Belostomatidae include some of the largest heteropteran species, which are general predators that play an important role as biological agents in aquatic environments (Menke 1979, Smith 1997, Saha et al. 2010). This family has a cosmopolitan distribution in tropical and subtropical areas in the world (Schnack 1976, Polhemus and Polhemus 2008). In South America, the genus Belostoma Latreille, 1807 is the most diverse and includes 61 species mainly distributed from Colombia and Brazil to Argentina and Chile (Heckman 2011). Nevertheless, cytogenetic reports in Belostoma from South America comprise the male chromosome complement of 15 species and male meiosis analysis of 13 species (Table 1). Ten of these species show a modal diploid chromosome number 2n = 29 = 26 + X₁X₂Y (male) and five species possess reduced chromosome numbers and a simple sex chromosome system XY/XX (male/female) (Table 1).

All species of Belostoma analyzed possess holokinetic chromosomes, i.e. chromosomes without a primary constriction and therefore without a localized centromere. Autosomal bivalents are synaptic and chiasmatic, whereas sex chromosomes are asynaptic and achiasmatic, and behave as univalents in first male meiotic division. However, at metaphase II sex chromosomes associate end-to-end through the so called touch-and-go pairing, forming a pseudo-bivalent or pseudo-multivalent. In the first meiotic division, autosomal bivalents segregate reductionally while sex chromosomes divide equationally (Ueshima 1979, Papeschi and Bidau 1985, Suja et al. 2000, Papeschi and Bressa 2006, Bardella et al. 2012). During meiosis, the kinetic activity is restricted to the chromosome ends and the chromosomes can be regarded as telokinetic (Motzko and Ruthmann 1984).

Most hypotheses on karyotype evolution in Heteroptera include both autosomal and sex chromosome fusions and fragmentations (Ueshima 1979, Manna 1984, Thomas 1987, Papeschi 1994, 1996, Pérez et al. 2004). The cytogenetic data available for Belostoma allow to hypothesize that current karyotypes with a multiple sex chromosome system (X_nY/X_nX_n, male/female) are derived through fragmentation of the ancestral X from an XY sex chromosome system. On the other hand, reduced chromosome complements with simple sex chromosome system (XY/XX, male/female) have probably originated through several chromosomal fusions (Papeschi 1996, Papeschi and Bressa 2006, Bardella et al. 2012).

The aim of this study was to perform a detailed comparison of male meiosis behaviour and examine the structure of the holokinetic chromosomes by means of C- and fluorescent bandings, and fluorescent in situ hybridization (FISH) with 18S rDNA probes in Belostoma dentatum (Mayr, 1863), Belostoma elongatum Montandon, 1908 and Belostoma gestroi Montandon, 1903. The female complement and the male meiosis of Belostoma elongatum and Belostoma gestroi are described for the first time. These results allowed us to distinguish morphologically similar species and, also, led us to propose a scenario of karyotype evolution in the genus Belostoma.

The Belostoma species analyzed here shared apparently similar karyotypes, since they possess the same chromosome complement (2n = 29 = 26 + X₁X₂Y, male), with chromosomes progressively decreasing in size. In Belostomatidae, this 2n is the modal diploid chromosome number and is present in 10 species of Belostoma (Papeschi and Bressa 2006, Chirino and Bressa 2011). On the other hand, in Belostoma elongatum and Belostoma gestroi the male meiotic behaviour followed a similar pattern as previously described for other species of this genus (Papeschi and Bidau 1985, Papeschi 1996, Bardella et al. 2012). Both Belostoma elongatum and Belostoma gestroi, as well as Belostoma dentatum, showed the following cytogenetic characteristics: a) synizesis observed, b) a multiple sex chromosome system (X₁X₂Y, male), c) Y chromosome negatively heteropycnotic at metaphase II, d) variation in the TCL and in the C-banding pattern, and e) a single pair of NOR-autosomes. Within Heteroptera, the synizesis stage was described in a few species of Belostoma (Papeschi and Bidau 1985, Papeschi 1992) and Dysdercus Guérin-Méneville, 1831 (Bressa 2003, Bressa et al. 2003). In this stage the chromatin condenses eccentrically in the nucleus and chromosome pairing begins.

Heterochromatin characterization in the three species revealed differences in the amount, distribution and location of the constitutive heterochromatin in autosomes and both X chromosomes: i) terminal C-positive bands in Belostoma dentatum, ii) conspicuous C-positive bands at terminal and interstitial positions in Belostoma elongatum, and iii) very scarce C-positive bands terminally located in Belostoma gestroi. This variation in the constitutive heterochromatin of these three species could imply changes in the DNA content in the karyotype evolution in the genus, which could modify the size of the chromosome complement. In accordance with this suggestion, the analysis of TCL showed a significant variation among the three species, which means that certain chromosomal changes, must have taken place during their evolution.

Papeschi (1992) found a great interspecific variation in DNA content as well as differences in C-positive heterochromatin percentage among Belostoma dentatum (1.93 pg, 58.54 %), Belostoma elongatum (1.75 pg, 59.47 %) and Belostoma gestroi (1.13 pg, 37.24 %). Taking into account the data previously described by Papeschi (1992), together with the results obtained from the analysis of the TCL in these three species, we propose the existence of positive relationships between TCL and DNA content and between TCL and C-positive heterochromatin percentage (Fig. 6). Thus, it is apparent that differences in the TCL in the three species might represent changes or variations in DNA content since the accumulation/addition of satellite DNA in all chromosomes of the complement. The comparison between the DNA content and the C-positive heterochromatin percentage of the three species shows a general trend, i.e. an increase in the DNA content is accompanied by an increase in the amount of C-positive heterochromatin. However, Belostoma dentatum and Belostoma elongatum have a very different DNA content and a similar percentage of C-positive heterochromatin. On the other hand, Belostoma gestroi shows the lowest DNA content and the lowest C-positive heterochromatin percentage. In accordance with the earlier reports on six other species of Belostoma (Papeschi and Bidau 1985, Papeschi 1988, 1991, 1992), the genome size differences between Belostoma dentatum and Belostoma elongatum could be due to a proportionate variation of both C-positive heterochromatin and C-negative chromatin occurred during evolution. The chromosomes of Belostoma gestroi could have gained low amount of C-positive heterochromatin, or else during evolution some C-positive bands became lost.

In Heteroptera, the classical distribution pattern of C-positive heterochromatin is terminal in some or all chromosomes. Interstitial C-positive bands are described in a few species and some of them correspond to secondary constrictions and NORs. In concordance with these cytogenetic features, the C-banding pattern observed in Belostoma elongatum with respect to both terminal and interstitial C-positive regions agrees with most previous reports within Belostoma (Papeschi 1995) and Heteroptera (Camacho et al. 1985, Panzera et al. 1995, Grozeva and Nokkala 2001, Angus et al. 2004, Ituarte and Papeschi 2004, Bressa et al. 2005, 2008, Franco et al. 2006).

The results with fluorescent banding indicate that all C-positive bands in the species analysed were not enriched with AT or CG base pairs, as all chromosomes were stained homogeneously with both DAPI and CMA₃ fluorochromes, except for the C- positive band observed in the medium-sized autosomal bivalent of Belostoma dentatum and Belostoma elongatum and in one of the large-sized of Belostoma gestroi, which was DAPI negative/CMA₃ positive. Therefore, the CMA₃ bright band is enriched in GC base pairs and could represent an NOR (see below). The presence of a CMA₃ bright band was also reported not only in other species of Belostoma (Papeschi and Bressa 2006) but also in other heteropteran species, at interstitial or terminal position, either on autosomes or sex chromosomes, and they are generally associated to NORs (González-García et al. 1996, Papeschi et al. 2001, 2003, Rebagliati et al. 2003, Cattani et al. 2004, Grozeva et al. 2004, Poggio et al. 2011).

In Belostomatidae, the location of NORs was previously analysed by FISH with 18S rDNA probe in Belostoma oxyurum (Dufour, 1863) (2n = 6 + XY, NOR in sex chromosomes), Belostoma micantulum (Stål, 1860) (2n = 14 + XY, NOR in sex chromosomes), Belostoma elegans (Mayr, 1871) (2n = 26 + X₁X₂Y, NOR in a pair of autosomes) (Papeschi and Bressa 2006), and Lethocerus patruelis (Stål, 1854) (2n = 22 + 2m + XY, NOR in sex chromosomes) (Kuznetsova et al. 2012). The present paper presents information about the number and chromosomal location of ribosomal clusters in Belostoma dentatum, Belostoma elongatum and Belostoma gestroi, which have a single cluster located in an autosomal pair. In these three species the NOR is associated with a CMA₃-bright band. The results of the fluorescent banding and FISH in these species agree with those described for Belostoma oxyurum, Belostoma micantulum and Belostoma elegans, in which the NOR regions are colocalized with a CMA₃-positive band and, therefore, the repeating unit of rDNA is GC-rich (Papeschi and Bressa 2006). Taking into account the data on the number and location of rDNA clusters along with the type of sex chromosome systems in Belostomatidae, we can observe two different patterns of rDNA distribution. The NOR is located at terminal position on both sex chromosomes in species that have a simple sex chromosome system (XY), or, in contrast, the NOR is placed at terminal position on an autosomal pair in the species with a multiple sex chromosome system (X₁X₂Y). Hence, our present results led us to propose that in Belostomatidae the location of rDNA genes could be associated with variants of the sex chromosome systems. Moreover, this relationship between the number and location of the NOR and the sex chromosome system has only been observed in this family of Heteroptera.

Previous cytogenetic data on worldwide Belostomatidae species allowed Papeschi and Bressa (2006) to propose an ancestral male karyotype 2n = 28 = 26 + XY, from which the karyotypes with multiple sex chromosome systems (2n = 26 + X₁X₂Y) and those with a low 2n (6 + XY, 14 + XY, 13 + XY, 22 + XY) would have originated by fragmentation of the ancestral X chromosome and chromosomal fusions, respectively. It is generally accepted that multiple sex chromosome systems in Heteroptera are the result of fragmentation (s) of the X and/or Y chromosome (s) of an ancestral simple sex chromosome system (Heizer 1950, Hughes-Schrader and Schrader 1961, Ueshima 1979, Manna 1984, Papeschi 1996, Papeschi and Bressa 2006). The holokinetic nature of heteropteran chromosomes and the achiasmatic male meiosis of sex chromosomes are the main facts that support this hypothesis and may account for the variability (Ueshima 1979, Manna 1984, Thomas 1987). In most cases of multiple sex chromosomes, the increase in the number of sex chromosomes is not accompanied by a reduction in the number of autosomes. The analysis of different populations of Belostoma orbiculatum Estévez and Polhemus, 2001 (Papeschi 1996), Belostoma plebejum (Stål, 1858) (Papeschi 1994), Belostoma dilatatum (Dufour, 1863) (Bardella et. al 2012), Oechalia pacifica (Stål, 1859) (Heizer 1950) and Banasa zeteki Sailer, 1959 (Pentatomidae) (Schrader and Hughes-Schrader 1958), polymorphic for the sex chromosome systems, lend support to the hypothesis of a fragmentation origin of a multiple sex chromosome systems. In all these examples one chromosome of the simple system (XY) was replaced by two chromosomes of smaller size in the mutant individuals. Therefore, these interspecific polymorphisms of sex chromosomes represent a direct evidence of the origin of multiple sex chromosome system through fragmentation in Belostoma. On the other hand, the species of this genus with reduced chromosome numbers are characterized by a larger chromosome size, a low DNA content and very scarce C-positive heterochromatin (Papeschi and Bressa 2006). These karyotypes probably originated from the ancestral chromosome complement through chromosome fusions. The possibility of their occurrence is supported by the fact that the autosomal fusions have been found in heterozygous condition in natural populations of Belostoma plebejum (Papeschi 1994), Triatoma infestans (Klug, 1834) (Poggio et al. 2013) and Mepraia gajardoi Frías, Henry and González, 1998 (Pérez et al. 2004) (Reduviidae).

Published data on karyotype evolution in species of this genus (Papeschi and Bressa 2006) along with the present results of rDNA-FISH support the hypothesis that in the ancestral male karyotype (2n = 28 = 26 + XY) the NOR would have been located in a pair of autosomes (Fig. 7). A fragmentation of the single X chromosome in the ancestral karyotype resulted in multiple X chromosomes and led to a karyotype with 2n = 29 = 26 + X₁X₂Y, but keeping the ancestral NOR-autosome pair, as represented by Belostoma dentatum, Belostoma elegans, Belostoma elongatum and Belostoma gestroi (Fig. 7). On the other hand, autosomal fusions and the fusion of the ancestral sex chromosome pair (XY) with the autosomes carrying the NOR would result in the reduction in the diploid chromosome (2n = 8, 16), increasing the chromosomes size and led to the presence of rDNA genes in both X and Y chromosomes. Alternatively, the rDNA gene cluster could have been translocated from an autosomal location to the X and Y chromosomes, as it has been proposed in Dysdercus albofasciatus Berg, 1878 (Bressa et al. 2009). Furthermore, future studies within Belostoma about this regular pattern will shed light on the karyotype evolution within the genus and also support the mechanisms involved in the karyotype evolution.

Conventional taxonomy of water bugs has focused almost entirely on adult specimens. There are relatively few publications on interspecific differences among the larvae, and fewer still concern South American species. The literature of the genus Belostoma includes much confusion because, in many cases, the species are very similar in coloration and appearance and only males or rarely only females can be identified (Heckman 2011). In Belostoma dentatum, Belostoma elongatum and Belostoma gestroi, the karyotype analyses allow us to get a detailed characterization and a better knowledge of their chromosomal structure. Hence, we conclude that the cytogenetic studies provide valuable features that can be used to solve problems on taxonomic identification, at least for this genus.