Recorded monologs

From May to July of 1998, monologs about the natural history of local mammals were elicited from Matses men from four Peruvian Matses villages and recorded on digital minidisk. All monologs were recorded in the Matses language. To elicit these texts, informants were asked to talk about a single mammalian taxon (e.g., a species of monkey, or bats in general), which was mentioned only once by the interviewer (Fleck). Informants were asked to say as much as they liked about any topic relating to the taxon in question, and were not interrupted or asked to continue, regardless of the length of their monolog (see appendix 1 for a transcription and translation of one of these recorded monologs; the appendix also contains an explanation of the Matses orthography used in the accounts). Each informant's monologs were recorded with no other adults present in order to achieve independence of response. For each taxon, the interview was replicated a total of 7 or 8 times with different informants. These recordings were subsequently transcribed and translated by Fleck and literate Matses assistants, and checked for linguistic accuracy with several other Matses speakers. The texts were then checked with speakers other than the narrators to obtain second opinions on the validity of some of the less commonly asserted natural history details.

Sentences in the translated texts were sorted by topic (physical appearance and anatomy, habitat preference, social behavior, vocalizations, daily activities, and food), and then combined to obtain composite essays for each taxon. These essays, which are presented under the heading “Matses natural history” in the accounts that follow, are supplemented by editorial comments (in square brackets) only as necessary to interpret otherwise obscure passages or to identify botanical taxa corresponding to Matses plant names. To provide a more complete ethnographic picture, topics concerning Matses nomenclature, classification, hunting strategies, and cultural significance (summarized under the heading “Ethnobiology” in each species account) were elaborated using data from additional sources, including interview-style question-and-answer sessions and participant-observation, including more than 600 hours that Fleck spent hunting with the Matses.

Plant identifications

Most of the plants mentioned by the Matses in their monologs were identified by various means. Palms were collected by Fleck in 1998 and 1999 from the area surrounding Nuevo San Juan with the help of Matses assistants, who named the palms while in the forest, prior to collection. Palm specimens were identified in the field using published identification guides (Henderson, 1994; Henderson et al., 1995), and voucher material was subsequently deposited at the herbarium at the Museo de Historia Natural de la Universidad Nacional Mayor de San Marcos (Lima) and in the New York Botanical Garden (New York). Other plants, particularly dicotyledonous trees, were identified by Fleck and two Matses assistants at the arboretum of the Instituto de Investigaciones de la Amazonía Peruana. This arboretum is maintained at the Centro de Investigaciones Jenaro Herrera, a forestry research station located about 80 km northwest of Nuevo San Juan, where Swiss botanists have identified all trees >10 cm at breast height (Spichiger et al., 1989, 1990). Plants and trees in the proximity of Nuevo San Juan were also identified by Fleck using identification keys in Gentry (1993).

Botanical taxa corresponding to Matses plant names are provided in square brackets (under the heading “Matses Natural History” for each primate species account) using the following conventions: (1) The generic name alone is provided when the Matses plant name corresponds to all the species in a locally polytypic genus (e.g., bin [Castilla (Moraceae)]). (2) The generic name is followed by “sp.” if the Matses plant name corresponds to just one unidentified local species in a genus (e.g., ichibin [Matisia sp. and Eriotheca sp. (Bombacaceae)]). (3) The generic name is followed by “spp.” if the Matses plant name corresponds to two or more unidentified species, but not to all the local species of that genus (e.g., şhankuin [Pourouma spp. (Moraceae)]). Where alternative technical names for plant families are current in the literature (e.g., Palmae vs. Arecaceae, Clusiaceae vs. Guttiferae), the nomenclature used here follows Gentry (1993).

Voucher material (total  =  3)

Marupa (AMNH 98321, 98610), Nuevo San Juan (MUSM 11108).

Unvouchered observations

Actiamë (Amanzo, 2006), Choncó (Amanzo, 2006), Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993; Heymann and Aquino, 1994), Río Orosa (Freese et al., 1982), Río Tapiche (Bennett et al., 2001), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí-Mirím (Salovaara et al., 2003), Tapiche (Jorge and Velazco, 2006).

Identification

Morphological revisions of howler monkeys have recognized as few as six valid species (Hill, 1962), but as many as 14 might be recognized on the basis of coat-color phenotypes (Groves, 2001; Gregorin, 2006). Cytogenetic and molecular research provides compelling support for some, but not all, of the species recognized in the recent taxonomic literature and underscores our still incomplete understanding of this geographically widespread genus. In the absence of any published synthesis of phenotypic and genetic data, the following remarks summarize the emerging taxonomic consensus, suggest where future research could usefully be focused, and explain our preferred binomen for the red howlers of the Yavarí-Ucayali interfluve.

Phylogenetic analyses of mitochondrial DNA sequence data (Cortés-Ortiz et al., 2003) support the recognition of just two trans-Andean howler species: Alouatta palliata (Gray, 1849), the mantled howler of Mesoamerica and northwestern South America, and A. pigra Lawrence, 1933, the black howler of Guatemala and southeastern Mexico.4 The distinctness of these species, which was first clearly indicated by conspicuous cranial and pelage differences that are maintained in sympatry (Smith, 1970), is additionally supported by karyotypic data (Steinberg et al., 2008) and by patterns of allelic variation at eight microsatellite loci (Ellsworth and Hoelzer, 2006). By contrast, the nominal taxon coibensis Thomas, 1902 (from the Azuero Peninsula and Coiba Island, Panama), which was recognized as a valid species by Groves (2001, 2005), seems to be genetically indistinguishable from A. palliata (see Cortés-Ortiz et al., 2003).

The reciprocal monophyly of several morphologically distinctive cis-Andean species or species complexes is likewise supported by mitochondrial sequence data (Cortés-Ortiz et al., 2003), including: (1) the brownish howlers of the Atlantic Forest, for which the oldest available name is Alouatta fusca (E. Geoffroy, 1812; see Gregorin, 2006); (2) the sexually dichromatic howlers of eastern Bolivia, central Brazil, Paraguay, and northern Argentina, for which the oldest available name is A. caraya (Humboldt, 1812); (3) the red-handed howlers of southeastern Amazonia, for which the oldest available name is A. belzebul (Linnaeus, 1766); and (4) the reddish howlers of northern Colombia, Venezuela, the Guiana Region, and western Amazonia, for which the oldest available name is A. seniculus. The latter complex is the one represented in the Yavarí-Ucayali interfluve.

In addition to their reddish coloration, members of the Alouatta seniculus complex are uniquely distinguished from other simiiform primates by the presence of microchromosomes in all published karyotypic preparations (Consigliere et al., 1996; Sampaio et al., 1996).5 Eight nominal taxa of red howlers referable to the seniculus complex have been regarded as valid by recent authors (table 1). Although subtle differences in pelage color, sexual dichromatism, and hyoid morphology are said to exist among some of these forms (Gregorin, 2006), chromosomal comparisons provide the most compelling evidence for species recognition. Unfortunately, published karyotypes have only been reported for arctoidea (see Stanyon et al., 1995), macconnelli (including “straminea”; Lima et al., 1990; Lima and Seuánez, 1991; Vassart et al., 1996), sara (see Minezawa et al., 1985; Stanyon et al., 1995), and seniculus (see Yunis et al., 1976; Lima and Seuánez, 1991). Among other chromosomal differences described in this literature, a shared Y-autosome translocation in arctoidea, macconnelli, and sara has resulted in a X₁X₂Y₁Y₂/X₁X₁X₂ X₂ sex-chromosome system, whereas seniculus has the standard XY/XX system. Chromosome painting (Consigliere et al., 1996; Oliveira et al., 2002), mtDNA sequence comparisons (Bonvicino et al., 2001; Cortés-Ortiz et al., 2003), and morphometric analyses (Bonvicino et al., 1995) provide additional evidence of genetic divergence among these four taxa.

TABLE 1

Nominal Taxa of Red Howlers in the Alouatta seniculus Complex Recognized as Valid by Recent Authors^a

The red howlers of the Yavarí-Ucayali interfluve are referable to Alouatta juara sensu Gregorin (2006), but the validity of this taxon seems questionable. Gregorin applied the name juara to western Amazonian (Brazilian and Peruvian) specimens collected south of the Amazon and west of the Purus, and he suggested that the name might also apply to unexamined Brazilian material collected north of the Amazon and west of the Negro. However, western Amazonian specimens geographically assignable to juara (sensu Gregorin, 2006) closely resemble typical seniculus in pelage coloration (Hill, 1962) and cranial morphology (Bonvicino et al., 1995). Likewise, the single available karyotype attributed to juara—obtained from an animal collected at Tefé (on the south bank of the Amazon between the Juruá and the Purus)—was found to be “basically the same” (Lima and Seuánez, 1991: 154) as the karyotype of typical seniculus from Colombia. Morphological material that we examined from the Yavarí-Ucayali interfluve falls within the range of phenotypic variation (coat color and craniodental dimensions) present in large comparative series of typical seniculus from northern Colombia; in the absence of any compelling data to the contrary, we conclude that the populations in question are conspecific, and that juara is a synonym of A. seniculus.

Craniodental measurements of the few available specimens of howler monkeys from the Yavarí-Ucayali interfluve are provided in table 2. The only specimen accompanied by external dimensions is MUSM 11108, an adult male that measured 528 × 589 × 137 × 34 mm; unfortunately, no weight was obtained from this animal.

TABLE 2

Craniodental Measurements (mm) of Alouatta seniculus from the Yavarí-Ucayali Interfluve

Ethnobiology

The Matses have only one name for the howler monkey, achu, a word of uncertain etymology and one that is not synchronically segmentable. The pan-Panoan term for howler monkey is du/ru, a word with which the Matses are familiar, but which they do not consider a word in their language and therefore not a true synonym. Two or three varieties are recognized: achudapa “big howler monkey” (or achu piudapa “big red howler monkey”), tsusa (a small howler monkey), and achu çhëşhë “black howler monkey.” The first two varieties, achudapa and tsusa, are recognized by all Matses hunters, but the third, achu çhëşhë (which we have never seen) is not known by all. This darker variety is reported to be red, but its underparts and the insides of its arms and legs are said to be black. It is said to live only along the upper Gálvez. Only those Matses that recognize achu çhëşhë use the term achu piudapa to contrast the red and black varieties. The term tsusa, which is not analyzable, is used to refer to the small variety, which is said to live in larger troops and to be lazier. Some speakers say that achudapa and tsusa sometimes travel together. Achu çhëşhë is said to differ in habitat preference from the other two varieties because it is found in forest where the trees are all short (i.e., the white-sand habitats of the upper Gálvez).

Traditionally, only old people were permitted to eat howler monkeys. If a child or a young person eats howler monkey meat, the Matses believe they will become lazy. The small variety of howler monkey, tsusa, is worse in this respect. For this reason, Matses do not hunt howler monkeys as frequently as they do other monkeys. The lassitude induced by eating howler monkey meat can be cured with frog poison6 and by following a special diet. In smaller villages, where game is still plentiful, howlers are not hunted at all, while in larger villages, howlers are killed more frequently, for older people to eat. Nowadays, despite disapproval from older and more traditional Matses, some younger people are ignoring this dietary taboo. Their justification is that non-Indians eat howler monkeys without any ill effect. Alternatively, some Matses say that only howlers of the tsusa variety must not be eaten. Howler monkey canines are sometimes used to make men's tooth necklaces, and young howlers are sometimes raised as pets.

Matses kill howler monkeys by following the sound of their calls from far away. Once they stop howling, they are harder to locate, so a hunter chants “chimu tawin tsipun tsipun” (this is a phrase from a myth where people turned into howler monkeys by attaching a length of a vine called chimu as tails), as he walks toward the source of the howling, believing this will keep the monkeys from shutting up. When they see people, howlers that have not been exposed to hunting will come closer for a look, in which case they can be shot from the ground. Hunted animals will climb high up into a tree with vine tangles and hide in them. The hunter then climbs up to kill the monkeys with bow and arrow or shotgun. Less frequently, howlers are encountered at a mineral lick and killed there. They are easy to kill because they do not run off. Quite rarely, they are killed as they swim across a river. Matses do not imitate howler monkey calls or use dogs to hunt them.

Matses natural history

Howler monkeys are red, the color of a red brocket deer [Mazama americana], particularly on their backs. Red howler monkeys have large heads, large upper bodies, and small hindquarters. They have a large larynx in their throat, which makes their heads look elongated. They have a long tail that they use to grab onto branches. Their underparts are sparsely haired. Females are smaller than males. Males have long beards and a white scrotum.

Howler monkeys can be found in any primary forest habitat, but they do not come to abandoned swiddens. They are found mostly in floodplain forests and in swamp-palm [Mauritia flexuosa] swamps. They are also found in upland forest far from rivers, but not as frequently. They go to mineral licks to eat mud and drink muddy water, and they also descend to drink water from streams. Otherwise, they do not walk around on the ground. They sometimes forage very high up in trees. They can swim well, and troops cross rivers swimming all together, with the babies clinging onto the adults' backs.

Howler monkeys are found in small troops, always fewer than 10 individuals, counting the young. The troop is led by a large male that has a redder back than the other animals. This male has multiple females as mates. Males carry the young [there is some controversy among informants about this]. Eagles prey on smaller howler monkeys. The large male becomes aggressive when there is an eagle around. Jaguars can kill howler monkeys.

Howler monkeys wake up earlier than other monkeys. Howler monkeys howl saying “yooo” very early in the morning [from around 4 a.m.] until dawn [5–6 a.m.], stopping and then starting up again. The Matses can hear them from the village. They howl when it is going to be a nice day, so when Matses hear howler monkeys howl before daybreak, they know it will be a clear day. They also howl in the late afternoon [3–4 pm] and during full moons. During full moons they wake up to howl, but do not travel or feed. The big male howls first, and then the females and any other males join in. The males' howls are louder and deeper than females'. The troop congregates and may put their arms around each other as they howl. They scream a lot when they are hit with an arrow.

Right from where they slept, howler monkeys howl, defecating at the same place where they howl. At daybreak they start to move through the trees slowly, looking for food. They are lazy, and do not move around much, and when they do, they do so slowly. Because they stay in the same place for a long time, they defecate where they hang out and there are many flies and ants around where they are, and it stinks there. They go to sleep together at dusk in trees that have many vine tangles. They have two or three places to which they always return to sleep. These sleeping sites can be identified by the smell. They always come back to the same mineral licks, where they get all dirty eating mud. They do not go to mineral licks early in the day; they go in the afternoon. One monkey stays in the trees as a lookout, while the others make a hole in the bank of the mineral lick and eat inside the hole. They do not follow a fixed route, but have a fixed territory and always come back to the same sleeping trees and mineral licks.

Their favorite food is swamp-palm [Mauritia flexuosa] fruit. They also eat the fruit of other palms, including isan [Oenocarpus bataua]. They drink the liquid from some unripe palm fruits, like dapais [Attalea phalerata]. Howler monkeys are the only monkeys that eat tsadte [Apeiba aspera (Tiliaceae)] fruits. They also eat many types of dicot tree fruits, including këku [Couma macrocarpa (Apocynaceae)], diden këku [Parahancornia peruviana (Apocynaceae)], machishte [Rhigospira quadrangularis and ?Mucoa duckei (Apocynaceae)], mamuin [Rheedia longifolia (Guttiferae)], okodo mabis [an undetermined species of Guttiferae], moste [Hymenaea spp. (Leguminosae)], achu inkuente [Inga spp. (Leguminosae)], mannan tsipuis [Inga spp. and ?Pithecellobium (Leguminosae)], tankada [Parkia igneiflora, P. multijuga, Pithecellobium auriculatum (Leguminosae)], bin [Castilla (Moraceae)], dadain [Clarisia racemosa (Moraceae)], chiwish [Ficus spp., Coussapoa spp. (Moraceae)], piuşh bëchi [Helicostylis tomentosa and H. elegans (Moraceae)], şhankuin [Pourouma spp. (Moraceae)], bata [Pseudolmedia and Maquira spp. (Moraceae)], kose [Manilkara bidentata (Sapotaceae)], and taëpa [undetermined]. They also eat vine fruits, including poşhodi [Passiflora nitida (Passifloraceae)] and nënë pada [undetermined]. They also eat epiphyte fruits, including nënë pada [undetermined], and many types of young, soft leaves, including leaves of tote [Eschweilera spp. and Lecythis spp. (Lecythidaceae)] and iwise [?Capirona (Rubiaceae)] trees. They eat spiders, crickets [or katydids], beetle grubs, and other invertebrates for which they search in palm crowns, under tree bark, and in rotten wood. They gnaw on rotten wood, especially rotten swamp-palm trunks.

Ateles belzebuth (E. Geoffroy, 1806)

Voucher material (total  =  12)

Nuevo San Juan (MUSM 11109, 11110), Orosa (AMNH 74027–74031), Boca Río Yaquerana (FMNH 88839–88843).

Unvouchered observations

Actiamë (Amanzo, 2006), Choncó (Amanzo, 2006), Divisor (Jorge and Velazco, 2006), Itia Tëbu (Amanzo, 2006), Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993; Heymann and Aquino, 1994), Río Orosa (Freese et al., 1982), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí-Mirím (Salovaara et al., 2003), Tapiche (Jorge and Velazco, 2006). (Note that previous reports of spider monkeys from the Yavarí-Ucayali interfluve identified the local species as either Ateles chamek or A. paniscus for reasons explained below.)

Identification

Kellogg and Goldman's (1944) revision of the spider monkeys recognized four polytypic species, of which two (Ateles fusciceps, A. geoffroyi) were trans-Andean endemics, one (A. paniscus) was restricted to Amazonia, and another (A. belzebuth) included both trans-Andean and Amazonian subspecies. Although Kellogg and Goldman's revision was largely based on pelage traits that some researchers (e.g., Hernández-Camacho and Cooper, 1976) have interpreted as infraspecific geographic variation, recent studies based on other kinds of data (reviewed by Collins, 2008) support the recognition of several valid species. The following paragraphs summarize the empirical basis for recognizing distinct species in Amazonia.

Amazonian spider monkeys consist of four allopatric nominal taxa, variously recognized as valid species or subspecies by modern authors: paniscus Linnaeus, 1758, which occurs north of the Amazon and east of the Rio Negro/Rio Branco; belzebuth E. Geoffroy, 1806, which occurs north of the Amazon and west of the Negro/Branco; chamek Humboldt, 1812, which occurs south of the Amazon and west of the Tapajos; and marginatus E. Geoffroy, 1809, which occurs south of the Amazon and east of the Tapajos. Kellogg and Goldman (1944) recognized chamek as a subspecies of Ateles paniscus, and marginatus as a subspecies of A. belzebuth, but subsequent analyses of allozymic, cytogenetic, morphometric, and sequence datasets provide compelling evidence that paniscus (sensu stricto) is a distinct species, and that chamek is more closely related to belzebuth and marginatus. Although skins of belzebuth (with buffy or whitish underparts) and marginatus (with whitish facial markings) are easily distinguished from those of chamek (which are almost completely black; see below), these three taxa are craniometrically similar (Froehlich et al., 1991), have the same diploid number of chromosomes (Medeiros et al., 1997; Nieves et al., 2005), and have been recovered as a clade in phylogenetic analyses of mitochondrial sequence data (Collins and Dubach, 2000; Collins, 2008). Because molecular analyses that have included representative samples of belzebuth, chamek, and marginatus suggest that these coat-color phenotypes are not reciprocally monophyletic (Collins, 2008: fig. 3.3), and in the absence of any other morphological traits known to be correlated with coat-color differences in this complex, we see no justification for recognizing chamek and marginatus as distinct species (contra Groves, 2001, 2005).

Although some museum skins of spider monkeys from the Yavarí-Ucayali interfluve are completely black (consistent with most published descriptions of the chamek phenotype; e.g., Kellogg and Goldman, 1944; Emmons, 1997), patches of blond or buffy hairs are present along the inner thighs of other specimens (e.g., AMNH 74029, 74031), and the exposed skin of the nose and the central part of the face of large adults is often unpigmented (reddish in life, according to the Matses; see below). Cranial measurements of our material (table 3) fall within the range of variation in homologous dimensions of western Brazilian, northeastern Bolivian, and eastern Peruvian specimens previously reported by Kellogg and Goldman (1944 [as Ateles paniscus chamek]) and Lönnberg (1940a [as A. ater peruviensis]). Most specimens are unaccompanied by weight data, but an adult female from Nuevo San Juan (MUSM 11109) weighed 7400 g.

TABLE 3

External and Craniodental Measurements (mm) of Ateles belzebuth from the Yavarí-Ucayali Interfluve

Ethnobiology

The spider monkey is called çhëşhëid, a term that the Matses analyze as meaning “black one” (probably a valid etymology). The spider monkey is also called by the archaic synonyms mëshe (not analyzable, but apparently containing the prefix më- for “hand”) and çhuna wisu ( =  “black çhuna”; çhuna is one of the archaic names of the woolly monkey). In the language used in the Matses' komok ceremony, spider and woolly monkeys are called şhëmën kudu (these are the only primates to have a name in the ceremonial language). All Matses hunters recognize two varieties of spider monkey, which are said to never occur in the same troop. One is called chëshëidtapa (“big spider monkey”), and the other variety is simply called tsidu, a word that is not analyzable and is only used to refer to this variety of spider monkey. The larger spider monkey subtype has a red nose and central part of its face, and is the “normal” or prototypical type of spider monkey. The tsidu spider monkey is much smaller, skinnier, has an all-black face, has less or no yellow hair on its inner thighs, has a higher-pitched call, lives in much smaller troops, is less common, and is found mostly in deep rainforest very far from rivers. Some tsidu spider monkeys are believed to be demons instead of real animals, and can cause hunters to miss them repeatedly, or, if killed, can make hunters get lost on the way back home. If a tsidu spider monkey makes a hunter get lost, he must dump the dead monkey, and only then will he find his way back home.

The principal economic importance of spider monkeys for the Matses is as food. Spider monkeys and woolly monkeys are the preferred primate game of the Matses. As with sloths and other large monkeys, the arm and leg bones of spider monkeys are broken (to tenderize the meat) before the carcass is bound with plaited palm leaves into a basketlike package, to which a tumpline made of a strip of bark is then attached for carrying. The normal way to cook large monkeys (≥ ca. 2 kg) is to singe the hair off, remove the viscera, cut off the appendages, and boil them in a big pot. Large monkeys are also smoked when much game is killed, or on long (multiple-day) hunting trips. The canines are sometimes used to make necklaces, although the larger teeth of other atelid species are preferred. Young spider monkeys make good pets, but captive-raised animals can become aggressive when they reach adulthood. The Matses believe that hunters cannot eat or touch the intestines of spider monkeys (and other primate game species), lest they lose their marksmanship.

One way that the Matses hunt monkeys (and ungulates) is by walking quickly along forest paths listening for calls and rustling branches; smelling the air; and looking for eaten food, spoor, and other signs of game. How the Matses hunt spider monkeys depends on how wary the local animals are, which in turn depends on how much previous experience a troop has had with hunters. As he walks, every so often a hunter will alternately imitate the calls of spider monkeys, woolly monkeys, and capuchin monkeys, hoping that one species or another will come or at least respond by calling back. If a spider monkey troop that is not wary hears the imitations, the monkeys will respond and come to the hunter. Troops that have had no experience with people will shake branches above the hunter, bark, defecate, urinate and throw branches down at him (which they also do when they see jaguars and anteaters). Or they may simply all come and stare at the person from above. Spider monkeys that are a bit wary will come, but once the large male that leads the troop sees the hunter, it screams and all the monkeys turn around and flee. Troops that are somewhat more wary will respond vocally to the hunter's call but without approaching him, in which case the hunter seeks the monkeys in the direction from which they responded. Troops that are a bit more wary will not respond to loud imitations, so the hunter must call softly and cannot detect wary monkeys that are far away. Very wary troops simply do not respond, and the wariest will hide or move quietly and quickly in the opposite direction when they hear the hunter's call. If no spider monkeys respond, the hunter may try a brown capuchin monkey call, which may make spider (or other) monkeys vocalize, thinking that capuchins have found fruits. The closer to the village a hunter is, the more wary the spider monkeys are likely to be, so as a hunter travels away from the village at first he only listens; then further out he makes low-volume spider monkey calls and/or capuchin monkey calls; and then further out calls spider monkeys at full volume. The hunter may simply imitate spider monkeys (and woolly monkeys) while walking to check if they might answer. Or he may be motivated to try the imitation upon finding dropped half-eaten fruits. Or if he comes across a sleeping tree early in the morning, he may call. When he is ready to turn back to the village, a hunter will make loud spider-, woolly-, and capuchin-monkey calls and listen for a while. In addition to listening and imitating calls, hunters also go to mineral licks (where there may be other game) to kill spider monkeys.

Ideally, the hunter sees the monkeys before they see him. If the troop comes toward the hunter after hearing his imitation (they can be heard approaching because they scream and make noise as they swing through the treetops), then the hunter hides and shoots them when they arrive. If they answer but do not come, the hunter must stalk them. If the hunter is able to catch them by surprise, he shoots the biggest one he can get a clear shot at, and chases a second one after the rest run off. When spider monkeys flee from a hunter, they split up and go in different directions (rather than fleeing all together, as uakari monkeys do). They move through the trees extremely quickly so that the hunter must run at full speed, and try get ahead of one and shoot it as it passes overhead, or he must keep up with one until it tires out. Because females that are carrying young go the slowest, they are the most easily killed. A hunter hunting alone generally kills only one or two spider monkeys from a troop.

The Matses traditionally hunted spider monkeys with bows and arrows, but now they hunt them almost exclusively with shotguns (in some villages bows and arrows are still used when there is no ammunition). Spider monkeys are hard for archers to kill because a shot monkey will pull out the arrow and keep running, dripping blood along the way. If the monkey dies without falling to the ground, the hunter must climb into the canopy to recover the carcass. If a spider monkey falls injured to the forest floor, it will try to bite any person or dog that approaches it.

Matses natural history

Spider monkeys have very long limbs and tails. They hold their tails stretched out, unlike woolly and capuchin monkeys, which coil up their tails as they travel. They use their tails to grab onto things, and when they travel swinging under the branches, they use their tail as an additional limb. They hang by their tails when they eat fruit, sometimes all spread out with limbs and tail grabbing onto separate branches. Their hands have only four fingers—that is why a fork is called “spider monkey's hand” [a recently coined term]. In contrast to their limbs, their heads are small. Their heads have little meat to eat on them, unlike woolly monkeys'. The spider monkey's head hair looks like it was combed forward from the back; this along with its red nose makes it look silly as it sits there scratching its small head. Adult males are larger than females and have more yellow hair on their inner thighs.

Spider monkeys are found in all primary forest habitats, but more often in upland forest than in floodplain forest. Spider monkeys mostly use the middle of the canopy, but often climb higher to forage, escape predators, and sleep. When they hear a tinamou fly, they climb up higher. They do not walk around or forage on the ground, unlike white-fronted capuchin monkeys, except when they go to mineral licks [see below].

Spider monkeys are more common than woolly monkeys. Troop size varies from one to very many [some narrators say up to 40], but 10 to 20 is more usual. A large male leads the troop. Females carry their young on their backs, unless the young are very small, in which case they carry them ventrally. Spider monkeys suckle their young in the same way that people do. Harpy eagles take the smaller animals, and jaguars and pumas may attack them at mineral licks. Spider monkeys defend themselves by grabbing and biting.

Spider monkeys have several different vocalizations. They scream very loudly to communicate over long distances saying “eeeeEEee, eeeeEEee”; they bark like a dog when they are mad saying “aik, aik, aik”; they communicate among themselves with soft vocalizations, saying “oh, oh, oh”; and they scream like people when they are shot. They make a lot of noise rustling branches as they move across the treetops.

When there is good weather, spider monkeys wake up early. When it is rainy, they get up later. They wake up calling loudly, calling their companions who are sleeping nearby, usually in the same tree. Once they are all together, they go in search of food. They travel through the trees screaming. They follow a daily route, traveling far in one day, but do not advance very fast [some say they complete their route in one day, others say it takes them 2–3 days, others say they simply use the same general area with no set route]. When in flight, they travel faster than any other monkey. Large males go the fastest, and females carrying young go slower. They travel swinging under the branches using their arms and tails, except when there are no small branches, in which case they run across the tops of large branches. They are most active in the morning, and they rest when the sun is high and hot. They rest lying on branches, while their young play. In the afternoon they are active again, but not as much as in the morning. When it rains, they sit in sheltered places such as under a tangle of vines or under large-leafed epiphytes.

They sleep in the same big tree every night [some speakers say it is only at one locality, others say there are two or three trees on their route where they sleep]. They sleep very high up, splitting into small groups that huddle together. They defecate and urinate where they sleep. Where they sleep, there are a lot of feces and defecated seeds on the ground. Especially common are the seeds of the isan palm [Oenocarpus bataua]. These seeds can be found sprouting under their sleeping trees. Sleeping sites stink very badly, and the Matses can smell them from far away. Spider monkeys come out and yell at night when there is a full moon.

Spider monkeys go to muddy mineral licks to eat mud and drink muddy water. They always go to the same mineral lick. Although the troop may split up to forage and eat fruits, they go to the mineral licks together as a group. At the mineral lick, some drink the muddy water while others stay up in the trees nearby watching for jaguars and people. Then they switch places. They make a hole in the walls of the mineral lick, like a cave, where they always collect mud. They drink from the mineral lick with their tails wrapped around trees. They descend and ascend via thin trees next to the mineral lick, and these trees are always covered with mud. After climbing back to the trees after drinking from the mineral lick, they sit there a while scratching their heads and eating their lice.

Spider monkeys eat mostly fruits. Large troops may split up to forage. When one spider monkey finds fruits, it starts eating, and then the others come to join it. The one that comes first finishes eating first and rests while waiting for the others to finish. Unlike other monkeys, they eat fruits without peeling them, swallowing even isan palm seeds and other large seeds [up to about the size of an acorn]. They eat palm fruits, especially isan palm fruits [which ripen during the early rainy season] and swamp-palm [Mauritia flexuosa] fruits, which ripen during the dry season]. They also drink the liquid or eat the soft endosperm of unripe niste palm [Iriartea deltoidea] fruits. They also eat a lot of dicot tree and vine fruits, especially këku [Couma macrocarpa (Apocynaceae)], şhankuin [Pourouma spp. (Moraceae)], dadain [Clarisia racemosa (Moraceae)], tonnad [species of Myristicaceae], moste [Hymenaea spp. (Leguminosae)] and chiwish [Ficus and Coussapoa spp. (Moraceae)], which are available during the rainy season. Other dicot tree fruits they eat include: wesnid debiate [Anacardium giganteum (Anacardiaceae)], machishte [Rhigospira quadrangularis and ?Mucoa duckei (Apocynaceae)], diden këku [Parahancornia peruviana (Apocynaceae)], ichibin [Matisia sp., Eriotheca sp. (Bombacaceae)], mamuin [Rheedia longifolia (Guttiferae], okodo mabis [an undetermined species of Guttiferae], mannan tsipuis [Inga spp. and ?Pithecellobium (Leguminosae)], tankada [Parkia igneiflora, P. multijuga, and Pithecellobium auriculatum (Leguminosae)], bin [Castilla (Moraceae)], şhanned [?Brosimum (Moraceae)], piuşh bëchi [Helicostylis tomentosa and H. elegans (Moraceae)], kuşhu tëbin [Naucleopsis mello-barretoi and N. ternstroemiiflora (Moraceae)], bata [Pseudolmedia and Maquira spp. (Moraceae)], mabis mabiskid [Chrysophyllum prieurii (Sapotaceae)], and kose [Manilkara bidentata (Sapotaceae)].

Lagothrix lagothricha (Humboldt, 1812)

Voucher material (total  =  11)

Boca Río Yaquerana (FMNH 88827–88833), Nuevo San Juan (MUSM 11122–11124), Quebrada Esperanza (FMNH 88834).

Unvouchered observations

Actiamë (Amanzo, 2006), Choncó (Amanzo, 2006), Divisor (Jorge and Velazco, 2006), Itia Tëbu (Amanzo, 2006), Jenaro Herrera (Aquino, 1978), Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993; Heymann and Aquino, 1994), Río Orosa (Freese et al., 1982), Río Tapiche (Bennett et al., 2001), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí-Mirím (Salovaara et al., 2003), Tapiche (Jorge and Velazco, 2006). (Note that some recent reports of woolly monkeys from the Yavarí-Ucayali interfluve identify the local species as Lagothrix poeppigii; see below.)

Identification

The genus Lagothrix was last revised by Fooden (1963), who recognized four subspecies of L. lagothricha, of which three occur in lowland Amazonia: L. l. cana (E. Geoffroy, 1812), L. l. lagothricha (Humboldt, 1812), and L. l. poeppigii Schinz, 1844. Fooden noted that these forms are distinguishable only by coat color, explicitly remarking the absence of morphometric and qualitative osteological differences among them. To date, not a shred of additional evidence has been provided to support the notion that cana, lagothricha, and poeppigii are valid species as recently proposed by Groves (2001, 2005). The species epithet of the Amazonian woolly monkey is often misspelled “lagotricha,” but the correct original spelling was validly fixed by Fooden (1963: 227) acting as first revisor in the sense of the International Code of Zoological Nomenclature (ICZN, 1999: Article 24).

The FMNH series from Boca Río Yaquerana and Quebrada Esperanza were examined by Fooden (1963), who identified the local form as Lagothrix lagothricha poeppigii. However, Matses hyponyms for woolly monkeys (see below) suggest that brownish (poeppigii-like) and grayish (cana-like) coat color phenotypes may coexist in the Yavarí-Ucayali interfluve, as they do along the Rio Juruá (Peres, 1993a). External and craniodental measurements of adult vouchers that we examined are summarized in table 4. Most specimens are unaccompanied by weight data, but an adult female (MUSM 11123) from Nuevo San Juan weighed 5250 g, another (MUSM 11122) weighed 6600 g, and an adult male (MUSM 11124) from the same locality weighed 7500 g.

TABLE 4

External and Craniodental Measurements (mm) of Lagothrix lagothricha from the Yavarí-Ucayali Interfluve

Ethnobiology

The woolly monkey is usually called poshto (not analyzable in Matses, but in neighboring Panoan languages the cognate means “[pot] belly.”) Additionally, there are four archaic synonymous terms for woolly monkeys: çhuna (the pan-Panoan term, not analyzable), abukkid “one that is high up,” mëdante chued “one with big hands,” and maksinkid “woolly one.” Two to three subtypes (hyponyms) are usually recognized: poshto piu “red[-brown] woolly monkey,” poshto tanun “gray woolly monkey” (or poshto uşhu “white/light-colored woolly monkey”) and poshto çhëşhë “black/dark woolly monkey.” The poshto piu variety has a red-brown back; the poshto tanun variety has grayish hair, and the poshto çhëşhë variety has contrastingly black venter and insides of limbs. Not all Matses consider poshto çhëşhë to be a distinct class, observing that either red-brown or gray woolly monkeys can have black undersides. Unlike other cases of primate overdifferentiation discussed here, most Matses do not claim that there are additional differences, such as morphological features or habitat preferences, consistently associated with these different color variants. They do, however, assert that different subtypes of woolly monkey do not occur together in the same troop. The Matses do not have a general term for “monkey,” but woolly monkeys can be considered the prototypical monkey, and sometimes the word poshto is used generically to mean all game monkeys, or all large (noncallitrichine) monkeys in general.

All Matses eat woolly monkeys. It is a favorite food, and clearly the most appreciated primate game. Hunters cannot eat or touch the intestines of a woolly monkey (or any other monkey), lest they lose their marksmanship. As with any game species, the first time a hunter kills a woolly monkey, he cannot eat any of it, or he will lose his ability to kill woolly monkeys again. Young people (everyone under about 45 or 50) cannot eat the liver of woolly monkeys, lest their teeth fall out. Women must cook the intestines separately from the rest of the meat, and wash their hands thoroughly after touching the intestines in order to avoid contaminating the men's food. Woolly monkey canines are the preferred teeth for making men's tooth necklaces. Young woolly monkeys are often raised as pets.

Woolly monkeys were traditionally hunted with bow and arrow, but nowadays they are hunted mostly with shotguns. Before firearms, the mark of a top hunter was the ability to kill woolly monkeys, which are often high in the canopy and (because of their robust size) do not die unless hit in the right place. The most prestigious shot for a Matses hunter was to hit a woolly monkey in the eye with an arrow. For this reason, interviewees were all concerned with the fact that woolly monkeys are no longer killed with arrows. Furthermore, narrators expressed a common concern that woolly monkeys are now becoming scarce, especially in the vicinity of large permanent villages (traditionally, the Matses relocated their villages/longhouses every three to seven years).

A single Matses hunter can kill up to four woolly monkeys when he encounters a troop. Hunters imitate woolly monkeys by inhaling while making a very loud “choook” call, with hands cupped around the mouth to amplify the sound. When they find partially eaten fruit on the ground, hunters call to see if any monkey responds. If the woolly monkeys have never been hunted, they respond and come. If they are a bit wary, they respond but do not come. And if they are very wary, they do not respond or come, thought they may respond to the imitated call of a brown capuchin monkey. According to the Matses, larger troops are more confident and therefore more likely to respond than smaller troops. Those troops that have not been hunted, especially larger troops, do not run off when they see hunters. Nonhunted animals may come to look at people, and may shake branches and defecate and urinate on them. Hunters prefer to kill large adult males, and will spend some time trying to get a clear shot at the largest male. After the first shot, the monkeys sometimes split up and run in different directions, and only later rejoin. After killing the first animal, the hunter will follow a second woolly monkey, preferably a male, until it gets tired and stops to rest in a place where the hunter can get a shot. If they stop in an inaccessible place, the hunter will shoo it by shaking vines until it moves on. If a wounded animal will not move to where a hunter can get off a second shot, he will climb the tree to finish it off. A hunter will also climb up to dislodge a dead animal that is stuck up in the tree. Sometimes a woolly monkey falls to the ground while fleeing, in which case it will run on the ground a bit and then climb back up. If they are shot in the eye with an arrow or in the chest with a shotgun they fall to the ground right away. If a second or third animal is killed quickly enough, the hunter may be able to hear another one moving through trees, and will leave the killed monkey and pursue another. (Impressively, hunters always remember where they killed each monkey and take little time to gather them up and pack them into a palm-leaf carrying pack.)

Matses natural history

Woolly monkeys have prehensile tails that they use for hanging onto branches. As they walk across the tops of branches, they hold their tails rolled up behind them. They are grayish or red-brown, and have black faces. They have many fleas in their genital area. Woolly monkeys are good-looking, especially the babies. The babies have sparsely haired underparts when they are first born.

Woolly monkeys are found in upland forest as well as floodplain forest. They prefer upland forest far from rivers, but go to riverside forest when the water level is high. They are most common in areas with good, productive soil. They can be found in hilltop forest and forest on hilly slopes. They use all levels of the canopy, including very high up. They do not walk around on the ground, but may drop to the ground to hide among buttressed roots when an eagle is spotted. They drink water up in the trees, from bromeliads growing in crotches of trees. They tear off the fruiting structure of the bromeliads to get at the water. They do not frequent mineral licks.

Some woolly monkey troops are large, up to 30 or 40 animals; other troops are small; and occasionally a solitary woolly monkey is found. Only females carry the young, the older ones on their backs and the younger ones ventrally. They suckle their young in the same way that people do.

To sleep, woolly monkeys spread out in small groups among the branches of different trees. At dawn they yell “choook” to each other from where they slept and then gather together. Unlike spider monkeys, they do not sleep in a fixed spot, but wherever they are when night falls. They also spread out into subgroups to find fruit, and if they find it, they call and gather to feed together. Large groups feed together, with the babies screaming and making a lot of racket. When they finish the fruits in the tree, they move on without resting, jumping across the branches, to look for more fruits. They travel far, without following a set route, jumping across the branches, noisily rustling the branches, calling to each other, with the young screaming. In the afternoon, when they are full and tired, they rest lying on the branches. The male that is the leader of the troops rests in the lower branches, keeping an eye out for hunters.

Woolly monkeys eat mostly dicot tree and vine fruits. They eat all kinds of dicot fruits, their favorites are këku [Couma macrocarpa (Apocynaceae)], ichibin [Matisia sp. and Eriotheca sp. (Bombacaceae)], şhankuin [Pourouma spp. (Moraceae)], and piuşh bëchi [Helicostylis tomentosa and H. elegans (Moraceae)]. Other dicot fruits they eat regularly include: şhupud [Spondias mombin (Anacardiaceae)], poshton tonte [?Macoubea guianensis (Apocynaceae)], machishte [Rhigospira quadrangularis and ?Mucoa duckei (Apocynaceae)], mamuin [Rheedia longifolia (Guttiferae)], moste [Hymenaea spp. (Leguminosae); this tree/fruit is also called poshton moste “one that is crunched by woolly monkeys”], mannan tsipuis [Inga spp. and ?Pithecellobium (Leguminosae)], tankada [Parkia igneiflora, P. multijuga, and Pithecellobium auriculatum (Leguminosae)], chichombid [Mouriri spp. (Melastomataceae)], shannëd [?Brosimum (Moraceae)], bin [Castilla (Moraceae)], dadain [Clarisia racemosa (Moraceae)], chiwish [Ficus spp. and Coussapoa spp. (Moraceae)], kuşhu tëbin [Naucleopsis mello-barretoi, and N. ternstroemiiflora (Moraceae)], bata [Pseudolmedia and Maquira spp. (Moraceae)], poşhodi [Passiflora nitida (Passifloraceae)], mabis mabiskid [Chrysophyllum prieurii (Sapotaceae)], kose [Manilkara bidentata (Sapotaceae)], poshto kaçho neste [Pouteria (Sapotaceae); this name means “woolly monkey back medicine”], taëpa [undetermined], and nënë pada [an undetermined epiphyte with lianalike stems]. They infrequently eat diden këku [Parahancornia peruviana (Apocynaceae)], and buku [Cecropia spp (Moraceae)]. They drink the fluid inside unripe niste palm [Iriartea deltoidea] fruits. Other palm fruits they eat include isan [Oenocarpus bataua], and swamp palm [Mauritia flexuosa]. They do not feed heavily on swamp-palm fruits, unlike uakari and howler monkeys. They also eat invertebrates including crickets and spiders. They look for insect larvae in rotten wood and under bark. They eat some leaf buds.

Family Cebidae Bonaparte, 1831

Cebids include night monkeys (Aotinae); capuchins and squirrel monkeys (Cebinae); and a distinctive clade of small, clawed species that includes Goeldi's monkey, marmosets, and tamarins (Callitrichinae). Eight cebid species in six genera are definitely known to occur in the Yavarí-Ucayali interfluve.

Subfamily Aotinae Elliot, 1913

Night (or owl) monkeys comprise eight currently recognized species in a single genus (Aotus) that ranges from Panama to northern Argentina (Groves, 2005). Seldom seen by day, and hard to see at night due to their weak eyeshine, night monkeys are more frequently detected by their distinctive vocalizations. Unfortunately, the taxonomy of night monkeys has never been comprehensively revised, and many questions of identification remain. In the absence of reliable criteria for field identification, night monkey identifications are usually justified by making assumptions about geographic range limits. Our attempt to identify collected specimens of night monkeys from the Yavarí-Ucayali interfluve, however, suggests that the current taxonomy is itself based on inadequately documented distributional assumptions.

Aotus nancymaae Hershkovitz, 1983

Voucher material (total  =  11)

Marupa (AMNH 98330, 98331), Nuevo San Juan (AMNH 268238; MUSM 11111–11113), Orosa (AMNH 73701, 73702, 74035), Quebrada Esperanza (FMNH 88868, 88869).

Unvouchered records

Actiamë (Amanzo, 2006), Choncó (Amanzo, 2006), Divisor (Jorge and Velazco, 2006), Jenaro Herrera (Aquino, 1978), Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993; Heymann and Aquino, 1994), Río Cochiquinas (Aquino and Encarnación, 1988), Río Orosa (Aquino and Encarnación, 1988), Río Tahuayo (Aquino and Encarnación, 1988), Río Tapiche (Bennett et al., 2001), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí-Mirím (Salovaara et al., 2003), Tapiche (Jorge and Velazco, 2006). (Note that some early reports of night monkeys from the Yavarí-Ucayali interfluve identified the local species as Aotus trivirgatus; see below.)

Identification

Night monkeys exhibit a limited range of phenotypic variation, mostly in pelage traits, and all of the named forms in this genus were once considered to belong to a single geographically variable species, Aotus trivirgatus (see Hershkovitz, 1949; Cabrera, 1958). However, subsequent karyotypic research by R.A. Brumback and N.S.F. Ma (reviewed by Hershkovitz, 1983) revealed unexpected geographic variation in diploid numbers (2n) and fundamental numbers (FN) that they interpreted as evidence for multiple species. The hypothesis that at least some night monkey karyomorphs represent valid taxa is supported by a variety of data, including correlated pelage traits, comparative serology, and clinical responses to experimental infection with malaria parasites (Hershkovitz, 1983); evidence of sympatry without apparent hybridization in the wild (Pieczarka et al., 1992); substantial mtDNA sequence divergence (Ashley and Vaughn, 1995; Plautz et al., 2009; Menezes et al., 2010); and reduced fertility in at least some hybrid pairings (Kumamoto and Houck, 2001). Together, these and other results (Ford, 1994; Defler and Bueno, 2007) effectively refute earlier notions that Aotus is monotypic, but many taxonomic issues remain unresolved.

The most frequently cited phenotypic character in the recent literature on Aotus concerns the coloration of the fur of the neck. The so-called “red-necked” and “gray-necked” phenotypes are defined by the presence or absence, respectively, of a lateral extension of the reddish ventral coloration along the neck below and behind the ear (Hershkovitz, 1983: fig. 1). Allegedly, the gray-necked forms of Aotus (for which the oldest available name is trivirgatus Humboldt, 1811) occur north of the Amazon, and red-necked forms (for which the oldest available name is azarae Humboldt, 1811) occur south of the Amazon, but Hershkovitz (1983) discussed some exceptions to this geographic pattern and others are indicated by specimens that we examined (see below). Although neck coloration is commonly used as the basis for recognizing species groups of Aotus (e.g., by Hershkovitz, 1983; Ford, 1994; Groves, 2001), recent sequencing studies do not support the reciprocal monophyly of gray-necked and red-necked night monkeys (Plautz et al., 2009; Menezes et al., 2010).

The original description of Aotus nancymaae was largely justified on the basis of karyotypic data, but no morphological voucher material of known geographic origin is apparently available for any chromosomal preparation attributed to this species, and existing vouchers of karyotyped laboratory animals do not consistently exhibit diagnostic pelage traits.7 Instead, only indirect evidence is available to correlate chromosomes with phenotypes. According to Hershkovitz (1983), unvouchered night monkey karyotypes prepared by N.S.F. Ma from blood samples obtained at Peruvian localities on the south (right) bank of the Amazon (including the Yavarí-Ucayali interfluve) all have 54 chromosomes and 72 autosomal arms (2n  =  54, FN  =  72), whereas most karyotypes from the north (left) bank of the Peruvian Amazon have 2n  =  46 and FN  =  60. Hershkovitz (1983) attributed divergent pelage traits to these karyotypes by examining skins from north-bank and south-bank Peruvian localities. He identified the north-bank 2n  =  46 karyotype as belonging to the gray-necked species Aotus vociferans (Spix, 1823), and he described a new species of red-necked night monkey, A. nancymaae (originally misspelled nancymai; see Groves, 2001) for the south-bank 2n  =  54 karyotype. However, an allegedly unique enclave of red-necked night monkeys with 2n  =  54 chromosomes (also referred to A. nancymaae) occurs along the lower Río Tigre, a north-bank tributary of the Amazon (Hershkovitz, 1983: fig. 9). Although Aquino and Encarnación (1988) believed that Aotus nancymaae and A. vociferans are allopatrically distributed, recent reports of sympatry from several localities near Leticia (Pieczarka et al., 1992) and a few specimens of nancymaae from north-bank localities where only vociferans should occur (e.g., AMNH 74382, from Apayacu) suggest otherwise. Future fieldworkers should be alert to the possibility that both species could be present along either bank, where they might occupy adjacent but distinct habitats.

All of the night monkey specimens that we examined from the Ucayali-Yavarí interfluve match the pelage characters attributed to Aotus nancymaae by Hershkovitz (1983). In particular, the reddish ventral coloration extends up the side of the neck below and behind the ear (conforming to the red-necked phenotype of other south-bank Amazonian forms), the interscapular whorl (Hershkovitz, 1983: fig. 1) is absent, the flanks are predominantly grizzled-grayish, and the distal half of the tail is blackish above and below. A distinct middorsal blackish stripe is present on the proximal half of the tail in some specimens (e.g., AMNH 73701, 73702), but not in others (e.g., AMNH 74035, 268238). Craniodental measurements of these specimens (table 5) compare closely with homologous dimensions of topotypic material (Hershkovitz, 1983: table V). Most specimens from the Yavarí-Ucayali interfluve are unaccompanied by external measurements and weights, but an adult male (MUSM 11111) from Nuevo San Juan measured 307 × 343 × 89 × 32 mm and weighed 820 g; an adult female (AMNH 268238) from the same locality measured 308 × 390 × 96 × 32 mm and weighed 804 g.

TABLE 5

Craniodental Measurements (mm) of Aotus nancymaae from the Yavarí-Ucayali Interfluve

Pending a critical revision of Aotus, we follow the currently accepted taxonomy (Groves, 2005) in referring our material to A. nancymaae, but it is not clear that this taxon is really diagnosable from A. miconax Thomas, 1927, a geographically adjacent species from the foothills of the eastern Andes. Hershkovitz (1983) admitted that nancymaae and miconax might be conspecific, and subsequent morphological analyses (Ford, 1994) have failed to discover any phenotypic trait that distinguishes them, but the karyotype of miconax is unknown. Although it would be reasonable to conclude on the basis of morphology that miconax is a senior synonym that should replace nancymaae as the oldest available name for the night monkeys of the Yavarí-Ucayali interfluve, we prefer to maintain existing usage and await karyotypic data from Andean foothill populations to support or refute this hypothesis.

Puertas and Bodmer (1993) suggested that Aotus nigriceps (another allegedly distinct red-necked species) might occur in the Reserva Comunal Tamshiyacu-Tahuayo—from which they also reported A. nancymaae—but noted that the species had only been recorded from vocalizations. In the absence of corroborating details, their identification is difficult to evaluate, but sympatry between congeneric species of night monkeys is not implausible (as noted above). Future primatological fieldworkers in the area should take care to record unusual vocalizations and (ideally) collect voucher specimens of locally co-occurring taxa.

Remarks

One of our voucher specimens (AMNH 268238) was shot at 1500 h on the afternoon of 31 May 1996 as it followed a troop of Saimiri sciureus at a height of 15 m above the ground in secondary upland forest. Two other night monkeys were part of this mixed-species troop, an unusual association not reported by the Matses (see below) nor, apparently, one previously reported in the literature.

Ethnobiology

The Matses call the night monkey dide, but another Matses name for the night monkey is diku. The term dide is the one usually overheard in everyday conversation, and diku is used mostly in myths. There is some controversy among the Matses as to whether diku is a synonym of dide or a less-commonly encountered subtype (different from dide).

The Matses eat night monkeys, but because the Matses traditionally did not hunt at night, and night monkeys are rarely encountered in the day, they were infrequently killed. Now that the Matses hunt at night with flashlights and shotguns (mainly for pacas, deer, tapirs, and caimans), they could kill night monkey more easily, but since night monkeys are small, the Matses are seldom willing to expend a shotgun shell to kill one. The Matses kill night monkeys opportunistically when the animals are detected at dawn entering tree holes, or when a monkey sticks its head out of its hole as a hunter passes by during the day. In the latter case, the hunter will shake vines and saplings to see if the animal exits the hole; if it does, the hunter will shoot from the ground. If the monkey stays inside its hole, the hunter may climb up a neighboring tree, make a bit of noise, and then shoot it when it sticks its head out again. If not, the hunter might climb the tree, plug the hole, and later return later with an axe to fell the tree or enlarge the hole to get the monkey(s) out.

The Matses believe that night monkeys “inform” people with their calls by revealing that large terrestrial game (especially white-lipped peccaries) or a jaguar is approaching or in the vicinity. The Matses consider night monkeys good and helpful because they provide this information.

Matses natural history

Night monkeys have big eyes, flat snouts, and striped foreheads. They have a gray body with a red chest, and a black tail-tip. They are the size of a squirrel monkey.

They live in small troops of about four individuals. They call saying “ii ii ii ii” [the vocalization attributed to night monkeys in traditional Matses ceremonial chants, not a phonetically accurate rendition of actual calls]. They are active at night and sleep in the day. They can see in the dark and forage for fruits in the dark. They sleep in holes in dicot trees and palm trees. They do not always sleep in the same hole. When they see that the day is dawning, they go inside their hole. Then at dusk they come out again.

They eat mostly dicot tree fruits, including diden këku [Parahancornia peruviana (Apocynaceae); this term means “night monkey's këku fruit”; këku is another fruit in the family Apocynaceae]. They surely eat other sweet dicot fruits that other monkeys eat, but it is hard to be certain which ones because they are seldom seen feeding.

Voucher material

None.

Unvouchered records

Nuevo San Juan (D.W. Fleck, unpublished), Río Blanco (Izawa, 1979), Santa Cecilia (Hershkovitz, 1977), Tapiche (Jorge and Velazco, 2006).

Identification

No specimen of Callimico goeldii is known from the Yavarí-Ucayali interfluve, but a skin and skull from Contayo (on the left bank of the upper Río Tapiche; AMNH 98281)8 agrees in all qualitative external and craniodental characters with published descriptions of the species (Thomas, 1904, 1913; Hershkovitz, 1977). The collector, Harvey Bassler (a petroleum geologist; Myers, 2000), recorded no external measurements, and the skull has been bisected longitudinally for anatomical study; as a result, transverse cranial measurements cannot be taken. However, other craniodental dimensions of this specimen, including condyloincisive length (CIL, 41.0 mm) and length of the maxillary toothrow (C1–M3, 15.4 mm) are within the published range of variation for Callimico goeldii (see Hershkovitz, 1977: appendix table 2).

Remarks

No other western Amazonian callitrichine is completely black, so unvouchered reports by competent observers have high credibility. Our sighting at Nuevo San Juan (of a small troop moving through secondary vegetation close to the ground on the right bank of the Río Gálvez) was unambiguous, and several other sight records from the region seem trustworthy.

Izawa (1979: 5) reported that “Callimico occurs as a relatively high population from the head to the upper basin of the Río Blanco, a [right-bank] tributary of the Río Tapiche.” Izawa also observed Callimico along the upper Río Tapiche, but did not specify on which bank(s) his observations were made. (Additional information from Izawa's primatological survey of the Tapiche-Blanco is apparently available in a Japanese-language publication that we have not seen [Izawa, 1978]). Recently, however, Jorge and Velazco (2006) reported Callimico at an inventory site on the right bank of the upper Tapiche.

Hershkovitz (1977: 928) attributed the Santa Cecilia sighting to the late Pekka Soini, whose correspondence (dated 28 June 1970) is still preserved in the FMNH Division of Mammals' archives. In fact, Soini did not explicitly mention Santa Cecilia, but stated that Callimico goeldii “seems to be definitely known … to some natives along the Maniti river.” Santa Cecilia is a small community on the right (east) bank of the lower Manití (Robbins et al., 1991), a site where Soini might plausibly have interviewed local informants. Subsequent primate census work along the Río Manití, however, has not produced additional records of this elusive species (Tapia et al., 1990).

Ethnobiology

A few Matses talk about a monkey called sipi çhëşhë “black tamarin,” which is said to be different from tamarins, extremely rare, all black (including the area around its mouth), and to forage on the ground and very close to the ground. However, most Matses have never seen this species.

Callithrix pygmaea (Spix, 1823)

Voucher material (total  =  12)

Boca Río Yaquerana (FMNH 88997–88999), Nuevo San Juan (AMNH 272828, MUSM 13301), Orosa (AMNH 73751, 74054–74056), Santa Cecilia (FMNH 87135–87137).

Unvouchered reports

Choncó (Amanzo, 2006), Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993; Heymann and Aquino, 1994), Río Aucayo (Castro and Soini, 1977), Río Tapiche (Bennett et al., 2001), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí-Mirím (Salovaara et al., 2003).

Identification

The pygmy marmoset (formerly Cebuella pygmaea) was last reviewed in substantive detail by Hershkovitz (1977), who recognized only Spix's epithet as valid, treating niveiventris 145Lönnberg, 1940, as a strict synonym. However, other authors (e.g., Cruz Lima, 1945; Cabrera, 1958; Napier, 1976; Roosmalen and Roosmalen, 1997; Groves, 2001, 2005) have treated C. p. pygmaea and C. p. niveiventris as valid subspecies distinguishable by ventral fur color. Roosmalen and Roosmalen (1997) mapped these nominal taxa as allopatrically distributed north and south of the Amazon, respectively.

Spix's original material of pygmaea (the holotype, preserved in the Zoologischen Staatssamlung München; Kraft, 1983) was collected at Tabatinga, on the north (left) bank of the Amazon just upstream from the mouth of the Yavarí,9 whereas Lönnberg's material of niveiventris (two syntypes in the Naturhistoriska Riksmuseet, Stockholm) was collected at Lago Ipixuna, about halfway between the Juruá and the Purus on the south (right) bank of the river. We have not examined the type material of either taxon, but specimens from the Yavarí-Ucayali interfluve more closely resemble niveiventris by virtue of their mostly whitish underparts, by contrast with specimens from the adjacent north bank of the Amazon (e.g., AMNH 74360–74370), which have grizzled-brownish underparts (resembling those of typical pygmaea according to Hershkovitz, 1977). Ventral fur color does not, however, consistently distinguish specimens from opposite banks of the Amazon, and we agree with Hershkovitz (1977) that subspecific nomenclature does not usefully describe any currently recognized pattern of geographic variation among museum specimens of pygmy marmosets. Unfortunately, molecular data suggesting a high degree of genetic divergence within C. pygmaea were all obtained from captive individuals of unknown geographic origin (Tagliaro et al., 2000). External and craniodental measurements of specimens we examined from the Yavarí-Ucayali interfluve are provided in table 6. Most are unaccompanied by weights, but an old adult male from Nuevo San Juan (AMNH 272828) weighed 115 g, and a young adult female from the same locality (MUSM 13301) weighed 99 g.

TABLE 6

External and Craniodental Measurements (mm) of Callithrix pygmaea from the Yavarí-Ucayali Interfluve

Ethnobiology

The Matses call the pygmy marmoset madun sipi, which means “demon's tamarin.” The Matses have no synonymous names for pygmy marmosets, and no subtypes are recognized by them.

The Matses do not hunt or kill pygmy marmosets. They do not keep them as pets, since they believe they can cause illness. Recently, some people have caught them to sell to non-Indians as pets. All Matses know about pygmy marmosets, but many have not observed one closely.

Matses natural history

Pygmy marmosets are very cute. They are small and gray. They have a flat face and a mane that makes their head look big. They are the size of a dwarf squirrel [Sciurillus pusillus], but they are most similar to tamarins.

They climb on trees and vines in any forest type. They always return to the same large, smooth-barked trees where they feed. They spend most of their time hanging on the trunks of these large trees. They can be found high or low in the forest. They vocalize saying “pss pss” as if they were giggling. When they see people, they hide on the opposite side of the trunk. They only move about in the daytime.

Their food is the bark of large smooth-trunked dicot trees. [Other speakers simply say they gnaw on the trees without being specific about what they eat, but none mention explicitly that they consume sap.] They rarely or never eat fruits [one speaker said he has seen them eat some fruits, but others disagree].

Genus Saguinus Hoffmannsegg, 1807

Two species of Saguinus (tamarins) occur sympatrically throughout the Yavarí-Ucayali interfluve: the saddleback tamarin, S. fuscicollis, and the moustached tamarin, S. mystax. Although they are morphologically distinctive and are recognized as different species by the Matses, many aspects of their ethnobiology and natural history are similar, so it is convenient to summarize these topics in a joint account.

Ethnobiology

Tamarins (and possibly Goeldi's monkey, see above) are called sipi (probably of onomatopoetic origin). Tamarins are also called by the synonym pishtadan (clearly of onomatopoetic origin, see below). Two to three types are recognized: sipi ëksed “white-lipped tamarin” ( =  moustached tamarin, Saguinus mystax), sipi kabëdi “spotted-back tamarin” ( =  saddleback tamarin, Saguinus fuscicollis), and sipi çhëşhë “black tamarin” ( =  Goeldi's monkey, Callimico goeldii, see above).

Traditionally, Matses hunters occasionally killed tamarins with arrows, sometimes just to test their marksmanship, but nowadays hunters will not expend a shotgun shell to kill one. Boys, however, sometimes kill them with arrows to roast and eat as a snack. They make good pets, so hunters will often try to catch young tamarins to give to their children. When they see a troop of tamarins, they wait until a mother sets her offspring on a branch to feed it, and then they spook the animals by yelling, and by shaking vines and saplings. The youngster will then often either fall or get left behind.

Matses natural history

Moustached tamarins have white teeth and white around their mouths. They are mostly black and have long black tails. Saddleback tamarins have spotted backs, black heads, and black tails. Moustached tamarins are larger than saddleback tamarins.

Moustached tamarins use higher levels of the forest than saddleback tamarins, but neither uses the highest levels of the canopy. Saddleback tamarins sometimes forage on the ground. They can both be found in all habitats, but moustached tamarins are more closely associated with primary forest, where there are many tall trees, whereas saddleback tamarins prefer thick, viney vegetation in secondary forest, in abandoned swiddens, and in old blowdowns. Saddleback tamarins are often seen at the edges of swiddens. [One narrator went further, claiming that moustached tamarins are never found in secondary forest, and that saddleback tamarins are only found in secondary forest.] Saddleback tamarins sometimes come to swiddens where they eat ripe plantains or papayas.

Tamarins live in medium-sized troops of about eight animals. When they see people, they say “pishtadan pishtadan.” When they get separated from the rest of the troop, they say “sii sii sii” and others in the troop answer back. They make a lot of noise when they see people, jaguars, or peccaries. They only move about in the daytime. Both species of tamarin sleep in tree holes or on branches in thick vine tangles.

Tamarins eat ripe dicot tree fruit such as këku [Couma macrocarpa (Apocynaceae)], tinte [Garcinia macrophylla (Guttiferae)], mannan tsipuis [Inga spp. and ?Pithecellobium (Leguminosae)], şhannëd [?Brosimum (Moraceae)], chichombid [Mouriri spp. (Melastomataceae)], piuşh bëchi [Helicostylis tomentosa and H. elegans (Moraceae)], and bata [Pseudolmedia and Maquira spp. (Moraceae)]. When they eat şhankuin fruits [Pourouma spp. (Moraceae)] they do not swallow the seeds (unlike spider monkeys). They never eat unripe fruits, and they eat little or no palm fruits. They also eat crickets [or katydids; the Matses term is taxonomically ambiguous], caterpillars, other insects, and spiders, which they find on the undersides of leaves and elsewhere.

Saguinus fuscicollis (Spix, 1823)

Voucher material (total  =  28)

Marupa (AMNH 98292, 98294–98296), Nuevo San Juan (AMNH 268235, 268236, 272796; MUSM 11102), Orosa (AMNH 73742, 73746, 73748, 73749, 73984, 74038, 74040–74043, 74045, 74046, 74048, 74051, 74053), Quebrada Esperanza (FMNH 88874), San Fernando (FMNH 88873), Santa Cecilia (FMNH 86958, 86964, 86965).

Unvouchered reports

Actiamë (Amanzo, 2006), Choncó (Amanzo, 2006), Divisor (Jorge and Velazco, 2006), Itia Tëu (Amanzo, 2006), Jenaro Herrera (Aquino, 1978), Orosa (Freese et al., 1982); Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993; Heymann and Aquino, 1994), Río Aucayo (Castro and Soini, 1977), Río Tapiche (Bennett et al., 2001), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí-Mirím (Salovaara et al., 2003), Tapiche (Jorge and Velazco, 2006).

Identification

The genus Saguinus was revised by Hershkovitz (1977), who referred specimens from Marupa, Orosa, Quebrada Esperanza, San Fernando, and Santa Cecilia to the subspecies S. fuscicollis nigrifrons (I. Geoffroy, 1850). Our material from Nuevo San Juan is indistinguishable from the specimens examined by Hershkovitz and confirms his inference that a single coat-color phenotype (table 7) occurs throughout most of the Yavarí-Ucayali interfluve.10 External and craniodental measurements from representative specimens are summarized in table 8. Most saddleback tamarin specimens from the Yavarí-Ucayali interfluve are unaccompanied by weight data, but an adult female (AMNH 268236) from Nuevo San Juan weighed 436 g, and an adult male (AMNH 272796) from the same locality weighed 420 g.

TABLE 7

Diagnostic Coat-color Differences Among Subspecies of Saguinus fuscicollis (sensu Hershkovitz, 1977) from the Yavarí-Ucayali and Adjacent Interfluvial Regions

TABLE 8

External and Craniodental Measurements (mm) of Saguinus fuscicollis from the Yavarí-Ucayali Interfluve

Remarks

Currently recognized species limits within the Saguinus nigricollis group of Hershkovitz (1977) are difficult to justify on the basis of recent mtDNA sequence analyses and new information about geographic ranges. As revised by Hershkovitz, this group included two polytypic species—S. nigricollis and S. fuscicollis—which were thought to occur sympatrically in Ecuador and Peru. A third species of this group, S. tripartitus, was recognized by Thorington (1988) on the basis of distributional data suggesting that it occurs (or once occurred) sympatrically with S. fuscicollis (contra Hershkovitz, 1977), and Groves (2001, 2005) additionally recognized S. graellsi and S. melanoleucus (subspecies of S. nigricollis and S. fuscicollis, respectively, according to Hershkovitz, 1977) as valid species. However, recent primate surveys have failed to find any locality where two or more nominal taxa of the S. nigricollis group occur sympatrically (de la Torre, 1996, 2000; Heymann, 2000; Heymann et al., 2002; Rylands et al., 2011)11, and there is at least one well-documented case of natural hybridization between distinct coat-color phenotypes currently ranked as species (Peres et al., 1996). Additionally, S. fuscicollis and S. nigricollis are not reciprocally monophyletic (whether or not tripartitus, graellsi, and/or melanoleucus are also recognized as full species), and pairwise genetic distances are generally lower within this complex than they are among most other currently recognized tamarin species (Cropp et al., 1999; Matauschek et al., 2011).

Pending a taxonomically comprehensive revision of the Saguinus nigricollis group that takes newly discovered facts about genetic sequence variation, geographic distributions, and hybridization into account, we follow Hershkovitz's (1977) nomenclature, but two plausible outcomes of future revisionary work could affect the nomenclature of populations in the Yavarí-Ucayali interfluve. If all of the tamarins in this complex were judged to be conspecific, the appropriate binomen for populations throughout the region would be S. nigricollis. Alternatively, if all morphologically diagnosable haplotype clades were recognized as full species (Matauschek et al., 2011), then the voucher material we examined would be referrable to S. nigrifrons and the coat-color phenotype reported as occurring between the Río Tapiche and the Río Blanco (see above) would be called S. fuscicollis.

Six specimens of Saguinus fuscicollis (sensu Hershkovitz, 1977) are labeled as having been collected at Marupa (on the south bank of the Amazon) by Harvey Bassler; these include four examples of the nigrifrons phenotype and three examples of the lagonotus phenotype. The first four are listed above as vouchers, but the latter two (AMNH 98286, 98287) were alleged by Hershkovitz (1977) to have been collected on the opposite (north) bank of the Amazon. In the absence of any other material of lagonotus from south-bank localities, Hershkovitz's hypothesis that AMNH 98286 and 98287 were mislabeled seems plausible because other vertebrate specimens that passed through the hands of Harvey Bassler are likewise associated with problematic locality data (Wiley, 2010).

Saguinus mystax (Spix, 1823)

Voucher material (total  =  35)

Marupa (AMNH 98288, 98289), Nuevo San Juan (AMNH 268237; MUSM 11105, 11106, 13303), Orosa (AMNH 73741, 73743–73745, 73747, 73750, 73985, 74039, 74044, 74047, 74049, 74050, 74052), Quebrada Esperanza (FMNH 88870–88872), Santa Cecilia (FMNH 86951–86957, 86959–86962, 87138–87140).

Unvouchered observations

Actiamë (Amanzo, 2006), Choncó (Amanzo, 2006), Itia Tëbu (Amanzo, 2006), Jenaro Herrera (Aquino, 1978), Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993; Heymann and Aquino, 1994), Río Aucayo (Castro and Soini, 1977), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí-Mirím (Salovaara et al., 2003), Tapiche (Jorge and Velazco, 2006).

Identification

Hershkovitz (1977) examined the series from Marupa, Orosa, Quebrada Esperanza, and Santa Cecilia for his taxonomic revision of Saguinus, wherein this material was referred to S. mystax mystax. Our material from Nuevo San Juan is indistinguishable from that examined by Hershkovitz, conforming to his diagnosis of the nominotypical race by lacking both the reddish coronal fur of S. m. pileatus (listed as a distinct species by Groves, 2001, 2005) and the whitish ventral tail base of S. m. pluto. Craniodental measurements from representative specimens are summarized in table 9. Most moustached tamarin specimens from the Yavarí-Ucayali interfluve are unaccompanied by weights, but an adult male (AMNH 268237) from Nuevo San Juan weighed 522 g, another (MUSM 13303) weighed 585 g, and an adult female from the same locality weighed 650 g.

TABLE 9

External and Craniodental Measurements (mm) of Saguinus mystax from the Yavarí-Ucayali Interfluve

Subfamily Cebinae Bonaparte, 1831

The subfamily Cebinae includes capuchins (Cebus), medium-sized monkeys with prehensile tails, and squirrel monkeys (Saimiri), which are much smaller and have nonprehensile tails. Two species of capuchins and one species of squirrel monkey are represented in the voucher material that we examined from the Yavarí-Ucayali interfluve.

Genus Cebus Erxleben, 1777

The white-fronted capuchin (Cebus albifrons) and the brown capuchin (C. apella) occur sympatrically throughout the region treated in this report. Their ethnobiology is conveniently summarized in a joint account.

Ethnobiology

Capuchin monkeys are called bëchun by the Matses. They are also known by the archaic synonym koekid “one that says ‘ko.’ ” The monomorphemic term chidu is used in a Matses myth to refer to capuchin monkeys, and chidu is also the principal term for capuchin monkey in the closely related language of the Kulinas, whom the Matses raided and from whom they incorporated many captives. Most Matses speakers do not list chidu as part of the “official” game synonym set for capuchins (Fleck and Voss, 2006), but consider it a synonym reserved for telling myths. Other Matses insist that, because it is a legitimate and archaic name, it should not be excluded when listing game synonyms.

Two types of capuchin monkey are universally recognized by the Matses: bëchun çhëşhë “black/dark-colored capuchin” (the brown capuchin, Cebus apella) and bëchun uşhu “white/light-colored capuchin” (the white-fronted capuchin, Cebus albifrons). The term bëchun (unmodified by çhëşhë or uşhu) can refer generically to both local species of capuchin monkeys, or specifically to the brown capuchin. The brown capuchin is called by its full name, bëchun çhëşhë, only when there is a need to be specific. Some hunters recognize two varieties of brown capuchin: bëchun çhëşhëdapa “large dark capuchin monkey” and bëchun çhëşhëmpi “small dark capuchin.” The large variety is said to travel in smaller troops, and be found more commonly in upland forest; the smaller variety is said to be the more common one and is found more frequently along rivers. Most hunters, however, do not recognize these subtypes of brown capuchin monkey.

Old people can eat capuchin monkeys, but young people believe they will become lazy and unenergetic if they eat them. As with the howler monkey (see above), this lassitude can be cured by application of frog poison to burns on the arms or chest. Some Matses say that young people will also age prematurely if they eat capuchin monkeys. Young capuchin monkeys are sometimes raised as pets, but most Matses prefer not to raise them because they are too mischievous, breaking chicken eggs, masturbating, and constantly getting into everything. The canines of brown capuchin monkeys are considered good for making tooth necklaces.

Both capuchin monkey species are hunted in the same way. Hunters usually do not make much of an effort to find capuchin monkeys because of the partial dietary taboo, especially now that mostly shotguns are used and hunters do not wish to expend valuable ammunition on food that only old people can eat. Matses hunters detect capuchin monkeys when they hear them knocking hard fruits against branches, or when they hear vocalizations or the noise of rustling branches as the monkeys travel through the canopy. When a Matses hunter imitates capuchin calls, the monkeys usually respond loudly but seldom actually come toward the hunter. When capuchins are being chased, they sometimes drop to the ground and run on the ground. When this happens, dogs will try to catch them, but capuchin monkeys are strong and easily kill dogs by biting them. Hunters must be ready to kill capuchin monkeys fighting with their hunting dogs, but capuchins also bite people if they are carelessly approached.

Cebus albifrons (Humboldt, 1812)

Voucher material (total  =  7)

Boca Río Yaquerana (FMNH 88854, 88855, 89173); Nuevo San Juan (MUSM 11119); Orosa (AMNH 73723, 74034); Santa Cecilia (FMNH 86932).

Unvouchered observations

Actiamë (Amanzo, 2006), Choncó (Amanzo, 2006), Divisor (Jorge and Velazco, 2006), Jenaro Herrera (Aquino, 1978), Orosa (Freese et al., 1982), Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993; Heymann and Aquino, 1994), Río Tapiche (Bennett et al., 2001), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí-Mirím (Salovaara et al., 2003).

Identification

The last morphological revision of Cebus was by Hill (1960), who reviewed the tediously complex and unedifying nomenclatural history of this difficult genus. Most students of capuchin taxonomy (e.g., Elliot, 1913; Hershkovitz, 1949; Hill, 1960) have emphasized the distinction between “tufted” and “untufted” species based on the presence or absence of tufts or crests of coronal fur in adult males (additional characters diagnosing these groups were tabulated by Hershkovitz, 1949). As far as known, no local fauna anywhere in the Neotropics has more than one species of tufted capuchin or more than one untufted species, whereas tufted and untufted capuchins occur sympatrically throughout much of Amazonia. Thus, the distinction seems important, and Silva (2002) recently proposed that these groups be formally recognized as subgenera: Cebus Erxleben, 1777, for the untufted species; and Sapajus Kerr, 1792, for the tufted species. At least some genetic evidence suggests that these might, in fact, be reciprocally monophyletic taxa (Casado et al., 2010).

The untufted capuchin of western Amazonia, Cebus albifrons, has been divided into a number of subspecies by authors. Based on mapped geographic ranges in Hershkovitz (1949) and Hill (1960), the population inhabiting the Yavarí-Ucayali interfluve is assignable to C. a. unicolor (Spix, 1823), the type locality of which is Tefé (on the south bank of the Amazon between the Juruá and the Purus), Amazonas, Brazil. However, Defler and Hernández-Camacho (2002) argued that Spix's unicolor is phenotypically indistinguishable from the nominotypical race, the type locality of which (fixed by neotype selection) is “about 10 km north of Maypures … Vichada, Colombia” (op. cit.: 54). Our voucher material is somewhat darker and duller than the color illustration of the neotype of C. a. albifrons (in Defler and Hernández-Camacho, 2002), but we remain unconvinced of the need for trinomial nomenclature to distinguish minor pelage differences among Amazonian populations of this species. Measurements of specimens collected in the Yavarí-Ucayali interfluve (table 10) are slightly larger than most homologous dimensions of topotypical C. a. albifrons (in Defler and Hernández-Camacho, 2002), but sample sizes are too small for confident inference.

TABLE 10

External and Craniodental Measurements (mm) of Cebus albifrons from the Yavarí-Ucayali Interfluve

Matses natural history

White-fronted capuchin monkeys are light-colored, except for the tops of their heads, which are brown. Their backs and tails are darker than their faces, chests, and bellies. Like brown capuchin monkeys, white-fronted capuchin monkeys raise their eyebrows repeatedly when they see people. Males touch their penises when they see people.

White-fronted capuchin monkeys can be found in any habitat, but they are more common in upland forest than in riverside forest, whereas brown capuchin monkeys have the opposite preference. They sometimes come to the edge of swiddens. White-fronted capuchin monkeys spend more time foraging on the ground than any other type of monkey. They play on the ground in small clearings. They use mostly the lower canopy, but go high to feed when they find food up high.

Troops are medium-sized, numbering up to about 20 or 25 animals, though more frequently about 8 to10. They have many young in their troops, which scream a lot as they pass from tree to tree. Sometimes white-fronted capuchin monkeys travel with woolly monkeys, but never with squirrel monkeys. They are less commonly encountered than brown capuchin monkeys.

They call differently from brown capuchin monkeys, calling: “kooo kooo kooo,” among other vocalizations. Like brown capuchin monkeys, they break off dead branches and palm fronds and pull off wasp nests and throw them down. They sleep together in a group up in a tree, unlike woolly monkeys, which spread out in different trees.

Their food is the same as that of brown capuchin monkeys [see below; however, unlike brown capuchins, they have not been observed to kill and eat titi monkeys]. They have the same ability to break open hard dicot fruits and palm nuts.

Cebus apella (Linnaeus, 1758)

Voucher material (total  =  8)

Boca Río Yaquerana (FMNH 88846), Nuevo San Juan (AMNH 268240, MUSM 11120), Orosa (AMNH 73989, 74032, 74033), Quebrada Esperanza (FMNH 88847, 88848).

Unvouchered observations

Actiamë (Amanzo, 2006), Choncó (Amanzo, 2006), Divisor (Jorge and Velazco, 2006), Itia Tëbu (Amanzo, 2006), Orosa (Freese et al., 1982), Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993; Heymann and Aquino, 1994), Río Tapiche (Bennett et al., 2001), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí Mirím (Salovaara et al., 2003), Tapiche (Jorge and Velazco, 2006).

Identification

The many nominal taxa of tufted capuchins—which collectively range from northern Colombia to northern Argentina—were all treated as synonyms or subspecies of the brown capuchin (Cebus apella) by Hershkovitz (1949) and Hill (1960), but some are now recognized as valid species by authors (e.g., Groves, 2001, 2005; Silva, 2002; Rylands et al., 2005). Unfortunately, the revisionary studies alleged to support the recognition of distinct species of tufted capuchin have never been published, the only published diagnoses (based on trivial sample sizes; Groves, 2001) are not useful for specimen identification, “undeniable evidence of some natural interbreeding” is admitted to exist (Groves, 2001: 152), and genetic distances among some “species” are unimpressive (e.g., 1.3% between cytochrome-b sequences of C. apella and C. “cay”; Casado et al., 2010). Therefore, a compelling case for recognizing multiple valid species among the nominal taxa traditionally regarded as synonyms or subspecies of C. apella has yet to be made. In particular, we are not convinced that any tufted species other than the brown capuchin inhabits Amazonian rainforests.

According to Hill (1960), the subspecies of Cebus apella in the Yavarí-Ucayali interfluve is macrocephalus Spix, 1823 (type locality “Lago Cactuá,” presumably somewhere on the Brazilian Amazon), but geographically adjacent forms include juruanus Lönnberg, 1939 (type locality João Pessôa on the Rio Juruá, Brazil) and peruanus Thomas, 1901 (type locality Huaynapata, Cusco, Peru). In fact, representative skins from the Yavarí-Ucayali interfluve plausibly fit Grove's (2001) descriptions of each of these nominal taxa. Some (AMNH 268240), for example, have more or less distinct blackish middoral stripes (as described for macrocephalus and juruanus), whereas others (AMNH 73989) lack any trace of a black middorsal stripe (resembling peruanus in this and other respects). Cabrera (1958) treated both juruanus and peruanus as synonyms of C. a. macrocephalus, an arrangement that is consistent with our suspicion that all western Amazonian tufted capuchins are taxonomically indistinguishable.

Measurements of examined voucher material are summarized in table 11. Most specimens of Cebus apella from the Yavarí-Ucayali interfluve are unaccompanied by weights, but an adult male (AMNH 268240) from Nuevo San Juan weighed 5200 g and a lactating adult female from the same locality (MUSM 11120) weighed 2510 g.

TABLE 11

External and Craniodental Measurements (mm) of Cebus apella from the Yavarí-Ucayali Interfluve

Ethnobiology

See the preceding account for capuchin ethnobiology.

Matses natural history

Brown capuchin monkeys have bald temples, especially the large males. Unlike woolly monkeys, they do not hang by their tails. They walk around on branches with their tails rolled up very tightly. Males are larger than females. Capuchin monkeys, including pets, raise their eyebrows repeatedly when they see people.

Brown capuchin monkeys are found in primary or secondary forest, in upland forest or in riverside forest. They are common along large streams. Sometimes they forage on the ground, but not as much as white capuchin monkeys do. They generally do not forage too high up.

Brown capuchin monkeys are in medium-sized troops, numbering up to about 15 animals. Sometimes a male is found alone. They very commonly travel and forage together with squirrel monkeys. Brown capuchin monkeys bully squirrel monkeys when they feed together with them. When capuchin and squirrel monkeys are together, they make a lot of noise. Like white-fronted capuchin monkeys, brown capuchin monkeys make a lot of noise banging hard dicot fruits and palm nuts to open them. They may spread out to forage, but when they find a lot of fruit they eat together.

Brown capuchin monkeys call out saying “bëtsiton” when they see people or jaguars. They also call out saying “ko-o ko-o ko-o ko-o.”

Brown capuchin monkeys travel through the forest being mischievous: they pull down wasp and termite nests and toss them down to the ground. They also throw down palm fronds and dead branches as they go. In abandoned swiddens they throw down peach-palm [Bactris gasipaes] fruits without eating them. At night they sleep in any tree that has vine tangles.

Capuchin monkeys eat all kinds of dicot tree fruits, including, wësnid dëbiate [Anacardium giganteum (Anacardiaceae)], këku [Couma macrocarpa (Apocynaceae)], diden këku [Parahancornia peruviana (Apocynaceae)], machishte [Rhigospira quadrangularis and ?Mucoa duckei (Apocynaceae)], kapan çhëşhte [Matisia bracteolosa and Quararibea ochrocalyx (Bombacaceae)], mamuin [Rheedia longifolia (Guttiferae)], okodo mabis [undetermined, Guttiferae)], moste [Hymenaea spp. (Leguminosae)], mannan tsipuis [Inga spp. and ?Pithecellobium (Leguminosae)], tankada [Parkia igneiflora, P. multijuga, and Pithecellobium auriculatum (Leguminosae)], bin [Castilla (Moraceae)], buku [Cecropia spp. (Moraceae)], dadain [Clarisia racemosa (Moraceae)], piuşh bëchi [Helicostylis tomentosa and H. elegans (Moraceae)], kuşhu tëbin [Naucleopsis mello-barretoi and N. ternstroemiiflora (Moraceae)], şhankuin [Pourouma spp. (Moraceae)], bata [Pseudolmedia and Maquira spp. (Moraceae)], tonnad [all species of Myristicaceae], mabis mabiskid [Chrysophyllum prieurii (Sapotaceae)], kose [Manilkara bidentata (Sapotaceae)], dadan dëso [Theobroma cacao (Sterculiaceae)], tonkodo [Theobroma (Sterculiaceae)], kuëte mëdiad [undetermined], and taëpa [undetermined]. They also eat vine fruits. They eat swamp-palm [Mauritia flexuosa] fruits regularly and isan palm [Oenocarpus bataua] fruits occasionally. Because they can break open fruits by banging them on a branch, they eat some fruits that other monkeys cannot eat, including dadan dëso and other cacaolike fruits. They can break open unripe palm fruits and consume the liquid and/or soft endosperm inside. Some of the palm fruits they feed on include: niste [Iriartea deltoidea], budëd [Attalea butyracea], di pinchuk [Astrocaryum chambira], şhukkate pinchuk [Astrocaryum murumuru], and akte pinchuk [Astrocaryum jauari]. They bite open cabbage palm [Euterpe precatoria] seeds and eat the soft endosperm, but do not eat the ripe mesocarp of cabbage palm fruits. They pick and wastefully throw down many fruits that they do not eat. They pull off palm leaves to get at the heart. When there are no fruits to be found, they eat any little thing they can find. They eat beetle larvae they find in rotten wood that they break apart. They eat invertebrates including scorpions, millipedes, crickets [or katydids; the Matses term is taxonomically ambiguous], etc. They also eat baby birds. They occasionally kill and eat titi monkeys [Callicebus cupreus].

Saimiri sciureus (Linnaeus, 1758)

Voucher material (total  =  22)

Boca Río Yaquerana (FMNH 88866), Marupa (AMNH 98269, 98270), Nuevo San Juan (AMNH 268245, 268246; MUSM 11155, 11156, 11158, 11159, 11161–11169), Quebrada Esperanza (FMNH 88867), San Fernando (FMNH 88865), Santa Cecilia (FMNH 86977).

Unvouchered observations

Actiamë (Amanzo, 2006), Choncó (Amanzo, 2006), Itia Tëbu (Amanzo, 2006), Jenaro Herrera (Aquino, 1978), Río Orosa (Freese et al., 1982), Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993; Heymann and Aquino, 1994), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí-Mirím (Salovaara et al., 2003).

Identification

Despite several decades of behavioral, chromosomal, morphological, and molecular research (reviewed by Hershkovitz, 1984; Thorington, 1985; Costello et al., 1993; Boinski and Cropp, 1999; Lavergne et al., 2010), the taxonomy of Saimiri (squirrel monkeys) remains profoundly unsatisfactory. The last formal taxonomic treatment of the genus (Hershkovitz, 1984) recognized a Central American species, S. oerstedi (Reinhardt, 1872) and three Amazonian species. The latter (in Hershkovitz's classification) consist of S. boliviensis (I. Geoffroy and de Blainville, 1834), S. sciureus (Linnaeus, 1758), and S. ustus (I. Geoffroy, 1843). A fourth Amazonian species, S. vanzolinii Ayres, 1985, was also recognized as valid by Groves (2001, 2005).

Saimiri boliviensis and S. sciureus have both been reported from the Yavarí-Ucayali interfluve. Based on his examination of museum specimens, Hershkovitz (1984) reported that S. boliviensis and S. sciureus occur sympatrically west of the Río Tapiche, whereas only S. sciureus occurs east of the Tapiche. Puertas and Bodmer (1993), however, reported unvouchered sightings of both species from the Reserva Comunal Tamshiyacu-Tahuayo (along the Quebrada Blanco and the Río Yavarí-Mirím), and Bennett et al. (2001) reported unvouchered sightings of S. boliviensis from both banks of the Tapiche. Given the discrepancy between Hershkovitz's specimen-based observations and the unvouchered observations of fieldworkers, the external characters by which S. boliviensis and S. sciureus can be distinguished are obviously relevant for assessing the reliability of sight identifications.

The only external characters by which Peruvian forms of Saimiri boliviensis and S. sciureus can be reliably distinguished are facial markings and caudal pelage: (1) As illustrated by Hershkovitz (1984: fig. 1), a broad arc of white fur extends onto the forehead above each eye in S. sciureus, separating the blackish superciliary vibrissae from the dark (greyish or blackish) fur of the crown; by contrast, the dark fur of the crown extends ventrally to the superciliary vibrissae in S. boliviensis, such that the vibrissae are harder to see. (2) The black tuft of long hairs at the end of the tail is thicker and more conspicuous in S. sciureus than in S. boliviensis. Unfortunately, neither of these characters is easily scored at a distance or in the absence of comparative material, and relevant illustrations in some field guides are inaccurate.12 Because primatologists have seldom described the phenotypic criteria they used to identify squirrel monkeys in the field, most unvouchered observations of Saimiri from northeastern Peru cannot be reliably assigned to species.

All of the specimens of Saimiri that we examined from east of the Río Tapiche in the Yavarí-Ucayali interfluve are unambiguously referable to S. sciureus (sensu Hershkovitz, 1984). Among other diagnostic traits, all examined specimens have broad arcs of white fur above each eye (separating the blackish superciliary vibrissae from the grayish or blackish coronal fur), and the black tail tufts of these specimens are longer and thicker than those in comparative series of S. boliviensis. Males have grayish-agouti coronal fur and grayish sideburns, whereas the coronal fur and sideburns of females are conspicuously darker (sometimes blackish in part; e.g., AMNH 268246, FMNH 88867, MUSM 11166). Although the Matses report variation in arm pigmentation among local troops of squirrel monkeys (see below), all of the specimens we examined have grayish-agouti upper arms and yellowish forearms. Measurement and weight data are summarized in table 12.

TABLE 12

External and Craniodental Measurements (mm) and Weights (g) of Saimiri sciureus from the Yavarí-Ucayali Interfluve

As recognized by authors (e.g., Hershkovitz, 1984; Groves, 2001, 2005), Saimiri sciureus is polytypic, with several subspecies that are distinguishable by pelage pigmentation and karyotypes; specimens from the Yavarí-Ucayali interfluve are usually referred to S. s. macrodon (Elliot, 1907), the type locality of which is in eastern Ecuador. Although neither S. sciureus nor S. s. macrodon are consistently recovered as monophyletic groups in recent phylogenetic analyses of cytochrome-b sequence data (Lavergne et al., 2002, 2010), the taxonomic interpretation of these results is unclear. Absent a comprehensive revision of Saimiri based on geographically dense sampling, multiple genes, and due attention to diagnostic morphological traits, we follow existing usage in referring our material to S. sciureus in its currently accepted sense.

Ethnobiology

Squirrel monkeys are called tsanka. They are also called tsankekid “one that says tsan.” The term tsanka appears to be a shorted form of tsankekid. Two subtypes are recognized: tsankadapa “big squirrel monkey” and tsankampi “little squirrel monkey.” The larger variety is said to have whitish arms and to live in smaller troops (or alone), while the small variety is said to have bright yellow arms and to live in larger troops. The small variety is said to be more common in river-edge forest than the larger variety. The smaller variety is said to be the only one found to travel together with brown capuchin monkeys.

Very long ago, the Matses did not eat squirrel monkeys, lest they become thin (like a squirrel monkey). More recently, perhaps even before contact with missionaries, Matses began to eat squirrel monkeys, but as with titi monkeys, hunters usually do not make a special effort to find them because they are small. A Matses hunter may listen to their noisy rustling of branches, and then try to get a shot off before they see him. After the first shot, the hunter may chase the troop making them run through the trees until one pauses on a relatively low branch, when the hunter can shoot it with an arrow or shotgun. The Matses consider squirrel monkeys to be very abundant, easy to kill, and tasty, but nevertheless hunters do not often kill them.

The canines of large males are sometimes used to make men's tooth necklaces, but the teeth of larger monkeys are preferred. Squirrel monkeys are frequently kept as pets and are considered very good pets by some Matses. Attractive women are called “squirrel monkey” because squirrel monkeys, especially their babies, are considered cute.

Matses natural history

Squirrel monkeys are cute. They have white fur around the eyes. The area around their mouth is black, as if painted with genipap dye. They have a long tail with a black tail tip, like a night monkey does. They have a strong smell.

They are found in all types of habitats, but not too far from rivers [within a few kilometers]. They prefer riverside forest and also like swamp-palm [Mauritia flexuosa] swamps. They like riverside forest because there are a lot of legume pods and insects there. They like swamp-palm swamps, both for the swamp-palm fruits and because there are many insects there.

They live in very large troops, and stay together without splitting up. Sometimes they eat, travel, and even sleep with brown capuchin monkeys. They may forage together with woolly monkeys, but do not stay with them as long as they do with brown capuchin monkeys. The females carry the young on their back or underside.

They vocalize saying “sii sii chan chan chan chan sii sii.” When they hear brown capuchin monkeys, they answer. The young also go “sii sii sii.” They yell like that as they travel noisily through the trees.

Squirrel monkeys are active during the day. They wake up early and start vocalizing right away. They always travel spastically, as if they are always hungry. They descend to the ground to drink water. They are usually found low in the forest, but they climb very high up to eat fruits or to flee from humans or predators. They sleep mostly in the fronds of palm trees such as budëd palms [Attalea butyracea]. They even sleep on the thorny fronds of pinchuk palms [Astrocaryum spp.]. They all sleep together, not spread out. They bed down making a lot of noise at dusk, when it is almost dark. Sometimes they are awake during moonlit nights.

Squirrel monkeys eat dicot tree fruits, especially legume pods such as mannan tsipuis [Inga spp. and ?Pithecellobium (Leguminosae)] and achu inkuente [Inga spp. (Leguminosae)]. Other dicot fruits they eat include: wesnid dëbiate [Anacardium giganteum (Anacardiaceae)], këku [Couma macrocarpa (Apocynaceae)], mamuin [Rheedia longifolia (Guttiferae)], okodo mabis [undetermined sp. (Guttiferae)], moste [Hymenaea spp. (Leguminosae)], chichombid [Mouriri spp. (Melastomataceae)], şhannëd [?Brosimum (Moraceae)], bin [Castilla (Moraceae)], dadain [Clarisia racemosa (Moraceae)], piuşh bëchi [Helicostylis tomentosa, H. elegans (Moraceae)], kuşhu tëbin [Naucleopsis mello-barretoi, N. ternstroemiiflora (Moraceae)], bata [Pseudolmedia and Maquira spp. (Moraceae)], poşhodi [Passiflora nitida (Passifloraceae)], mabis mabiskid [Chrysophyllum prieurii (Sapotaceae)], and kose [Manilkara bidentata (Sapotaceae)]. They eat swamp-palm [Mauritia flexuosa] fruits in the dry season. They also eat a lot of spiders, crickets [or katydids; the Matses term is taxonomically ambiguous], and other insects.

Family Pitheciidae Mivart, 1865

The family Pitheciidae comprises four genera, of which three, each represented by a single species, occur in the Yavarí-Ucayali interfluve: Cacajao Lesson, 1840; Callicebus Thomas, 1903; and Pithecia Desmarest, 1804.

Cacajao calvus (I. Geoffroy, 1847)

Voucher material (total  =  25)

Nuevo San Juan (AMNH 272795; MUSM 11114, 11115, 13300), Orosa (AMNH 73716–73720), Quebrada Esperanza (FMNH 88810–88825).

Unvouchered observations

Jenaro Herrera (Aquino, 1978), Quebrada Blanco (Bartecki and Heymann, 1987), Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993), Río Tapiche (Bennett et al., 2001), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí-Mirím (Salovaara et al., 2003), Tapiche (Jorge and Velazco, 2006). For additional records, see Heymann and Aquino (2010).

Identification

The genus Cacajao was reviewed by Hershkovitz (1987b), who identified the specimens from Orosa and Quebrada Esperanza as C. calvus ucayalii Thomas, 1928. Our new material from Nuevo San Juan is morphologically indistinguishable from the Orosa and Quebrada Esperanza series. However, in none of the material that we examined (including specimens identified as C. c. ucayalii by Hershkovitz) is the dorsal surface of the tail blackish as he described it (op. cit.: 34); instead, the dorsal surface of the tail is densely covered with long reddish fur like that on the middle and lower back. External and craniodental measurement data from specimens collected in the Yavarí-Ucayali interfluve are summarized in table 13. Most specimens are unaccompanied by weights, but three adult females (AMNH 272794; MUSM 11114, 13300) from Nuevo San Juan weighed 2.9, 3.0, and 3.6 kg.

TABLE 13

External and Craniodental Measurements (mm) of Cacajao calvus from the Yavarí-Ucayali Interfluve

Ethnobiology

The Matses call the red uakari monkey senta, a term that is not etymologically analyzable. Some also call it bëşhuidkid piu, but most Matses do not consider this a true synonym. Unlike most other large monkeys, no subtypes of uakari monkey are recognized by any of the Matses.

There are no taboos against eating uakari monkeys (other than the general rule that a hunter should not eat or touch monkey intestines, or eat any of their first kill of a particular species). However, Matses who come from places where there are no uakari monkeys (e.g., along the Chobayacu), do not eat them saying “My father did not eat those,” as their justification. Many Matses that live outside the range of uakari monkeys are not familiar with them, and it is likely that, as the Matses moved around in the past, they did not continuously live in areas inhabited by uakaris. This may explain the lack of linguistic attention and the absence of uakari monkeys in Matses myths. The canines of large male uakari monkeys are used to make men's tooth necklaces.

The Matses do not imitate uakari monkey calls. Encountering uakari monkeys is relatively rare, so the Matses make no special effort to find them. Upon hearing their calls or rustling of branches, a Matses hunter will approach the troop and start shooting when he sees a large animal. When animals from a hunted troop see humans, they slowly move toward the upper canopy and/or move slowly away. When they hear a firearm report, they move away faster, but even then they do not seem to be too scared. Unlike spider monkeys, uakaris do not split up when they run away from hunters, but run in the same direction. Therefore, a good, solitary Matses hunter can easily kill several animals at a single encounter with a troop. They are packed for carrying and cooked in the same way as woolly and spider monkeys.

Matses natural history

Uakari monkeys are strikingly red, as if they had painted themselves with annatto dye. They have red body fur and bald, bright red heads and faces. Even the young are red. Their faces are swollen, as if they had been shot in the eye with an arrow. Their tails are markedly short, in contrast to all other monkeys. They can't grasp with their tails; instead, they eat fruits while hanging by their feet. In appearance they are most similar to saki monkeys, particularly with respect to their long pelage and bald foreheads. With respect to other characteristics, they are like woolly monkeys: their calls are similar to woolly monkeys', large males are strong like large male woolly monkeys, and their canines are like woolly monkeys' [and therefore desirable for making necklaces]. They have a strong smell, both in their fur and their meat, but they cannot be detected by hunters by smelling because they do not leave their odor.

Uakari monkeys do not occur in all of Matses territory. They are not found east of the Gálvez, on the Brazilian side of the lower Yavarí, or along the upper reaches of the Chobayacu [a left-bank tributary of the Yaquerana]. Uakari monkeys are found in primary upland and floodplain forest. They are frequently found in swamp-palm [Mauritia flexuosa] swamps and in flooded forest. After feeding for a while in floodplain forest, a uakari monkey troop may travel very far into upland forest. Sometimes one goes for many months without seeing any uakari monkeys, and then they appear again. They use mostly the middle canopy layer, lower than spider and howler monkeys, but they may climb higher when they see people, or when they find fruit there. They seldom descend to the ground. They call out saying “kakaka kakaka kakaka.”

Uakari monkeys are found in huge troops [>100 individuals, according to some Matses who can count], with many females carrying young on their backs. They travel as a group, making much noise rustling the branches. In flooded forest they drop dead branches into the water. They like to stay together. They may split up to find fruit, and rejoin to eat together when a large amount of fruit is found. They sleep near each other in the branches of dicot trees. When they are not hunted, they ignore humans, or come down closer to the ground to look at them. Even troops that are hunted are not very scared of humans.

Uakari monkeys eat mostly dicot fruits. They eat many different kinds of sweet dicot fruits. These include wesnid dëbiate [Anacardium giganteum (Anacardiaceae)], këku [Couma macrocarpa (Apocynaceae)], diden këku [Parahancornia peruviana (Apocynaceae)], machishte [Rhigospira quadrangularis and ?Mucoa duckei (Apocynaceae)], ichibin [Matisia sp. and Eriotheca sp. (Bombacaceae)], mamuin [Rheedia longifolia (Guttiferae)], okodo mabis [undetermined sp. (Guttiferae)], moste [Hymenaea spp. (Leguminosae)], tankada [Parkia igneiflora, P. multijuga, and Pithecellobium auriculatum (Leguminosae)], şhannëd [?Brosimum (Moraceae)], bin [Castilla (Moraceae)], dadain [Clarisia racemosa (Moraceae)], piuşh bëchi [Helicostylis tomentosa and H. elegans (Moraceae)], kuşhu tëbin [Naucleopsis mello-barretoi and N. ternstroemiiflora (Moraceae)], şhankuin [Pourouma spp. (Moraceae)], bata [Pseudolmedia and Maquira spp. (Moraceae)], poşhodi [Passiflora nitida (Passifloraceae)], mabis mabiskid [Chrysophyllum prieurii (Sapotaceae)], kose [Manilkara bidentata (Sapotaceae)], and taëpa [undetermined]. They bite the fruits, and if they are unripe they throw them away. They come to the edges of rivers to eat legume pods: mannan tsipuis [Inga spp. and ?Pithecellobium (Leguminosae)] and achu inkuente [Inga spp. (Leguminosae)]. They also eat fruits of a few types of palms, most remarkably, the endosperm of unripe swamp-palm [Mauritia flexuosa] fruits. They also eat the mesocarp of swamp-palm and isan [Oenocarpus bataua] fruits. They also eat some insects and a few leaf buds. Their diet is similar to that of woolly monkeys.

Callicebus cupreus (Spix, 1823)

Voucher material (total  =  16)

Nuevo San Juan (AMNH 268239; MUSM 11116, 11117), Orosa (AMNH 73703–73708), Quebrada Esperanza (FMNH 88858–88860), San Fernando (FMNH 88856, 88857), Santa Cecilia (FMNH 86989, 86990).

Unvouchered observations

Actiamë (Amanzo, 2006), Choncó (Amanzo, 2006), Itia Tëbu (Amanzo, 2006), Jenaro Herrera (Aquino, 1978), Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993; Heymann and Aquino, 1994), Río Orosa (Freese et al., 1982), Río Tapiche (Bennett et al., 2001), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí-Mirím (Salovaara et al., 2003), Tapiche (Jorge and Velazco, 2006). (Note that some reports of Callicebus from the Yavarí-Ucayali interfluve have identified the local species as C. caligatus or C. moloch; see below.)

Identification

The Amazonian forms of Callicebus (titi monkeys) were revised by Hershkovitz (1963), who recognized only two valid species, C. moloch and C. torquatus. However, Hershkovitz (1990) and Groves (2001) subsequently recognized 12 Amazonian species, and Roosmalen et al. (2002) recognized 22 Amazonian species that they assigned to four species groups.13 Although most taxonomic diagnoses in this literature are based on coat-color differences, there is some evidence that the species groups recognized by Roosmalen et al. (2002) are morphometrically differentiated and have distinctive karyotypes (Kobayashi, 1995). Unfortunately, the karyotypes of many alleged species are unknown, and molecular evidence supporting the recognition of distinct taxa within species groups is almost nonexistent.

Hershkovitz (1963) referred all of the Callicebus specimens he examined from the Yavarí-Ucayali interfluve to the subspecies C. moloch cupreus, but he later (Hershkovitz, 1990) identified some specimens of titi monkeys from the Yavarí-Ucayali interfluve as C. cupreus and others from the same region as C. caligatus. A small series of AMNH specimens from Orosa was listed among the material examined for both species in the latter publication, and an accompanying map (Hershkovitz, 1990: fig. 36) showed C. cupreus and C. caligatus as occurring sympatrically at Orosa. Later, Hershkovitz (personal commun., 1995) explained that this was an error caused by listing original and revised identifications of the same specimens in separate species accounts.

In fact, titi specimens from the Yavarí-Ucayali interfluve could be referred either to Callicebus cupreus or to C. caligatus as those species were diagnosed by Hershkovitz (1990). For example, the forehead is grayish or reddish agouti with only a narrow blackish brow-line (as described for cupreus) in most of our specimens (e.g., AMNH 73708, 268239), but in several others (e.g., AMNH 73705, 73706) the entire forehead and lower crown is blackish (as described for caligatus). Similarly, some of our specimens (e.g., AMNH 268239) have clear reddish hands, feet, and lower limbs (as in cupreus), whereas others (e.g., AMNH 73703, 73706) have dark reddish-brown or even blackish extremities (as in caligatus).

According to Roosmalen et al. (2002), only Callicebus cupreus inhabits the Yavarí-Ucayali interfluve, but some of our specimens more closely resemble their diagnosis of C. caligatus, particularly with respect to the presence of a blackish forehead and lower crown as described above. Additionally, the basal part of the tail is blackish in some of our specimens (e.g., AMNH 73703) and others (e.g., AMNH 268239) have extensively pale-tipped tails, both traits associated with C. caligatus rather than C. cupreus in their diagnoses. In effect, our material combines traits of both species as recognized by these authors, who do not list any material examined, and whose illustrations do not correspond closely with their text.14

We are not convinced that Callicebus caligatus or any of the other nominal taxa belonging to the Cupreus Group of Kobayashi (1995) and Roosmalen et al. (2002)—including C. caquetensis, C. discolor, C. dubius, C. ornatus, C. stephennashi, and probably C. aureipalati—are valid species distinct from C. cupreus. Although these forms collectively range across most of western Amazonia, they resemble one another in pelage markings (see Roosmalen et al., 2002: fig. 49; Defler et al., 2010: fig. 5; Wallace et al., 2006: figs. 4, 5), and at least some allegedly diagnostic differences seem to break down when large series of specimens (as opposed to just a few exemplars) are examined. Additionally, nonpelage differences among these taxa appear to be wanting: all analyzed forms are morphometrically similar (Kobayashi, 1995), have the same diploid number of chromosomes (2n  =  46; Roosmalen et al., 2002; Defler et al., 2010), and seem to occupy the same ecological niche (inhabiting forest edges and disturbed vegetation; Roosmalen et al., 2002). Because no verifiable examples of sympatry are known among nominal species in this group, it is plausible that these taxa represent nothing more than geographic coat-color variants of the same biological species, a hypothesis that merits future evaluation with molecular sequence data and with targeted collecting in areas where different coat-color phenotypes might be expected to come in contact.

External and craniodental measurements of Callicebus cupreus specimens that we examined from the Yavarí-Ucayali interfluve are summarized in table 14. Most specimens are unaccompanied by weight data, but an adult male (MUSM 11116) from Nuevo San Juan weighed 1020 g, and an adult female (AMNH 268239) from the same locality weighed 1042 g.

TABLE 14

Measurements (mm) of Callicebus cupreus from the Yavarí-Ucayali Interfluve

Ethnobiology

The Matses call the titi monkey wadë, which is not linguistically analyzable. It has three other supposedly ancient synonyms, tsokon, sëdë, and masoko. The first synonym is onomatopoetic and unique to Matses. The other two are of uncertain etymology: both occur in neighboring Mayoruna languages, and therefore could easily be either borrowings or truly archaic words. Some Matses assert that masoko is not archaic, but a borrowing; there is no controversy about the other two names. Two subtypes are recognized, wadë piu “red titi monkey” and wadë çhëşhë “black (dark-colored) titi monkey.” The first is the commonly seen variety, which is found close to rivers. The second variety is reported to live in upland forest far from rivers, where the red variety is absent. Not all Matses hunters are familiar with the darker variety or accept that two varies exist. Those that claim to have seen the darker variety do not know much about it. (Despite a considerable effort to find it, the darker variety was not collected during our mammal inventory at Nuevo San Juan; it probably represents another case of overdifferentiation of a single biological species.)

Matses eat titi monkeys, but because they are small, hunters usually do not make a special effort to find them. Traditionally, as larger game became scarce in a locality, more titi monkeys would be killed. Nowadays, most hunters will not expend a shotgun shell to kill one despite the shortage of game around the larger and more permanent villages. One way to kill titi monkeys is to simply shoot them from the ground. Another way is to climb up at tree where the titi monkeys are hiding. When the titi monkeys see the hunter up in the tree, they will usually drop and flee running on the ground, where waiting hunting dogs and/or a second hunter will chase them down and kill them. Matses do not imitate their call to hunt them.

As with all monkeys, titi monkey are dressed for the pot by singing off the fur and then gutting them. Unlike large monkeys, titi monkeys (and other small monkeys) may be roasted instead of boiled. Pregnant women or men with pregnant wives cannot eat titi monkeys, lest their children grow up to have bad teeth [titi monkeys have black teeth]. Young people will not eat the liver of a titi monkey, lest it make their teeth rot.

Matses natural history

Titi monkeys have red underparts, a black brow, black teeth, and a whitish tail. They are small, the size of a squirrel monkey.

Titi monkeys can be found in any forest type, but they strongly prefer habitats that have open canopies [i.e., where light shines through to the understory], such as abandoned swiddens, riverside forest, and secondary forest at blowdowns. When there are abundant ripe legume pods growing along rivers [November–January], titi monkeys are especially common there. They are not found very high up in the trees. They descend to the ground occasionally to drink water, eat insects or fruits, or to cross a swidden.

They live in small troops that number only up to about five or six animals. Sometimes one is found alone. The male always carries the babies. They carry the baby on their back or their belly, and give it to the female to suckle. They call out saying “tsokon, tsokon, tsokon, on, on, on” shortly before dawn [around 5 a.m.], later than howler monkeys, and then again later in the morning [around 8 a.m.], and sometimes at midday. The male sings first, and then the rest join in. They sing very loudly, and Matses always hear them from their homes.

Titi monkeys sleep hugging each other on a tree branch among thick vegetation, in vine tangles. Right from where they sleep they sing in the morning. They spread out to eat, but not too far from each other. After eating, they rest at midday. They do not travel far. They are curious, and animals that have not been hunted will stare from the forest edge at people passing by in boats.

Titi monkeys are the only monkeys that eat the ripe mesocarp of cabbage palm [Euterpe precatoria] fruits, and that is why they have black teeth [capuchins bite unripe cabbage palm fruits to eat the soft endosperm, but don't eat the ripe mesocarp]. They eat all types of dicot fruits, especially mannan tsipuis [Inga spp. and ?Pithecellobium (Leguminosae)] and achu inkuente [Inga spp. (Leguminosae)]. Other dicot fruits they eat include: wesnid dëbiate [Anacardium giganteum. (Anacardiaceae)], këku [Couma macrocarpa (Apocynaceae)], diden këku [Parahancornia peruviana (Apocynaceae)], machishte [Rhigospira quadrangularis and ?Mucoa duckei (Apocynaceae)], okodo mabis [undetermined sp. (Guttiferae)], mamuin [Rheedia longifolia (Guttiferae)], chichombid [Mouriri spp. (Melastomataceae)], şhannëd [?Brosimum (Moraceae)], piuşh bëchi [Helicostylis tomentosa and H. elegans (Moraceae)], bata [Pseudolmedia and ?Maquira spp. (Moraceae)], mabis mabiskid [Chrysophyllum prieurii (Sapotaceae)], kose [Manilkara bidentata (Sapotaceae)], and taëpa [undetermined].

Pithecia monachus E. Geoffroy, 1812

Voucher material (total  =  34)

Nuevo San Juan (AMNH 26841–26844; MUSM 11126–11133, 11136, 11138, 11140–11151, 11153, 11154), Orosa (AMNH 73714, 74036, 74037), Quebrada Esperanza (FMNH 88862), San Fernando (FMNH 88861), Santa Cecilia (FMNH 87002).

Unvouchered observations

Actiamë (Amanzo, 2006), Choncó (Amanzo, 2006), Itia Tëbu (Amanzo, 2006), Jenaro Herrera (Aquino, 1978), Reserva Comunal Tamshiyacu-Tahuayo (Puertas and Bodmer, 1993; Heymann and Aquino, 1994), Río Orosa (Freese et al., 1982), Río Tapiche (Bennett et al., 2001), Río Yavarí (left bank below Angamos; Salovaara et al., 2003), Río Yavarí-Mirím (Salovaara et al., 2003), Tapiche (Jorge and Velazco, 2006).

Identification

Saki monkeys from the Yavarí-Ucayali interfluve were referred to Pithecia hirsuta by Hershkovitz (1979), but subsequent study of type material led Hershkovitz (1987c) to identify the same specimens as P. monachus. Although all of our material is referable to P. monachus as that taxon was briefly distinguished in Hershkovitz's (1987c) key, the taxonomic significance of phenotypic variation among western Amazonian populations of sakis remains to be convincingly established. Because our series from Nuevo San Juan (N  =  26 adult specimens) is the largest sample available to document morphological variation within a local population of any Pithecia species, we provide a detailed description below.

The 11 adult male skins from the Río Gálvez exhibit limited variation in pelage characters. The dorsal body pelage and the pelage of the outer surfaces of the arms and legs is predominantly blackish, but the individual hairs have pale tips that produce a frosted mass effect; specimens with long, whitish hair tips (e.g., MUSM 11126) appear more heavily frosted than individuals with shorter, yellowish, or brownish terminal bands (MUSM 11142). The distinctive ruff or mane surrounding the head does not contrast in coloration with the rest of the body fur. The hands are distinctly paler than the arms, appearing whitish in most individuals, but the bases of the individual hairs on the hands are always dark. Most specimens also have whitish feet, but a few (e.g., MUSM 11128, 11129, 11145) have grizzled pedal fur that is not conspicuously paler than the fur of the thigh and shank.

Whereas the body pelage is long and lax, especially over the shoulders and nape (60–100 mm), the coronal pelage (from between the ears to the brow line) is abruptly shorter (<10 mm) and stiffer, forming a dense mat; the cheeks are densely and more or less uniformly covered by hairs that are similar in length and texture to those of the coronal pelage. This pelage of the crown and cheeks is also abruptly paler than the body fur, ranging in mass effect from whitish to yellowish or brownish gray; the individual facial hairs always have dark bases and/or dark tips, however, so even the palest faces appear finely grizzled on close inspection. The muzzle is more sparsely haired than the cheeks and crown, and the exposed skin is blackish; a small patch of blackish skin is also exposed on the center of the forehead just above the brow line in most specimens. The throat is naked and glandular, but the chest has a prominent patch of long, pale-tipped fur; the abdominal pelage is sparse and uniformly black.

The female pelage is similar to the male condition, with a few exceptions. The most obvious difference is that the transition from the long, lax nuchal fur to the shorter fur of the head is not abrupt as it is in males; instead, the cranial hairs gradually decrease in length and increase in stiffness from occiput to forehead. The coronal pelage of females is also much more chromatically variable than that of males, with some females having conspicuously pale (whitish or beige) crowns and foreheads whereas others have much darker (brownish) cranial fur; the individuals with the palest (MUSM 11141) and the darkest (MUSM 11136) cranial fur in our series are both young adults as judged by tooth wear. Although the throat of females is mostly bare, the exposed gular skin is less obviously glandular than it is in males. The pectoral tuft of long, pale-tipped fur is also less well developed in females than it is in males.

Head-and-body length and several osteological measurements show strong sexual dimorphism in our sample (table 15). In general, the craniodental pattern is typical of that seen in most primates (Schultz, 1962) with dimorphism in the facial skull (rostrum, orbits, zygoma) and the canine teeth versus monomorphism in the braincase and postcanine dentition. Three qualitative cranial traits exhibit marked variation in our sample, which we analyzed for association with age (“young” adults versus “old” adults, as determined by toothwear and basicranial suture closure) and sex using Fisher's exact test. (1) The sagittal crest (coded as present or absent) is highly significantly associated with sex and with age (p ≪ 0.01 for both tests), being present only in males and more frequently in old adults than in young adults. (2) The mesopterygoid fossa is quite variable in its anterior extent, terminating behind the toothrows in some individuals (e.g., AMNH 268244, MUSM 11145) and penetrating deeply between the molars in others (MUSM 11127, 11128), apparently independently of age and sex. (3) The pterygoid fossae are shallow and unenclosed by bone in most individuals but more deeply excavated in others (MUSM 11133, 11142, 11145), variation that also appears to be unassociated with age or sex.

TABLE 15

External and Craniodental Measurements (mm) and Weights (g) of Pithecia monachus from the Yavarí-Ucayali Interfluve

In qualitative cranial characters, the Río Gálvez sakis all agree with Hershkovitz's (1987c: table IV) diagnosis of the Pithecia monachus group, an assemblage that includes P. aequatorialis, P. albicans, P. irrorata, and P. monachus. According to Hershkovitz, taxa within this group are distinguishable only by pelage traits. In the comparative material we examined, the distinguishing external characters of all the saki species recognized by Hershkovitz are readily apparent, at least in adult males; by his classification, our vouchers are unambiguously referable to P. monachus. By contrast, P. irrorata (which occurs south of the Amazon between the Juruá and the Tapajós; Hershkovitz, 1987c) has an almost-naked head, the exposed blackish skin of which is conspicuously set off by an encircling white-fringed ruff. Another geographically adjacent form, P. aequatorialis (which occurs north of the Amazon between the Tigre and the Napo; Bravo and Ríos, 2007; Aquino et al., 2009) has long reddish or red-based fur on the cheeks (resembling the mutton-chop whiskers of 19th-century gentlemen) and a conspicuously reddish chest. Pending a comprehensive analysis of character variation in the monachus group, these and other character differences seem sufficient to maintain the nominal taxa recognized by Hershkovitz as distinct, but whether or not these should be recognized as full species remains to be convincingly demonstrated.15 In the event that a trinomial classification is judged to better represent the situation, our material from Nuevo San Juan would be referable to the nominotypical form P. monachus monachus.

Ethnobiology

Saki monkeys are called bëşhuidkid, a term that is perhaps etymologically analyzable as meaning “white-faced one.” Two subtypes are recognized by the Matses: mamu (with no meaning in Matses, but in closely related languages it means “pitch,” a black adhesive made of beeswax and tree latex) and bëshudu (not analyzable, except possibly the prefix bë- “face/forehead”). The mamu variety is said to be larger, have a whiter face, whiter hands, and a bushier coat, while the bëshudu variety is said to be smaller, to have a black or darker face, less white on its hands and a thinner coat. Troops of mamu are said to be found more frequently in upland habitats far from large rivers, and bëshudu more frequently along rivers. A minority of Matses hunters do not recognize two varieties and consider mamu and bëshudu to be synonyms of bëşhuidkid. Fleck et al. (1999) discussed this ethnobiological issue in detail and summarized morphological and genetic evidence that the Matses overdifferentiate saki monkeys.

All Matses eat sakis. As with other monkeys, sakis are cooked by first singing the hair off and then boiling them; likewise, men cannot eat or touch the intestines. Children usually eat the head. The Matses believe that if they give a dog a saki monkey bone, the dog will die; therefore, they are always mindful of this when disposing of the bones.

The Matses hunt saki monkeys by listening for their grunts and calls or the noise they make rustling branches as they move through the trees, or by looking for their dropped half-eaten fruits. When a hunter with dogs encounters saki monkeys and the monkeys do not see him, he imitates their raptor-warning call, upon which all the monkeys drop to the ground. The hunter then looses his dogs, which chase the monkeys down as they run away. Hunters also run after monkeys on the ground and club them or slam them against a tree when they catch them. (Saki monkeys will bite and sometimes even kill dogs that are trying to catch them. They will also bite people, if they are not careful when grabbing them.) If saki monkeys are familiar with this trick, they will not drop to the ground when a hunter imitates their raptor-warning call.

Now that most Matses are using shotguns, hunters without dogs will often ignore saki monkeys early in the hunt because the amount of meat they have is hardly worth a shell. Likewise, if hunters have dogs but the saki monkeys will not drop to the ground, hunters will probably not shoot at them early in the hunt. If hunters encounter sakis on the way back from an unsuccessful hunt, however, they will often shoot.

Sakis in trees are shot from the ground with arrows or shotguns. When they see people, sakis will not drop to the ground but run away very fast over the tops of branches and jumping from tree to tree. Hunters may chase one down until it gets tired and then shoot it (the Matses do not usually shoot at moving targets). Sometimes a hunter will climb a tree to kill a tired-out saki monkey that he cannot shoot from the ground, but the monkey may evade the hunter by dropping to the ground and running off.

Matses natural history

Saki monkeys have a lot of hair. Their bodies are actually small, but they look big because of all the hair. Their coats look like their hair has started to go gray, like an old person's hair. Their faces and hands are white. The tail is furry and looks big, but it does not have much meat on it and cannot be used to grab onto things. Their head hair is also long, but not on the forehead. They are strong and run very fast through the trees without falling down. They run fast on the ground as well. They run fast even with their young on their backs. Saki monkeys can drop to the ground without hurting themselves.

Saki monkeys are found in all types of habitats, in upland forest and floodplain forest, in primary forest and in blowdowns. They travel high or low in the trees.

They are found in small troops, fewer than 10 animals, or sometimes alone. The troop usually stays together to travel, feed, and sleep, though occasionally they split up to search for food. Saki monkeys wake up just before dawn. When they wake up in the morning they go “she she she aa aa aa.” When they see a jaguar, they say “waa waa waa.” Sometimes they feed in the same tree with woolly or uakari monkeys. The whole troop sleeps in the same tree, wherever they are when night falls [i.e., they have no permanent sleeping sites].

Saki monkeys eat mostly dicot fruits, including wesnid dëbiate [Anacardium giganteum (Anacardiaceae)], këku [Couma macrocarpa (Apocynaceae)], diden këku [Parahancornia peruviana (Apocynaceae)], machishte [Rhigospira quadrangularis and ?Mucoa duckei (Apocynaceae)], mamuin [Rheedia longifolia (Guttiferae)], okodo mabis [undetermined sp. (Guttiferae)], moste [Hymenaea spp. (Leguminosae)], mannan tsipuis [Inga spp. and ?Pithecellobium (Leguminosae)], tankada [Parkia igneiflora, P. multijuga, and Pithecellobium auriculatum (Leguminosae)], şhupuduşh [Brosimum parinarioides (Moraceae)], şhupud [Brosimum and Poulsenia (Moraceae)], şhannëd [?Brosimum (Moraceae)], bin [Castilla (Moraceae)], dadain [Clarisia racemosa (Moraceae)], piuşh bëchi [Helicostylis tomentosa and H. elegans (Moraceae)], kuşhu tëbin [Naucleopsis mello-barretoi and N. ternstroemiiflora (Moraceae)], şhankuin [Pourouma spp. (Moraceae)], bata [Pseudolmedia and Maquira spp. (Moraceae)], poşhodi [Passiflora nitida (Passifloraceae)], mabis mabiskid [Chrysophyllum prieurii (Sapotaceae)], and kose [Manilkara bidentata (Sapotaceae)]. They eat some of the dicot fruits before they are ripe. They throw down a lot of fruits as they eat. They eat palm fruits, especially swamp-palm [Mauritia flexuosa] fruit and some isan [Oenocarpus bataua] fruit. They also eat some leaves and insects.

Spider monkey natural history

We analyzed recorded monologs about spider monkeys by enumerating units of ecobehavioral information that could be checked for consistency against the scientific literature (after Townsend, 1995). Because there have been relatively few field studies of the local species, Ateles belzebuth, we regarded Matses observations as corroborated by the scientific literature if similar observations had been published by field biologists for any spider monkey species. Also, because some field studies provide incomplete taxonomic identifications of ingested fruits, we assessed dietary information about food plants at the generic level.

Of the 86 ecobehavioral information units tallied from Matses monologs about Ateles belzebuth, two concerned unidentified plant taxa in its diet: şhanned (possibly corresponding to one or more species of Brosimum [Moraceae]) and okodo mabis (an undetermined species of Guttiferae). As no confident assessment of these dietary items is possible in the absence of positive botanical matches, they were excluded from consideration. Of the remaining 84 ecobehavioral information units, 63 (75%) were corroborated by published information from scientific field studies (table 19). These corroborated observations include traits that are common to all spider monkey species, such as tail-assisted brachiation (item 3) and frugivory (item 59), as well as behaviors that could be species or site specific. Examples of species-specific traits might be the complex of behaviors associated with visiting mineral licks (items 48–56), which are apparently unknown in other well-studied congeners; another might be the habit of eating unpeeled fruits (item 61), a behavior that is apparently not shared with A. paniscus (see Roosmalen, 1985). An example of a site-specific behavior might be use of the middle level of the canopy (item 7): as suggested by Mittermeier (1978), use of different canopy levels could represent behavioral plasticity in response to differences in forest structure among study sites rather than evolved species-specific traits.

TABLE 19

Ecobehavioral Information Recovered from Matses Monologs about Spider Monkeys (Ateles belzebuth)

Of the remaining 21 uncorroborated ecobehavioral observations that the Matses reported for Ateles belzebuth, we judged 20 to be plausible because they (1) are reasonable extrapolations from published scientific observations, (2) could be expected to depend on local conditions, (3) are behaviors that only monkey hunters might think to mention, and/or (4) represent subjective judgments of similarity. An example of a plausible observation in the first category is that small spider monkeys are taken by harpy eagles (item 19). Although harpy eagle (Harpia harpyja) predation on spider monkeys has not been reported in the scientific literature, harpy eagles are known to take howler monkeys (Peres, 1990; Sherman, 1991) and there is a report of a crested eagle (Morphnus guianensis) taking a young spider monkey (Julliot, 1994). Another plausible observation of the first kind is that spider monkeys climb higher when they hear a tinamou fly (item 11). Although this also seems not to have been reported in the literature, Roosmalen (1985) notes that Guianan spider monkeys (A. paniscus) become upset by the flight of alarmed terrestrial animals (including tinamous), and climbing higher would be a prudent response to the arrival of an unseen predator.

An example of a plausible observation in the second category is that spider monkeys are more common than woolly monkeys. Although we have not encountered any published study in which this has been reported (indeed, most reports suggest that Lagothrix is more common than Ateles where both taxa are present; e.g., Izawa, 1975), it is reasonable to assume that some aspect of the local flora or phenology might favor Ateles in the Yavarí-Ucayali interfluve. Examples of plausible observations in the third category include items 35 and 36, neither of which is addressed in the scientific literature, but which are consistent with inferences from spider monkey kills by other native Amazonian hunters (Shepard, 2002; da Silva et al., 2005). A plausible observation of the fourth kind is the remark that spider monkeys suckle their young like people do (item 18), a simile that does not appear in the scientific literature, and one that we are unable to evaluate in the absence of any explicit, published description of nursing in Ateles.

In effect, only one Matses observation about Ateles belzebuth seems potentially problematic: that a large male leads the troop (item 15). This behavior does not appear to have been reported in any published field study of spider monkeys. Indeed, at least one study suggests that a dominant female leads most foraging parties (Roosmalen, 1985). Perhaps the Matses, who do not carry binoculars, are confused about the sex of leading animals, or they may assume that behavioral dominance in agonistic encounters between adult males and females implies a leadership role for males. A third possibility, of course, is that the Matses are correct, and that the local population of A. belzebuth differs in this respect from populations at other spider monkey research sites.

Community patterns

Although we have not undertaken similarly detailed analyses of Matses natural history accounts for other monkeys, we tabulated interspecific comparisons of diet, habitat use, and other behaviors to illustrate various aspects of primate community structure that can be reconstructed from recorded monologs. These are interesting topics in themselves, and they provide another context for ethnobiological fact-checking. Because there have been no long-term synecological primate studies in the Yavarí-Ucayali interfluve, Matses observations are a potentially useful source of information about patterns of resource use by the local primate community.

Matses information about feeding behavior (table 20) is broadly consistent with the results of many field studies that have identified frugivory as the dominant trophic characteristic of Neotropical primate faunas (Izawa, 1975; Mittermeier and Roosmalen, 1981; Soini, 1986; Terborgh, 1983; Rylands, 1987; Peres, 1993b). Additionally, the Matses accurately describe many taxon-specific dietary features (described or referenced by the same authors), such as leaf consumption by Alouatta, geophagy by Alouatta and Ateles, vertebrate predation by both species of Cebus, “bark” (gum) consumption by Callithrix, and consumption of unripe fruits by Pithecia. However, Matses informants failed to mention several aspects of primate diets documented by scientific field studies. The Matses admit that they know little about the foods eaten by night monkeys, but other notable omissions from recorded monologs include arthropod consumption by Callithrix pygmaea (see Soini, 1988) and vertebrate predation by both local species of Saguinus (see Heymann et al., 2000). The general pattern apparent in these data is that Matses dietary observations are less complete for the smaller primates than they are for larger monkeys, but there are a few omissions even for the largest species.16

TABLE 20

Diets of Local Primate Species Based on Matses Natural History Monologs^a

Information gleaned from Matses monologs about community use of specific food plants (table 21) indicates that some fruits are eaten promiscuously, affording nourishment to many species of monkeys, whereas other fruits are eaten by just a few species. Fruits in the first category include some that are popular items in primate diets elsewhere in western Amazonia (Terborgh, 1983; Stevenson et al., 2000), notably including several genera of Moraceae. However, other locally important food-plant taxa are rarely mentioned in the primate literature. The family Apocynaceae, for example, is seldom represented among lists of important primate fruits at other Amazonian localities, and the swamp palm (Mauritia flexuosa) has only been identified as an important food resource for primates at one other site in northeastern Peru (Aquino and Bodmer, 2004). Some higher-taxonomic absences are also noteworthy. Among others, the families Annonaceae and Sapindaceae, both commonly eaten by primates in southeastern Peru (Terborgh, 1983) are not represented among the fruits positively identified as local monkey fodder by the Matses. Because there are noteworthy floristic, climatic, and edaphic differences between Matses hunting grounds in the Yavarí-Ucayali interfluve and many other Amazonian study sites, environmental factors may be responsible for some local peculiarities of primate fruit consumption.17 Additionally, almost all of the plant taxa mentioned in Matses monologs are trees, and important primate food plants that have other growth forms (e.g., Sapindaceae, many of which are vines) might be underrepresented because they are seldom named by the Matses.

TABLE 21

Fruits Eaten by Local Primate Species Based on Matses Monologs

Other interactions between the local primate fauna and particular plant taxa raise some interesting questions. Matses informants definitely assert that only howler monkeys eat the fruit of Apeiba aspera (Tiliaceae), but elsewhere in Amazonia the fruit of A. aspera is one of the preferred foods of capuchins and is strongly avoided by howlers (Stevenson et al., 2000). Similarly, the Matses assert that only titi monkeys eat the fruit of the palm Euterpe precatoria (capuchins are said to eat just the unripe seeds), but the fruit of E. precatoria is eaten by spider monkeys at many other localities (Roosmalen, 1985; Castellanos and Chanin, 1996; Iwanaga and Ferrari, 2001; Wallace, 2006). Whether such anomalies are attributable to intersite differences in environmental factors alleged to influence primate food preferences (Russo et al., 2005) or to observer error is unknown.

Community patterns of habitat and substrate use reconstructed from Matses monologs (tables 22, 23) are generally consistent with the scientific literature. All of the atelids, for example, are accurately portrayed as using only the canopy of primary forest; howlers and squirrel monkeys are said to strongly prefer floodplain forest; Cebus apella and C. albifrons are said to prefer floodplain and upland forest respectively; titis are said to strongly prefer forest edges and secondary growth; the locally sympatric tamarins are said to differ in their preference for primary forest versus secondary growth; etc (Castro and Soini, 1977; Rylands, 1987; Peres, 1993c, 1997; Haugaasen and Peres, 2005; Parry et al., 2007). Significantly, the Matses say that uakaris use both floodplain and upland forests, in agreement with recent field studies conducted within the Yavarí-Ucayali interfluve but in contrast with reports from other regions, where these monkeys are said to be flooded-forest specialists (Aquino, 1998; Heymann and Aquino, 2010).

TABLE 22

Habitat Utilization by Local Primate Species Based on Matses Monologs^a

TABLE 23

Substrate Use by Local Primate Species Based on Matses Monologs^a

Matses observations about interspecific interactions (table 24) are not extensive, but they include some that are familiar from the scientific literature, such as the common association between Cebus apella and Saimiri sciureus (see Podolski, 1990). Interestingly, the Matses observation that C. albifrons does not associate with Saimiri conflicts with the results of field studies in southern Peru (where Saimiri seems to associate with both C. apella and C. albifrons; Terborgh, 1983), but it is easier to reconcile with field studies from adjacent regions in northern Peru (where Saimiri associates much more frequently with C. apella than with C. albifrons; Soini, 1986). Most of the other interspecific interactions reported by the Matses are unremarkable, but to our knowledge there has been only one previous report that Cebus apella kills and eats titi monkeys (Sampaio and Ferrari, 2005). The lack of Matses observations about interactions involving Ateles and Alouatta is consistent with published generalizations that these genera rarely associate with other primates (Neville et al., 1988; Roosmalen and Klein, 1988). However, unreported interactions between Saguinus fuscicollis and S. mystax (which often form mixed troops; Heymann, 1990; Peres, 1993b; Heymann and Buchanan-Smith, 2000) may reflect cultural inattention to the habits of small monkeys.

TABLE 24

Interspecific Interactions among Local Primates Based on Matses Monologs^a

In summary, most Matses observations about primate behavior and natural history seem to be accurate (to the extent that they overlap with the results of published field studies) or to be plausible extensions of previously reported facts. Although the Matses are clearly biased observers, providing much more extensive and detailed information about large game species than about small species of little or no cultural importance, the format-free context of our recorded monologs may have been responsible for the omission of some relevant information even about culturally important taxa. A great deal more information about primate natural history could presumably have been obtained from question-and-answer interviews, but there is an obvious risk of eliciting misleading responses to leading questions when interviewees are eager to please.

Primatologists may be encouraged by these results to explore the ethnobiological knowledge of other native Amazonians, which might be a cost-effective alternative to lengthy fieldwork for some research objectives. However, there are significant linguistic barriers to effective cross-cultural communication. Some degree of bilingual fluency is obviously essential on the part of the interviewer or the interviewee, but it is also important to understand relevant aspects of ethnobiological classification and nomenclature. Among the Matses, these include a surprisingly large number of zoologically redundant names (synonyms and hyponyms) as explained below.

Principal term

The name for each species that we designate the “principal term” is (1) the term most frequently overheard; (2) the one that everyone knows, including children; (3) the first term offered when eliciting names using field guides or specimens; and (4) the term that speakers themselves say is the principal name when more than one name applies to a species. There is no informant disagreement about principal terms for primates. Intervillage lexical variation for principal names of other mammals is restricted to poorly known and culturally unimportant taxa, such as short-eared dogs (Atelocynus microtis, which are extremely rare), and opossums (which are seldom observed and of no cultural significance). Apparently, the pygmy marmoset (Callithrix pygmaea) is not quite obscure enough to be among the taxa for which there is disagreement about principal names.

“Archaic” synonyms

Many game species have, in addition to the principle name, several synonyms that every good Matses hunter is expected to know (Fleck and Voss, 2006). Fathers and grandfathers teach their sons and grandsons these synonym sets, and visiting older men routinely quiz young men on these terms. Although these synonyms are said to be tsusiobon onkete “old men's speech,” there are etymological reasons to doubt that all are genuinely archaic terms. Polymorphemic (linguistically segmentable) synonyms that are obviously descriptive (e.g., abukkid “one that is high up” for woolly monkeys) or onomatopoetic (such as pishtadan for tamarins), for example, seem likely to be of recent origin. By contrast, monomorphemic (unsegmentable) terms with cognates in other Panoan languages (e.g., çhuna, another synonym for wooly monkeys) might have been inherited from some ancestral protolanguage, or they might be modern borrowings from a sister tongue.

There is a clear pattern in Matses folk-biological nomenclature that culturally important species (e.g., Lagothrix lagothricha) tend to have more synonyms than culturally insignificant species (e.g., Callithrix pygmaea), but this correlation is less apparent among species of intermediate cultural salience. For example, one would expect the uakari monkey (Cacajao calvus) to have at least one or two synonyms given that the smaller and less important titi monkey (Callicebus cupreus) has three. Historical factors might explain anomalies like this if synonymy reflects long-term rather than current cultural importance. (It is clear from Matses testimonials and from known historical accounts that the Matses have not been in continuous contact with uakari monkeys in the past.) The saki monkey (Pithecia monachus) might also be expected to have synonyms, but we note that a minority of Matses informants who do not recognize varieties of saki monkeys consider the terms mamu and bëshudu to be synonyms of bëşhuidkid. We will return to this distinction between synonyms and hyponyms (named subtypes or varieties) in our discussion of overdifferentiation below.

Ceremonial terms

This category applies to words from a secret language that was formerly chanted or spoken only in the komok or “singing souls” ceremony (Romanoff 1984; Romanoff et al., 2004), when invented, archaic, or borrowed words were substituted for common usage in Matses. The pattern noted above, of giving special linguistic attention to culturally important animals, can also be seen in the komok ceremonial language. Of the 56 documented substitute terms in the komok language, 19 are animal names, of which 14 apply to game species and two to other culturally significant animals (jaguar/dog and bushmaster).

Only the two most culturally important primates, the woolly monkey and the spider monkey, have a komok ceremonial term, and the same substitute name (şhëmën kudu) is used for both species. This follows a general trend in the komok lexicon: ceremonial terms are often less specific (and never more specific) than non-ceremonial terms. Two other examples of this type of underdifferentiation occur among the ceremonial terms for mammals. The term mëndu is used for both the jaguar and the domestic dog (a pattern also found in everyday usage in other Panoan languages, but not with a cognate term), and the term ana pachi is used for both the paca and the agouti, which in everyday usage have completely different Matses names.

Overdifferentiated variety

This category refers to one of three logical possibilities that can result from comparisons of “folk species” and biological species: (1) one-to-one correspondence; (2) underdifferentiation; and (3) overdifferentiation. One-to-one correspondence is the result when a single folk species (e.g., senta) corresponds to exactly one biological species (Cacajao calvus; table 25); underdifferentiation results when a single folk species (e.g., cuesban) corresponds to more than one biological species (at least 57 species of bats; Fleck et al., 2002); and overdifferentiation refers to cases where a single biological species (e.g., Pithecia monachus) is denoted by two or more non-synonymous folk species names (bëshudu and mamu; Fleck et al., 1999).

With the exception of the komok ceremonial term that applies to both spider monkeys and woolly monkeys, as discussed above, there are no cases of underdifferentiation in the Matses primatological lexicon, and there are only five unambiguous examples of one-to-one correspondence (Cacajao calvus, Cebus albifrons, Saguinus fuscicollis, S. mystax, and Callithrix pygmaea each have unique names; table 25). By contrast, overdifferentiation is relatively common. Although overdifferentiation results in multiple names that apply to the same biological species, these names are hyponyms rather than synonyms because they apply to different real or alleged varieties.

As with the synonym sets described above, good hunters are expected to know hyponyms for different game animals and how to identify the corresponding overdifferentiated varieties, whereas this knowledge is absent or incomplete for children, women, and “worthless men.” Also, as noted by Fleck and Voss (2006), there is a tendency for overdifferentiated species to be the same culturally important ones that receive multiple synonyms. Part of the explanation for this may be that hunters have frequent opportunities to inspect the physical details of killed game animals, and this personal experience allows them to detect subtle phenotypic differences among individuals or age cohorts that might be interpreted as analogous to the discontinuities that exist between closely related species. In fact, Matses hyponyms for saki monkeys (bëshudu and mamu) appear to be based on age-correlated variation in size and pelage traits rather than genetic differences (Fleck et al., 1999). Information summarized in the preceding species accounts suggest that Matses hyponyms for other primates may also reflect ontogenetic or individual variation in size and pelage, but we cannot rule out two alternative explanations for some names: (1) the possibility, however remote, that additional primate taxa remain to be discovered in the Yavarí-Ucayali interfluve; or (2) that the Matses retain information in their lexicon about species previously encountered by the tribe before they settled in their current territory.

By contrast with game synonym sets, there seems to be no conscious attempt by the Matses to standardize overdifferentiation (e.g., by ritually quizzing young men and boys; Fleck and Voss, 2006), perhaps because overdifferentiation is the result of observing biological variability and is therefore subject to individual interpretation. In effect, hyponyms denote purported discontinuities between named varieties that some hunters may eventually find to be false. Accordingly, some hyponyms are topics of speaker disagreement (with corresponding majority and minority views; table 25).

Several additional aspects of Matses ethnobiological overdiffertiation are noteworthy. One is that speaker disagreement about hyponyms for primates (and other taxa) usually involves species of intermediate cultural importance. Part of the reason for this is that small and/or culturally insignificant species are not overdifferentiated by anyone, perhaps because they are infrequently observed (e.g., Cebuella pygmaea) and seldom talked about. Large, culturally important species, on the other hand, are often seen and frequently discussed, so that hyponyms become widely known and accepted among hunters.

Another interesting aspect of Matses overdifferentiation is that there is often one superordinate term and two hyponyms. The three exceptions among the primate examples are woolly and howler monkeys, each of which has three hyponyms, and the night monkey (Aotus), which has no superordinate term. As if motivated by such anomalies, we find interspeaker disagreement about the classification of these species. Nomenclature for folk “genera” with two local species—bëchun (Cebus) and sipi (Saguinus)—follow this same pattern, suggesting that these biologically legitimate classifications may have been the model for the apparently superfluous practice of overdifferentiating game species. With respect to this labeling pattern, we note that sometimes the hyponyms are descriptive modifications of the superordinate term (e.g., achu çhëşhë “black howler monkey”), while in other cases one or two hyponyms will be monomorphemic.

In three of the five cases where there is disagreement about the status of an “archaic” synonym, the disagreement hinges upon whether the term is really a synonym or a hyponym; examples include the two saki monkey (Pithecia) names bëshudu and mamu. By contrast, the disputed “archaic” synonyms masoko (for the titi monkey, Callicebus cupreus) and chidu (for capuchin monkeys, Cebus spp.) are simply not recognized as legitimate by a subset of hunters who claim they are a borrowing and a term used only in myths, respectively. Closer inspection reveals that all cases of disagreement about whether a hyponym exists at all as a legitimate term involve descriptive polymorphemic terms, whereas disagreements about whether a term is a hyponym or a synonym involve monomorphemic terms. The Matses are conscious of this linguistic distinction. When they argue about animal nomenclature, the nature of the term in question will be focused on. If a speaker is trying to discredit an overdifferentiated variety that is designated by a descriptive hyponym, he may say the term is simply an ad hoc expression that can be used to describe individuals in a variable but continuous species (the high phenotypic variability of domestic dogs is often mentioned as an example). Alternatively, when the hyponym in question is monomorphemic, the person discrediting the overdifferentiated category is unlikely to suggest discarding the term. Rather, he will typically suggest it is actually a synonym (as with saki monkeys) or less frequently, a borrowing (as with titi monkeys).