Introduction

Titi monkeys, Callicebus Thomas 1903 (Pitheciidae), one of the most speciose platyrrhine genera, are distributed in the tropical forests of the Amazon and Orinoco basins, in the Atlantic forest of northeastern and southeastern Brazil, in the Chaco and in dry forests of Paraguay and Bolivia. The southern limits are the Pilcomayo and Paraguay rivers. The distribution of Callicebus is generally limited by river barriers (van Roosmalen et al. 2002). Callicebus dubius was described as a hybrid form by Hershkovitz (1988) and rearranged by van Roosmalen et al. (2002) as a valid species of the C. cupreus species group. The range of C. dubius is still uncertain. Hershkovitz (1990) assumed the type locality to be the right (east) bank of the Rio Purus, opposite to Lake Ayapuá. Two other species, Callicebus caligatus and Callicebus cupreus, occur in this area (van Roosmalen et al. 2002). Some specimens of C. dubius, deposited in the British Museum, were obtained in nearby Humaitá, a town on the left bank of the Rio Madeira. The holotype is an adult female (skin and skull), deposited in the Field Museum of Natural History, Chicago, number 38886, collected by Carl Lako in June 1931 (van Roosmalen et al. 2002).

According to van Roosmalen and colleagues (2002) the distribution of C. dubius corresponds to the “south of the Rio Ituxí, or maybe even the Rio Mucuím, both right bank tributaries of the Rio Purus, eastern limit the Rio Madeira south of the town of Humaitá, and west to the Rio Purus, southern limit unknown”. This description is partially inconsistent with or at least not logically represented by the map in van Roosmalen et al. (2002) that shows the Rio de las Piedras (Bolivia) as the southern limit, with the Madre de Dios and Madeira defining the eastern limit of the species distribution. Rowe and Martinez (2003), however, have registered Callicebus brunneus in that region. Rowe and Martinez (2003) surveyed titi monkeys and found that their distribution in northern Bolivia is not consistent with the possible southern limit suggested by van Roosmalen et al. (2002) for C. dubius, but the distribution of C. brunneus coincides with reports by Anderson (1997) and Hershkovitz (1990). In addition, Robert Wallace (pers. comm.) has recorded a different species, which is not C. dubius, in the Department of Pando, Bolivia.

The map in van Roosmalen et al. (2002) indicates the Rio Mucuim as eastern limit for C. dubius, although there is a sampling gap between this river and the Rio Ituxí. In this paper we provide additional data on the geographic distribution of C. dubius, confirming (a) its occurrence between the rivers Ituxi and Mucuim (Fig. 1) and (b) the latter river as the eastern limit. We also report an observation of individuals of this species crossing eastward towards the right bank of the Rio Mucuim using a man-made wooden bridge, and discuss the implications of this observation.

Methods

The study area was located on the left bank of the Rio Madeira in the municipality of Canutama, state of Amazonas, Brazil. This is a transitional environment between grassland (‘Campinas’), ‘Campinaranas’, and ‘Terra firme’ forest, with a high density of understory babaçu-palms (Attalea speciosa) and typical savanna (‘Cerrado’) and grassland (‘Campinas’) vegetation in open areas (M. Hopkins, pers. comm.). About 13 km of trail routes were used for a mammal survey for 15 days in April 2007. Rapid survey methods used for mammals (Fonseca, 2001; Young et al., 2003) were applied, including linear transects, search for indirect evidence (tracks, scats/regurgitations, bones, etc.) and interviews with local inhabitants. However, only C. dubius sighting or vocalization areas were used for mapping purposes. One voucher specimen (INPA 5671) of C. dubius was collected to guarantee species identification on the basis of external features (such as pelt color and pattern) according to Hershkovitz (1988, 1990) and van Roosmalen et al. (2002). The site coordinates were obtained with a GPS Garmin unit, and maps were produced using ArcView GIS 3.2.

Results and Discussion

Nine groups of C. dubius were recorded in April 2007 (Table 1), eight of them within the Mucuim-Ituxi interfluvium. This confirms the distribution of this species (Fig. 1, 2) between the Ituxi and Mucuim rivers, as suggested by van Roosmalen et al. (2002) but without proof through locality records. The ninth group was recorded on the right bank of the Rio Mucuim (Table 1). C. dubius occupied areas of ‘Campinarana’ and ‘terra-firme’ open canopy forests covered by palms. During our survey, we did not observe Callicebus stephennashi, confirming its restriction to the area between the left bank of the Rio Ipixuna and the right bank of the Rio Mucuim.

Table 1.

New records of Callicebus dubius. Callicebus dubius invasion on the right bank of the Rio Mucuim, which is in the range of Callicebus stephennashi (Record 1), and Callicebus dubius on the left bank of the Rio Mucuim (Records 2–8).

On 26^th April 2007 we observed several C. dubius during an intraspecific agonistic interaction on an unpaved road that crosses the Rio Mucuim, in the hydrographic basin of the Rio Purus. The animals were on the right bank of the Rio Mucuim, close to the wooden bridge that crosses this river. The event was observed, photographed, and reported by M. Hopkins and P. Assunção, botanical researchers from INPA, during the first Geoma Madeira-Purus Expedition. This observation could imply a possible range expansion of C. dubius into the distributional range of C. stephennashi. Bridge crossing might be a more common event than expected as other primate species have also been observed exhibiting this behavior. On November 2008 a group of more than 10 individuals of Mico intermedius was seen and three of them photographed crossing a small bridge on an unpaved road over the Rio Água Branca (9° 09′ 41.4″ S and 60 °28′ 03.7″ W, road MT-206 connecting Colniza - Mato Grosso State and Machadinho D'oeste, Rondonia State), a small tributary of the Rio Guariba in northern Mato Grosso (I. Theobald, pers. comm.). These bridges over the rivers Mucuim and Ãgua Branca do not exceed 30 and 15 m in length, respectively. These observations indicate that man-made structures may break down natural geographic barriers and thus interfere with biogeographic processes. The implications of such interference, e.g. potential for hybridization or displacement of one species by another, remain to be determined.

Figure 1.

Geographic distribution of Callicebus dubius. The gray area represents van Roosmalen's et al. (2002) description of the distribution (not their map). The hatched polygon is the distribution between the rivers Mucuim and Ituxi confirmed in our study The question marks indicate the inconsistencies in van Roosmalen's and colleagues' map noticed by Rowe & Martinez (2003) and R. Wallace (pers. comm.).

Figure 2.

Sites of new records of Callicebus dubius on the left bank of the Rio Mucuim. The left star represents sites 2 to 8 and the right star represents site 1.