Anochetus Mayr, 1861 is a monophyletic genus and a sister genus of Odontomachus Latreille, 1804 (Larabee et al. 2016, Fernandes 2017). Morphologically, they belong to the subtribe Odontomachiti of trap-jaw ants (Brown-Jr 1976).

Anochetus targionii has 2n = 30 chromosomes (Fig. 1a), which is considered as a modal number according to Santos et al. (2010). Anochetus chromosome numbers range from 2n = 24–46, which represents higher karyotype diversity than that found in Odontomachus (2n = 32–42) (reviewed in Mariano et al. 2019). However, only 12 morphospecies out of 113 valid species of Anochetus have been cytogenetically analyzed: nine from the Indo-Malayan and three from the Neotropics, A. altisquamis Mayr, 1887 (2n = 30), A. horridus Kempf, 1964 (2n = 46), and A. emarginatus (Fabricius, 1804) (2n = 28) (Santos et al. 2010, Mariano et al. 2015).

Figure 1.

Karyotypes of the tribe Odontomachiti (Ponerinae): a Anochetus targionii (2n = 30) b Odontomachus haematodus (2n = 44). Scale bar: 5µm.

Since Anochetus diversified earlier than Odontomachus (Larabee et al. 2016, Fernandes 2017), higher karyotypic variation in Anochetus would be expected (Santos et al. 2010). Anochetus targionii has the same chromosome number as A. altisquamis, A. modicus Brown, 1978, and A. graeffei Mayr, 1870. It seems that 2n = 30 chromosomes is the plesiomorphic condition since it is found throughout the genus Anochetus and is present in A. altisquamis, which is considered a phylogenetically “basal” clade (Larabee et al. 2016, Fernandes 2017). Anochetus species also share a constant number of acrocentric chromosomes. Within the lineage of A. horridus, chromosome fission seems to have played an important role in recent karyotype evolution, increasing the number of chromosomes in the karyotype. According to Larabee et al. (2016), A. horridus diversified around 25 million years ago (MYA), whereas A. targionii diversified less than 10 MYA.

Odontomachus haematodus has 2n = 44 chromosomes, of which 18 are acrocentric (Fig. 1b), confirming information provided by Santos et al. (2007, unpublished data) and Aguiar et al. (2012, unpublished data) (reviewed in Mariano et al. 2019). Similar to Anochetus, Odontomachus species seem to have characteristic karyotypes that are slightly different between Odontomachus phylogenetic clades.

The Indo-Pacific species, O. rixosus Smith, 1857 and O. latidens Mayr, 1867, have 2n = 30 chromosomes but no further information about their karyotypes is available (reviewed in Lorite and Palomeque 2010). The other known karyotype of Odontomachus species belongs to the haematodus group according to molecular phylogeny (Larabee et al. 2016). All known karyotypes from the haematodus group (reviewed in Santos et al. 2010) have 44 chromosomes, including O. haematodus, whose karyotype is described in this study. Odontomachus chelifer (Latreille, 1802) has plesiomorphic traits and the highest number of acrocentric chromosomes (40) among Indo-Pacific species (reviewed in Santos et al. 2010). The species O. meinerti Forel, 1905 and O. bauri Emery, 1892 have 34 and 14 acrocentric chromosomes out of 44 respectively (Teixeira 2018).

This suggests that heterochromatin growth at telomeric regions of shorter arms of acrocentric chromosomes may be significant in Odontomachus karyotype evolution. This is in accordance with the Minimum Interaction Theory proposed by Imai et al. (1994), which proposes that reduced interactions between different chromosomes inside the nucleus increases the fitness of the individual.

The genus Pseudoponera Emery, 1900 has six valid species (Bolton 2019). Two species Pseudoponera gilberti and P. stigma are near-identical morphologically. Conflicting cytogenetic analyses (Mariano et al. 2012) due to misidentification have recently been resolved and the two species distinguished in samples from the Atlantic forest (Correia et al. 2016). The chromosome number for P. gilberti is 2n = 12 (10m + 2sm) and for P. stigma was 2n = 14, all of them metacentrics. In spite of minor differences in chromosome morphology of P. stigma between Atlantic forest (Correia et al. 2016) and Amazonia, both karyotypes share the same chromosome number.

Figure 2.

Karyotypes of the genus Pseudoponera (Ponerinae): a P. gilberti (2n = 12) b P. stigma (2n = 14). Scale bar: 5µm.

The fungus-growing ants from the genus Acromyrmex form a sister group of the genus Atta and together are believed to be monophyletic. There are 33 valid species of Acromyrmex and 18 species of Atta (Bolton 2019), with wide distributions throughout the Neotropics (Delabie et al. 2011). The relationship between Atta and Acromyrmex became clearer under a combined approach using morphological, molecular, and cytogenetic tools (Cristiano et al. 2013). Cytogenetic data are available for five Atta species (Barros et al. 2011, 2014, 2015) from three of the four monophyletic groups according to the molecular phylogeny proposed by Bacci et al. (2009) and 13 species of Acromyrmex (reviewed in Barros et al. 2011, 2016, Teixeira et al. 2017).

The leaf-cutter ant Atta sexdens has 2n = 22 (Fig. 3a), and chromosome morphology is the same (18m + 2sm + 2st) to that of other Atta species from the Brazilian savannah and Atlantic Forest (Barros et al. 2014, 2015). The Amazonian population of Acromyrmex balzani analyzed in this study has 2n = 38 chromosomes and the same karyotype (Fig. 3b) as that of the Brazilian savannah and Atlantic forest populations (Barros et al. 2016). The largest metacentric pair of A. balzani is large, about twice the length of the largest subtelocentric chromosome previously identified in other Brazilian populations of this species (Barros et al. 2016). In all other Acromyrmex species studied so far, the former pair of chromosomes is of similar length. Based on the recent checklist of the ants of French Guiana (Franco et al. 2019) this is the first record for A. balzani in French Guiana and also the first cytogenetic analysis of this species in the region.

Figure 3.

Karyotypes of fungus-growing ants (Myrmicinae, Attini: Attina): a Atta sexdens (2n = 22) b Acromyrmex balzani (2n = 38) c Cyphomyrmex transversus (2n = 24) d C. transversus (n = 12, male karyotype) e Myrmicocrypta sp. (2n = 30). Scale bar: 5µm.

So far, all karyotype analyses showed that Atta spp. have 2n = 22 chromosomes and Acromyrmex spp. have 2n = 38 chromosomes (Barros et al. 2016, Teixeira et al. 2017). Acromyrmex striatus (Roger, 1863) is an exception, with 2n = 22 chromosomes, the same as Atta spp. and is considered the sister group of leaf-cutter ants (Cristiano et al. 2013). There are variations between the morphological features of certain chromosomes in Acromyrmex due to heterochromatin growth (Barros et al. 2016). Interpopulation cytogenetic studies for ants are scarce (e.g., Mariano et al., 2006, 2012, Aguiar et al. 2017) and none are available for leaf-cutter ants.

The fungus-growing attine Cyphomyrmex transversus has 2n = 24 and n = 12 with mostly metacentric and submetacentric chromosomes (Fig. 3c, d) which differs from 2n = 42, all of them acrocentric, observed by Mariano et al. (2019), highlighting the importance of detailed cytogenetic studies in this species. It has a range from northern Brazil to central Argentina including the northeastern regions of Brazil (Kempf 1965). Cyphomyrmex transversus and the three other Cyphomyrmex species which have been karyotyped (see Sosa-Calvo et al. 2017 for recent taxonomic changes) have chromosome numbers ranging between 2n = 20 and 2n = 42 (reviewed in Mariano et al. 2019). It seems that the high proportion of metacentric chromosomes is characteristic of this genus. In spite of morphological affinity of C. transversus (present study) to C. rimosus (Spinola, 1851), observed by Kempf (1965), the karyotype divergence between them (2n = 24 and 2n = 32, respectively; Murakami et al. 1998) is puzzling because of the numerical difference of their karyotypes coupled with similar morphology of chromosomes. These findings merit further study using advanced chromosome banding and molecular phylogenetic techniques.

The fungus-growing species Myrmicocrypta sp. had 2n = 30 chromosomes, 18 of them metacentric (Fig. 3e). The studied colony was collected from the cavities of a rotten log. Myrmicocrypta Smith, 1860 is the sister genus of Mycocepurus Forel, 1893, and both are members of the clade Palleoattina (Sosa-Calvo et al. 2017). Myrmicocrypta is widely distributed in the Neotropics, from Mexico to Argentina and includes 27 valid species (Bolton 2019). A recent study by Sosa-Calvo et al. (2019) suggests that only two species can nest in rotten logs: M. spinosa Weber, 1937 and the undescribed species, M. JSC001. This is a derived characteristic for Myrmicocrypta and therefore this clade is apparently monophyletic. The only extant cytogenetic data available for this genus are from the Montagne des Singes area, French Guiana (Mariano et al. 2011), about 60 km from where the samples from the present study were collected.

Since the studied sample was identified as an undescribed species, it is possible that the present species is M. JSC001. Myrmicocrypta spinosa has not been recorded in French Guiana: the samples studied by Sosa-Calvo et al. (2019) included only M. JSC001. The species studied by Mariano et al. (2011) had a slightly different karyotype, probably as a result of variation in the chromosome condensation. These results highlight the importance of good chromosome preparations for studying karyotype configuration.

This study represents the first cytogenetic analysis for the genus Allomerus. A. decemarticulatus and the A. octoarticulatus species complex had 2n = 28 and 2n = 44, respectively (Fig. 4). The Allomerus species are specialist ants inhabiting diverse obligate myrmecophytic plants in South America. Allomerus decemarticulatus and A. octoarticulatus, have been intensively studied in French Guiana from an ecological perspective: they build galleries using the fungus Trimmatostroma cordae Sharma & Singh, 1976 (see Dejean et al. 2005, Ruiz-González et al. 2011). The molecular phylogeny of the genus showed that A. octoarticulatus is a complex of two species that cannot be separated morphologically (Orivel et al. 2017). However, these two species are associated with different plants: A. octoarticulatus var. demerarae inhabits only Cordia nodosa, while A. octoarticulatus can be associated with several myrmecophytic plant species throughout its distribution range.

Figure 4.

Karyotypes of the genus Allomerus (Myrmicinae): a A. decemarticulatus (2n = 28) b A. octoarticulatus var. demerarae (2n = 44) associated with Cordia nodosa c A. octoarticulatus (2n = 44) associated with Hirtella sp. Scale bar: 5µm.

The number of acrocentric chromosomes is highly different between these two species, even though meta/submetacentric and acrocentric chromosomes predominate in A. decemarticulatus and A. octoarticulatus, respectively. According to the Minimum Interaction Theory, centric fissions may have played an important role in the chromosome evolution of Allomerus; however, the karyotypes of additional species should be investigated to support this conclusion. A comparison between the two species of A. octoarticulatus, which nest in different plant species, was also made (Fig. 4b, c). However, basic cytogenetic techniques (chromosome number and morphology) could not differentiate between the two. Additional banding techniques with molecular probes may further illuminate this question in the future.

Strumigenys diabola has 2n = 40 (Fig. 5a) with many chromosomes having short arms (submeta/subtelocentrics). The genus Strumigenys Smith, 1860 harbors small cryptic species specialized in preying on collembolans. There are currently more than 800 valid species of Strumigenys (Bolton 2019) of which 190 are from the Neotropics. Strumigenys diabola are reported in northern and northeastern Brazil and in French Guiana (Janicki et al. 2016). For the Neotropics, cytogenetic data was previously only available for Strumigenys louisianae Roger, 1863, which has 2n = 4 (Alves-Silva et al. 2014) and for eight species from Asia and Oceania (Lorite and Palomeque 2010). This is the second cytogenetic record in Neotropics and the absence of data in the Strumigenys mandibularis group sensu Bolton (2000) make comparisons with other species impossible. Further studies of this genus will help understanding chromosome evolution and phylogeny of the group.

Figure 5.

Karyotypes of four genera of Myrmicinae: a Strumigenys diabola (2n = 40) b Wasmannia auropunctata (2n = 32) c Solenopsis geminata (2n = 32) d S. geminata (n = 16, male karyotype) e Crematogaster longispina (2n = 24). Box shows the chromosome pair with size heteromorphism. Scale bar: 5µm.

In this study, workers of W. auropunctata had 2n = 32 (Fig. 5b), with one chromosome pair showing considerable size heteromorphism in all individuals analyzed. The genus Wasmannia Forel, 1893 includes 11 species and is endemic to the Neotropics. The “little fire ant” W. auropunctata is notable because of its reproductive mechanism (Fournier et al. 2005). It has three different genetic systems: haplodiploidy, male clonality, and thelytoky (Foucaud et al. 2006). In this study, the same chromosome number and a similar karyotype from colonies from Ilhéus and Una, southeast Bahia, Brazil were observed (Souza et al. 2011), although there are differences in chromosome classification. The heteromorphic pattern was not described for the Atlantic forest population and therefore needs to be investigated further.

Our analysis found 2n = 32 in female Solenopsis geminata and n = 16 in males with most chromosomes (26) being metacentric or submetacentric (Fig. 5c, d). The genus Solenopsis Westwood, 1840 is difficult to identify at the species level, although these species form obvious natural groups (Pacheco and Mackay 2013). The chromosome number for this species in our analysis is the same as that observed in five previously described fire ant species including S. geminata, (reviewed in Lorite and Palomeque 2010) which belong to the subgenus Solenopsis (Pacheco and Mackay 2013).

We compared our data with those from colonies of S. geminata from the USA (Crozier 1970) and India (Imai et al. 1984). The karyotype from French Guiana is similar to that from India, despite certain differences in chromosome classification. Differences in karyotypic formula among various localities and colonies were reported by Imai et al. (1984) based on their observation of the presence/absence of the short arm in some chromosomes as a result of C-band polymorphisms. Those patterns demonstrate the importance of understanding heterochromatin dynamics at the population level for analyzing karyotype evolution of ants.

The ant genus Crematogaster is a global, widespread, and species-rich clade. It currently comprises 498 valid species and is divided into two subgenera, Crematogaster sensu stricto and Crematogaster (Orthocrema) Santschi, 1918 (Blaimer 2012a, b). The subgenus Orthocrema is more complex, and numerous clades exist within this group. Crematogaster longispina, which belongs to the subgenus Orthocrema, has 2n = 24, and all chromosomes are meta/submetacentrics (Fig. 5e). This is the first New World Crematogaster karyotype ever described, which makes reasonable comparisons difficult. Karyotype data is available for 17 morphospecies of Crematogaster from Malaysia, Indonesia, India, Japan, and Australia (reviewed in Lorite and Palomeque 2010).

Within Crematogaster spp., the chromosome number ranges from 2n = 24–58, with 10 morphospecies having 2n = 24 or 26. Increasing the number of studied species in the Neotropics may help to understand the chromosome evolution of the group.

Dolichoderus Lund, 1831 is the most speciose ant genus in the subfamily Dolichoderinae, with 130 valid species (Bolton 2019). Chromosomal data is available for seven species collected from the Atlantic forest (Santos et al. 2016) and four species from the Indo-Malayan region (reviewed in Lorite and Palomeque 2010). The genus demonstrates high chromosome variation, 2n = 10–52, and is the most cytogenetically diverse genus within Dolichoderinae. According to the molecular phylogeny produced by Santos et al. (2016), this species occupies a less derived position, which agrees with previous conclusions that suggest that this species belongs to a separate species group (Mackay 1993). The chromosome number already known for this species is 2n = 38, and it also has many meta/submetacentric chromosomes (Santos et al. 2016). However, in this study, additional acrocentric chromosomes were observed in D. imitator (2n = 46). Chromosomal intraspecific variation in Dolichoderus has not previously been reported. This again emphasizes the importance of karyotypic studies at the level of certain populations, which may represent either geographic clines or a species complex. Enhancing population studies for this species may have important implications for our understanding of both taxonomy and chromosome evolution of Formicidae.

Figure 6.

Karyotype of Dolichoderus imitator (2n = 46) (Dolichoderinae). Scale bar: 5µm.