Evidence of in situ hybridization between trans-Andean tamarin species (Callitrichidae, Oedipomidas) in northwestern Colombia Authors: Ana María Pérez-Estrada, Iván Darío Soto-Calderón Published: 2026 ABSTRACT We provide morphological and genetic evidence of the first known case of hybridization between the Cotton-top Tamarin (Oedipomidas oedipus) and the Geoffroy’s Tamarin (O. geoffroyi), which was incidentally found in the wild in northwestern Colombia, with evolutionary and conservation implications. Species diversity in primates is usually founded on the Phylogenetic Species Concept, in which the definition of species is based on diagnosable units “as defined in terms of their evolutionary differentiation from other such forms” (Cracraft 1983:169), rather than on reproductive isolation, as the latter is often violated in primates (Groves 2012). Natural hybridization in primates has been widely reported among sympatric or parapatric species in contact zones, especially among recently diverged taxa (Cortés-Ortiz et al. 2007; Zinner et al. 2011; Malukiewicz 2019). In addition to the lack of complete reproductive barriers, interbreeding and genetic introgression between different primate populations may be facilitated by human-induced processes such as recent range expansions due to habitat conversion and climate change, and the introduction of species into areas outside their natural ranges, making primates even more prone to hybridization and genetic introgression (Malukiewicz 2019). Moreover, ex situ hybridization events in primates have also occurred due to poor management conditions and even experimental crosses in controlled environments (Coimbra-Filho et al. 1993; Soto-Calderón et al. 2018; Malukiewicz 2019). In the family Callitrichidae, both natural and anthropogenic hybridization has been documented in Callithrix, Cebuella, and Leontocebus (Alonso et al. 1987; Cheverud et al. 1993; Neusser et al. 2005; Malukiewicz et al. 2014). In particular, there was a recent report of ex situ hybridization between species of Oedipomidas (formerly in the genus Saguinus), consisting of a case of interbreeding between the White-footed Tamarin (Oedipomidas leucopus) and the Cotton-top Tamarin (O. oedipus) and 2 independent crosses of O. leucopus and the Geoffroy’s Tamarin (O. geoffroyi) held in different wildlife refuges (Soto-Calderón et al. 2018). These were the first known hybridization events between recognized species of Oedipomidas, demonstrating the incidence of incomplete prezygotic reproductive barriers between species of this genus. The genus Oedipomidas forms a clade that includes 3 species of trans-Andean tamarins (O. leucopus, O. oedipus, and O. geoffroyi) with parapatric ranges partially delimited by riverine barriers (Groves 2001; Brcko et al. 2022). The species O. oedipus and O. geoffroyi diverged around 1.18 Ma (Lopes et al. 2023) and were considered subspecies of the same species (Saguinus oedipus) until the 1980s, when they were elevated to species status due to their distinct distribution, morphology, and behavior (Hanihara and Natori 1987). Despite their parapatric distribution, natural contact zones of O. oedipus and O. geoffroyi are, to our knowledge, undocumented. However, the introduction of individuals beyond their natural ranges may promote their contact and create opportunities for hybridization. Here, we report the first case of in situ hybridization between Oedipomidas species and provide morphological and molecular evidence for O. oedipus × O. geoffroyi hybridization. We also discuss the possible existence of a contact zone in the wild and the conservation implications of this case of hybridization. Methods Study area and subjects The study site is located in the Tulenapa Reserve, a 150 ha tropical humid secondary forest within the distribution range of O. oedipus, embedded in an extensive matrix of plantain and oil palm plantations. This reserve is part of the Tulenapa Campus of the Universidad de Antioquia in the municipality of Carepa, Antioquia, Colombia (7°46′42″N 76°40′10″W; Fig. 1)—it also once harbored the facilities of an animal refuge managed by a regional environmental authority (Corpourabá), where smuggled wildlife was held for rehabilitation. Intentionally or not, some animals were released into this area after being rehabilitated. Geographic distribution of Oedipomidas oedipus in green (IUCN 2020) and O. geoffroyi in red (IUCN 2019) in northwestern Colombia, and location of the Tulenapa study site (black dot). In 2021, during a population study of cotton-top tamarins, a mixed group was found by chance in the Tulenapa forest. The group consisted of 11 individuals, including 8 cotton-top tamarins, 1 Geoffroy’s Tamarin, and 2 apparent hybrids with pelage color patterns and crest sizes intermediate between these 2 tamarin species. Data collection The group was monitored and baited for 11 wk. Tamarins were captured using a hand-activated multi-compartment trap (Savage et al. 1993). Animals were weighed and sedated with a dose of 12 to 15 mg/kg intramuscular Ketamid (Holliday, Buenos Aires, Argentina). Sex, age class, and body measurements were recorded for each individual. Frontal and lateral photographs of the head and dorsal and lateral photographs of the whole body of the captured tamarins were taken and used for morphological comparisons. Samples of blood (<1% of body mass), tissue (2 × 2 mm from the apical area of the ear), hair (20 to 30 bulbs), and feces (if available) were collected. After sampling, subcutaneous saline solution was administered for rehydration, and the tamarins were released at the same capture site after full recovery. Four individuals of the target group were captured, consisting of an adult female and an infant male of O. oedipus, an O. geoffroyi adult male, and a potentially hybrid juvenile female (FH). Feces were also collected opportunistically from an unidentified group member. The O. oedipus adult female (FO) was deemed to be reproductively active, as evidenced by the voluminous appearance of her genitalia, and she was subsequently tested as the candidate dam of FH. The only O. geoffroyi adult male (MG) was then tested as the potential sire of FH. A set of 8 samples of O. geoffroyi and 19 samples of O. oedipus were also included in this study as references for the 2 species and processed for hybridization analyses. All O. geoffroyi and 1 O. oedipus were confiscated tamarins of unknown origin, and 18 O. oedipus samples were collected in the wild from across their range. Genomic DNA was extracted from tissue samples using a Monarch Genomic DNA Purification Kit (New England Biolabs, Ipswich, Massachusetts, United States), according to the manufacturer’s protocol. Ten microsatellite loci (Table 1), previously optimized for trans-Andean tamarins, were amplified using fluorescent-labeled primers and genotyped as in Soto-Calderón et al. (2018). Raw microsatellite data were visualized and edited using Geneious v.7.1.9 (Biomatters Ltd.), and the allele assignment was performed using FlexiBin (Amos et al. 2007). A 700 bp fragment of the mitochondrial Hypervariable Region I (HVI) was also amplified and sequenced using a Sanger protocol as detailed in Acevedo-Garcés et al. (2021). Table 1Identification of allelic origin in 10 microsatellite loci of the putative hybrid (FH).IndividualSB2SB7SB8SB10SB19SB24SB30SB31SB37CJ12 FO 221221129123234242217219205205144144106106202202182196251249 FH 221223129123234240217205205199144144106100202204182196251253 MG 221223123123242240203205197199138144104100204204182196255253Alleles consistent with inheritance from the candidate dam Oedipomidas oedipus (FO) and the candidate sire O. geoffroyi (MG) correspond to the first and second allele, respectively, for each locus. Identification of allelic origin in 10 microsatellite loci of the putative hybrid (FH). Alleles consistent with inheritance from the candidate dam Oedipomidas oedipus (FO) and the candidate sire O. geoffroyi (MG) correspond to the first and second allele, respectively, for each locus. Data analysis The mean number of alleles, the effective number of alleles and the probability of identity (PI) were estimated using GenAIEx 6.503 (Peakall and Smouse 2006). A paternity analysis was performed matching the multilocus microsatellite genotypes of the individuals of interest to identify maternal and paternal alleles inherited by FH at each locus. A Bayesian assignment test using the microsatellite data was run in the program Structure v2.3.4 (Pritchard et al. 2000) to estimate the ancestry coefficients of FH, defined here as the fraction of the genome assigned to O. oedipus and O. geoffroyi. The test was run with 10 iterations for K number of clusters between 1 and 3, using 50,000 Markov Chain Monte Carlo steps and a 10% burn-in. The Evanno method implemented in StructureSelector (Li and Liu 2018) was used to infer the most likely number of K clusters. In addition, the program NewHybrids (Anderson and Thompson 2002) was used to estimate the probability that FH was assigned to 1 of 6 categories (species 1, species 2, F1, F2, backcross with species 1, and backcross with species 2). Finally, a haplotype network was constructed in PopArt using the median-joining algorithm (Leigh and Bryant 2015) to determine the species origin of the FH maternal lineage. Ethical note This research complied with the Colombian regulations for working with wildlife. The Universidad de Antioquia holds a permit for non-commercial scientific research issued by the Ministry of the Environment (Resolution No. 0524, 27 May 2014). The capture and sampling procedures implemented here were approved by the Animal Research Ethics Committee of the Universidad de Antioquia on 3 May 2013. In addition, since O. oedipus is a critically endangered species, the project was granted the Permit for Collection of Endangered and Endemic Species (8201-2-32320, 19 December 2019). These methods followed the American Society of Mammalogists guidelines to handle live animals (Sikes et al. 2016). Results A comparison of the pelage color and crest patterns of the potential dam FO, the potential sire MG, and the presumptive hybrid FH allowed us to recognize that FH exhibited an intermediate phenotype between the 2 Oedipomidas species (Fig. 2). FH had a white medium-sized crest running from the forehead to the nape as in O. oedipus, without a reddish portion compared to O. geoffroyi. The dorsal pelage of FH showed an intermediate mottled color pattern, while the rump and proximal third of the tail were rusty red in color, more like O. oedipus. The upper arms and the legs of FH exhibited a white pigmentation—similar to that observed in O. oedipus—whereas the color pattern of the facial skin, ventral pelage, and limbs were quite similar in both Oedipomidas species and FH. The phenotype of both potential parents, FO and MG, is typical of O. oedipus and O. geoffroyi, respectively, as there is no significant variation in the pelage characteristics within each species. Phenotypes of the presumptive hybrid and the candidate parents of the species Oedipomidas geoffroyi and O. oedipus. The photographs correspond to frontal and lateral views of the head, and dorsal and lateral views of the entire body. The microsatellite loci were found to be polymorphic, exhibiting a mean number of alleles per locus of 6.9, an effective number of alleles per locus of 4.4 and a combined PI of 8.5 × 10−10 and 2.2 × 10−12 for O. oedipus and O. geoffroyi, respectively. FH shared 1 allele with each putative parent for each of the 10 loci, consistent with a hybridization event between O. oedipus and O. geoffroyi (Table 1). Specifically, FH was heterozygous at the SB10 locus, with alleles that have been identified as being exclusive to O. oedipus and O. geoffroyi, as evidenced by non-overlapping allele ranges in an extensive genetic database of these 2 species (data not shown). The Bayesian assignment test yielded 2 as the most likely K number of gene clusters corresponding to the parental tamarin species (Fig. 3). As expected, MG was assigned to O. geoffroyi (99.4 ± 0.1%), while FO showed 86.6 ± 1.1% of O. oedipus, and FH was 87.2 ± 0.7% of O. geoffroyi and 12.8 ± 0.7% of O. oedipus. The NewHybrids analysis robustly assigned all O. oedipus and O. geoffroyi in our sample to the corresponding species, including the potential parents FO (0.988 O. oedipus) and MG (0.993 O. geoffroyi; Supplementary Data SD1). This analysis also yielded a higher probability of FH being an F1 hybrid (0.572), followed by backcross with O. geoffroyi (0.204), and similar probabilities for pure O. geoffroyi and F2 hybrid (0.084 and 0.087, respectively). Bayesian ancestry assignment using 10 microsatellite loci. Values on the y-axis are ancestry coefficients corresponding to the parental species Oedipomidas oedipus (green) and O. geoffroyi (red). FH is the presumptive hybrid, and FO and MG are the putative dam and sire, respectively. The other individuals correspond to 8 confiscated O. geoffroyi of unknown origin and 19 O. oedipus from various locations within their ranges. The mitochondrial haplotype network showed a clear distinction of O. oedipus and O. geoffroyi haplotypes with 30 mutational steps between them (Supplementary Data SD2). The haplotypes of FH and FO were identical and assigned to O. oedipus, while the haplotype of MG was assigned to the O. geoffroyi cluster. Discussion The morphological and genetic data collected in this study provide evidence for the first documented case of hybridization between the species O. oedipus and O. geoffroyi in the wild. Some of FH’s physical traits exhibit an intermediate phenotype between the parental species, such as a medium-sized crest and an intermediate dorsal color pattern. These traits reinforce the hybrid nature of FH, given the absence of variation in pelage among individuals of each species. Nevertheless, FH’s overall appearance more closely resembles that of O. oedipus, to the point that it could easily be mistaken for a specimen of that species. Similar observations were recorded in hybrids of the Brazilian marmosets Callithrix penicillata and C. geoffroyi, whose pelage exhibited intermediate to parental-like morphologies (Fuzessy et al. 2014). The multilocus genotype of FH and the mitochondrial data are consistent with the assignment of MG and FO as the parents of FH, thereby confirming that FH is a hybrid of O. oedipus and O. geoffroyi. There are several possible scenarios to explain the ancestry of FH (Fig. 4). First, FH may correspond to a first-generation hybrid (F1 hybrid; Fig. 4A) which is supported by the typical phenotypes of O. geoffroyi and O. oedipus of MG and FO, respectively, and the higher probability of F1 assigned to FH by NewHybrids. Genealogical representation of 3 mutually exclusive alternative hypotheses to explain the hybridization history of the hybrid (FH), given the cross of the Oedipomidas oedipus dam (FO) and the O. geoffroyi sire (MG). A) First-generation hybridization (50:50); B) Backcross between an O. oedipus × O. geoffroyi hybrid and O. geoffroyi (25:75); and C) Double backcross between a hybrid and O. geoffroyi (12.5:87.5). As an alternative hypothesis, one could consider the possibility that FH is the product of a backcross between an F1 hybrid and O. geoffroyi (Fig. 4B) which is supported by the Structure analysis that assigned a percentage of O. geoffroyi ancestry to FO greater than 10% and to FH greater than 85%. However, NewHybrids assigned a relatively low probability (0.2) to this scenario. In a third possible scenario, FH could be the offspring of a double backcross with O. geoffroyi, as the observed ancestry proportions of FH (87.2% O. geoffroyi and 12.8% O. oedipus) resemble those expected for a second-generation backcross (87.5:12.5; Fig. 4C). Nonetheless, this is an unlikely explanation, as FO would have to show a predominant O. geoffroyi ancestry (75%), which contradicts both genetic and phenotypic evidence. The most plausible explanation for this observation would be that FH is a first-generation hybrid (as postulated in the aforementioned first hypothesis) exhibiting an expected 50:50 ancestry ratio but empirically biased toward O. geoffroyi. Given the use of only 10 microsatellite markers and the expectation that the 2 parental species have shared ancestral polymorphisms due to their recent diversification, these markers may have limited power to assign ancestral proportions with respect to the 2 tamarin species. This explanation differs from the other 2 hypotheses in that it aligns with previous empirical evidence of reproductive isolation between trans-Andean tamarin species, which was indicated by an O. oedipus × O. leucopus hybrid that was an XY female presumably affected by gonadal dysgenesis (Soto-Calderón et al. 2018). Consequently, a greater number of markers would be required to obtain more accurate ancestry estimates and to discern whether reproductive barriers may be operating between O. oedipus and O. geoffroyi. Since these 2 species have parapatric distributions delimited by a river barrier, there is a possibility that they may have a contact zone along the Atrato River basin, particularly in the headwaters of the river where natural hybridization may be occurring, as it has been documented in saddle-back tamarin subspecies in the Brazilian Amazon (Peres et al. 1996). Current hybridization between O. oedipus and O. geoffroyi in the wild may lead to the formation of a stable hybrid zone with introgression between these species (Cortés-Ortiz et al. 2019a). Natural hybridization of other Callitrichidae species has been reported in contact zones along river barriers and in areas of ecological transition, particularly between different species of marmoset (Callithrix) in Brazil (Malukiewicz 2019). In the case of C. jacchus × C. penicillata hybridization along the São Francisco River, mitochondrial data have suggested a unidirectional gene flow from C. penicillata to C. jacchus, whereas nuclear data indicated bidirectional gene flow between the 2 species (Malukiewicz et al. 2015). Natural hybrid zones have also been documented in different howler monkeys. For instance, evidence of presumably ancient bidirectional hybridization has been found between Alouatta caraya and A. guariba clamitans in narrow contact zones of southern Brazil (Mourthe et al. 2019). Additionally, A. palliata × A. pigra hybrids in Mexico show differential patterns of introgression among mitochondrial, autosomal, and sexual genetic markers (Cortés-Ortiz et al. 2019b). Thus, it is imperative to collect empirical evidence of O. oedipus × O. geoffroyi hybridization in the wild to assess the potential introgression between these species. The hybridization event documented in this study seems to be the result of introducing a specimen of O. geoffroyi into an area within the O. oedipus distribution. This may not be an isolated incident, as the illegal wildlife trade can lead to the translocation and introduction of tamarins to areas outside their natural range, thereby promoting their contact and interbreeding. To illustrate, other hybrids of Oedipomidas species have been identified in an urban park in Cartagena, Colombia, due to their intermediate phenotypes (AMPE and IDSC, personal observation, May 2025), despite being far from the distribution limits of these species. Anthropogenic hybridization between several Callithrix species has been widely reported in southeastern Brazil (Malukiewicz 2019). Many of these cases involve marmosets that were kept as pets and then released into urban areas. Consequently, the Brazilian illegal pet trade has led to the formation of well-established exotic populations of marmosets and the emergence of hybrid swarms. This pattern underscores the importance of finding effective solutions to the wildlife trade as well as carefully managing rescued/smuggled primates and ex situ populations and making thoughtful decisions about reintroduction to avert anthropogenic interspecies contact and hybridization. Hybridization has been shown to alter the genetic composition and average fitness of natural populations, which could have both evolutionary and conservation implications. For instance, different outcomes have been identified in Callithrix hybridization. One hypothesis suggests that hybrids of the jacchus group may be undergoing adaptation to novel habitats in comparison to their parental species. In contrast, experimental work has revealed that jacchus × aurita group hybrids may have reduced fitness due to the greater divergence between the jacchus and aurita groups (Malukiewicz 2019). Although the fitness of the Oedipomidas hybrids is still unknown, their proliferation may have an impact on the conservation of native species which is especially relevant for the persistence of O. oedipus, which is a critically endangered (CR) species endemic to northwestern Colombia (Rodríguez et al. 2021), and a conservation priority in the country (MADS 2024). Even though hybrids and other specimens with taxonomic uncertainty are not specifically considered as conservation targets by the Colombian legislation, natural hybrid zones may constitute an integral component of a historical natural process and warrant alternative management considerations. Further studies are necessary to better understand the evolutionary processes and possible outcomes of hybridization during the speciation of trans-Andean tamarin species. It is also important to evaluate the plausible interaction between these species in a natural contact zone and the impact of human activities on their hybridization. Supplementary Material