Rondon’s Marmoset, Mico rondoni sp. n., from Southwestern Brazilian Amazonia
- First Online:
- Cite this article as:
- Ferrari, S.F., Sena, L., Schneider, M.P.C. et al. Int J Primatol (2010) 31: 693. doi:10.1007/s10764-010-9422-6
- 264 Views
We describe Rondon’s marmoset (Mico rondoni sp. n.) from the Rio Jamari in the Brazilian state of Rondônia and differentiate the species from other Amazonian marmosets on the basis of morphological, genetic, and zoogeographic characteristics. Mico rondoni sp. n. is a typical bare-eared marmoset of predominantly grayish coloration and unpigmented face and cheiridia, which contrasts clearly with its more brownish and pigmented parapatric congeners Mico melanurus and M. nigriceps. Genetically, the new species is unquestionably distinct from Mico emiliae, with which it had been synonimized previously. The geographic range of the new species is defined as the area delimited by the Mamoré, Madeira, and Jiparaná rivers to the west, north, and east, respectively, and the Serra dos Pacaás Novos, to the south, where it is substituted by Mico melanurus. Depending on its exact limits, this range may cover an area of little more than 100,000 km2. A characteristic of the ecology of Mico rondoni sp. n. is its range-wide sympatry with the similarly sized tamarin Saguinus fuscicollis weddelli, which may be a key factor determining the rarity or absence of the new species from many areas within its geographic range. Its apparent scarcity may make the new species exceptionally vulnerable to the effects of the intense human colonization occurring throughout much of its known range, and provoke cause for concern for its long-term survival.
KeywordsBrazilian AmazoniaCallitrichidaeconservationMico rondoni sp. n.molecular geneticszoogeography
The marmosets (Mico Lesson, 1840, cf. Rylands et al.2000) of the southern Amazon basin constitute one of the most diverse platyrrhine groups, encompassing ≥13 species, 6 of which were described during the final decade of the 20th century (Ferrari and Lopes 1992a; Mittermeier et al.1992; Silva and Noronha 1998; van Roosmalen et al.1998, 2000). Despite 2 major revisions based on morphology (Hershkovitz 1977; Vivo 1991), and a growing body of genetic data (Nagamachi et al.1996; Sena et al.2002; Tagliaro et al.2000), the systematics of the group is still controversial. This is at least partly due to the considerable lacunas in our knowledge of the distribution of the genus in many areas.
The Brazilian state of Rondônia is a case in point. Up until 1990, only 4 localities of Mico were available (Hershkovitz 1977; Martins et al.1987; Vivo 1985) for an area of ca. 240,000 km2. Hershkovitz (1977) originally placed the whole of Rondônia (south of the Madeira) within the distribution of Callithrix (= Mico) argentata melanura, whereas Vivo (1991) identified newly collected material as Mico emiliae, a taxon not recognized by the former author, even at the subspecific level. Before this, Mico emiliae was known only from southern Pará (Rylands et al.1993; van Roosmalen et al.1998, 2000), east of the Tapajós River, the region’s primary zoogeographic barrier (Sena et al.2002).
Ferrari (2004) mapped the distribution of Mico in Rondônia and identified 2 principal zoogeographic barriers, the Rio Ji-Paraná and the Pacaás Novos mountains, which further reinforced the isolation of the local Mico emiliae from the original, eastern population. Given these considerations, we reevaluated the morphology and zoogeography of the marmosets from northern Rondônia, and analyzed sequences of the mitochondrial cytochrome oxidase II gene. The combined results indicate that the population found in the northern-northwestern extreme of the state is clearly distinct from both the eastern form of Mico emiliae and parapatric congeners. This population is thus reclassified as a new species, Mico rondoni sp. n., which is presented here.
Mico rondoni sp. n.
MPEG-21893, adult female, was collected at the type locality by Fernando da Silva Braga and Arlindo Pinto de Souza Júnior on November 25, 1987. Stuffed skin, skull, and skeleton. Deposited in the zoological collection of the Goeldi Museum in Belém, Brazil.
Paratypes (all from the type locality)
MPEG-21366, MPEG-21646, MPEG-21659, MPEG-21660, MPEG-21886, MPEG-21894, MPEG-21896, and MPEG-21897 (adult males); MPEG-21367, MPEG-21647, and MN-28486 (adult females).
Cachoeira Samuel (now the Samuel Hydroelectric reservoir) on the Rio Jamari, a right bank tributary of the Rio Madeira, in the municipality of Candeias do Jamari, Rondônia, Brazil (08°45′S, 63°28′W).
Mico melanurus occurs at a number of sites south of the Serra dos Pacaás Novos and east of the upper Ji-Paraná (Ferrari 2004). National Museum (Rio de Janeiro) specimen 23827 is from the Rio Urupá (Vivo 1985), which originates in the Serra dos Pacaás Novos and flows into the Ji-Paraná at the town of Jiparaná. Weddell’s saddleback tamarin, Saguinus fuscicollis weddelli, which is sympatric with Mico rondoni sp. n. throughout its range, is also found only on the left or west bank of the Rio Ji-Paraná, including some areas south of the Pacaás Novos (Ferrari 2004).
The new species appears to be absent from many sites within the limits of its geographic range, including those at which Saguinus fuscicollis weddelli is present (Fig. 1). Ferrari et al. (1998) recorded Mico rondoni sp. n. at only 3 of the 20 sites they surveyed, whereas they recorded Saguinus fuscicollis weddelli at 13. Mico rondoni sp. n. was also relatively rare where it was syntopic with Saguinus fuscicollis weddelli (Appendix I). The occurrence of Saguinus fuscicollis weddelli at most sites would seem to contradict the idea that they lack habitat appropriate habitat for callitrichids. Mico rondoni sp. n. has been recorded at only 5 sites within the range proposed here, although it likely occurs in other areas (Ferrari pers. obs.).
Description of the Holotype
The face is unpigmented, but presents small spots of melanic pigmentation around the lips and nostrils. The ears are virtually naked, except for short, sparse hairs at the front of the posterior surface, unpigmented on the attached portion and mottled with eumelanin on the free portion. The forehead pelage is blackish-gray, with a small whitish patch at the center, contrasting with the crown.
The nape and mantle are silvery-gray, the latter indistinct from the back and proximal portions of the arms, darkening gradually toward the center of the back, becoming dark silvery-gray. The basal portion of hairs on the nape and mantle is silvery-white, darkened abruptly in the medial band and apex. The proximal portion of the dorsal surface of the arms is similar to the mantle, darkening gradually to blackish-gray on the forearms, darkening further at the wrists and the back of the hands. The lower dorsum is grayish-brown, darker at the tail base, with hairs yellowish-brown basally, darkening abruptly in the medial portion, but lighter again at the apex.
The tail is completely black, except for the presence of sparse, dark brown hairs at the base. The dorsal surface of hands is moderately hirsute, dark gray, with a few yellowish-chestnut and grayish-chestnut hairs. The dorsal surface of the proximal portion of the legs is slightly lighter in color than the lower dorsum, with a diffuse, lighter patch on the side of the thighs. The pelage of the legs darkens gradually toward the ankles, which are blackish toward the heel.
The throat and chest are light gray, lighter than the mantle. The ventrum is yellowish-brown; the pelage of the flanks is light yellowish-brown, virtually indistinct from the thighs. The ventral surface of the thighs is yellowish-brown, darkening to reddish-brown on the shins and to black at the ankles.
The coloration of the majority of the paratypes is similar to that of the holotype. The only infant specimen (MPEG-21892) is distinct from the adults. The lightly colored forehead patch is ampler, and extends sideways to the temporal region; the crown is darker, despite a larger quantity of grayish hairs. The ears are moderately hairy, unlike those of the adult. The pelage is longer on the sides of the neck and in front of the ears; the pelage is distinctly more agouti over all the dorsal surface of the body, especially the back. The extremities are much lighter in color than in the adult; the arms are more uniformly agouti, and lack the darker wrists, ankles, and dorsal surfaces.
Among the adults, specimen MPEG-21366 is considerably darker than all other individuals, with a blacker coloration of the crown, which contrasts markedly with the white forehead patch. There are whitish-gray hairs on the sides of the forehead, in front of the ears; the mantle is dark gray, darkening along the dorsum, to blackish-gray on the lumbar region and legs. The thorax is dark gray, similar to the mantle, darkening to blackish-gray on the abdomen and grayish-brown on the lower ventrum. The ventral surface of the arms is dark gray, gradually getting darker toward the extremities, where it becomes blackish-gray. The ventral surface of the legs is a dark reddish-brown proximally, turning abruptly blackish-brown from the knee to the ankles.
Specimens MPEG-21646, 21660, and 21897 do not have the yellowish tone of the lower back and thighs. Specimens MPEG-21894 and 21897 have slightly lighter mantles. In specimen MPEG-21897, the basal band of the mantle hairs is less well defined. The crown is relatively dark in specimen MPEG-21646. The mantle of specimen MPEG-21659 is more silvery in color than in the other individuals. Specimen MPEG-21896 is slightly more yellowish coloration than other individuals from the type locality.
Additional Chromatic Variation
Additional morphological variants were found in the supplementary material. Specimen MPEG-21889 is a juvenile, and presents a coloration pattern intermediate between that of the infant (MPEG-21892) and adult specimens. The crown is lighter than that of the infant, and the general agouti coloration is less extensive; the extremities of the arms are slightly darker than those of the infant, but much lighter than those of the adults. The coloration of the lower back and legs is more similar to that observed in adults.
Specimens MPEG-21651 and 21655 present a much darker longitudinal dorsal line than all other individuals. In specimen MPEG-21649, the external surface of the thigh is grayish, similar to the coloration of the mantle, and there is a darker patch at the center of the nape of the neck (also seen in MPEG-21653). Specimens MPEG-21650, 21652, and 21887 all have an area of lighter gray between the shoulders. Specimens MPEG-21657, 21658, and 21885 present a lighter gray coloration on the lower dorsum, extending as far as the base of the tail. The ankles of specimen MPEG-21658 are relatively lightly colored. Specimen MPEG-21895 is blackish-brown in color, especially on the legs, which lack the reddish tones observed in all other specimens.
Chromatic Differences Among Taxa
A comparison of Mico rondoni sp. n. with specimens of M. emiliae collected in Alta Floresta, Mato Grosso, west of the Rio Teles Pires, revealed a number of major differences, which included, in M. emiliae, the presence of more ample pigmentation in the circumbuccal region, and hairs on the external surface of the auricular pavilion, with small tufts on the anterior part of the ears. The crown of Mico emiliae is lighter in color, and less distinct. The pelage on the sides of the face and neck is lighter in Mico emiliae, which also has a much darker, well-defined triangular patch on the lower forehead, just above the middle of the eyes, contrasting with the lighter, more discrete patch in M. rondoni sp. n., due to the presence of a large number of grayish hairs. The extremities of the arms of Mico emiliae are similar in color to the proximal portion, rather than contrasting. The back of the hands in Mico emiliae is a light yellowish-brown. The ventral surfaces of the arms and the belly of Mico emiliae are yellowish-beige, whereas the ventral surface of the legs is light yellowish-brown. The ankles of Mico emiliae are a more pronounced yellowish-brown, barely distinct from the more proximal pelage, although the dorsum of the foot is a stronger tone of yellowish-brown than that of the ankle. The base of the tail of Mico emiliae is a distinct yellowish-brown, especially on the ventral (perianal) surface. The pelage of the ventrum and members of Mico emiliae is longer and denser, with more contrast between the dorsal and ventral coloration along the flanks.
The coloration of Mico rondoni sp. n. is also markedly distinct from that of its geographical neighbors, M. melanurus and M. nigriceps (Fig. 4). Mico melanurus is clearly different from M. rondoni sp. n. due to the almost completely pigmented face and ears, and a well-defined yellowish-brown patch on the rump and frontal portion of the thigh. Specimens from Mato Grosso lack the lighter patch on the forehead, which is uniformly blackish. The crown is brown, darker toward the forehead and lighter toward the nape. The dorsum is brown, darkening gradually toward the rump. The proximal portion of the arms is brown, similar to the mantle. The frontal surface of the legs is reddish-brown, shifting to blackish-brown on the posterior surface. The chest and throat are light brown, contrasting with the rest of the ventrum, which is reddish-brown. The ventral surface of the thighs is a lighter reddish-brown, darkening on the shins and ankles. There is a strong contrast between the coloration of the dorsal and ventral surfaces of the trunk.
Specimens from Rondônia (São Francisco do Guaporé) present a more agouti pelage than those from Mato Grosso, and are distinct from both of these specimens and Mico rondoni sp. n. in the broader, lighter apical band of the hairs, which contrasts with the darker medial band. The crown is blackish-brown, which contrasts with the light brown nape. There is a tiny, ill-defined grayish patch on the forehead. The yellowish-brown thigh stripe is larger and more well-defined than that in the specimens from Mato Grosso. The pelage is lighter at the extremities of the members. The back of the hands is yellowish-brown; the dorsal surface of the feet is yellowish-brown, speckled with red hairs. The chest, throat, and ventral surface of the arms are beige, contrasting subtly with the yellowish-brown of the abdomen. The ventral surface of the thighs is like the abdomen, darkening gradually to reddish-brown on the legs and ankles.
The species distributed to the east of the Rio Ji-Paraná, Mico nigriceps, is characterized by an almost completely pigmented face, except for the nasal and supraorbital area. The light-colored forehead patch is outlined more distinctly; the ears have small hairs, denser than those in Mico rondoni sp. n., especially in the front. The crown is blackish, and extends down to the nape and sides of the head. The dorsum presents much darker pelage from the region of the shoulders, hairs with blackish-brown basal and subapical bands. The hairs of the lower back have a distinct light gray apical band. The flanks are yellowish-brown, with less contrast relative to the ventral surface. The back of the hands is light gray, with a more agouti pelage. The legs are yellowish-brown, darkening on the back of the foot; there is no blackish hair on the heels. The throat is light gray, contrasting with the rest of the ventrum, which is light yellowish-brown. The ventral surface of the arms is light brown. The anogenital area is more hirsute, with a strong reddish-brown coloration, which does not extend as far as the base of the tail.
The general coloration of a juvenile of this species (MPEG-22956) is darker than that of the adults, and differs from that of the juvenile specimen of Mico rondoni sp. n. (MPEG-21889) in its darker pelage, which is also less agouti. The crown is blackish; arms are brown colored; back of the hand as in the adult; neck and thorax dark gray, hairs with wide light gray basal band. The ventrum is light reddish-brown. The extremities of the arms are relatively dark; the backs of the hands and feet are reddish-brown, similar to the ventral surface of the legs.
External measurements of Mico rondoni sp. n., holotype and adult paratypes from the Goeldi Museum (MPEG)
Body mass (g)
Length (mm) of:
A comparison of mean measurements recorded for adult Mico rondoni sp. n., with values available for other bare-ear marmosets
Mean body mass in g (n)
Mean length in mm (n specimens) of:
Mico rondoni sp.n.a
A comparison of mean craniometric measurements of adult Mico rondoni sp. n. with those of other bare-eared marmosets
Mean dimension in mm (n specimens)
Brain case width
Width across upper molars
Length of mandible
Height of auricular process
Length of upper molar series
Width across upper canines
Mico rondoni sp. n.
M. emiliae (MT)
M. emiliae (PA)
Ecology and Conservation
Mico rondoni sp. n. presents the characteristic specializations of the digestive tract related to the dietary exploitation of plant exudates, including both chisel-like lower incisors, used to gouge the bark of gum-producing trees, and the relatively complex and enlarged cecum necessary for the digestion of exudates (Ferrari and Martins 1992). At the type locality, the new species has been observed regularly gouging the bark of gum-producing trees with its lower incisors to provoke exudate flow, and feeding on this gum (Lopes and Ferrari 1994). Other data on the ecology of the new species are scant, although it seems likely that it is similar to that of other Amazonian marmosets (Rylands 1986; Tavares and Ferrari 1999).
One fundamental ecological difference in comparison with other species is widespread sympatry with a tamarin (Saguinus fuscicollis weddelli). Sympatry with Saguinus occurs in 2 other marmosets, Mico argentatus (Ferrari and Lopes 1996) and M. melanurus, but is limited to a small portion of their geographic ranges. In the case of Mico melanurus, sympatry with Saguinus fuscicollis weddelli appears to be restricted to a contact zone south of the Pacaás Novos (Fig. 1). Sympatry between Mico argentatus and Saguinus niger is limited to the northern extreme of the Tocantins-Xingu interfluvium. The more widespread distribution of Saguinus niger within this area indicates that the larger tamarin has a competitive advantage, although Mico argentatus may be more successful in disturbed habitat (Tavares and Ferrari 1999).
By contrast with these other marmosets, the range of Mico rondoni sp. n. appears to be wholly encompassed by that of Saguinus fuscicollis weddelli. Saguinus fuscicollis is noted for the formation of interspecific associations with other callitrichids, in particular Saguinus (Heymann and Buchanan-Smith 2000), and also Callimico (Porter 2004). Lopes and Ferrari (1994) observed similar associations at Cachoeira Samuel.
As niche separation in these species is related to body size differences, it may be relevant that Saguinus fuscicollis weddelli is closely similar in size to Mico rondoni sp. n. The tamarin is also known to “parasitize” the gum-feeding sources of Mico rondoni sp. n. (Lopes and Ferrari 1994), which may provide an additional competitive advantage. Mico rondoni sp. n. is relatively rare or absent throughout much of its probable geographic range, which is characterized by extensive deforestation, primarily along the main BR-364 interstate highway. Despite its apparent ability to tolerate habitat disturbance, the new species may be relatively vulnerable to the effects of human colonization. Mico rondoni sp. n. may in fact be the member of the genus most threatened with extinction, although for the time being, we would recommend allocating it to the IUCN (2001) category of data deficient.
The genetic characteristics of Mico rondoni sp. n. are relatively well known owing to the collection of specimens from the type locality during the flooding of the Samuel reservoir. Like other Mico species, the karyotype of Mico rondoni sp. n. is somewhat homogeneous, although a certain amount of variation is seen in some chromosomes (Barros et al. 1990), and Nagamachi et al. (1996) identified relatively large amounts of distal constitutive heterochromatin.
In their electrophoretic study, Meireles et al. (1992) identified distinct polymorphisms on opposite banks of the Rio Jamari. Whereas specimens from the right bank were polymorphic for 5 protein systems, those from the left bank were variable for only 1 system (PGD), indicating that the Jamari has acted as a zoogeographic barrier between local populations of Mico rondoni sp. n. in recent times.
Sena et al. (2002) were able to include a wider sample of species, including Mico emiliae, in their phylogenetic analysis based on the nucleotide sequences of the mitochondrial gene cytochrome oxidase II (COII). For this, we processed samples from 26 individuals belonging to 12 species, including 4 specimens of Mico rondoni sp. n. and 1 of M. emiliae from northern Mato Grosso, via the dideoxyterminal method, resulting in the sequencing of 549 base pairs. We constructed maximum parsimony and maximum likelihood phylogenetic trees and obtained pairwise maximum likelihood distance values among the specimens studied (Sena 1998).
Cytochrome oxidase II provided relatively conservative estimates of divergence among Mico species (Sena et al.2002) in comparison with the mitochondrial D-loop (Tagliaro et al. 1997). Even so, there is a clear pattern of divergence in nucleotide sequences—2.84%—between Mico rondoni sp. n. and M. emiliae, while that between M. emiliae and M. argentatus (1.35–1.37%) is consistent with their geographic proximity.
Perhaps surprisingly, Mico rondoni sp. n. was only slightly less divergent from its geographical neighbor, M. melanurus (2.62%), and similar values were recorded for other western species. This may reflect the relative isolation of Mico rondoni sp. n. populations, due to a combination of its outlying position, and the ecological and zoogeographic barriers that separate it from neighboring species. It would nevertheless be important to include M. nigriceps in future analyses, to clarify the genetic relationships among these geographically proximate species.
The local names for this marmoset—“sagüi-branco” or “macaquinho-branco”—refer to its whitish or silvery coloration (branco = white), in contrast with the similarly sized, but distinctly blackish Saguinus fuscicollis weddelli, with which it is invariably syntopic. This tamarin is known locally as “sagüi-preto” or “macaquinho-preto” (preto = black). In English, the species should be referred to as Rondon’s marmoset.
In the absence of any distinctive traits, the new species is named after the Brazilian explorer Cândido Rondon, who commanded the first terrestrial expeditions through southwestern Amazonia—traversing the geographic distribution of Mico rondoni sp. n.—in the early 20th century. The new species is endemic to the Brazilian state—Rondônia—which was named after Rondon in 1958, so its name also alludes to its geographic distribution.
In addition to its morphological and genetic distinctiveness, the new species is arguably the most geographically isolated member of the genus, with no congeners to the north or west. To the east, the range of Mico rondoni sp. n. is isolated effectively from that of M. nigriceps by the lower Ji-Paraná, which plays an important role in the zoogeography of a number of other platyrrhines, as varied as Saguinus fuscicollis, Cebus albifrons, Chiropotes albinasus and Lagothrix cana (Ferrari 2004; Ferrari et al. 1999; Iwanaga and Ferrari 2002), as well as Callicebus (van Roosmalen et al. 2002).
The southern limit of the range of Mico rondoni sp. n. is less well defined, but appears to be determined by a combination of factors, including both physical (the Serra dos Pacaás Novos) and ecological barriers. A contact zone may nevertheless exist between the new species and Mico melanurus, most probably either to the west of the Pacaás Novos, close to the Rio Mamoré, or to the east, in the vicinity of the town of Jiparaná. At the present time, information on the distribution of either species within this area is sketchy and, though probable, contact between their ranges has not been confirmed.
By contrast, Saguinus fuscicollis weddelli appears to be more widespread within this area than either marmoset species (Ferrari 2004). Even so, the identification of the exact limits of the distribution of the 2 species, and associated ecological and genetic patterns, should be a priority for future research.
Priority should also be given to the identification of the populations of Mico rondoni sp. n. existing within the limits of its geographic range, as proposed here, and the collection of more detailed data on their ecological characteristics, in particular, their relationship with sympatric Saguinus fuscicollis weddelli. As suggested in the preceding text, the combination of interspecific competition with Saguinus fuscicollis weddelli, a limited distribution within a relatively small range, and ongoing anthropogenic habitat disturbance within this range may pose a significant threat to the survival of the new species over the long term. It is thus essential to obtain more detailed information not only to define better the conservation status of the new species, but also to contribute to the development of management strategies.
Fieldwork was supported by Tecnosolo S.A./DHV Consultants, SEDAM-Rondônia, and the John T. and Catherine D. MacArthur Foundation. Genetic analyses were supported by CAPES. S. F. Ferrari and M. P. C. Schneider thank CNPq for research stipends (302747/2008-7 and 311121/2006-3). We thank Mário de Vivo (Museu de Zoologia), João Alves Oliveira (Museu Nacional), the National Primate Centre, and the Rio de Janeiro Primate Centre for access to specimens. We especially thank Aderson Avelar, Anthony Rylands, Cida Lopes, Ernesto Cruz, Horacio Schneider, Inocêncio Gorayeb, Izaura Muniz, Marcelo Vallinoto, and Simone Iwanaga. Special thanks to Stephen Nash for the drawing of the new species.