were assumed, respectively, as the complementary values, in the discussion.

We also stress that many times, in older works, the authors only made reference to previous analyses, without reanalysing the species and therefore, only the original work was considered. In addition, species cited as sp., i.e., without species-specific identification, were not included in Table 1, except for one Trigonisca species, which was the only species of this genus analysed cytogenetically until now. When a species with a doubtful identification presented a different chromosome number of other species of its genus (as in the case of one Trigona with 32 chromosomes), it was only reported in the text.

On the other hand, it must be considered that there are specific names that are possibly being used for a group of sibling species, such as Frieseomelitta varia, Scaptotrigona postica and Tetragonisca angustula (Camargo and Pedro 2013). This can also occur with other widely distributed groups, such as Plebeia droryana.

Results and discussion

Karyotype diversity in Meliponini

The compiled data show that only 104 species of stingless bees (19.77% of the species described to date), representing 32 of the 54 living genera (59.26%) have been studied cytogenetically. Despite this restriction, a detailed analysis of Table 1 shows that the haploid chromosome number of Meliponini ranges from n = 8 to n = 20, and it is possible to recognize three main groups, with n = 9, 15 and 17.

The first group comprises the species of the genus Melipona. Among the 74 described species (Camargo and Pedro 2013), 23 were analysed cytogenetically. Of these, 22 species have the chromosome number of n = 9 (2n = 18), while the two subspecies of Melipona seminigra (M. seminigra merrillae and M. seminigra pernigra) have karyotypes consisting of 22 chromosomes in the normal complement (Francini et al. 2011; Silva et al. 2014). Other numerical chromosomal variations were found in Melipona rufiventris and Melipona quinquefasciata, but in these cases, the changes were due to the presence of B chromosomes (see below). It is also noteworthy that the chromosome number of n = 9 was not found in any of the other genera analysed to date.

Karyotypes with n = 15, in turn, have been detected in 11 species belonging to the genera Celetrigona (01 species), Duckeola (01), Frieseomelitta (06), Geotrigona (one of the two species examined), Leurotrigona (one of the two species analysed) and Trigonisca (01). Considering these genera, Celetrigona, Trigonisca and Leurotrigona are species that belong to the same clade (Trigonisca) in the phylogeny proposed by Rasmussen and Cameron (2010). Duckeola and Frieseomelitta are also close to each other but phylogenetically distant from the Trigonisca clade (Rasmussen and Cameron 2010). Geotrigona, in turn, has no close phylogenetic relationships with any of the genera referred to above, which suggests that the chromosome number of n = 15 appeared several times, independently in the evolution of Meliponini. Worth noting are the case of two species of Leurotrigona, Leurotrigona muelleri and Leurotrigona pusilla, with n = 8 and 15 chromosomes, respectively, and that the two species of Geotrigona already karyotyped also have different chromosome numbers, being n = 15 (Geotrigona mombuca) and n = 17 (Geotrigona subterranea).

Moreover, karyotypes with n = 17 (2n = 34) have already been verified in species of Camargoia, Cephalotrigona, Dactylurina, Friesella, Geotrigona, Meliponula (one of the three species analysed), Mourela, Nannotrigona, Oxytrigona, Paratrigona, Partamona, Plebeia, Scaptotrigona, Scaura, Schwarziana, Tetragona, Tetragonisca and Trigona. In all, 54 species have this chromosome number which, as already observed by Rocha et al. (2003a), is the most common in the Meliponini studied to date.

Regarding chromosome number, three other genera, not phylogenetically related, should be highlighted, Lestrimelitta (Neotropical), Hypotrigona (Afrotropical) and Tetragonula (Indo-Malayan/Australasian). Whereas in Lestrimelitta, data indicate a haploid number of 14 chromosomes, the analyses conducted by Kerr (1969, 1972) and Kerr and Silveira (1972) in Hypotrigona are unclear, indicating a range between 13 and 15 chromosomes. So, if n = 14 is the chromosome number of Hypotrigona, this feature also appeared more than once in the evolution of Meliponini. The genus Tetragonula, in turn, deserves to be mentioned because it presents n = 20 chromosomes, the largest haploid chromosome number ever recorded between Meliponini.

Cytogenetic information on Meliponini (Table 1) also shows that, with few exceptions (previously mentioned), chromosome number is constant within a genus. For Partamona, for example, the 11 already karyotyped species have 2n = 34 chromosomes and the same happens for 23 of the 24 Melipona species (as discussed above). Such a numerical constant has also been observed in Frieseomelitta, Plebeia, Scaptotrigona, Scaura, Tetragona and Trigona, all with at least three species analysed cytogenetically.

In this respect, numerical variations caused by the presence of B chromosomes in Meliponini has attracted attention (Costa et al. 1992; Brito et al. 1997; Brito 1998; Martins et al. 2009, 2014; Barth et al. 2011; Silva et al. 2013a, b) and for the study of these chromosomes, both classical and molecular cytogenetic techniques have been used (Costa et al. 1992; Tosta et al. 2004, 2007; Marthe et al. 2007; Martins et al. 2009, 2013, 2014; Barth et al. 2011; Correa et al. 2014).

The first report about the existence of B chromosomes in Meliponini was made by Costa et al. (1992) in Partamona helleri. Since then, analyses revealed the presence of up to 12 different B chromosomes in this species (Martins et al. 2014), and have demonstrated B distribution patterns that indicate the occurrence of geographical variation between populations (Martins et al. 2009). These analyses also showed that these chromosomes are usually heterochromatic and have many AT-rich sequences (Brito et al. 1997, 2005; Brito-Ribon et al. 1999).

The combination of cytogenetic analyses with specific sequences of B chromosomes allowed the partial sequencing of a RAPD marker associated with the presence of these chromosomes in P. helleri (Tosta et al. 2004) and subsequently, the transformation of this into a SCAR marker (Tosta et al. 2007). Further analysis using the SCAR marker as a probe in dot-blot experiments, amplification of the SCAR marker and sequencing confirmed the presence of these chromosomes in P. helleri and Partamona rustica larvae and indicated their presence in adults of Partamona cupira and Partamona criptica (Tosta et al. 2014). With regard to P. cupira, the presence of B chromosomes had already been detected by Marthe et al. (2010) in some individuals from two colonies from Guimarânia/Minas Gerais. These data show that the SCAR marker can facilitate the study of B chromosomes in this genus, because it eliminates the need to work with larvae, which sometimes preclude analyses.

In addition, B chromosomes have also been identified in species of two other genera of stingless bees, Melipona and Tetragonisca. In populations of M. quinquefasciata, a variation of 0 to 4 B chromosomes between individuals of the same colony, as well as between individuals in different locations has been verified (Rocha 2002; Silva et al. 2013a, b). Lopes et al. (2008), in turn, described the presence of a small chromosome B in some individuals (male and female) of an M. rufiventris colony from Guimarâmia/Minas Gerais and Barth et al. (2011) detected the presence of 0–3 B chromosomes in some individuals from 10 of the 11 Tetragonisca fiebrigi colonies from Tangará da Serra/Mato Grosso do Sul. It is emphasized that these chromosomes were not detected in any of the T. angustula colonies analysed in different works (Rocha et al. 2003a; Barth et al. 2011; Miranda 2012).

We can then see that, to date, the B chromosomes of P. helleri were the best studied. However, similar studies are beginning to be developed in other species.

Karyotypic evolution

Kerr and Silveira (1972) were the first to propose a hypothesis to explain the chromosomal variation in Hymenoptera in general, and in Meliponini, in particular. Analysing the cytogenetic data available for several species of bees, these authors found that their chromosome numbers ranged from 6 to 9 and from 12 to 20. They then suggested that these numbers could have arisen as a result of polyploidy events, from a basic number.

Later studies, however, have shown that several species of Meliponini had an intermediate number of chromosomes in addition to differences related to the quantity and location of heterochromatic regions along the chromosome arms, which could not be explained by polyploidy (Pompolo 1992, 1994). Thus, the Minimum Interaction Theory that was initially postulated to explain the karyotype evolution of Australian ants by Imai et al. (1986, 1988, 1994) also began to be used to explain the karyotype evolution of other organisms, including some genera of stingless bees.

The Minimum Interaction Theory predicts the occurrence of changes in karyotypes to minimize deleterious interactions between chromosomes. These modifications would be through a cycle in which centric fissions increase the number of chromosomes, reducing their size and consequently, reducing the interactions between them. After fission, the in tandem growth of heterochromatin occurred in one of the chromosomal arms, restoring the stability of the telomeres. According to this theory, therefore, the karyotypes of ancestral species would consist of a small number of large chromosomes and over time, the descendants’ species would present karyotypes with a larger number of smaller chromosomes with telocentric and acrocentric morphology, resulting from fissions. Accordingly, Pompolo (1992, 1994), Costa et al. (1992), Hoshiba and Imai (1993), Pompolo and Campos (1995), Caixeiro (1996, 1999), Brito (1998), Rocha and Pompolo (1998), Mampumbu (2002), Rocha et al. (2003a), Krinski et al. (2010) and Godoy et al. (2013) used this theory to explain the changes observed in various groups of Hymenoptera.

Pompolo and Campos (1995), for example, used the Minimum Interaction Theory to explain the karyotypic differences (number and morphology) found between L. muelleri (2n = 16) and L. pusilla (2n = 30). According to the authors, in L. muelleri predominate metacentric chromosomes with both arms predominantly euchromatic (only the centromeric region of a metacentric pair and the short arm of an acrocentric pair showed heterochromatic regions). The karyotype of L. pusilla, which has a predominance of acro/submetacentric chromosomes and a greater amount of heterochromatin in one arm, would have emerged from centric fissions and the addition of heterochromatin.

However, the Minimum Interaction Theory does not explain the variations found in species of Melipona (Rocha and Pompolo 1998) or Euglossa (Fernandes et al. 2013) and not even in other hymenopteran groups, e.g. the parasitoid superfamily Chalcidoidea (Gokhman and Gumovsky 2009) and the tribe Epiponini (= Polybiini) of the family Vespidae (Menezes et al. 2014). In Melipona, for example, karyotype evolution does not seem to be related to fission centric events, because although most species present n = 9 chromosomes, a large variation has been observed in the amount of heterochromatin and location of heterochromatic blocks (Rocha and Pompolo 1998; Rocha et al. 2007). The karyotype of M. seminigra, with n = 11, however, could have originated from an ancestor with n = 9, through the fission of two chromosomes.

Therefore, as the chromosomal variations observed in karyotypes of different Meliponini groups do not seem to be explained by a single theory or a single type of structural change, in the present study, we aimed to analyse the karyotype variations, verifying how these variations are related to the phylogenetic relationships among the different genera of this tribe.

In this sense, phylogenetic analysis based on molecular data (Rasmussen and Cameron 2010), considering the different taxa of Meliponini, reinforced the separation of this tribe into three monophyletic lineages: Afrotropical, Indo-Malayan/Australasian and Neotropical. These authors also recognize the presence of three distinct clades in the Neotropical lineage: Trigonisca sensu lato (including Celetrigona, Dolichotrigona, Trigonisca, and Leurotrigona - called in this work “Group 1”), Melipona sensu lato (“Group 2”) and the other Meliponini (“Group 3”). Cytogenetic analyses, in turn, show that the most common karyotypes discussed above (n = 9, 15 or 17) are well separated in these three groups of Neotropical Meliponini (Fig. 1).

Fig. 1
figure 1

Distribution of the three main chromosomal numbers (n = 9, 15 and 17) found in the Meliponini on the phylogenetic tree proposed for the Neotropical Meliponini (modified of Rasmussen and Cameron 2010)

Full size image

In this scenario, it is noteworthy that the Melipona clade has been extensively studied from different biological aspects. Taxonomically, the species of this clade can be grouped into four subgenera: Melikerria, Melipona, Michmelia and Eomelipona (Kerr et al. 1967; Moure 1992). Considering the amount and distribution of heterochromatin between different species of this clade, Rocha and Pompolo (1998) and Rocha et al. (2002) proposed dividing it into two groups. Group I comprises species that have less than 50% heterochromatin in their chromosomes (M. marginata, M. quadrifasciata, M. bicolor, M. asilvae, M. subnitida, M. mandacaia, M. puncticolis and the A complement of M. quinquefasciata). All of these species, except M. quinquefasciata, belong to the subgenera Eomelipona and Melipona. Group II, in turn, includes species of the subgenera Michmelia and Melikerria, which have a high content of heterochromatin, spread over almost the entire length of the chromosomes (M. scutellaris, M. fuscopilosa, M. capixaba, M. captiosa, M. crinita, M. rufiventris, M. mondury, M. fasciculata, M. flavolineata and M. fuliginosa).

Analysis of the genome sizes of these species (Tavares et al. 2010, 2012) confirmed the classification proposed by Rocha and Pompolo (1998) and Rocha et al. (2002), showing that species with low amounts of heterochromatin also have a lower amount of DNA per haploid nucleus (subgenera Melipona and Eomelipona), whereas those having high amounts of heterochromatin have a greater amount of DNA (subgenera Melikerria and Michmelia). Together, these data show that differences in DNA content represent genomic changes that have occurred through the addition or deletion of heterochromatin. It is noteworthy, however, that the resulting grouping of cytogenetic analysis and DNA quantification differ from the phylogenetic proposal of Rasmussen and Cameron (2010), which shows that Eomelipona and Michmelia are the two closest subgenera, although the four subgenera form a single clade.

Furthermore, cytogenetic data available in the literature showed the presence of species with higher chromosome numbers in the Afrotropical (n = 17–18—Kerr and Araújo 1957; Kerr 1969, 1972) and Indo-Malayan/Australasian lineages (n = 18–20—Hoshiba and Imai 1993; Travenzoli et al. 2015) as well as in Bombus, the sister genus of the stingless bees (n = 18–19—Owen et al. 1995), Apis (n = 16—Hoshiba and Kusanagi 1978) and solitary species such as Euglossa (n = 21—Fernandes et al. 2013) and Eulaema (n = 21—Pompolo et al. 1986).

Therefore, considering that karyotypes with high chromosome numbers are the most common among the above groups, including Neotropical stingless bees (n = 17), it is suggested that the ancestor of the karyotype of all Meliponini had a high chromosome number (n = 17–20). Thus, in some groups, the ancestral karyotype may have been maintained. For example, if the ancestral karyotype was comprised of 17 chromosomes, this number would have been maintained in most current genera that comprise “Group 3” (Fig. 1). Karyotypes with lower haploid numbers than 17, such as those found in the genera Duckeola (n = 15), Frieseomelitta (n = 15), Geotrigona (n = 15), Lestrimelitta (n = 14), Parapartamona (n = 16) and Ptilotrigona (n = 11), in turn, must have originated through additional fusions that occurred subsequent to the initial event that gave rise to the current characteristic chromosome number of this group. Centric fusion events could also help explain the current karyotypes found in species of Melipona (“Group 2”), Celetrigona, Leurotrigona and Trigonisca (“Group 1”). These fusion events occurred very early in the separation of Neotropical Meliponini, as they seem to have contributed to the karyotype organization of various genera.

In some cases, more than one type of chromosomal rearrangement has been used to account for these variations. Domingues et al. (2005), for example, found that Trigona fulviventris had 2n = 32 chromosomes, differently from other Trigona species already karyotyped (2n = 34). As the first chromosome pair of this species is much larger than the others and has a metacentric morphology and heterochromatin restricted to the pericentromeric region, the authors attributed the reduction of chromosome number to a centric fusion of two pseudoacrocentric chromosomes. In this case, however, it is noteworthy that the species analysed was possibly Trigona braueri because T. fulviventris is not found in Brazil.

A similar situation was proposed by Domingues (2005) when analysing the karyotype of Scaura latitarsis. In this case, the author proposed that a pericentromeric inversion of a pseudoacrocentric chromosome (AM), followed by elimination of heterochromatin could explain the presence of one metacentric pair in the karyotype of this species, which was not detected on the karyotype of two other species of the genus (S. longula and S. atlantica).

In conclusion, cytogenetic analysis performed to date with Neotropical Meliponini has identified three different groups that gather most species and are congruent with the clades Melipona (n = 9), Trigonisca (n = 15) and other Meliponini (n = 17) as proposed by Rasmussen and Cameron (2010). It is noteworthy, however, that in the latter clade, although species with n = 17 predominate, there is also a sub-group with n = 15 consisting by the genera Frieseomelitta and Duckeola, that are phylogenetically close to each other. Therefore, the karyotypes with n = 15 must have originated independently, because Trigonisca and Frieseomelitta/Duckeola are phylogenetically distant and most genera of the group with which Frieseomelitta/Duckeola is associated have n = 17 (Fig. 1). So, after its origin, each clade/group followed their own route of karyotype evolution, i.e., several rearrangements occurred during the chromosomal karyotype evolution of different groups of the tribe Meliponini. Consequently, the chromosomal karyotype variations observed in different groups seem not to be explained by a single theory or a single kind of structural change. In many cases, more than one type of rearrangement is necessary to explain the variations detected.

Concluding remarks and prospects

The data presented show that, in spite of the early cytogenetic studies being restricted to description of the chromosome number, current studies involve the content and location of the heterochromatic regions as well as the molecular characterization of chromatin (AT and CG bases content), localization of genes and even whole chromosomes by molecular cytogenetic techniques (Brito et al. 2005; Martins et al. 2013; Lopes et al. 2014). In situ hybridization, for example, has been used for the location of ribosomal DNA sequences (Rocha et al. 2002; Mampumbu 2002; Brito et al. 2005; Ferreira 2015) and other repetitive sequences (Ferreira et al. 2015; Travenzoli et al. 2015) whereas microdissection has allowed the construction of probes from specific regions or entire chromosomes, thereby contributing to studies on chromosome evolution of Meliponini (Fernandes et al. 2011; Martins et al. 2013). It is worth noting that these advances have been accompanied by increases in informatics technology, with the development of capture and image analysis software. These software applications have allowed more precise measurements in addition to permitting more refined analyses of the patterns of chromosomal banding and composition of chromatin.

These studies have shown that although chromosome number in general is conserved among species of a certain genus, other aspects, such as chromosome morphology, quantity, distribution and composition of heterochromatin, may vary between them, which can aid in making taxonomic and evolutionary inferences from this tribe (Brito 1998; Rocha et al. 2002; Duarte et al. 2009; Barth et al. 2011; Miranda 2012; Miranda et al. 2013; Lopes et al. 2014). Thus, the emergence of techniques, such as molecular cytogenetic, may help clarify aspects not yet revealed by classical techniques.