AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 88:483498 (1992)
Morphological and Behavioral Adaptations for Foraging in Generalist Primates: The Case of the Cebines CHARLES H. JANSON ANDSUE BOINSKI Department of Ecology and Evolution, State University of New York, Stony Brook, New York 11794 (C.H.J.); Laboratory of Comparative Ethology. N K H D , NIH, Animal Center, P.O. Box 289, Poolesuille, Maryland 20837 (S.B.)
KEY WORDS Body size
Cebinae, Saimiri, Cebus, Omnivory, Foraging,
ABSTRACT
In addition to being frugivorous, Cebus and Saimiri stand out among the New World primates of similar body size in being heavily dependent on animal matter for protein (faunivory). A detailed description of the morphology and behavior of the two genera is presented with the object of evaluating the interaction and respective contributions of morphological and behavioral adaptations to foraging patterns. Our conclusions include the following: First, body size is extremely important in explaining the observed variation in diet. Second, the emphasis on faunivory is facilitated more by behavioral than by morphological specialization. Third, whatever morphological specializations are present, particularly in Cebus, are probably favored by diet a t the most food-depauperate time of year. Fourth, although morphology may well reveal what a primate may potentially eat, to map this potential onto actual diet requires a detailed knowledge of its natural ecosystem. Finally, we consider whether the behavioral data support the tenuous morphological evidence for grouping Cebus and Saimiri within the clade Cebinae. o 1992 Wiley-Liss, Inc.
Despite the considerable emphasis in physical anthropology on relating morphological features of primates to their ecology, current correlational studies between morphology and diet leave significant dietary variation unexplained (Kay, 1984). Part of the difficulty in relating morphology to diet is that morphology can at best reveal only the broad limits of what diet can be (Scheine and Kay, 1982; Milton, 1984). However, in nature, foods with distinct processing needs are not equally abundant. Thus, primate populations living in different habitats may differ in diet despite similarities in morphology (Sussman 1987). Another reason that morphology and diet may not always match is that food processing requires both morphology and behavior: some primate species may use specialized behaviors to overcome physical limitations that would otherwise hinder the ingestion of certain foods; for ex0 1992 WILEY-LISS. INC.
ample, Cercopithecus species with low molar shearing capacity consume many insects, but discard the chitin (Kay, 1984). The relative importance of behavioral components in determining food choice is likely to be greatest in species that can derive some nutritional benefit from almost any type of food. Without strong constraints on what such species can eat, food choice may well be determined by how quickly individuals can learn to find or process different food items. Niche specializations in this situation may be largely determined by behavioral adaptations. In this paper, we discuss behavioral and morphological aspects of foraging in squirrel monkeys and capuchins,
Received July 11,1990; accepted December 3,1991. Dr. Boinsky is now a t the Department of Anthropology, 1350 Turlington, University of Florida, Gainesville, FL 32611.
484
C.H. JANSON AND S. BOINSK-I
cific coast, and C. oliuaceus (formerly nigriuittatus)north of the Amazon and east of the Rio Negro (Hershkovitz, 1949). The two species groups differ in that C. apella are slightly heavier and considerably more “robust” in jaw morphology than are the remaining species (Kmzey, 1974). In contrast to Cebus, the infrageneric taxonomy of Saimiri is unsettled. It is probably best to consider Saimiri a s a old species complex with extensive local adaptation. Some authors (e.g., Napier and Napier, 1967; Thorington, 1985) divide all Saimiri populations into only two species (sciureusin South America, oerstedi in Central America), with TAXONOMY AND PHYLOGENY marked differences among populations. OF THE CEBINAE Hershkovitz (1984), however, uses these The subfamily Cebinae contains two gen- same population differences to argue for the era: Saimiri, the squirrel monkeys, and Ce- existence of at least four species in two spebus, the capuchin monkeys. Yet, it should be cies groups: a boliviensis group consisting of stressed that the position of Saimiri with S. boliuiensis, and a sciureus group consistrespect to Cebus is equivocal. Until the early ing of s. sciureus, s. oerstedi, and s. ustus. 1970s, Saimiri was considered more closely The two groups are also commonly described related to Cebus than to any other platyr- as, respectively, Roman and Gothic squirrel rhine genus (Rosenberger, 1981). Since monkeys based on facial pelage (MacLean, then, Hershkovitz (1977) has argued on 1964). Although animals from different morphological grounds that Saimiri and Ce- parts of South America differ in many genetbus are separate lineages, a s have Sarich ically based traits, designation of Saimiri and Cronin (1980) using molecular data. In populations as species or subspecies is a detailed examination of platyrrhine sys- largely arbitrary, as they mate readily in tematics, Ford (1986) concluded that Cebus, captivity (cf. Moore et al., 1990)and they are Saimiri, and the callitrichids were unlikely allopatric in the wild (Hershovitz, 1984). to be closely related. However, Rosenberger Since most detailed ecological field data (1977, 1981) has argued that these three come from two widely separated sites taxa are closely related, with Saimiri and (Parque Nacional Corcovado in Costa Rica, Cebus being each other’s closest relatives Parque Nacional del Manu in Peru), we foland together forming a higher taxon with low the “traditional” taxonomy and refer to the callitrichids. Our report is not premised the former population as oerstedi and the on the assumption that Cebus and Saimiri latter a s sciureus. form a clade. We consider Cebus and Saimiri MORPHOLOGICAL FEATURES together because they have in the past been OF CEBINES grouped together taxonomically, and they Morphologically, Saimiri and Cebus are share a proclivity for faunivory that is unbroadly similar compared with other platyrusual among platyrrhines. Generally, four species of Cebus are recog- rhines. Unlike other Neotropical primates of nized in two species groups: one containing their body size (0.6-1.1 kg for Saimiri, 2-4 only C. apella, which ranges throughout kg for Cebus),the cebines obtain virtually all tropical South America east of the Andes, their protein from insects, rarely ingesting and the other containing three apparently substantial quantities of leaves or other vegparapatric species, C. capucinus in Central etation (Terborgh, 1983). Both genera have America north of the Isthmus of Panama, C. short digestive tracts (Hill, 1960) with a relatbifrons in western Amazonia and the Pa- atively small absorptive surface area (Chiv-
platyrrhines of the subfamily Cebinae. Behavior has, of course, a neural, hence a morphological, basis. Yet we discuss behavior and its neurological proximate mechanisms as distinct from “conventional” skeletal and dental morphology. Because these monkeys are relatively similar in regard to morphology and ecology, particularly faunivory (see below), they provide a n appropriate context in which to examine the relative importance of morphological and behavioral adaptation for foraging. We also employ the behavioral data to examine the hypothesis that Cebus and Saimiri comprise a phylogenetic clade.
FORAGING ADAPTATIONS OF CEBINES
ers and Hladik, 1980), suggesting that they should not be efficient in digesting structural carbohydrates. Moreover, neither genus possesses dental adaptations for folivory. All cebines use predominantly quadrupedal walking or running for locomotion, with the only major difference being in the frequency of leaping (higher in Saimiri than in Cebus, Fleagle and Mittermeier, 1980;Boinski, 1989a;Janson, unpubl. data). Cranial characters shared by Saimiri and Cebus include specializations of the visual cortex, centrally placed foramen magna, short nasal bones, reduced pteromastoids (Rosenberger, 1977; Fleagle and Rosenberger, 1983). Despite these similarities, there are numerous distinctive differences between Saimiri and Cebus, and to a much smaller degree among Cebus species. We suspect that many of these differences are a direct or indirect result of differences in body size; female squirrel monkeys weigh 0.5-0.7 kg, females of the gracile Cebus about 2.2 kg, and those of C. apella about 2.7 kg; adult males weigh 25-35% more than females in all cebines (Janson, 1984a). This fourfold range in body mass affects many factors important to foraging: metabolic demand for energy and protein, absolute strength, agility (the larger species are less agile), and leaping ability (cf. Demes and Gunther, 1989). Body size is certainly not the only feature that distinguishes the three groups of cebines. First, the teeth of Saimiri are more complex, possessing a pronounced cingulum on most teeth; by contrast, in Cebus, the lower cusps have lost all but a small part of the upper lingual cingulum (Kinzey, 1973; Rosenberger, in press). This difference is likely related t o the greater reliance of Saimiri on insects and Cebus on fruits (Terborgh, 1983,1986; Boinski and Timm, 1985; Boinski, 1989b). The relative molar dimensions of Saimiri do not differ from those of Cebus, and both are characteristic of frugivores (Kay, 1975). However, Cebus (especially apella) possess very thick dental enamel, relatively thicker in fact than any other living primate (Kay, 1981; E. Dumont, personal communication). This thick
485
enamel is likely related to the extensive use of very hard or tough plant tissues (espe-
cially palm nuts, Izawa and Mizuno, 1977; Leland and Struhsaker, 1977). Among Cebus, the use of tough plant tissues is most frequent in apella (Terborgh, 19831, which differs from the other Cebus in having a distinctively more robust molar dentition (Kinzey, 1974). Second, the mandible of Saimiri is thinner (relative to its body size) than that of Cebus, suggesting that the former cannot generate as large a biting force (Bouvier, 1986). Compared with other Cebus species, apella has a distinctively deeper, more buttressed mandible, larger zygomatic arches, and sometimes a sagittal crest (Kinzey, 1974; Bouvier, 1986). These differences are all consistent with apella’s greater use of dead branches for insect foraging and of palm nuts as an alternative food to fruits (Terborgh, 1983; Janson, 1985). Third, the manipulative abilities of the two genera differ substantially. Although both genera have better prehension abilities than either the smaller callitrichids or the larger atelines, only Cebus has varied forms of precision grip (Costello and Fragaszy, 1988). The digits of Saimiri are not capable of independent movement nor can they generate much pressure between the thumb and other digits (Fragaszy, 1983). Capuchins have pseudo-opposable thumbs and are capable of moving individual digits (especially the thumb) relatively independently of others (Costello and Fragaszy, 1988). In addition, Cebus have relatively longer digits and hands than those of Saimiri (Fragaszy et al., 1989). Finally, postcranial skeletal differences between Cebus and Saimiri are present. The most obvious of these is the hair-covered prehensile tail of Cebus, from which it can suspend itself for brief periods. The tail of Saimiri, although not prehensile, is relatively large and heavy and seems to act as an effective counterbalance (Boinski, 1989a). The fact that all gracile Cebus use the ground more often than any known population of apella (Robinson and Janson, 1987) or Saimiri (
Morphological and Behavioral Adaptations for Foraging in Generalist Primates: The Case of the Cebines CHARLES H. JANSON ANDSUE BOINSKI Department of Ecology and Evolution, State University of New York, Stony Brook, New York 11794 (C.H.J.); Laboratory of Comparative Ethology. N K H D , NIH, Animal Center, P.O. Box 289, Poolesuille, Maryland 20837 (S.B.)
KEY WORDS Body size
Cebinae, Saimiri, Cebus, Omnivory, Foraging,
ABSTRACT
In addition to being frugivorous, Cebus and Saimiri stand out among the New World primates of similar body size in being heavily dependent on animal matter for protein (faunivory). A detailed description of the morphology and behavior of the two genera is presented with the object of evaluating the interaction and respective contributions of morphological and behavioral adaptations to foraging patterns. Our conclusions include the following: First, body size is extremely important in explaining the observed variation in diet. Second, the emphasis on faunivory is facilitated more by behavioral than by morphological specialization. Third, whatever morphological specializations are present, particularly in Cebus, are probably favored by diet a t the most food-depauperate time of year. Fourth, although morphology may well reveal what a primate may potentially eat, to map this potential onto actual diet requires a detailed knowledge of its natural ecosystem. Finally, we consider whether the behavioral data support the tenuous morphological evidence for grouping Cebus and Saimiri within the clade Cebinae. o 1992 Wiley-Liss, Inc.
Despite the considerable emphasis in physical anthropology on relating morphological features of primates to their ecology, current correlational studies between morphology and diet leave significant dietary variation unexplained (Kay, 1984). Part of the difficulty in relating morphology to diet is that morphology can at best reveal only the broad limits of what diet can be (Scheine and Kay, 1982; Milton, 1984). However, in nature, foods with distinct processing needs are not equally abundant. Thus, primate populations living in different habitats may differ in diet despite similarities in morphology (Sussman 1987). Another reason that morphology and diet may not always match is that food processing requires both morphology and behavior: some primate species may use specialized behaviors to overcome physical limitations that would otherwise hinder the ingestion of certain foods; for ex0 1992 WILEY-LISS. INC.
ample, Cercopithecus species with low molar shearing capacity consume many insects, but discard the chitin (Kay, 1984). The relative importance of behavioral components in determining food choice is likely to be greatest in species that can derive some nutritional benefit from almost any type of food. Without strong constraints on what such species can eat, food choice may well be determined by how quickly individuals can learn to find or process different food items. Niche specializations in this situation may be largely determined by behavioral adaptations. In this paper, we discuss behavioral and morphological aspects of foraging in squirrel monkeys and capuchins,
Received July 11,1990; accepted December 3,1991. Dr. Boinsky is now a t the Department of Anthropology, 1350 Turlington, University of Florida, Gainesville, FL 32611.
484
C.H. JANSON AND S. BOINSK-I
cific coast, and C. oliuaceus (formerly nigriuittatus)north of the Amazon and east of the Rio Negro (Hershkovitz, 1949). The two species groups differ in that C. apella are slightly heavier and considerably more “robust” in jaw morphology than are the remaining species (Kmzey, 1974). In contrast to Cebus, the infrageneric taxonomy of Saimiri is unsettled. It is probably best to consider Saimiri a s a old species complex with extensive local adaptation. Some authors (e.g., Napier and Napier, 1967; Thorington, 1985) divide all Saimiri populations into only two species (sciureusin South America, oerstedi in Central America), with TAXONOMY AND PHYLOGENY marked differences among populations. OF THE CEBINAE Hershkovitz (1984), however, uses these The subfamily Cebinae contains two gen- same population differences to argue for the era: Saimiri, the squirrel monkeys, and Ce- existence of at least four species in two spebus, the capuchin monkeys. Yet, it should be cies groups: a boliviensis group consisting of stressed that the position of Saimiri with S. boliuiensis, and a sciureus group consistrespect to Cebus is equivocal. Until the early ing of s. sciureus, s. oerstedi, and s. ustus. 1970s, Saimiri was considered more closely The two groups are also commonly described related to Cebus than to any other platyr- as, respectively, Roman and Gothic squirrel rhine genus (Rosenberger, 1981). Since monkeys based on facial pelage (MacLean, then, Hershkovitz (1977) has argued on 1964). Although animals from different morphological grounds that Saimiri and Ce- parts of South America differ in many genetbus are separate lineages, a s have Sarich ically based traits, designation of Saimiri and Cronin (1980) using molecular data. In populations as species or subspecies is a detailed examination of platyrrhine sys- largely arbitrary, as they mate readily in tematics, Ford (1986) concluded that Cebus, captivity (cf. Moore et al., 1990)and they are Saimiri, and the callitrichids were unlikely allopatric in the wild (Hershovitz, 1984). to be closely related. However, Rosenberger Since most detailed ecological field data (1977, 1981) has argued that these three come from two widely separated sites taxa are closely related, with Saimiri and (Parque Nacional Corcovado in Costa Rica, Cebus being each other’s closest relatives Parque Nacional del Manu in Peru), we foland together forming a higher taxon with low the “traditional” taxonomy and refer to the callitrichids. Our report is not premised the former population as oerstedi and the on the assumption that Cebus and Saimiri latter a s sciureus. form a clade. We consider Cebus and Saimiri MORPHOLOGICAL FEATURES together because they have in the past been OF CEBINES grouped together taxonomically, and they Morphologically, Saimiri and Cebus are share a proclivity for faunivory that is unbroadly similar compared with other platyrusual among platyrrhines. Generally, four species of Cebus are recog- rhines. Unlike other Neotropical primates of nized in two species groups: one containing their body size (0.6-1.1 kg for Saimiri, 2-4 only C. apella, which ranges throughout kg for Cebus),the cebines obtain virtually all tropical South America east of the Andes, their protein from insects, rarely ingesting and the other containing three apparently substantial quantities of leaves or other vegparapatric species, C. capucinus in Central etation (Terborgh, 1983). Both genera have America north of the Isthmus of Panama, C. short digestive tracts (Hill, 1960) with a relatbifrons in western Amazonia and the Pa- atively small absorptive surface area (Chiv-
platyrrhines of the subfamily Cebinae. Behavior has, of course, a neural, hence a morphological, basis. Yet we discuss behavior and its neurological proximate mechanisms as distinct from “conventional” skeletal and dental morphology. Because these monkeys are relatively similar in regard to morphology and ecology, particularly faunivory (see below), they provide a n appropriate context in which to examine the relative importance of morphological and behavioral adaptation for foraging. We also employ the behavioral data to examine the hypothesis that Cebus and Saimiri comprise a phylogenetic clade.
FORAGING ADAPTATIONS OF CEBINES
ers and Hladik, 1980), suggesting that they should not be efficient in digesting structural carbohydrates. Moreover, neither genus possesses dental adaptations for folivory. All cebines use predominantly quadrupedal walking or running for locomotion, with the only major difference being in the frequency of leaping (higher in Saimiri than in Cebus, Fleagle and Mittermeier, 1980;Boinski, 1989a;Janson, unpubl. data). Cranial characters shared by Saimiri and Cebus include specializations of the visual cortex, centrally placed foramen magna, short nasal bones, reduced pteromastoids (Rosenberger, 1977; Fleagle and Rosenberger, 1983). Despite these similarities, there are numerous distinctive differences between Saimiri and Cebus, and to a much smaller degree among Cebus species. We suspect that many of these differences are a direct or indirect result of differences in body size; female squirrel monkeys weigh 0.5-0.7 kg, females of the gracile Cebus about 2.2 kg, and those of C. apella about 2.7 kg; adult males weigh 25-35% more than females in all cebines (Janson, 1984a). This fourfold range in body mass affects many factors important to foraging: metabolic demand for energy and protein, absolute strength, agility (the larger species are less agile), and leaping ability (cf. Demes and Gunther, 1989). Body size is certainly not the only feature that distinguishes the three groups of cebines. First, the teeth of Saimiri are more complex, possessing a pronounced cingulum on most teeth; by contrast, in Cebus, the lower cusps have lost all but a small part of the upper lingual cingulum (Kinzey, 1973; Rosenberger, in press). This difference is likely related t o the greater reliance of Saimiri on insects and Cebus on fruits (Terborgh, 1983,1986; Boinski and Timm, 1985; Boinski, 1989b). The relative molar dimensions of Saimiri do not differ from those of Cebus, and both are characteristic of frugivores (Kay, 1975). However, Cebus (especially apella) possess very thick dental enamel, relatively thicker in fact than any other living primate (Kay, 1981; E. Dumont, personal communication). This thick
485
enamel is likely related to the extensive use of very hard or tough plant tissues (espe-
cially palm nuts, Izawa and Mizuno, 1977; Leland and Struhsaker, 1977). Among Cebus, the use of tough plant tissues is most frequent in apella (Terborgh, 19831, which differs from the other Cebus in having a distinctively more robust molar dentition (Kinzey, 1974). Second, the mandible of Saimiri is thinner (relative to its body size) than that of Cebus, suggesting that the former cannot generate as large a biting force (Bouvier, 1986). Compared with other Cebus species, apella has a distinctively deeper, more buttressed mandible, larger zygomatic arches, and sometimes a sagittal crest (Kinzey, 1974; Bouvier, 1986). These differences are all consistent with apella’s greater use of dead branches for insect foraging and of palm nuts as an alternative food to fruits (Terborgh, 1983; Janson, 1985). Third, the manipulative abilities of the two genera differ substantially. Although both genera have better prehension abilities than either the smaller callitrichids or the larger atelines, only Cebus has varied forms of precision grip (Costello and Fragaszy, 1988). The digits of Saimiri are not capable of independent movement nor can they generate much pressure between the thumb and other digits (Fragaszy, 1983). Capuchins have pseudo-opposable thumbs and are capable of moving individual digits (especially the thumb) relatively independently of others (Costello and Fragaszy, 1988). In addition, Cebus have relatively longer digits and hands than those of Saimiri (Fragaszy et al., 1989). Finally, postcranial skeletal differences between Cebus and Saimiri are present. The most obvious of these is the hair-covered prehensile tail of Cebus, from which it can suspend itself for brief periods. The tail of Saimiri, although not prehensile, is relatively large and heavy and seems to act as an effective counterbalance (Boinski, 1989a). The fact that all gracile Cebus use the ground more often than any known population of apella (Robinson and Janson, 1987) or Saimiri (
