Neotropical primate evolution and phylogenetic reconstruction using chromosomal data

Abstract Platyrrhini are a group of Neotropical primates living in central and south America, and have been extensively studied through morphological and molecular data in order to shed light on their phylogeny and evolution. Agreement on the main clades of Neotropical primates has been reached using different approaches, but many phylogenetic nodes remain under discussion. Contrasting hypotheses have been proposed, presumably due to different markers and the presence of polymorphisms in the features considered; furthermore, neither Neotropical primate biodiversity nor their taxonomy are entirely known. In our perspective, a cytogenetic approach can help by making an important contribution to the evaluation of the phylogenetic relationships among Platyrrhini. In this work, molecular cytogenetic data regarding the principal nodes of the Neotropical monkey tree have been reviewed; classical cytogenetic data have also been considered, especially when other data have proven elusive, permitting us to discuss highly derived karyotypes characterized by a wide range of diploid numbers of chromosomes and variable chromosomal evolution with different rearrangement and polymorphism rates.

In the following discussions of each family (Cebidae, Atelidae and Pitechidae), their principal features as well as phylogenetic relationships, according to recent molecular data, are reported. Indeed, Platyrrhini have also been studied in depth at the intra-family level, analyzing intergenus and intragenus relationships. The principal works on Platyrrhini phylogeny, using different molecular markers for their analyses, at various levels are reported below (Table I). The Cebidae family includes many genera and is marked by a close relationship between Cebus (capuchin monkeys) and Saimiri (squirrel monkeys) (Schneider et al. 2001;Perelman et al. 2011) which share many characteristics at the morphological level that indicate a common origin, such as a high ratio of brain/body weight, pronounced sexual dimorphism, and the ability to gradually change their diet which is based mainly on fruits and insects. However, they also have some differences; for example, capuchin monkeys (Cebus and Sapajus) have a semi-prehensile tail which is unable to sustain weight, whereas Saimiri are born with prehensile tails but lose grasping ability as they age. Among Cebidae it is also possible to distinguish Callithrix (Atlantic marmosets), Cebuella (marmosets), Saguinus (tamarins), and Leontopithecus (lion tamarins); the latter, also known as callitrichids, are characterized, with some exceptions, by small body size, claw-like nails on all digits but the hallux, two molars instead of three, and dietary exploitation of plant exudates (gums and saps). In addition, there is a peculiarity in marmosets in that they are known to give birth to twins. Another Cebidae genus is the monotypic Callimico (Goeldi's monkeys), with the species C. goeldii characterized by small body size and clawlike nails like marmosets, as well as by other features more typical of larger bodied Platyrrhini, such as a third molar and single births. For this reason, its phylogenetic position has been highly debated, although it is now considered to be resolved (Canavez et al. 1999b;Schneider et al. 2001). Among callitrichids, indeed, molecular phylogenetic reconstructions present Saguinus as the sister taxa of Leontopithecus and Callimico as well as the Callithrix and Cebuella clades (Schneider et al. 2001;Perelman et al. 2011).  (Schneider et al. 1996(Schneider et al. , 2001Canavez et al. 1999a;Opazo et al. 2006); (b) with Pitheciidae as sister clade to that of Atelidae and Cebidae (Ray et al. 2005;Osterholz et al. 2009;Wildman et al. 2009;Perelman et al. 2011;Kiesling et al. 2015). Table I. List of platyrrhine phylogenetic studies based on molecular markers and references; letters C, A and P stand for Cebidae, Atelidae Moreira and Seuánez (1999) Moreover, Cebidae includes the genus Aotus (owl/night monkeys), the only representative platyrrhine species having nocturnal habits, characterized by monogamous social organization in small groups. Aotus is primarily frugivorous, but these primates also consume leaves and insects. The phylogenetic position of Aotus has been highly debated and still remains unclear; it is supposed to be a sister clade to either Cebus/Saimiri (Opazo et al. 2006;Wildman et al. 2009) or callitrichines (Perelman et al. 2011;Kiesling et al. 2015). Moreover, other research has led to conflicting phylogenetic reconstructions on the basis of the markers analyzed (Perez et al. 2012).

Comparative cytogenetics and phylogenetic reconstructions
Classic cytogenetic studies using banding analysis allowed researchers to demonstrate that primate chromosomes have been conserved during evolution (Dutrillaux 1979(Dutrillaux , 1988Dutrillaux & Couturier 1981;Dutrillaux et al. 1986). Since the 1990s (Wienberg et al. 1990), the karyotypes of different primate species have also been compared at the molecular level, applying fluorescent in situ hybridization (FISH) with human chromosomal probes. This molecular cytogenetic approach is known as "chromosome painting" and consists of the hybridization of the human DNA probes of a whole chromosome (labeled with a fluorescent substance) with the DNA of a target species, taking advantage of their complementary nature. Chromosome painting allows researchers to determine chromosomal homologies at the level of whole or partial chromosomes, as well as interchromosomal rearrangements (translocations, fissions and fusions) that have occurred during evolution; chromosomal painting permits the determination of which chromosomes, or chromosomal syntenies (the localization of two or more genes on the same chromosome), have been conserved or reshaped, identifying syntenic associations in the genomes of the species being compared. Subsequently, it has been possible to hybridize not only human probes but even other animal probes (zoo-FISH) made through flow sorting. Human and other primate probes have been reciprocally hybridized through chromosome painting, permitting the detection of real homologies in two reciprocal experiments and rearrangement breakpoints Dumas et al. 2007). In recent years, sub-regional or locus-specific probes, produced by microdissection or by cloning DNA within vectors, have also been used for FISH. These probes have demonstrated a high resolving power, identifying intrachromosomal rearrangements and breakpoints that are not detectable through painting (Stanyon et al. 2008).
Chromosomal data obtained through comparative cytogenetics have been used for phylogenetic reconstructions using the cladistic approach and the principle of parsimony. The first step in this field is making the distinction between homology due to shared ancestry and homoplasy due to parallel or convergent evolution; thus, among homologies, it is necessary to distinguish ancestral chromosomal syntenies (synapomorphies) from new shared syntenic associations (symplesiomorphies) formed as a consequence of chromosomal rearrangements Rokas & Holland 2000). Since rearrangements are rare events in mammals (two for every 10 million years, Froenicke 2005;Murphy et al. 2005), the common derivative syntenic associations between two species are useful for phylogenetic reconstructions. Through this analysis, in a comparative perspective, it has been possible to reconstruct the hypothetical ancestral karyotype of all primates and of the main nodes of the primate evolutionary tree (Stanyon et al. 2008). To distinguish conserved from derived characteristics, a comparison with an outgroupa closely related species that is considered external to the group under examinationis used. According to the principle of parsimony, the interpretation that involves the least likely number of steps is preferred among the various possible interpretations of a phenomenon (chromosomal organization). Another aspect to take into account in phylogenetic reconstructions, and which complicates the analysis, is the distinction of hemiplasy due to the phylogenetic sorting of a genetic polymorphism Robinson et al. 2008).

Discussion
The aim of this work is to review the molecular cytogenetic data available in the literature for any major lineages of the platyrrhine tree (Table II) while also considering useful classical cytogenetic data. We report human associations and evolutionary rearrangements characterizing the principal nodes of the Neotropical primate tree. The tree adopted is the one proposed by Perelman et al. (2011), in terms of the families recognized, but with some modifications made in order to take into account chromosomal data at the inter-and intrageneric levels. Furthermore, we also point out principal cytogenetic data that are in conflict with molecular data. In particular, the New World monkey data gathered are discussed below, for each family (Cebidae, Atelidae and Pithecidae) and for each genus within it, following the tree reported in Figure 2(a-c). This phylogenetic tree was drawn using Mesquite, a software program for evolutionary biology designed to help biologists organize and analyze comparative data. A previous review of the same topic (De Oliveira et al. 2012) has been published, explaining conflicting features through traditional interpretative hypotheses, taking into account the distinction between homologies and homoplasy. In addition to this, we explain the possible evolutionary scenarios, considering hemiplasy in addition to homology and homoplasy. Some discordance in evolutionary interpretation can occur when a tree constructed through chromosomal data is not in accordance with a species tree due to the phylogenetic sorting of a genetic polymorphism; this kind of evolutionary event is termed hemiplasy.

Cebidae family
Cytogenetic analysis of the data found in the literature shows low variability within and between species, in both the number and structure of chromosomes in the Cebidae family (Dutrillaux & Couturier 1981;Seuánez et al. 1988;Nagamachi et al. 1997aNagamachi et al. ,b, 1999, showing highly conservative genomes with diploid numbers of chromosomes ranging between 44 and 54. This has led researchers to assume that the adaptive radiation of Cebidae was characterized by a limited number of chromosomal rearrangements. Chromosomal painting data enabled us to show that all of the syntenies in the putative ancestral platyrrhine karyotype (i.e., 3a/21, 5/7a, 2b/16b, 8a/ 18, 14/15a, and 10a/16a) were conserved in Cebus, and Mico. C. capucinus, C. albifrons and even Sapajus apella, previously known as C. apella (Richard et al. 1996;Garcia et al. 2002;Amaral et al. 2008), share a pericentric inversion of a submetacentric chromosome formed by 14/15a association, resulting in the form 14/15a/14. In particular, C. capucinus presents the most conserved karyotype among all Platyrrhini (Richard et al. 1996;Garcia et al. 2002;Amaral et al. 2008), while the other gracile/un-tufted Cebus and the robust-tufted Sapajus species are more derived (Figure 2(a)). The karyotype of C. albifrons differs from that of C. capucinus by another pericentric inversion in the 14/15a human association which results in a metacentric chromosome with 15a/14/ 15a/14 in tandem, and by a fusion followed by a pericentric inversion involving the homologous-tohuman chromosomes 15b and 8b (8/15/8) (Amaral et al. 2008). The C. olivaceus (also known as C. nigrivittatus) and S. apella subspecies group are linked by a chromosomal inversion homologous to human synteny 20; they differ through another diverse pericentric inversion in the association 14/ 15a/14, resulting in a metacentric chromosome, and an apomorphic robertsonian rearrangement in the chromosomes homologous to human 12 and 15b (12/15) in C. olivaceus; the interchromosomal rearrangements mentioned above (not shown in Figure 2) in C. albifrons and C. olivaceus (2n = 52) explain their different diploid number when compared with that of the other Cebus capuchin species (2n = 54) (Richard et al. 1996;Garcia et al. 2002;Amaral et al. 2008).
Chromosome painting on Saimiri sciureus shows the 2a/15b human association that could represent a link (synapomorphy) between marmosets and tamarins (Figure 2(a)), Dumas et al. 2005Dumas et al. , 2007. Moreover, all species of the Saimiri genus possess the same diploid number of chromosomes, 2n = 44, although pericentric inversions characterize three geographically distinct karyotypes (Ma et al. 1974;.
Chromosomal painting data on both marmosets and tamarins (callitrichids), specifically Callithrix jacchus (Sherlock et al. 1996;Neusser et al. 2001), Cebuella pygmaea, Mico argentatus , Saguinus oedipus  and Leontopithecus chrysomelas (Gerbault-Serreau et al. 2004), allow researchers to show the chromosomal associations phylogenetically linking these species (13/17/20, 13/9/22, 2a/15b), later confirmed by reciprocal chromosome painting (Dumas et al. 2007) (Figure 2(a)). In particular, comparative analysis permits us to identify S. oedipus as the sister group of the remaining callitrichids, having human synteny 1a and 10b not fused ); the other callitrichid species are Table II. List of platyrrhine species analyzed using a molecular cytogenetic approach and considering the principal published classical cytogenetic data. Chromosome painting studies with human probes, reciprocal chromosome painting, and multidirectional probes (New World monkey probes used are reported in parentheses) are highlighted in bold. Notes: Homo sapiens (HSA), Aotus nancymaae ( linked by the 1a/10b association. Particularly noteworthy is Mico argentatus, a species of the newly recognized genus, which shows the same chromosomal syntenies with respect to Cebuella pygmaea ), but has a large amount of heterochromatin at the terminal ends of two chromosomes; the C. jacchus karyotype differs from that of other species by a single fission. In particular, a comparison of classical cytogenetic data on tamarins, Saguinus and Leontopithecus, shows  that they have similar karyotypes (2n = 46) and differ only by para-and pericentric inversions detected on at least four acrocentric chromosomes (data not shown in Figure 2) (Nagamachi et al. 1997b;Neusser et al. 2001).
The Callimico genuswith one species, Callimico goeldii, whose phylogenetic position has been highly discussedis characterized by a translocation involving the Y chromosome and an autosome. Consequently, males may have a diploid number of 47 or 48 chromosomes Margulis et al. 1995). Chromosome painting permitted researchers to definitely demonstrate that Callimico is phylogenetically linked with Callithrix and Cebuella (marmosets) by sharing the same human chromosomal association 1a/10b , which characterizes all callitrichids (Figure 2(a)), eliminating any previous doubts.
The taxonomy of the Aotus genus of owl monkeys has been debated since it was first described by the Spanish naturalist Félix de Azara in 1802, especially in terms of the number of species and subspecies recognized. Initially, just one species was recognized, Aotus trivirgatus; subsequently, on the basis of chromosomal characteristics and geographical distribution, nine species and four subspecies have been recognized due to the presence of sibling species (species seemingly identical from a morphological point of view but which possess divergent karyotypes). The karyotypes of these species are characterized by many polymorphisms and a variable diploid number, ranging from 46 to 59 chromosomes (Galbreath 1983;Torres et al. 1998), with differences as well between males and females in the diploid number due to a translocation between chromosome Y and an autosome (Ma et al. 1976;Pieczarka & Nagamachi 1988). It has been suggested that the karyotypes of these species originated from the ancestral platyrrhine karyotype (2n = 54), passing through fissions, translocations and inversions (De Boer 1974;Ma et al. 1976;Mudry et al. 1984;Figure 2. Platyrrhine molecular phylogenetic tree, modified from Perelman et al. (2011); this tree, drawn using the Mesquite program, reports human ancestral and new associations characterizing each principal node: Cebidae (a), Atelidae (b), and Pitheciidae (c). Chromosomes are numbered according to their homology with human chromosomes. We used Neusser and colleagues' nomenclature (2001) for segment identification; note that not all associations represent real homologies, but only the ones confirmed by reciprocal chromosome painting. For example, peculiar associations indicated by (*), such as 2a/15b, 10/11, 16a/10a and 13/17/20, that link some taxa, conflict with molecular reconstructions and need to be analyzed through Bacterial Artificial Chromosome mapping in order to test if they constitute true homologies. The inversion breakpoints of the human 10a/16a association, presumably linking Atelidae and Pitheciidae, also have to be better analyzed. Pieczarka et al. 1992Pieczarka et al. , 1993Torres et al. 1998). Chromosome painting was first applied to metaphases of Aotus nancymaae, showing that the karyotype of owl monkeys is highly derived. Successively, two other Aotus karyotypes have been reconstructed: one of a karyomorph (Ruiz-Herrera et al. 2005) as well as that of A. lemurinus griseimembra (Stanyon et al. 2011). These three Aotus samples share the following derived associations: 1/3, 1/16, 2/20, 4/15, 7/11, 10/11, 14/15 twice and 16/22, and the loss of the ancestral New World monkey associations 2b/ 16b, 10a/16a (Figure 2(a)). Presumably the loss of the 10a/16a association occurred by fusion with synteny 22 and a successive inversion to give 10a/22/ 16a. Aotus l. griseimembra has the least derived karyotype, while the karyomorph and A. nancymaae share four derived associations (2/12, 5/15, 9/15, 10/22) indicating a sister-clade relationship between them (Stanyon et al. 2011). A peculiarity arises in the syntenic association 10/11 shown in the karyotypes of Aotus and Callicebus (Pitheciidae) (Dumas et al. 2005); this association could be either a real homology phylogenetically linking the two genera or a homoplastic result of convergent evolution, or even a hemiplasy; thus, further analyses are needed through Bacterial Artificial Chromosome mapping in order to test these possible explanations.
While painting data support the monophyly of Atelidae, they do not help in resolving the branching genera sequence. In support of the previous approach, classical cytogenetic data analysis on Atelidae permitted researchers to identify a derived inversion involving the 10a/16a association linking Alouatta, Brachyteles and Lagothrix, resulting in 16a/ 10a/16a/10a, which was presumably lost in Ateles which instead presents the association 16a/10a/16a; moreover, researchers have formulated an evolutionary tree with four branches in the following order, Alouatta, Brachyteles, Lagothrix, and Ateles, due to two inversions of human synteny 8b linking Brachyteles, Lagothrix and Ateles, and of human synteny 13 linking Lagothrix and Ateles (De Oliveira et al. 2005). These inversions need to be further tested with BAC probes to check whether they share the same breakpoints and can be considered real homologies.
Neither the taxonomy nor the phylogenetic relationships within the genus Alouatta are clear, and there is no agreement among researchers; indeed, from nine to 19 species have been recognized. The classical cytogenetic studies allowed researchers to highlight in this genus a large variation in the diploid number, from 2n = 44 to 2n = 58, and two unusual features: a system of multiple sex chromosomes involving a translocation between Y chromosomes and an autosome Armada et al. 1987;Mudry et al. 1998Mudry et al. , 2001Steinberg et al. 2008), and the presence of various microchromosomes (Lima & Seuánez 1991), probably composed of repetitive DNA. In the Alouatta genus, Alouatta sara and A. seniculus arctoidea were the first New World monkey species to be analyzed through chromosome painting (Consigliere et al. 1996). They are characterized by high chromosomal variability; indeed, the chromosomal rearrangements responsible for the differences between the karyotypes of these two species are two robertsonian translocations, five tandem translocations and five intrachromosomal rearrangements. Later, human chromosomal probes hybridized on metaphases of Alouatta belzebul (Consigliere et al. 1996) permitted researchers to show a less rearranged karyotype than the species mentioned above. Chromosome painting has been performed on more species: A. fusca (guariba) (De Oliveira et al. 2002;Stanyon et al. 2011), A. caraya, A. seniculus macconnelli (De Oliveira et al. 2002) and A. g. clamitans (Stanyon et al. 2011). Through these works, it has been shown that the Alouatta monophyletic group is linked by a Y-autosomal translocation (Y/15b) as well as by the loss of the ancestral association 2b/16b and the presence of the association 3c/15b (except in A. belzebul). Through these studies, it has also been possible to distinguish two species subgroups, one formed by A. caraya and A. belzebul linked by human associations 2a/20, 5b/7a/5a/7a and 4c/16, while A. seniculus arctoidea, A. s. macconnelli, A. sara and A. g. guariba are linked by the 2a/4b/15a2 association and the fission resulting in the double 10a/16a associations (Consigliere et al. 1996(Consigliere et al. , 1998De Oliveira et al. 2002;Stanyon et al. 2011). Moreover, A. seniculus arctoidea, A. s. macconnelli and A. sara are linked by the associations 20/1c, 8b/7a/5a/7a, while A. guariba guariba and A. g. clamitans are linked by the syntenic associations 1/14, 6/15, 7/15, 10/22 and 17/18 (De Oliveira et al. 2002) (Figure 2(b)).
Human chromosome probes hybridized to Chiropotes utahicki and C. israelita gave the same pattern of hybridization, and comparison with the ancestral hypothetical platyrrhine karyotype indicates the Chiropotes karyotype is very conserved (Stanyon et al. 2004). Classical and molecular cytogenetics have also been applied to Pithecia irrorata (2n = 48) and Cacajao calvus rubicundus (2n = 45 in males, 2n = 46 in females) using human and Saguinus oedipus whole chromosome probes (Finotelo et al. 2010). These analyses indicate that the chromosomal differences found among these three taxa are consequences of centric fusions and fissions, pericentric and paracentric inversions, tandem fusions and a Y-autosome translocation; furthermore, these three species are linked by the 2a/10b human association, and Chiropotes and Cacajao are linked by a fission of the ancestral New World association 5/7a (giving association 5a/7a and synteny 5b) and by a fusion leading to the association of human synteny 20/15a/ 14 (Figure 2(c)). It should be noted that the 5/7a fission in pitheciide species and in Atelidae is a homoplasy since it has already been shown that they have different breakpoints (De Oliveira et al. 2005;Finotelo et al. 2010). In addition, an inversion of the human association 10a/16a has been found in Chiropotes, Pithecia and Cacajao through G banding data analysis; this result apparently is a synapomorphic feature linking all pithecids (including Callicebus) (Finotelo et al. 2010), so this association is worthy of further investigation in order to test whether the same breakpoints are shared by the different species.
For Callicebus, from 28 to 32 species have been recognized (Van Roosmalen et al. 2002;Van Roosmalen & Van Roosmalen 2013), with chromosomal diploid numbers ranging from 16 chromosomes in C. lugens  to 50 chromosomes in C. donacophilus pallescens, C. pallescens and C. hoffmannsi (De Boer 1974;Minezawa & Borda 1984;Stanyon et al. 2000;Rodrigues et al. 2001). Callicebus lugens is the species with the most derived karyotype and lowest diploid number of chromosomes found among all primates. Chromosome painting has been applied to six Callicebus species -Callicebus moloch, C. lugens, C. cupreus, C. pallescens, C. d. pallescens and C. personatus (Stanyon et al. 2000(Stanyon et al. , 2003Barros et al. 2003;Dumas et al. 2005;Rodrigues et al. 2011) -showing that fusions are the predominant rearrangements responsible for the karyotype evolution of these species; moreover, it has been shown that three new human associations, 7/15, 10/11 and 22/2, and two inversions involving the 2/22 (22/2/22) and 16/2 human associations (16/2/16/2), characterize the hypothetical ancestral Callicebus karyotype (Figure 2 (c)). Apart from these ancestral associations shared by all of the species analyzed, further comparison permits the identification of other specific associations and arrangements, such as: 12/19 linking all species but C. lugens; 13/17 present in C. cupreus, C. pallescens, and C. d. pallescens; 17/20 association present in all species except C. d. pallescens, C. personatus and C. moloch; 9/7/5a shared by C. cupreus and C. d. pallescens (not reported in Figure 2). These analyses also permitted the demonstration that C. pallescens is a different taxon if compared with C. d. pallescens, and therefore it is possible to assume that they are two different species (Dumas et al. 2005). The data discussed highlight very highly rearranged karyotypes among Callicebus species; however, from the data gathered so far, it has not been possible to find any human syntenic associations, apart from the above-mentioned inversion of human association 10a/16a, linking Callicebus to other pithecids. On the other hand, the 13/17 association found in Callicebus needs further investigation since a similar association has been found in callitrichids (Cebidae), although, as unpublished data suggest, this could be the result of convergent evolution (De Oliveira et al. 2012). Moreover, as mentioned previously, the 10/ 11 associations found in Callicebus could be a cytogenetic link with Aotus (Cebidae), so they must be better analyzed in order to test breakpoints and real homology.

Conclusion
The taxonomic and phylogenetic relationships of Platyrrhini have been difficult to reconstruct on the basis of morphological characteristics because of problems in distinguishing homology from convergence. On the other hand, at the molecular level, difficulties in finding accurate relationships within and among taxa have been probably due to a rapid separation of lineages during radiation and the low number of nucleotide differences between species. Indeed, even if phylogenetic reconstructions agree regarding the identification of three main branches -Pitheciidae, Atelidae, Cebidaetheir relationships are still debated, as are some unresolved nodes at inter-and intrageneric levels as well. In this perspective, classical and molecular cytogenetics are useful tools to help in the phylogenetic reconstruction of New World monkeys.
In this work, we review the molecular cytogenetic data available in the literature for principal nodes of the platyrrhine tree, also considering informative chromosomal banding patterns; this analysis has permitted us to report the main objectives reached so far through this approach and to discuss divergent data in respect to recent molecular claims. The principal issues are listed below: 1. Classical and molecular cytogenetics have shown that Neotropical primates are karyologically significantly variable and derivative with respect to the average of primates (one rearrangement for every 10 million years), but with clear differences between families.
2. Comparative cytogenetics indicates that the biodiversity of this group of species is not entirely known, for example as has been demonstrated with the description of diverse species of owl monkeys (Aotus), howler monkeys (Alouatta) and titi monkeys (Callicebus). This situation may occur because New World monkeys often present a condition of "sibling species". 3. Chromosome painting has permitted researchers to show the monophyly of New World primates since all share the syntenic associations 8a/18, 10a/16a, 2b/16b, 5/7a ) characterizing the hypothetical platyrrhine ancestral karyotype (2n = 54); these results are also supported through reciprocal chromosome painting applied to Lagothrix lagotricha , Saguinus oedipus , Aotus nancymaae (Stanyon et al. 2004), Callicebus pallescens (Dumas et al. 2005), Mico argentatus (Callithrix argentata), Cebuella pygmaea, Callimico goeldii and Saimiri sciureus (Dumas et al. 2007). 4. The Cebus (Cebidae) karyotype is the most similar to that of the hypothetical ancestral platyrrhine, but Chiropotes (Pitheciidae) also shows a conserved karyotype. On the other hand, among Atelidae highly derived karyotypes have been found, expecially in Ateles and Alouatta. Moreover, Callicebus species (Pitheciidae) also show highly derived karyotypes. 5. Tamarins and marmosets (callitrichids, Cebidae) constitute a monophyletic group sharing the following derived chromosomal associations: 13/17/ 20, 13/19/22 and 2a/15b. The 2/15b human syntenic association of marmosets and tamarins has also been found in Saimiri sciureus, indicating a possible link between them; since molecular data instead link Saimiri to Cebus, BAC mapping is required to test whether this 2a/15b association may represent a real synapomorphy, or homoplasy (the result of a convergence event). 6. Chromosome painting has resolved the debate on the phylogenetic placement of Callimico goeldii (Cebidae). The presence of the human association 1a/10b phylogenetically links Callimico to marmosets , in agreement with molecular data. 7. Even if the position of Callicebus among Pithecidae is supported at the molecular level, only an inversion of the 16a/10a ancestral platyrrhine association permits the inference of a cytogenetic link with other Pithecids, while on the other hand the syntenic associations 13/17 and 17/20 present in some Callicebus species could link this genus with callitrichids (Cebidae). As seen in the case of Saimiri, all of these associations (inv 10a/16a, 13/17, 17/20) are worthy of further investigation to test for real homologies, or the presence of homoplasy, or if they could be explained as consequences of hemiplasy. 8. The phylogentic relationships of Aotus among Cebidae has always been controversial when reconstructed through molecular and morphological data. Even from a cytogenetic point of view, no synapomorphies have been detected to link it to the Cebus/Samiri clade or to the callitrichids. On the contrary, the 10/11 associations instead link owl monkeys (Cebidae) and Callicebus (Pitheciidae), even if molecular analyses do not provide similar evidence. In addition, this association needs to be checked in order to verify whether it shares identical breakpoints and could, then, represent a real synapomorphy, or whether it could be homoplasy or hemiplasy. 9. The 10a/16a inversion present in Atelidae and Pithecidae also needs further investigation in order to test whether it shares the same breakpoints and there could thus be a synapomorphy between the two families, or if it could instead be a homoplasy. 10. No human associations have been provided to corroborate the molecular phylogenetic recognition of two new genera, Mico and Sapajus, apart from some intrachromosomal rearrangements. The same is true for Leontocebus (S. fuscicollis), as no molecular cytogenetic mapping has so far been performed on it.
In conclusion, we would like to stress the importance of recognizing hemiplasy because "phylogenetic discordance" due to chromosomal traits could be explained through evolutionary interpretations that take into account homology or homoplasy not only due to convergence or parallelism but also due to polymorphisms in random lineage sorting. For example, the human associations reported above which link Cebidae and Pithecidae (10/11) or Pithecidae and Atelidae (inv 10a/16a), apparently in discordance, could be considered a consequence of polymorphic lineage sorting rather than a contrast, as previously hypothesized (De Oliveira et al. 2012). Moreover due to the demonstrated complexity of the evolutionary radiation in Neotropical monkeys, we emphasize the necessity of employing multidisciplinary and comparative approaches in order to clarify phylogenetic assessments.