Introduction

Amazonia, particularly its western part, is one of the most species-rich regions of the world (for example, Gentry 1988; Voss and Emmons 1996; Myers et al. 2000). The Río Napo, a major tributary of the upper Río Amazonas, has been identified as a center of species richness for four major taxa—vascular plants, amphibians, birds, and mammals (Bass et al. 2010)—indicating the importance of this region for global, regional and local conservation efforts. However, the forests there are coming under increasing threat, particularly due to oil and gas exploration and drilling (Finer et al. 2008; Soto et al. 2010). Like other areas in Peru, the Río Napo region is part of the country-wide concessions for oil drilling (Perú Petro 2007); a continuously growing threat to habitats and species. Primates, particularly the larger species of the family Atelidae, are sensitive to habitat disturbance and fragmentation (Cowlishaw and Dunbar 2000). Even small-scale disturbance such as seismic explorations preceding actual oil drilling may affect their populations (Kolowski and Alonso 2012). Furthermore, bushmeat hunting, which often increases when remote areas become more accessible due to commercial, large-scale exploitation of oil, gas and timber, is also a major threat to primate populations in Amazonia (Peres 1990).

With this background, we conducted a survey of primate populations along the upper Río Curaray, a southern tributary of the Río Napo. This is a relatively remote area (300 km from the city of Iquitos) but may become a focus of oil exploitation. Although a few general or taxon-specific primate surveys have been conducted in the Río Curaray basin (Aquino et al. 2005, 2013; Heymann 2000; Heymann et al. 2002; Kolowski and Alonso 2012), the status of the primate fauna of this area is little known.

Here we present the results of a survey conducted in November and December 2012 on both banks of the upper Río Curaray. We were interested in evaluating the diversity and abundance of primate populations and their relationship to habitat type, and examining whether this river is a species boundary in its upper reaches, as suggested by previous studies on Saguinus and Pithecia (Heymann et al. 2002; Aquino et al. 2009a).

Methods

Study area

The Río Curaray is in the extreme northwest of Peruvian Amazonia, towards the border with Ecuador, and is included in Lote 39 of the oil drilling concession to Repsol Exploración Perú (Perú Petro 2007). So far, forests along both sides of the Curaray show only slight disturbance, mainly due to sporadic logging and subsistence hunting. They will, however, be exposed to the threats emerging from petroleum prospecting and the infrastructure and personnel that accompany it.

Climate data from the nearest meteorological station (Arica, 1°36′01″S, 75°12′01″W, at the confluence of the ríos Nashiño and Curaray; Fig. 1) are available only for the period between December 1976 and July 1982 (SENAMHI 2008). Mean annual rainfall exceeds 2200 mm per year, with January receiving <150 mm per month.

For our censuses, we identified four different areas, two on each bank of the Curaray (Fig. 1; for geographic coordinates and habitat types see Table 1). We distinguished the following habitat types:

High forest (monte alto): vegetation composed almost entirely of trees of generally 20–25 m height, with some emergents above 30 m; open understorey, compact soils. Common tree species: Macrolobium angustifolium (Fabaceae; common name: “pashaco”), Eschweilera spp. (Lecythidaceae; “machimango”), Couma macrocarpa (Apocynaceae; “leche huayo”), Chrysophyllum spp. (Sapotaceae; “caimitillo”), Cedrelinga cateniformis (Fabaceae, “tornillo”), Parahancornia sp. (Apocynaceae; “naranjo podrido”), Pouteria spp. (Sapotaceae; “caimitillo”), Hymenaea courbaril (Fabaceae; “azúcar huayo”) and Vantanea spp. (Humiriaceae; “añuje rumo”). The few palms (Arecaceae) are mainly Astrocaryum murumuru (“huicungo”) and Iriartea sp. (“pona”). This vegetation type is common on low hills and high terraces.

Figure 1.

Location of survey sites on the Río Curaray. Numbers correspond to the list in Table 1.

Low forest (monte bajo): trees with heights of 15 to 20 m, the majority covered by epiphytes and lianas. Dense understorey characterized by the presence of herbaceous plants such as Calathea sp. (Marantaceae; “bijao”), Bactris sp. (Arecaceae; “ñejilla”) and Costus sp. (Costaceae; “cañagria”); includes riparian vegetation. Common tree and liana species are Couroupita guianensis (Lecythidaceae; “ayahuma”), Inga spp. (Mimosaceae; “shimbillo”), Cecropia spp. (Cecropiaceae “cético”), Rheedia sp. (Clusiaceae; “charichuelo”), Annona sp. (Annonaceae; “anona”) and Passiflora sp. (Passifloraceae; “granadilla”). This habitat type is common on low and medium terraces, and is subject to inundation on the lower terraces.

Palmal de altura: dominated by palms of 20–25 m height, intermingled with emergent trees of >30 m height such as M. angustifolium and Eschweilera spp. The most common palms are Oenocarpus bataua (“ungurahui”), Socratea sp. (“huacrapona”), Iriartea sp., Astrocaryum chambira (“chambira”), A. murumuru, Phytelephas macrocarpa (“yarina”), Scheelea cephalotes (“shapaja”) and Scheelea sp. (“shebón”). The understorey is generally open and sometimes abundant in small Lepidocaryum tenue (“irapay”) palms or perennial herbs. This habitat type is found mainly on low hills and high terraces.

Palmal de planicie: dominated (>70% of individuals) by Mauritia flexuosa (Arecaceae; “aguaje”), associated with Mauritiella sp. (Arecaceae; “aguajillo”), Euterpe sp. (Arecaceae; “chonta”) and some fig trees Ficus (Moraceae; “renaco”). Common in medium and low terrace forests. Abundance of stilt roots and frequent flooding with black water can make access difficult. Subject to inundation on the low terraces.

Table 1.

Census areas and their predominant forest types at the Río Curaray.

Varillal: dense vegetation with trees and small trees between 10 and 25 m and few emergents above 30 m. Trees bolt upright with sclerophyllic leaves, similar to typical white-sand forests of the Peruvian Amazon. The soils differ from typical white-sand forest, however, by being sandyclayey and rarely entirely sandy. Emergent trees represented mainly by Parkia spp., Manilkara spp. and Eschweilera spp. Uniquely present south of the Curaray in so-called manchales, located between peaks of the low hills, and on high terraces.

Tree swamp (pantano arbóreo): composed of trees of 20–25 m with an open understorey and ground covered by standing water and abundant stilt roots. The vegetation is dominated by Ficus spp. (“renaco”), intermingled with Tachigali sp. (Mimosaceae; “tangarana”), Guarea sp. (Meliaceae; “requia”), and few palms, mainly Euterpe sp. This habitat type is common on low terraces where it is subject to inundation.

Results

We obtained 304 sightings of 13 primate species. Most were of Lagothrix poeppigii (N = 49), followed by Callicebus discolor (N = 32) and Ateles belzebuth (N = 31); the least sightings were logged for C. pygmaea (N = 8), and Pithecia napensis ¹ (N = 16; Table 2). The smallest groups were those of C. discolor, Aotus vociferans and Pithecia napensis, and the largest were of Saimiri macrodon (previously Saimiri sciureus) and L. poeppigii (Table 2). The range of observed group sizes generally matched those recorded in other areas of northeastern Peruvian Amazonia, except for A. belzebuth which had larger groups than in other areas (Table 2).

Saguinus tripartitus and P. napensis were recorded only north of the Río Curaray, and Saguinus lagonotus, P. aequatorialis and Sapajus macrocephalus (previously Cebus apella macrocephala) only south of the Curaray (Fig. 2). The number of sightings was highest for A. vociferans, L. poeppigii and S. lagonotus, and the highest population densities were those of S. macrodon and L. poeppigii (Table 3).

All primate species combined, the majority of sightings were in palmal alto and high forest; 12 of the 13 primate species were encountered in these habitat types (Table 4). Only four and five species, respectively, were sighted in varillal and tree swamps (Table 4). Atelids and cebids were most frequently observed in high forest, palmal de altura and palmal de planicie. Cebuella pygmaea was encountered exclusively and C. discolor mainly in low forest. Results of the cluster analysis reflect the uneven community composition over habitat types (Fig. 3). Primate communities of varillales and tree swamps cluster closely together and, more distantly, with palmales de planicie. High forest clusters with palmales de altura (Fig. 3). Low forest clearly sticks out, which is due to the lack or scarcity of sightings of large and medium-sized primates (atelids, Cebus, and Sapajus) there, and the frequent sightings of small primates (callitrichids, pitheciids, and Saimiri).

On 30 occasions we saw two species associated with each other. Two-thirds were of squirrel monkeys Saimiri macrodon travelling with the capuchin monkeys S. macrocephalus (13 cases) or C. yuracus (seven cases).

Discussion

The number of primate species encountered during our survey (13) is higher than that reported by Heymann et al. (2002), who did not record A. belzebuth and C. pygmaea. It matches the number of species found in the Manú National Park (Terborgh 1983) and in the Reserva Comunal Tamshiyacu-Tahuayo (now: Area de Conservación Regional Comunal Tamshiyacu-Tahuayo, ACRCTT) (Puertas and Bodmer 1993)². It is higher than the number of primate species in the Yasuní National Park, Ecuador, where 10 species are found (Bass et al. 2010; Marsh 2004). However, not all species occur syntopically, and the maximum number of species at any survey site was 11 (on the south bank of the Curaray). This supports the prediction (Heymann et al. 2002) of a maximum of 10–11 species per site, and is in line with the findings of Palminteri and co-workers, who found a maximum of 10 species out of a pool of 13 at all of their survey sites in southeastern Peru (Palminteri 2010). While on a large (continental) scale, forest cover and rainfall are the major predictors of primate species richness (Peres and Janson 1999), habitat type is strongly predictive on the regional/local scale, with terra firme forests harboring richer primate communities than flooded forests (Palminteri et al. 2011; Peres 1997). Our survey corroborates these findings, with fewer primate species in those habitats that are subject to inundation. The lowest number was found in varillal, which might be explained by the low floristic diversity (and probably productivity) of white-sand forests (Fine et al. 2010; Oñate Calvín 2012).

Figure 2.

Primate communities north and south of the Rio Curaray at the four localities surveyed (see Table 1, Fig. 1). Saguinus tripartitus and Pithecia napensis were observed only north of the river (left bank), and Saguinus lagonotus, Sapajus macrocephalus, and Pithecia aequatorialis were observed only south of the river (right bank).

Figure 3.

Single-linkage cluster analyses of the similarity of primate communities in the different habitat types.

Table 2.

Primate species and their group sizes recorded during transect censuses.

Table 3.

Sighting rates and population density estimates.

Our survey confirms previous observations that the Río Curaray forms a distributional limit for two species each of Saguinus and Pithecia (Aquino and Encarnación 1996; Heymann et al. 2002; Rylands et al. 2011). In line with Heymann et al. (2002), we encountered S. macrocephalus only on the south bank of the Río Curaray. However, in contrast to Heymann et al. (2002), we encountered Cebus yuracus (previously Cebus albifrons yuracus) on both banks. The restriction of S. lagonotus, S. macrocephalus and P. aequatorialis to the south bank of the Río Curaray accounts for the higher number of primate species compared to the Yasuní National Park, located north of the Río Curaray (the eastern border of which is only about 25 km from our survey area).

Table 4.

Number of sightings of different primate species per habitat type.

That the Río Curaray is a barrier is quite surprising, as it is quite narrow (50–100 m wide) and strongly meandering, resulting in frequent river bend cut-offs of small islands that could transfer species from one bank of the river to the other. However, as meanders of the Río Curaray are extremely constricted (see Google Earth, 1°10′S–2°30′S. 74°05′W–75°35′W), these islands may simply be too small to accommodate a population large enough to persist until merging with a population on the opposite bank of the river (Heymann et al. 2002).

Our population density estimates are higher than those obtained for the Río Pucacuro and the upper Río Itaya (Aquino et al. 2000a, 2009b). For L. poeppigii, A. belzebuth and S. lagonotus they are also higher than those obtained by Kolowski and Alonso (2012) in the non-hunted forest of the upper reaches of Quebrada Arabela, about 50 km from our area. Since Kolowski and Alonso (2012) used the number of individuals seen upon encounter rather than complete counts to estimate group size for calculating population densities, their estimates are inherently smaller than ours, even if real population densities were actually very similar. More importantly, the fact that both our density estimates and those of Kolowski and Alonso (2012) are consistently higher than those for the Río Pucacuro and the upper Río Itaya supports the notion that human interference affects primate population densities. This effect is particularly strong for the large atelids which are preferred by hunters (Aquino et al. 2000b; Peres 1990; Puertas and Bodmer 1993), but may also be pertinent for medium-sized and smaller primates (Endo et al. 2010). Being closer to Iquitos (where bushmeat was, and still is, common in the markets Castro et al. 1990), and more accessible than the upper Río Curaray, hunting pressure is much stronger at Río Pucacuro and the upper Río Itaya.

For L. poeppigii and A. belzebuth our estimates are also higher than those for the Yasuní National Park (Dew 2005). Dew obtained his estimates by relating study group size to home-range size, so again results cannot be directly compared. Nevertheless, it is noteworthy that in these two studies and in our study, the density of L. poeppigii was always 2–3 times higher than the density of A. belzebuth. While it is tempting to speculate that interspecific competition might keep population densities of A. belzebuth lower than those of L. poeppigii (Dew 2005, Iwanaga and Ferrari 2002), a reverse pattern, i.e. higher population densities for Ateles, has been reported from four out of five non-hunted sites in the Manú National Park (Endo et al. 2010). Detailed, comparative, long-term, ecological studies and biogeographic analyses are needed to reveal whether populations of Ateles and Lagothrix affect each other, whether local ecological conditions favor one or the other species, or whether historical events or processes are responsible for current patterns.

Amongst the small species (body mass <1 kg), C. pygmaea and C. discolor stick out by either having been recorded exclusively or by strongly prevailing, respectively, in a single habitat type. Cebuella pygmaea is a highly specialized exudativore that prefers floodplain forest (Soini 1982; de la Torre et al. 2000). The only available ecological study of C. discolor (by Carillo-Bilbao et al. 2005) indicates that this species uses mainly the lower canopy and the understorey, which may facilitate its existence in low forest. However, S. tripartitus and S. lagonotus also prefer the lower forest strata (Heymann 2000, Heymann et al. 2002), but do not prevail in low forest. Additional ecological factors must play a role that we have yet to identify. In conclusion, our survey revealed that the upper Río Curaray harbors a species-rich primate fauna, which adds to the recognition of the Río Napo region as one of the most species-rich areas of the world. To conserve this biodiversity, the creation of a protected area that includes both banks of the upper Río Curaray and that adjoins the Yasuní National Park on the Ecuadorian side would be highly desirable.