AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 97:435-449 (1995) Effect of Life History on the Squirrel Monkey (Platyrrhini, Saimiri) Cranium WALTER CARL HARTWIG Z)epartmen.t of AnatomicaE Sciences, State University of New York, Stony Brook, New York 11794 KEY WORDS Saimiri, Orbit, Ontogeny, New World monkeys, Predation ABSTRACT Among primates, squirrel monkeys uniquely possess an interorbital fenestra, in which the midline bony orbitosphenoid septum is largely absent and the soft tissues of the orbits are separated only by a thin membrane. Neural development may contribute t o the approximation of the orbits to the midline in Saimiri, insofar as other platyrrhines with relatively large brains also have relatively narrow interorbital spaces compared to their near relatives. In Saimiri the narrow spacing of the orbits is further exacerbated by intense predation pressure on infants that may select for precocial neonates. The result is a large-headed neonate that is subject to unusual parturition constraints. These parturition constraints apply to the size and dolichocephalic shape of the squirrel monkey head in general, and to the relatively large eyes and approximated orbits in particular. The unique interorbital condition in Saimiri is an example of the effects of life history on skeletal morphology. o 1995 Wiley-Liss, Inc. Stereoscopic vision requires overlapping visual fields. This is accentuated by convergent orbits, and among mammals a decreased emphasis on olfaction has resulted in relatively closely apposed orbits in primates, particularly anthropoid primates. The squirrel monkey interorbital fenestra (Fig. 1) is unique among primates and is a rare, but not unknown, condition among mammals (Haines, 1950). Hershkovitz (1977) argued that thinning of the interorbital septum is typical of small New World monkeys such as Saguinus, and that the condition in Saimiri was not a discrete character but rather an extreme expression along a morphological continuum. Maier (1983)observed that the fenestra develops postnatally, presumably due t o secondary resorption of the septum by the closely apposed musculature of the eye bulbs. This study suggests that the unique approximation of the eyes to the midline results from the large relative brain size of squirrel monkeys and the precocial nature of their prenatal devel0 1995 WILEY-LISS, INC. opment. Cranial traits such as relative dolichocephaly and closely apposed orbits are most likely the result of the need to pass a precocial neonate, which already has a large neurocranium, through the transverse diameter of the birth canal. BACKGROUND Schultz (1940, 1960) published the first comparative, quantitative analysis of relative orbit and eye size in primates. In comparisons across major primate radiations, Schultz (1940) correctly noted that larger primates have relatively smaller orbits, and he explicitly noted that orbit size and eye size are not correlated in large-bodied animals; a measure of the bony orbit should not substitute for an estimation of eye bulb size Received June 30, 1994; accepted March 22, 1995. Address reprint requests to Waiter Carl Hartwig, Department of Anthropology, George Washington University, Washington, DC 20052. 2 - * * I .. H Saimiri 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. R ~ 15. 16. b GENUS Fig. 1. a: Oblique anterior view of Saimiri cranium, showing the fenestrated interorbital septum, which is lined by membrane during life. b: Box plots of interorbital breadth in New World monkeys. Saimiri (far right) has a narrower interorbital breadth than all modern anthropoids. Each box and whiskers represents 95% of the sample. Asterisks represent outliers. See Table 1 for genus codes. SQUIRREL MONKEY LIFE HISTORY in these groups. He did not, however, compare orbit size to a measure of overall cranial size, nor did he explore interspecific relationships within a radiation such as New World monkeys. Biegert (1963) laid further theoretical groundwork and suggested that the potential for midline approximation of the orbits was driven by the size of the intervening nasal capsule. Maier (1993) has recently advanced this work in primates and demonstrated that the ossified parts of the nasal capsule become the ethmoid bone and thus do not directly contribute to the formation of the interorbital septum, which is primarily an aspect of the orbitosphenoid (Maier, 1983).Considered broadly, primates with decreased reliance on olfaction develop smaller nasal capsules and thus express a relative narrowing of the interorbital space. It should be pointed out that there is nothing unusual about the food-processing habits or diet of the squirrel monkey that would suggest that the interorbital fenestra is due to a biomechanical advantage. Specific discussions of the interorbital fenestra in squirrel monkeys essentially began when a fossil cranium of Dolichocebus, a New World monkey from the Oligocene of Argentina, was fully prepared and appeared to exhibit a homologous interorbital fenestra (Rosenberger, 1979a; Hershkovitz, 1979). Rosenberger (1979a) suggested that the functional significance of the fenestra was dubious, but neither he nor Hershkovitz (1977, 1979) explained why the intriguing morphology might exist. The most detailed anatomical treatment of the unusual interorbital morphology in Saimiri is that of Maier (19831, who studied an ontogenetic series of prenatal through adult specimens. He demonstrated that the interorbital septum does ossify, as evidenced by fetal (Starck, 1960) and young postnatal specimens with intact septa. The fenestration is an epigenetic phenomenon of steady secondary resorption caused by the apposition of the relatively large eye bulbs and their attendant musculature. In reference to what is driving this phenomenon in squirrel monkeys, Maier (1983) cited Cartmill’s (1972) explanation that orbital convergence is primarily connected with predatory hab- 437 its, but he also recognized that Saimiri was not the only New World monkey that foraged for insects via visual predation. He concluded that a reduction in the posterior nasal capsule (Biegert, 1963;Maier, 1980), a characteristic of more encephalized primates (Maier, 19931, also contributed to the condition of secondary fenestration. This study agrees with Maier’s (1983) ontogenetic description of the fenestra, and attempts t o explain its development in a bio-behavioral context. Two hypotheses are considered here. As noted by Cartmill (19721, increased orbital convergence results in enhanced depth perception a t close range, but a t the expense of depth perception at long range. Increased convergence could lead to closer orbital approximation to the midline. Because squirrel monkeys practice manual capture of insects as a foraging behavior, a structural modification of the orbits that enhanced their ability to resolve objects in three dimensions a t close range would certainly be a selective advantage. This possibility is evaluated here in the context of laboratory studies of the squirrel monkey visual system. An alternative hypothesis is based on cranial morphometric data of other New World monkeys and a consideration of squirrel monkey life history. It is hypothesized that the uniquely narrowed interorbital space and fenestrated interorbital septum are due t o growth constraints on the orbits imposed by a relatively large neurocranium and the precocial condition of squirrel monkey neonates. This hypothesis predicts that the relative sizes of the braincase and the bony orbits will be large compared t o the birth canal, and requires that a condition of squirrel monkey life history explain the precocial nature of their prenatal development. MATERIALS AND METHODS Morphometric data were collected on over 1,000 New World monkey crania (Table 1) as part of a larger study (Hartwig, 1993). These data were examined in order to determine if cranial size features were related to orbital spacing across platyrrhines in general. Such comparison is facilitated by bivar- 438 W.C. HARTWIG TABLE 1. Samwle size and distribution of ulatvrrhines used for moruhometric analysis' Taxon Alouatta Aotus Ateles Brachyteles Cacajao Callicebus Callimico Callithrix Cebuella Cebus Chil-opotes Lagothrix Leontopithecus Pithecia Saguinus Saiiniri Total 'N = N M F 28 133 110 4 44 125 10 56 11 110 15 78 13 41 145 174 1,097 12 45 40 2 22 63 16 81 68 2 21 49 4 20 8 43 8 35 3 13 64 65 500 number of individuals; M = males; F = females; S 5 32 3 66 7 43 9 25 75 108 557 = subadults; Code iate regressions of cranial dimensions on a n estimate of size. The measurement of body size used in the following allometric regressions is a surrogate derived from a composite of three cranial measurements (total cranial length, basion-prosthion, opisthocranionprosthion), which was determined to have the best fit to a sample of 165 adult individuals across 14 genera for which known wildshot body weights were available. Reduced major-axis regression of this composite against the known body weights yielded a slope value of 0.991 and a loss value of 0.181; least squares regression on the same data yielded a n r2 value of .977 and a standard error of .047 (Hartwig, 1993). I t is recognized that a morphometric composite will not represent size equally effectively for each genus. The composite used here represents the closest match to known individual, wild-shot body weights in the dataset, and was determined from numerous iterations of measurement combinations. Neurocranium size is likewise estimated from a composite of linear dimensions of the neurocranium (maximum neurocranial length, breadth, and height, postorbital constriction and biasterionic breadth). This composite represents a reasonable surrogate for cranial capacity as determined by filling the braincase with sintered glass beads (Martin, 1990), though here it is termed neurocranial size to avoid confusion with the concept of cranial capacity as determined by = S Code 2 20 17 2 6 5 1 1 2 28 2 5 1 5 15 16 128 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 numerical genus code for Figure l b . brain weights. Other measurements used are defined in Table 2. Any explanation of a time-transgressive phenomenon such as the evolution of a n unusual morphological character state cannot be entirely deductive. The following analysis uses analogies to cause and effect relationships in cranial development in New World monkeys to interpret the most likely circumstances that have resulted in the squirrel monkey interorbital fenestra. These circumstances involve intrinsic mechanisms such a s the effect of relative brain size on the positioning of the bony orbits, and extrinsic mechanisms such as the size of the head relative to the birth canal. These extrinsic factors are mediated by aspects of squirrel monkey life history that can be observed but not directly tested. The results of the morphometric analysis are presented below, followed by a consideration of how those results satisfy the predictions of the observed life history causal mechanisms. RESULTS Orbital dimensions in New World monkeys Various measurements of the bony orbit were observed in the sixteen genera of New World monkeys in order to place the interorbital spacing of Saimiri in a comparative context. As demonstrated in Figure l b , squirrel monkeys have a uniquely narrow SQUIRREL MONKEY LIFE HISTORY 439 TABLE 2. Measurements used in this study Variable Definition Neurocranial size Body size surrogate Biorbital breadth Interorbital breadth Palate length (Maximum length, width, height of neuroeranium + biasterionic breadth maximum postorbital constriction)/5 (Total skull length + basion-nasion + basion-opisthocranion)/3 Maximum breadth measured along outer margin of bony orbit Dacryon-dacryon Orale-alveolon interorbital breadth. Within New World monkeys, however, the uakari monkey (Cacajao) appears to have a relatively narrow interorbital breadth, to the extent that its absolute measurement is in the range of platyrrhines of much smaller body size (Aotus, callitrichines). Because Cacajao belongs to the Pitheciinae, a well-defined subfamily of very closely related New World monkey genera (Kinzey, 1992; Schneider et al., 1993),a more detailed look at orbital spacing in this group may serve as a case study for changes in orbital positioning in the context of a known phylogeny. Orbital spacing in pitheciine New World monkeys The crania of pitheciine genera (Pithecia, Chiropotes, and Cacajo) are remarkably similar t o one another except for key features of the neurocranium and orbit, for which Cacajao is considered to represent the derived condition (Hershkovitz, 1977; Rosenberger, 1992; Hartwig, 1993). Pitheciines are characterized by a distinctive masticatory morphology (Rosenberger, 1992) reflecting an unusual and highly specialized diet common to all three genera (Kinzey, 1992). These morphological traits include parallel postcanine toothrows, large, projecting canines, and procumbent upper incisors. In the upper face and braincase, however, the differences between Pithecia and Cacajao are distinct (Fig. 2a; Chiropotes, which is similar to Cacajao in most aspects of cranial morphology, is excluded in the following comparisons due to small sample size). Orbit width is relatively greater than would be expected in the larger cranium of Cacajao, but biorbital breadth is interspecifically scaled with Pithecia (Fig. 2b-e). As expected, then, interorbital breadth is absolutely, and thus, relatively, narrower in Cacajao. Neurocranial differences are illustrated + through a composite measure of neurocranial size, which clearly demonstrates the relatively large neurocranium of Cacajao when compared to Pithecia (Fig. 3a). Uakaries have larger neurocrania than would be expected for a pitheciine of their overall cranial size. Across all platyrrhines, pitheciines as a subfamily express a typical range of values for neurocranial size (Fig. 3b). The fact that a relative increase in neurocranial size is correlated with a relative decrease in interorbital breadth among these closely related genera supports the hypothesis that the relatively large neurocranial size of Saimiri is responsible for its relatively closely spaced orbits. This extends a general trend in anthropoids in which increased relative temporal and frontal lobe size results in a repositioning of the orbital cavity inferior and medial to the anterior and middle cranial fossae (Rosenberger, 1985).According to Ravosa (1991, p. 389), “this places greater constraints on the position of the anthropoid orbit, which necessarily becomes more fully integrated with morphological variation in the basicranium.” Implications for Saimiri The data for Pithecia and Cacajao suggest that neural development may influence ultimate orbital positioning in other New World monkeys. Neural development in Sairniri has been reviewed from structural and behavioral perspectives by several authors (Bauchot and Stephan, 1969;Hemmer, 1971; Elias, 1977; Kaack et al., 1979; Manocha, 1979; Rosenberger, 1979b; Bauchot and Manocha, 1981; Leutenegger, 1982; Stephan et al., 1986; Harvey et al., 1987; Boinski and Fragaszy, 1989; Martin, 1990; Pucciarelli et al., 1990; Corner and Richtsmeier, 1992). Most studies suggest that Saimiri has a relatively capacious neurocranium, but differ in how this is measured and how they rank a 1 cm 1 cm Cucujao Pifheciu 3.90 3 82 3.58 30 b I I 3s 40 350 --L-- 4s SU - I I 412 424 448 460 SIZE Body size (from cranial composite) 0 50 I 436 ~ , 1 1 I 412 424 436 448 0, Q 2 e -2 !s 045 .s 040 * E 035 0 e 0 030 2 m 0 25 5 d Cacajao 21 400 14 Pithecia e Fig. 2 1 4.60 SIZE 441 SQUIRREL MONKEY LIFE HISTORY Saimiri against other New World monkeys. Clearly, the estimation and implications of relative brain size are fundamental problems in evolutionary biology. While recognizing these problems, the objective here is to assess the possible relationship between size of the neurocranium and approximation of the bony orbits in New World monkeys. I n a regression of mean values for brain weight on body weight in twelve New World monkey genera, Saimiri appears to have a relatively large brain compared to other genera (Fig. 4;Table 3). When more recent body weight estimates and data from Elias (1977) are used, Saimiri and Cebus both appear to have relatively large brains (Fig. 4b). Body weight estimates for the larger New World monkeys are currently in dispute (Peres, 19941, and figures used as mean body weights differ markedly across studies, as do exactly what aspects of the brain are retained for weighing. When Martin's (1990) modulus equation for inferring cranial capacity from linear craniometric data is applied to the present dataset (Table l),Saimiri stands out as a platyrrhine with a relatively large neurocranium (Fig. 3b). In the absence of a perfect measurement of relatively neurocranial size, data derived from measuring museum skeletal material do reliably demonstrate that Saimiri has a relatively large neurocranium among New World monkeys. Short of asserting that it has the relatively largest neurocranium, the morphometric data presented here are at least consistent with published studies of brain and body weight means (Bauchot and Stephan, 1969; Elias, 1977; Manocha, 1979; Stephan et al., 1986; Fig. 2. Comparisons of orbital size and spacing in Pithecia and Cacajao (a).For biorbital breadth (b),neither Pithecia nor Cacajao has a n unusual distribution compared to other New World monkeys or to each other (c) in log-log regressions against a body-size surrogate. Rather, the observation that Cacajao has relatively larger orbits is reflected in measurements of size-corrected interorbital breadth (d) and single orbit breadth (e). Both absolutely and relatively (d; P < .Ol), Cucujao has narrower interorbital breadth than Pithecia. Ellipses represent 90% of the sample. Figure 2a was drawn by Luci Betti-Nash. - 42 41 I I I ~ .-N 0 Y) ~ E 40 I e 5 2 ~ 39 38 - 31 L ~ 412 I 1 424 436 L.J 4.48 460 Body size (from cranial composite) 44 ~ 40 0 p! a 1, 2 I+ 36 - 2 2 2 32 ~ 35 40 45 50 Body size (from cranial composite) Fig. 3. Log-log regression ofneurocranial size, a composite measurement representing cranial capacity in Pithecia and Cucujuo (see Table 2). Unlike measurements of the face, Cucujuo has a relatively larger neurocranium than would he expected from extending the ontogenetic scaling of Pithecia (a). In order to verify that Pithecia is not unusually small-brained, thus defeating the comparison, a broad interspecific regression is shown (b), in which only Saimiri appears to have a relatively large neurocranium. Ellipses represent 90% of the sample. Harvey e t al., 1987; Martin, 1990; see Table 3). Relative neurocranial size is thus associated with orbital approximation in Saimiri and it fits a pattern established i n uakari monkeys. The interorbital distance in Saim- 442 W.C. KARTWIG iri, however, is absolutely narrow for anthro- I I poid primates, while its degree of relative brain size is not similarly extreme. Therefore, while large relative brain size may explain the baseline condition of interorbital spacing in Saimiri, other mechanisms probably underlie the extreme degree to which the orbits are approximated and the interorbital fenestra is manifested. I I C + * R x DISCUSSION Orbital approximation and vision in the squirrel monkey 0 * Improvedvisual performance would be the I I I I most direct, adaptive, selectionist explana5 6 I 8 4 9 Mean body weight I (g) tion for the unusual squirrel monkey interor- a I I I bital morphology. Orbital approximation to the midline, however, would not contribute to enhanced stereoscopy or visual acuity. If 00 squirrel monkeys displayed a n unusual level of visual performance, the size and disposic tion of their eye bulbs may be explained as c + a consequence or artifact of such a condition. .eo D * Data generated by studies of the squirrel 0 monkey visual system are particularly relevant to the hypothesis that enhanced visual c x performance may be a n underlying cause of 0 5 2 its closely approximated orbits. I In electrophysiological comparisons with * macaques, squirrel monkeys present similar central visual field capacities (Doty et al., 1 -1 I I 9 19641, which may indicate that they are ca4 5 6 I 8 pable of resolving a finer retinal image given b Mean body weight I1 (g) their smaller eye diameter and total extent Cebuella 0Aofirs of unilateral striate cortex (Cowey and Ellis, 0 Cuiliihrk 0Cebzu 1967, 1969). The same studies, however, Cuiiimico Pirhecia measure mean visual acuity in rhesus macaques as superior to that in squirrel monkeys, who moreover demonstrate poorer peSuimin Aides ripheral, parafoveal, and foveal acuity (Rolls Cullicebus - Lagofhrrr and Cowey, 1970). Another study (Ordy and Samorajski, 1968) ranked the thresholds at Fig. 4. Log-log regressions of species brain and body which separate objects could be discrimi- weight means in-selected New World monkey genera, nated in Saimiri, Callithrix, Lemur, and Tu- from Stephan et al. (1986) and Ford and Davis (1992). a: Data are taken directly from Stephan et al. (19861, and paia a s comparable to one another and to demonstrate that Saimiri has a relatively large brain values for catarrhines, including humans. weight for a given mean body weight. This regression To the extent that depth perception can be largely matches that derived morphometrically in Figure measured in laboratory primates, the 3. Due to differences in reported mean body weights, rethreshold for depth discrimination in squir- vised body weight means reported by Ford and Davis and data from Elias (1977) are compared to the rel monkeys appears to be the same as that (1992) Stephan et al. (1986) brain weight data in b. In this refor humans (Schmidt, 1968; Jacobs, 1985). gression Sairniri and Cebus clearly display relatively The possibility that depth perception in large brains. See Table 3 for raw data and species names. I ~ D 0 L I + a ;I 443 SQUIRREL MONKEY LIFE HISTORY TABLE 3. Brain and body weights in selected ptatyrrhine genera' Taxon Cebuetta pygmaea Callithrir jacchus Saguinus oedipus Catlimico goetdii Saimiri sciureus' Saimiri sciureus3 Aotus triuirgatus Callicebus rnoloch Pithecia monachus Cebus albifrons Cebus a t b i f r o d Lagothrix tagotricha Alouatta senicutus Ateles geoffroyi Brain weight (g) 4.15 7.60 10.00 11.00 24.12 23.30 17.10 19.00 35.00 71.00 66.8 101.00 52.00 108.00 Body weight I (g) Body weight I1 (g) Ratio I (brain wtbody wt I) Ratio I1 (brain wthody wt 11) 123 257 420 389 675 667 934 920 2,348 2,265 2,060 7,695 6,767 7,704 ,035 .027 .026 ,034 ,030 .024 120 280 380 480 660 830 900 1,500 3,100 - 5,200 6,400 8,000 ,023 .028 ,037 ,036 ,035 ,018 ,021 ,015 ,031 ,032 ,013 ,008 .014 ,021 ,021 ,023 ,023 ,019 .008 ,013 'Brain weight and body weight I from Stephan et al. (19861, body weight I1 from Ford and Davis (19921, except as specified. zBrain weight data for 8 females from Manocha (1979). 3Brain weight and body weight data from Elias (1977). Saimiri is qualitatively different from depth perception in other primates still exists, but laboratory measures of vision have failed to demonstrate this to date. Indeed, laboratory studies of the squirrel monkey indicate no unusual functional or performance features of the visual cortex, depth perception, or visual acuity (Allman and McGuinness, 1986). Saimiri ontogeny and behavioral ecology Field studies on Saimiri and other New World monkeys indicate that squirrel monkeys occupy a niche subject to unpredictable food supply and distribution ("erborgh, 1983), and to intense predation pressure (Boinski, 1987a). These conditions factor prominently in a life-history model of ecological risk aversion strategies that affect maturation patterns in primates (Janson and van Schaik, 1993).Aspects of squirrel monkey behavioral ecology suggest that growth trajectories resulting in precocial neonates may be advantageous. The following section details prenatal and postnatal neural growth trajectories in squirrel monkeys in the context of their life history and behavioral ecology. Ontogeny Squirrel monkeys appear to have a relatively long gestation period compared to other New World monkeys and to many Old World monkeys. The mean gestation length of 170 days for Saimiri is greater than that for all New World monkeys except the large- bodied atelines, and is comparable to the range of 162-187 days reported for the Old World monkey genera Macaca, Cercocebus, Papio, Cercopithecus, Nasalis, and Presbytis (Ardito, 1976; Brizzee and Dunlap, 1986; Harvey e t al., 1987). It should be noted that a wide range of gestation lengths in captive squirrel monkeys has been reported (Boinski, 1987a), and one study that controlled for precise date of conception reported gestation lengths in the 145-155 day range (Kerber et al,. 1977). Gestation length in Aotus, which is slightly larger than Saimiri (Table 31,is 133days, as would be predicted from a regression for anthropoid primates (Martin, 1992). The preponderance of published data, whether measured in captive animals or estimated in the wild, however, suggests relatively long gestation lengths in Saimiri (see compiled data in Brizzee and Dunlap, 1986). It seems that selection for relatively precocia1 neonates via extended gestation plays a role in determining neonatal size in Saimiri. The large size of the Saimiri neurocranium a t birth and its apparently rapid maturation after birth are unusual features for New World monkeys. Most New World monkeys display isometric ontogenetic scaling between sub-adults and adults for neurocranial size. This relationship does not obtain for squirrel monkeys (Fig. 5 ) . At birth, female squirrel monkey brains are fully 62% of adult brain weight (Elias, 1977; see also Manocha, 1979), whereas in Cebus neonatal W.C. HARTWIG 444 3.90 I I I I 3.82 Sub-adult 6) .A N m F0 3.74 s .r( E f! 0 s 3.66 2 0 0 3.58 3.50 3.6 I I I I 3.7 3.8 3.9 4.0 4.1 Body size (from cranial composite) Fig. 5. Log-log regression of neurocranium size on a composite measure of body size in subadult and adult Sairniri, illustrating the early onset of adult neurocranium size in squirrel monkeys. brain weight is less than 50% adult size (Elias, 1977). These studies also indicate that there is negligible brain weight increase in squirrel monkeys beyond 60 days postnatally, while in Cebus brain growth rate is slower and the duration of growth is much longer (Elias, 1977;Manocha, 1979;Bauchot and Manocha, 1981; Harvey et al., 1987). The obstetric complicationsof giving birth to a precocially large-brained neonate may play a role in its unusual, arguably compromised, interorbital morphology. Among anthropoids, Saimiri neonates have the largest neurocranial breadth dimensions as a percentage of transverse diameter of the female pelvic inlet (Leutenegger, 1982). Likewise, neonate neurocranial length is substantially greater (136%)than is the corresponding sagittal dimension of the pelvic inlet. Leutenegger (1982) suggested that during birth non-human primates may pass through the pelvic inlet in a “face-first”orientation, such that the head is extremely dorsiflexed and cranial height rather than cranial length must pass through the sagittal dimension of the pelvic inlet. Once past the pelvic inlet, the head “relaxes”to assume a typical delivery position of vertex distal, occiput posterior, in reference to the nonhuman primate pelvic outlet. One study described the presentation of two normal squirrel monkey births in captivity in this manner (Bowden et al., 1967). A reorientation of the neonatal cranium can mitigate stress on the sagittal axis of the pelvic inlet during birth, but no simultaneous repositioning in order to relieve the other pelvic inlet diameters is possible. Thus, the transverse axis of the pelvic inlet is still subject to the passage of the breadth of the neonatal cranium. A compromise between pelvic inlet transverse diameter and 445 SQUIRREL MONKEY LIFE HISTORY a neonate with a relatively large brain may be the relatively dolichocephalic shape of the Saimiri neurocranium, in which cranial breadth is minimized. The Saimiri brain is relatively larger than the brain of like-sized Aotus and Callicebus, but is packaged in a relatively longer, rather than relatively longer and wider, neurocranium (Hartwig, 1993). It is suggested here that extreme orbital approximation in the squirrel monkey is a function of the need to package the orbits of a relatively large neonate in a transverse dimension that is constrained by the size of the pelvic inlet. Data on biorbital breadth in neonatal squirrel monkeys are not available. Data collected by Leutenegger (1973, 1982) and Schultz (1940, 1960) demonstrate that the neurocranium of squirrel monkey neonates is in general quite large compared to the birth canal, and that adult squirrel monkeys have relatively capacious orbits and relatively voluminous eye bulbs compared to other anthropoids. Given that squirrel monkeys are precocial at birth as demonstrated by cranial morphometric and body weight data, it is safe to hypothesize that the relationship between biorbital width and neurocranial width in neonates is not radically different from that expressed in the subadults measured for this study. These data for subadults (Fig. 6) indicate that the biorbital breadth dimension in squirrel monkeys does increase postnatally in a typical pattern of ontogenetic scaling. That is, at birth, biorbital breadth in the squirrel monkey is slightly smaller than is neurocranial breadth, t o a degree that is relatively smaller than is the condition in adults. This observation satisfies the expectations of a hypothesis of parturition constraints. If the extremely narrow interorbital space in squirrel monkeys is a result of the limits imposed by the transverse diameter of the birth canal, then neonatal biorbital breadth would be expected to be narrower than the canal rather than narrower than the neurocranium, which can be deformed to a more narrow dimension during birth. Leutenegger (1982) reported that the squirrel monkey neonatal neurocranium was 121%wider than the transverse diame- 1 ' 06l- I 1 a 2. Subadult 3.60- 1 Adult I I I I 352 - 5 W m a m * * Adult 344 - .- 336 - 328 - 320 34 b I I I I I I I 35 36 31 38 39 40 41 42 Body size (from cranial composite) Fig. 6. Biorbital breadth in subadult squirrel monkeys is relatively more narrow compared to neurocranial breadth than is the condition in adults (a).Expansion of biorbital breadth ontogenitically is depicted in b. These comparisons suggest that neonatal biorbital breadth is on the order of 25-30% more narrow than is neonatal neurocranial breadth, which itself is reported to be 121% the size of the transverse diameter of the pelvic canal through which it must pass (Leutenegger, 1982). ter of the birth canal, thus necessitating its inward deformation during birth. Figure 6 indicates that biorbital breadth in subadult (and, by inference, neonatal) squirrel monkeys is roughly 25% narrower than neurocranial breadth. This value is almost identical with the minimal difference required if 446 W.C. HARTWIG the orbits are t o pass through the birth canal without deforming. It should be noted that these conditions do not apply to neonates of the only nocturnal anthropoid, Aotus, which are typically platyrrhine in relative body size, brain size, and altriciality (Harvey et al., 1987).Aotus neonates are absolutely smaller than Saimiri neonates, whileAotus adults are, on average, absolutely larger than Saimiri adults. Aotus adults display greatly enlarged orbits but not unusually narrow interorbital breadths (Fig. lb). Thus, despite having greatly enlarged orbits as part of their adaptation to nocturnality, night monkeys do not exhibit an interorbital fenestra. This is most likely due to lateral expansion of the orbital margin enabled by more relative postnatal growth and by the absence of any spatial constraint during parturition. Behavioral ecology Independent studies have indicated that Saimiri neonates have a relatively large brain and are born in a relatively precocial state after a relatively long gestation. These measurable characteristics of the reproductive biology of the squirrel monkey may be explained through their behavioral ecology and natural history. Terborgh (1983) described the Saimiri populations of Cocha Cashu, Peru, as temporary patch specialists who exploit ephemeral resources described as unpredictable in location, of fluctuating abundance, and generally found and eaten by single individuals (Boinski and Fragaszy, 1989). Their social organization seems to be one in which every individual forages for itself within a large group assembly that lacks hierarchy or unity. Squirrel monkey infants appear to forage independently at a relatively early age. In their field study in Costa Rica, Boinski and Fragaszy (1989) reported that squirrel monkeys foraged successfully at 7 weeks of age, first attempted to capture mobile prey at 11weeks, and were weaned by 16 weeks. Moreover, they observed very little social influence on foraging behavior and no food provisioning of infants by mothers or other troop members. Elias (1977) found that captive squirrel monkeys achieved locomotor and manipulative skills far earlier than did captive capuchins of similar chronological ages. The seasonal timing of foraging independence is also relevant here. Boinski (1987a) reported that squirrel monkey births at her field site were extremely synchronous within troops, and occurred such that infants matured t o foraging independence at a time when arthropod and fruitlflower abundance reached annual highs (see also Coe and Rosenblum, 1978). Relative resource abundance is likely the primary influence on birth synchrony in Saimiri, but Boinski (1987a) noted an important secondary influence, intense predation by raptors, that more directly relates to infant size and precociality. She noted that predation attempts by raptors were targeted almost exclusively at infants. Squirrel monkey infants are carried dorsally, and thus are particularly susceptible to predation by aerial strike. Relatively larger infants, as products of extended gestation, would be less obvious prey choices for smaller predacious birds, and would stand a better chance of surviving attacks by larger raptors (Janson and van Schaik, 1993;Boinski, personal communication). Given the combination of the low density and patchy distribution of resources in squirrel monkey habitats (Boinski, 1987b;Janson and Boinski, 1992), the “every individual for itself‘ foraging behavior (Terborgh, 19831, seasonal fluctuations in food availability (Boinski, 19891,and early onset of infant foraging independence (Boinski and Fragaszy, 19891, a developmentally precocial neonate would seem to be a logical component of Saimiri behavioral ecology and adaptation. The differential survival of precocial neonates during periods of intense infant predation pressure may further influence the reproductive biology of wild squirrel monkeys (Boinski, 1987a). The extended gestation and rapid maturation during an abbreviated infancy (Elias, 1977) follow the Janson and van Schaik (1993) model of ecological risk aversion. The Janson and van Schaik (1993) model argues that juvenile growth will be slowed in group living species subject to conditions of intense predation pressure and intraspecific competition for resources. Because risk of predation is greater at the periphery of a group, 447 SQUIRREL MONKEY LIFE HISTORY juveniles will tend to stay in the more protective center, where intraspecific competition for resources is greatest. According to the model, juvenile growth is adjusted to a slower rate in order to lower metabolic needs to a level that can be met by the inefficient foraging techniques of juveniles competing with adult conspecifics. It is suggested here that the period of infancy in squirrel monkeys is the period in which the risk of predation is most critical, and any mechanisms that reduced the duration of this period would be selectively advantageous. Squirrel monkeys appear to reduce this period via a combination of extended gestation and early onset of juvenility, at which time squirrel monkeys are already competent foragers and virtually full-sized. It is within this biobehavioral continuum that the unusual interorbital fenestra is explained most effectively. SUMMARY Ontogenetically, the sequence of events leading to the unusual interorbital morphology in Saimiri would seem to involve: a fetal brain that is relatively large at birth (Harvey et al., 1987); a developing eye that is relatively larger in Saimiri adults than in all diurnal anthropoids (Schultz, 1940; Biegert, 1963); obstetric constraints imposed by a relatively large natal neurocranium (Leutenegger, 1982) that are partially absorbed by the dolichocephalic neurocranial contour and closely apposed orbits of the Saimiri cranium; and secondary resorption of the interorbital septum postnatally due to approximation of the eye bulbs and attendant musculature to the midline (Maier, 1983). The preponderance of data suggest that squirrel monkey infants must be able to forage independently at a very young age, presumably to survive intraspecific competition for food and to reduce risk of predation during the dependent infancy period when they are carried dorsally. These pressures are incompatible with a primate life history strategy in which infants mature slowly and are a sustained energy cost to their mothers in terms of “typical” primate lactation and ma- ternal transport patterns. It would appear that squirrel monkeys enter the growth period ofjuvenility, here considered to be equivalent to the onset of foraging independence, as soon a s possible. The large but loosely knit squirrel monkey troops would seem to exist for the benefit of group defense, though clearly at the aggravation of intraspecific competition for food resources. To this extent squirrel monkeys conform to the ecological risk aversion model of Janson and van Schaik (19931, though with particular reference to the infancy, rather than juvenile, phase of life history. CONCLUSIONS Enhanced stereoscopy does not appear to be a cause or effect of close approximation of the eyes in the squirrel monkey. Its uniquely closely apposed orbits result proximately from the effects of having a relatively large brain and reaching a precocial state of development prenatally. These effects intersect with obstetric constraints imposed by the absolute limits of the pelvic inlet. Compromise of midline interorbital breadth and bony interorbital separation of the eyes reflects the potential of the midline craniofacia1 complex to absorb such growth outcomes. Ultimately, this derived perinatal growth trajectory results from survival pressures imposed by a niche in which a small-bodied primate exploits unpredictable, widely dispersed food resources and is subject to intense predation pressure. ACKNOWLEDGMENTS Alfred L. Rosenberger first encouraged me to consider life history aspects of Saimiri before attempting to explain its unusual cranial morphology. Conversations with him and John Fleagle have greatly improved my understanding of New World monkey natural history and my appreciation for a developmental approach to comparative anatomy. Their comments and those of two anonymous reviewers greatly improved the manuscript. Steve Leigh, Sue Boinski, Katie Milton, and Matt Ravosa also provided useful suggestions. For access to specimens under their care I thank Bruce Patterson of the Field Museum of Natural History, RDE MacPhee 448 W.C. HARTWIG Elias MF (1977)Relative maturity of Cebus and squirrel monkeys at birth and during infancy. Dev. Psychohiol. 10:519-528. Falk D (1979) Cladistic analysis of New World monkey sulcal patterns: Methodological implications for primate brain studies. J . Hum. Evol. 8:637-645. Ford SM, and Davis LC (1992) Systematics and body LITERATURE CITED size: Implications for feeding adaptations in New Allman J, and McGuinness E (1986) Visual cortex in World monkeys. Am. J . Phys. Anthropol. 88:415-468. primates. In HD Steklis, and J Erwin (eds.): Compara- Haines RW (1950)The interorbital septum in mammals. tive Primate Biology, Volume 4: Neurosciences. New J . Linnaean SOC.London 41585407. York Alan R. Liss, Inc., pp. 279-326. Hartwig WC (1993) Comparative Morphology, Ontogeny Ardito G (1976) Checklist of the data on the gestation and Allometry of the Platyrrhine Cranium. Unpublength of primates. J . Hum. Evol. 5:213-222. lished Ph.D. dissertation. Berkeley: University of CalBauchot R, and Manocha SL (1981) La croissance enifornia. cephalique chez Saimiri sciureus (Simiens, Cebides). Harvey P, Martin RD, and Clutton-Brock TH (1987) Life Mammalia 45:251-255. histories in comparative perspective. In BB Smuts, Bauchot R, and Stephan H (1969) Encephalisation et le DL Cheney, RM Seyfarth, RW Wrangham, and TT niveau evolutif chez les simiens. Mammalia Struhsaker (eds.) Primate Societies. Chicago: Univer33:225-275. sity of Chicago Press, pp. 181-196. Biegert J (1963) The evaluation of characteristics of the Hemmer B (1971)Beitrag zur Erfassung der progressioskull, hands and feet for primary taxonomy. In Washnen Cephalisation bei Primaten. Proc. 3rd Int. Congr. burn SL (ed.): Classification and Human Evolution. Primatol. 2:99-107. Chicago: Aldine, pp. 116-145. Hershkovitz P (1977) Living New World Monkeys, VolBoinski S (1987a) Birth synchrony in squirrel monkeys ume 1,with a n Introduction to the Primates. Chicago: (Saimiri oerstedi). A strategy to reduce neonatal predUniversity of Chicago Press. ation. Behav. Ecol. Sociobiol. 21:393-400. Hershkovitz P (1979) Supposed squirrel monkey affniBoinski S (1987b) Habitat use by squirrel monkeys ties of the late Oligocene Dolichocebus gainanensis. (Saimiri oerstedi) in Costa Rica. Folia Primatol. Nature 298:201-202. 49:151-167. Jacobs GH (1985) Visual system of the squirrel monkey. Boinski S (1989) The positional behavior and substrate In LA Rosenblum, and CL Coe (eds.): Handbook of use of squirrel monkeys: Ecological implications. J. Squirrel Monkey Research. New York Plenum Press, Hum. Evol. 18:659-678. pp. 271-293. Boinski S, and Fragaszy DM (1989) The ontogeny of foraging in squirrel monkeys, Saimiri oerstedi. Anim. Janson CH, and Boinski S (1992) Morphological and behavioral adaptations for foraging in generalist priBehav. 37:415-428. mates: The case of the cebines. Am. J . Phys. AnthroBowden D, Winter P, Ploog D (1967) Pregnancy and pol. 88:483-498. delivery behavior in the squirrel monkey (Saimiri sciJanson CH, and van Schaik C (1993) Ecological ureus) and other primates. Folia Primatol. 5:l-42. risk aversion in juvenile primates: Slow and steady Brizzee KP, and Dunlap WP (1986) Growth. In WR Duwins the race. In ME Pereira and LA Fairbanks (eds.): kelow, and J Erwin (eds.):Comparative Primate BiolJuvenile Primates: Life History, Development, and ogy, Volume 3: Reproduction and Development. New Behavior. New York: Oxford University Press, pp. York: Alan R. Liss, Inc., pp. 363413. 57-76. Cartmill M (1972) Arboreal adaptations and the origin ofthe order Primates. In Tuttle R (ed.): The Functional Kaack B, Walter L, and Brizze KR (1979) The growth and development of the squirrel monkey (Saimiri sciand Evolutionary History of Primates. Chicago: Alureus). Growth 43; 116-135. dine, pp. 97-122. Coe CL, and Rosenblum LA (1978) Annual reproductive Kerber WT, Conaway CH, and Moore-Smith D (1977) The duration of gestation in the squirrel monkey strategy of the squirrel monkey (Saimiri sciureus). (Saimiri sciureusj. Lab. Anim. Sci. 27:700-702. Folia Primatol. 29:19-42. Corner BD, and Richtsmeier J T (1992)Cranial growth in Kinzey WG (1992) Dietary and dental adaptations in the Pitheciinae. Am. J.Phys. Anthropol. 88r499-514. the squirrel monkey (Saimiri sciureus):A quantitative Leutenegger W (1973) Maternal-fetal weight relationanalysis using three dimensional coordinate data. Am. ships in primates. Folia Primatol. 20:280-293. J . Phys. Anthropol. 87167-82. Leutenegger W (1982) Encephalization and obstetrics Cowey A, and Ellis CM (1967)Visual acuity of the rhesus in primates with particular reference to human evoluand squirrel monkeys. J . Comp. Physiol. Psychol. tion. In E Armstrong, and D Falk (eds.):Primate Brain 64~80-84. Evolution, Methods and Concepts. New York: Plenum Cowey A, and Ellis CM (1969) The cortical representaPress, pp. 85-95. tion of the retina in the squirrel and rhesus monkeys and its relation to visual acuity. Exp. Neurol. Maier W (1980) Nasal structures in Old and New World 24:374-385. primates. In R Ciochon, andAB Chiarelli (eds.):EvoluDoty RW, Kimura DS, and Mogenson GJ (1964) Photitionary Biology of the New World Monkeys and Contically and electrically elicited responses in the central nental Drift. New York: Plenum Press, pp. 219-241. visual system of the squirrel monkey. Exp. Neurol. Maier W (1983) Morphology . -. of the interorbital region - of IOr19-51. Saimiri sciureus. Folia Primatol. 41:277-303. of the American Museum of Natural History, Richard W. Thorington of the National Museum of Natural History, and their respective staffs. SQUIRREL MONKEY LIFE HISTORY Maier W (1993) Zur evolutiven und funktionellen Morphologie des Gesichtsschadels der Primaten. Z. Morphol. Anthropol. 79:279-299. Manocha SL (1979) Physical growth and brain development of captive-bred male and female squirrel monkeys, Saimiri sciureus. Experientia 35:96-98. Martin RD (1990) Primate Origins and Evolution. Princeton: Princeton University Press. Martin RD (1992) Goeldi and the dwarfs: The evolutionary biology of the small New World monkeys. J . Hum. Evol. 22t367-393. Ordy J, and Samorajski T (1968)Visual acuity and ERGCFF in relation to morphologic organization of the retina among diurnal and nocturnal primates. Vision Res. 8:1205-1225. Peres CA (1994) Which are the largest New World monkeys? J . Hum. Evol. 26245-250. Pucciarelli HM, Dressino V, and Niveiro MH (1990) Changes in skull components of the squirrel monkey evoked by growth and nutrition: An experimental study. Am. J. Phys. Anthropol. 81.535-544. Ravosa MJ (1991) Interspecific perspective on mechanical and non-mechanical models of primate circumorbital morphology. Am. J. Phys. Anthropol. 86:369-398. Rolls ET, and Cowey A (1970) Topography of the retina and striate cortex and its relationship to visual acuity in rhesus monkeys and squirrel monkeys. Exp. Brain Res. 1Ot298-310. Rosenberger AL (1979a) Cranial anatomy and implications of Dolichocebus, a late Oligocene ceboid primate. Nature 279:416-417. Rosenberger AL (197913) Phylogeny, Evolution and Classification of the New World Monkeys (Platyrrhini, Primates). Unpublished Ph.D. dissertation. New York: City University of New York. 449 Rosenberger AL (1985) In support of the NecrolemurTarsier hypothesis. Folia Primatol. 45:179-194. Rosenberger AL (1992) Evolution of feeding niches in New World monkeys. Am. J. Phys. Anthropol. 88525-562. Schmidt U (1968) Untersuchungen zur visuellen Raumorientierung bei TotenkopfAffen (Saimiri sciureus). 2. Vergleich. Physiol. 60:176-208. Schneider H, Schneider MPC, Sapaio I, Harada ML, Stanhope M, Czelusniak J , and Goodman M (1993) Molecular phylogeny of the New World monkeys (Platyrrhini, Primates). Mol. Phylogenet. Evol. 2225-242. Schultz AH (1940) The size of the orbit and of the eye in primates. Am. J. Phys. Anthropol. 26:389-408. Schultz AH (1960) Age changes and variability in the skull and teeth of the Central American monkeys Alouatta, Cebus and Ateles. Proc. Zool. SOC.London 133~337-390. Starck D (1960) Das Cranium eines Schimpansenfetus (Pan troglodytes [Blumenbach 17991) von 71mm SchStlg., nebst Bemerkungen uber die Korperform von Schimpansenfeten. Gegenbaurs Morphol. Jahrb. lOOt559-647. Stephan H, Baron G, and Frahm HD (1986) Comparative size of brains and brain components. In HD Steklis, and J Erwin (eds.):Comparative Primate Biology, Volume 4: Neurosciences. New York Alan R. Liss, Inc., pp. 1-38. Terborgh J (1983) Five New World Primates: A Study in Comparative Ecology. Princeton: Princeton University Press.