Homework answers / question archive / What selection pressures drove the evolution of primates? You will provide a summary of each position, the evidence supporting it, and a discussion of which argument you -nd most compelling and why

What selection pressures drove the evolution of primates? You will provide a summary of each position, the evidence supporting it, and a discussion of which argument you -nd most compelling and why

Sociology

Share With

What selection pressures drove the evolution of primates? You will provide a summary of each position, the evidence supporting it, and a discussion of which argument you -nd most compelling and why. You should demonstrate your understanding of the material through your original response (do not just repeat what you read) to the debate). You must use and reference the two resources provided for you (if you choose from the list). In addition, I recommend including at least two additional resources relevant to the debate. questions to consider •Why did you chose the debate you chose? •What i the nature and history of the debate? •What are the two positions? •What is each author’s central claim/argument? •Do the authors make any assumptions or unsubstantiated claims? •What evidence do the authors of each position present? How does the evidence support their point? Are there alternate ways to interpret their evidence? •What are the strengths and weaknesses of their arguments? •Why is the debate important or interesting? make sure you: •Characterize the text’s arguments as fairly and accurately as possible. If you disagree, disagree with what the author is actually arguing. •Include your own voice by questioning arguments and evidence and by raising critical questions. •Be speci-c in your critiques and claims and use citations to back them up. how to structure your paper: •Introduction paragraph with thesis statement–Briey outline or summarize your paper and the arguments you will make. For example: “In this paper, I will argue/explain/explore...”. •Make sure to break up the body of your paper into paragraphs. Papers that consist of gigantic paragraphs will be marked down. Use transition language between paragraphs as necessary. •Conclusion paragraph with re-stated thesis: re-state your thesis and briey summarize what you did in the paper. •In text citations and bibliography are required. •Please refer to the APA style guide: https://pitt.libguides.com/citationhelp/apa7 •Include page numbers at the bottom of each page A brief overview of citations (but please also reference the APA site): If you use someone else's thoughts, ideas, words, etc you *must* cite them - even if you are not quoting them. In text citations should look like this: "According to Bailey (2004) Neanderthal upper molars have a skewed outline". Or "Crown outline of upper molars is skewed in Neanderthals (Bailey, 2004)." Then the Bibliography/Reference would look like this (or similar): Bailey SE (2004) A morphometric analysis of maxillary molar crowns of Middle-Late Pleistocene Humans. Journal of Human Evolution. 47: 183-198. American Journal of Primatology 75:95–106 (2013) REVIEW ARTICLE Rethinking Primate Origins Again ROBERT W. SUSSMAN1∗ , D. TAB RASMUSSEN1 , AND PETER H. RAVEN2 1 Department of Anthropology, Washington University, St. Louis, Missouri 2 Missouri Botanical Garden, St. Louis, Missouri In 1974, Cartmill introduced the theory that the earliest primate adaptations were related to their being visually oriented predators active on slender branches. Given more recent data on primate-like marsupials, nocturnal prosimians, and early fossil primates, and the context in which these primates first appeared, this theory has been modified. We hypothesize that our earliest primate relatives were likely exploiting the products of co-evolving angiosperms, along with insects attracted to fruits and flowers, in the slender supports of the terminal branch milieu. This has been referred to as the primate/angiosperm co-evolution theory. Cartmill subsequently posited that: “If the first euprimates had grasping feet and blunt teeth adapted for eating fruit, but retained small divergent orbits . . . ” then the angiosperm coevolution theory would have support. The recent discovery of Carpolestes simpsoni provides this support. In addition, new field data on small primate diets, and a new theory concerning the visual adaptations of primates, have provided further evidence supporting the angiosperm coevolution C 2012 Wiley Periodicals, Inc. theory. Am. J. Primatol. 75:95–106, 2013. Key words: angiosperms; coevolution; Plesiadapiformes; stereoscopy; primate diets THE ARBOREAL THEORY OF PRIMATE EVOLUTION Traditionally, certain morphological traits of primates, such as grasping extremities, loss of claws, reduced olfaction, convergence and frontation of the eyes, and enlargement of the brain, were thought to be related to the acquisition of an arboreal way of life by a preprimate, primitive “insectivore” mammal in the late Mesozoic [Smith, 1912]. In the earliest 20th Century, Cuvier’s order Insectivora was poorly understood phylogenetically but recognized adaptively as a primitive condition for mammals; Simpson [1945, p 175], among others, noted the grouping was phylogenetically unnatural, and was “something of a scrap basket” [see Rose, 2006]. To evolutionary theorists at the time, like Smith, the key point was that today’s small, insect-eating mammals provided a model of the primitive type of animal from which primates and other specialized orders must have come. Gregory [1910] had already proposed Archonta as a superordinal phylogenetic group containing primates, tree shrews, dermopterans, (and incorrectly) bats. For Smith, the term insectivore described a mammal that is tiny, olfactory-oriented, with high metabolism and sharp puncturing teeth adapted for piercing insect exoskeletons; the concept included today’s lipotyphlan shrews, moles, hedgehogs, and a tremendous diversity of extinct forms [Butler, 2010; Dunn & Rasmussen, 2009; Rose, 2006; Seiffert, 2010]. Perhaps English hedgehogs of the C 2012 Wiley Periodicals, Inc. modern family Erinaceidae, familiar to Smith, were important templates for his evolutionary model. Grafton Elliot Smith was an English anatomist who was an expert in the comparative study of the brain and senses [Armelagos, 1997]; he presented the first united theory of primate origins by stating that the reduction in the sense of smell and the elaboration of the senses of vision, touch, and hearing in early primates were adaptations to life amid the branches of trees. Frederick Wood Jones [1916], an early assistant to Smith, elaborated Smith’s theories. He focused on the postcranial features of the primate morphological pattern. Wood Jones emphasized the adaptive significance of the use of the forelimbs for touch and climbing in the trees, which led to the “emancipation of the forelimbs” and improved eye–hand coordination. Since these early papers, this theory was elaborated upon and referred to as the “arboreal theory” of primate evolution [Le Gros Clark, 1963; Howells, 1947]. According to the arboreal theory, it was simply the move from the ground to the trees that ∗ Correspondence to: R.W. Sussman, Department of Anthropology, Washington University, St. Louis, MO 63130. E-mail: rwsussma@wustl.edu Received 21 May 2012; revised 18 October 2012; revision accepted 18 October 2012 DOI 10.1002/ajp.22096 Published online 26 November 2012 in Wiley Online Library (wileyonlinelibrary.com). 96 / Sussman et al. explained everything unique about primates. Until the 1970s, this theory was generally accepted as the explanation for the adaptive significance of many unique primate morphological traits. These traits were considered to be adaptations of our earliest ancestors to the acquisition of an arboreal way of life. THE ORIGIN OF PRIMATES: A PALEONTOLOGICAL APPROACH The arboreal theory of primate evolution was developed by primate anatomists and behavioralists using traits found in living primates. Historically, paleontologists have also had to deal with a complex, natural radiation of Paleocene, and early Eocene mammals with some features that are unique and primate-like in the modern sense, and other features that are not. These early relatives of modern primates are classified as Plesiadapiformes [Bloch et al., 2007; Gunnell, 1989; Rose, 2006; Silcox et al., 2007; for reviews see Conroy, 1990; Fleagle, 1999]. The Plesiadapiformes stand in contrast to the later primates of the Eocene which are called “primates of modern aspect” or Euprimates (true primates) to avoid confusion with these archaic groups. Because paleontologists are working with forms close to the point of divergence between nonprimates and Euprimates that had not acquired all the specializations evident in today’s narrow sample of extant species, they often find it difficult to draw boundaries between the recognized higher taxonomic groupings based on living taxa. This point is self-evident because today’s higher taxonomic groups are a function only of time and extinction. In fact, many students of the order Primates (including plesiadapiforms) claimed that no major adaptive shift occurred in the earliest stages of primate evolution [Le Gros Clark, 1963; McKenna, 1966; Simpson, 1955, 1961]. These authors asserted that one would have to rely on specific “taxonomic traits,” the adaptive significance of which is not necessarily understood, to define the order Primates. One such trait is that the bubble of bone (the auditory bulla) covering the middle and inner ear spaces is formed from the petrosal part of the temporal bone, rather than from an independent ossification [MacPhee et al., 1983] but whether this attribute denotes a restricted clade to be tagged with the ordinal label Primates or only a subgroup within a broader clade sharing a common ancestor with living primates is a taxonomic issue. More importantly, detailed comparisons of plesiadapiform basicrania (of the genus Ignacius) demonstrated that plesiadapiforms show closer affinity to euprimates than to any other mammalian group, including dermopterans [Bloch & Silcox, 2001]. Genetic phylogenetic analyses resolve that Gregory’s [1910] century-old Archonta (sans bats) is Am. J. Primatol. a natural grouping [e.g., Adkins & Honeycutt, 1991], and paleontological analysis of plesiadapiforms show they are nested adjacent to Primates, but within the extant archontan group of dermopterans and treeshrews [Bloch et al., 2007; Silcox et al., 2007]. There have been no viable, long-lived departures from this hypothesis except for short episodes linking plesiadapiforms to dermopterans [Hoffstetter, 1977], like the one in which plesiadapiforms were linked with dermopterans based on misinterpreted fossil finger evidence [Beard, 1990, 1993; see Krause, 1991]. The reality of a plesiadapiform–primate clade has long been widely accepted among paleontologists and mammalogists [see Rose, 2006], but the idea seems less well known to the anthropological community because of confusion with the taxonomic and adaptive debates about whether or not to split this natural evolutionary radiation into two orders. A natural evolutionary grouping of extant primates and plesiadapiforms is now stronger than ever, based on the basicranium, the foot, and the molar dentition. From a paleontological perspective, the most important thing about plesiadapiforms is that they provide the opportunity to see character mosaics in an ancient outgroup of the modern euprimates. It is not surprising that there should be a wide and complex diversity of mammals in the morphological zone between “primate” as defined by living taxa and “nonprimate.” As an increasing number of plesiadapiform primates become known from fossil skeletons and crania the sequence of acquisition of euprimate characters has been clarified. Several shared-derived features linking plesidadapiforms to euprimates, with particular osteological structures of the grasping foot, can now be interpreted confidently as synapomorphy rather than parallelism [Bloch & Boyer, 2002; Bloch et al., 2007]. The early primate-like plesiadapiforms are usually grouped into several families (half a dozen to a dozen) including the better-known groups Microsyopidae, Micromomyidae, Paromomyidae, Carpolestidae, and Plesiadapidae [Bloch et al., 2007; Conroy, 1990; Fleagle, 1999; Rose, 2006]. The plesiadapiforms are found in Paleocene and Eocene deposits of Europe and North America, and range in size from that of a mouse to a large domestic cat [Gingerich, 1976; Kay & Cartmill, 1974; Silcox et al., 2007; Silcox & Gunnell, 2008]. In some ways, they did not look like modern primates. Some are viewed as persistently primitive in many respects (Microsyopidae) while others depart in a way indicating adaptation to angiosperm resources (Picrodontidae, Paromomyidae, Carpolestidae). The taxa known by cranial morphology had small brains, lacked orbital convergence and retained notable snouts and at least some sported large foramina to support olfactory senses [Gingerich, 1976; Kay & Cartmill, 1974]. At least one group of large-bodied forms had claws rather than nails, and limbs reminiscent of squirrel-like rodents Primate Origins Again / 97 [Gingerich, 1976]. But as more cranial and skeletal material of diverse plesiadapiforms has been recovered and analyzed, basicranial and limb structures distinctly similar to those of euprimates have been discovered [Bloch & Silcox, 2001; Bloch et al., 2007]. Scientists originally classified these animals as primates because they shared a number of molar characteristics with genera of Eocene euprimates. These distinctive features include the reduction in height and sharpness of the lower molar trigonids, and the broadening of the talonid basins for grinding. Even Purgatorius, the geologically oldest probable plesiadapiform shows these features [Clemens, 1974]. While these dental traits seem to represent a departure from a more insectivorous diet to one that includes more fruit or nuts, or other items requiring less piercing, slicing, and more grinding, it is also clear that there is no reason to interpret such dental departures as a single evolutionary event of phylogenetic significance without supporting evidence. From the start plesiadapiforms were viewed as less insectivorous and more frugivorous, or in some cases, flowerivorous or nectarivorous, than are primitive “insectivorous” mammals. It has been suggested, for example, that plesiadapiforms with these derived dental features and the earliest primates were the result of an evolutionary shift from mammals that were primarily insect eaters to those including more and more plant material in their diet [Szalay, 1968, 1972]. The teeth of very early euprimates clearly show a combination of frugivorous and insectivorous features [Strait, 1993]. Early Eocene euprimates do not have molars extremely specialized for insect diets, although some small ones like Teilhardina look more insectivorous than others [Ni et al., 2004], while larger ones like Cantius seem to have been more frugivorous [Gingerich, 1986]. Even in the most insectivorous of these early primates, the teeth depart from that of insectivores in having broader talonid basins (for grinding foods) and less trenchant trigonids (for piercing and slicing exoskeletons). This is dentally the most distinctive evolutionary trend of early primates and plesiadapiforms: that they move away from insectivory. These two groups shared a common ancestor to the exclusion of all other known groups of mammals, and we reiterate the point that the taxonomic and adaptive arguments about the primate status of plesiadapiforms are secondary to the phylogenetic likelihood that this is a single, natural, diverse radiation. THE DIETS OF SMALL-BODIED PRIMITIVE PRIMATES As a general rule, the diets of small extant primates derived from field studies follow a clear pattern: they are omnivores, eating a range of fruit, some insects, and other items. In our experience, there seems to be a perception among some colleagues that small-bodied primates are largely insectivorous. This may derive in part from comparative analyses that showed primates below 500 g would be expected to obtain their protein from insects, rather than leaves [Kay, 1984]. But just as large primates eat plenty of nonleaf foods, so small ones rely heavily on noninsect diets. The only primates known to be invertebrate specialists are members of the genus Tarsius and Loris [Nekaris & Rasmussen, 2003], which are also very specialized in other aspects of their locomotion and sensory systems [Nekaris, 2005; Rasmussen & Nekaris, 1998]. Other lorisines and galagines eat a broad range of plant foods [Nekaris & Bearder, 2007, Nekaris et al., 2010; Sussman, 2003a], cheirogaleids eat fruit, flowers, nectar, insect secretions, plant exudates [Atsalis, 2008; Sussman, 2003a], small anthropoids like callitrichids may be characterized as relying heavily on fruit, other angiosperm products, exudates, even fungus, along with a variable amount of insects or other invertebrates depending on species [Digby et al., 2011; Garber, 1993, Rehg, 2006; Sussman, 2003b]. Extant primates are simply not “insectivorous” in a strict sense except for two specialized lineages. In a broad review of the issue of size and diet drawing on the evidence from field studies, Ankel-Simons [2007, p 227] concluded that our knowledge of extant small primates “obviously contradicts the hypothesis that tiny primates must eat animal protein to survive.” Similar small-bodied mammals, like the arboreal phalangeroid marsupials of Australia and New Guinea, are distinctly not primary insectivores, but rather fruit and nectar specialists [Rasmussen & Sussman, 2007]. The same general dietary interpretation is true of the very primate-like Neotropical marsupial, Caluromys [Rasmussen, 1990]. The evidence from extant primates and, by inference from comparable dentitions of fossil primates, is overwhelmingly in favor of these animals being omnivores utilizing plant and animal resources. THE ORIGIN OF PRIMATES: TERMINAL BRANCH FEEDING INSECTIVORES Cartmill [1974, 1992] disagreed with the arboreal theory of primate evolution, and he rejected the notion that the plesiadapiforms were primates, based largely on two taxa, the large-bodied specialist Plesiadapis [Gingerich, 1976], and the primitive microsyopoid Palaechthon [Kay & Cartmill, 1977]. Other than teeth, practically nothing was known of Carpolestidae or other plesiadapiform families. Still, he considered plesiadapiforms to be close relatives of the earliest primates with whom they shared many traits, but he rejected the view that the plesiadapiforms exhibited an adaptive trend leading to Am. J. Primatol. 98 / Sussman et al. the primates (using an adaptive rather than phylogenetic definition of the order). Cartmill argued that, whereas some of these early mammals possess primate-like features, others do not display any significant adaptation that would justify including them in Primates. He argued further that a shift from eating primarily insects to eating primarily plants does not define a boundary between insectivores and primates. In fact, notable departures away from omnivory (insectivory/herbivory) toward specialized plant processing apparently did not appear among plesiadapiforms until as much as six million years after the earliest known members of the group appeared [Kay & Cartmill, 1977], although it is very important to note that the entire basis for recognizing plesiadapiform taxa in these first six million years was entirely based on broadening of the talonid and other characters associated with plant feeding. Additional intensive paleontological work on plesiadapiforms has demonstrated greater complexity. The earliest members of the group, what many call microsyopoids, were relatively similar to “insectivores,” but later forms representing subsequent evolutionary radiations show a range of adaptations to arboreal habitats and herbivorous diets. Even among arboreal plesiadapiforms, there is great variation in the nature of their arboreal adaptations [Bloch et al., 2007]. Cartmill asserted that an entirely different adaptive shift was responsible for the evolution of euprimates. The earliest primates were those characterized by the traits described by Le Gros Clark [1963] and Martin [1986, 1990]: for example, grasping ability of the hands and feet with flattened nails on the fingers and toes, orbital convergence and stereoscopic vision, and reduction of the snout and olfactory senses. Furthermore, Cartmill [1972, 1974] argued that if other mammals have been able to live in the trees and yet do not evolve primate morphology, there must be something wrong with the arboreal theory of primate evolution. Of the two dozen or so orders of terrestrial mammals, a dozen include arboreal forms. Many of these animals exhibit highly successful adaptations to life in the trees and yet do not possess characteristic primate traits. For example, tree squirrels (Sciurus) fill a specific set of niches in the forest canopy, but unlike primates, they lack orbital convergence, have small brains, reproduce rapidly, and bear claws on their digits instead of nails. Their behavior differs significantly from that of primates as well: they normally move vertically on trees, using their claws to run and climb on large branches and trunks and feeding on hard fruits and nuts [Garber & Sussman, 1984]. Another example of a successful group of arboreal mammals, the sloths, have large, nearly immobile, hook-like claws instead of prehensile hands [Montgomery & Sunguist, 1978]. Arboreal kangaroos also have no notable primate-like traits Am. J. Primatol. [Martin, 2005]. It is interesting to note that in those primates that have become secondarily adapted to feeding on sap that they extract from tree trunks, including callitrichids and the needle-clawed bush baby, Euoticus, the ancestral nails have evolved into claw-like structures [Burrows & Nash, 2010; Nash, 1986]. This provides an elegant example demonstrating that primate traits are context specific, and not simply a response to arboreality. We also have concrete evidence that the geologically earliest euprimates had nails instead of claws [Rose et al., 2011]. In an illuminating use of the comparative method, Cartmill [1974] compared the function of specific traits shared by primates and other animals in an attempt to determine the precise niche that early primates might have filled. He found that grasping hands and feet are common in animals that habitually forage in terminal branches, permitting the animals to suspend themselves by their hind limbs (and tail, if prehensile), while using the forelimbs to reach and manipulate food items. Some degree of prehensile hands and feet are widespread among shrub-layer insectivores and related herbivorous forms. Since Cartmill’s work, the prehensile hand and foot of primates has been intensively studied, often in the context of primate evolution [Hamrick, 1998, 2001; Lemelin, 1996, 1999]. Osteological correlates of primate prehensile hands have been identified, which allows us to evaluate fossil remains, and the behavioral correlates of prehensile structure have been studied in primates and primate analogs [Lemelin, 1999; Rasmussen & Sussman, 2007]. The grasping feet and hands of primates are unique in their anatomical details. Cartmill also compared animals with convergent orbits to those who did not have this feature. He found that outside of primates, this feature is largely but not entirely restricted to predators. Optic convergence is particularly marked in owls, hawks, and cats each of which depends on vision for the detection of prey. Given these facts, Cartmill [1972, 1974] argued that the earliest primate adaptation involved visually oriented predation on insects in the lower canopy and undergrowth of tropical forests. More recently, Allman [1977] suggested that forward-facing eyes did not enhance stereoscopic vision so much as they allowed an animal to see more clearly what is in front of it; a factor he speculated would be much more important to a nocturnal animal than to a diurnal one. This presumably explained why nocturnal predators such as owls and cats, as did early primates, had forward-facing eyes, whereas diurnal predators retained a more panoramic visual field. Thus, Cartmill [1992] modified his argument, claiming that the earliest primates were nocturnal visually oriented predators. The issue of whether or not the earliest euprimates (and for that matter, the plesiadapiforms) were nocturnal has recently come under scrutiny. Primate Origins Again / 99 While scholars have often accepted the idea that early primates must have been small and nocturnal (and eaten bugs), comparative studies of the opsin genes and comparative study of orbit size in fossil primates have raised the obvious possibility that ancestral primates were not nocturnal [see review by Ankel-Simons & Rasmussen, 2008]. After all, color vision seems especially suited to use for finding fruit and flowers in daylight rather than tracking insects at night. WERE THE EARLIEST PRIMATES NOCTURNAL VISUALLY ORIENTED PREDATORS? Although Cartmill’s argument is an elegant one, we must now ask, “Is it correct?” Are the features shared by living primates the result of the adaptive shift to the role of a nocturnal visually oriented predator that searched for insects on terminal branches? It is certainly true that many small mammals, including primates, eat insects. However, since Cartmill first presented his theory, the diet of a number of prosimians has been studied in detail. As mentioned earlier, the most heavily faunivorous primates are extreme specialists, such as Tarsius and Loris [Nekaris & Rasmussen, 2003]. The idea that cheirogaleids are the best living model of early or ancestral primates has been very influential [Charles-Dominique & Martin, 1970, volume edited by Charles-Dominique et al., 1980]. These primates are decidedly omnivorous [Ankel-Simons, 2007; Atsalis, 2008; Sussman, 2003a]. As we have reviewed, the entire idea of an ancestral, primate that is strictly insectivorous is unsupported by dental fossils or any other comparative evidence. For example, the general anatomy of the digestive tract of primates (e.g., the relative size of gut compartments including the caecum) reflects adaptations for an omnivorous diet (plant and animal food) [Martin, 1990]. The dietary pattern in the great majority of primates (over 95%) is omnivory [Harding, 1981]. Furthermore, the trend toward orbital convergence in primates culminates in the slow-moving lorises [Nekaris, 2005]. However, most lorises rely heavily on scent to detect their prey. Most of the insect prey captured by lorises are slow-moving and smelly [Rasmussen & Nekaris, 1998], though some do indeed seasonally spend a great deal of time eating floral nectar in terminal branches [Moore & Nekaris, 2012, personal communication]. Lorises detect insects with their highly developed sense of smell. Other nocturnal primates such as galagos, mouse lemurs, and tarsiers seem to use mainly hearing in hunting prey [Atsalis, 2008; Charles-Dominique, 1977; Crompton, 1995; Doyle, 1974; Martin, 1972; Neimitz, 1979; Oxnard et al., 1990; Pariente, 1979]. Some animals that have stereoscopy, such as sloths and koalas, are strictly plant eaters [Rasmussen & Sussman, 2007]. Many birds that do not have convergent eyes are highly insectivorous, detecting their prey visually. For example, the tyrannid flycatchers sit on high, exposed perches, visually identify insects flying by, dart out from their perch, and snap up the insects in mid-air. In fact, the only mammals that possess a complex visual system similar to that of primates are fruit bats (Megachiroptera), which rely on fruit and flower diets [Pettigrew, 1986]. It seems likely, therefore, that visual predation per se is not a sufficient explanation for the visual adaptations of the earliest primates. As we shall see, the difference between nocturnal and diurnal predation also is not a good explanation for the visual adaptations of modern primates. It is more likely that the primate visual system evolved in connection with feeding on fruits, flowers, and arthropods attracted to them in the small, terminal branches and dense foliage of forest trees. AN ALTERNATIVE THEORY OF PRIMATE ORIGINS: COEVOLUTION WITH ANGIOSPERMS The Paleocene–Eocene boundary was a period of rapid change involving notably the culmination of a warming trend and widespread tropical to subtropical conditions (the Early Eocene Climatic Optimum, ca. 56–48 Myr), when mean annual temperature increased by ca. 6? C over already very warm global conditions. This involved concurrent adaptive shifts in a number of plant and animal groups, including primates. These shifts seem to have begun earlier, before the temperature peak, at around 70 Ma (64– 78 Ma)[Steiper & Seiffert, 2012], a time in which the angiosperm component of floras had increased from 0% to 80% (80–125 Ma); this is considered a key event for the diversification of birds and mammals [Benton, 2010, Meredith et al., 2011]. It is in the context of the interrelationships among these groups that we might find an alternative hypothesis for the origin of primates [Go?mez & Verdu?, 2012; Rasmussen, 1990; Sussman, 1991; Sussman & Raven, 1978]. The uniqueness of the earliest primates of modern aspect (the euprimates) appears to involve a combination of the features described by both Cartmill and Szalay. The novel adaptive shift involved two aspects: (1) becoming well adapted to feed on small branches, and (2) including a high proportion of plant material in the diet. In this theory of primate evolution, the ecological resource providing a new basis for exploitation is identified—the fruit and flowers of flowering plants (angiosperms), and the insects attracted to these products. Am. J. Primatol. 100 / Sussman et al. Angiosperms are a monophyletic grouping of plants that contain all of the plant species that produce flowers to enhance their prospects of outcrossing and fruit to enhance their ability to disperse their seeds. In most flowering plants, both reproduction and dispersal require the cooperation of animals, often insects or birds. Flowers provide a reward of nectar to animals such as bees and hummingbirds that, in visiting the flower, will inadvertently rub off some pollen and carry it to the next flower. Fruits often are a sweet, nutritious offering to animals that will then swallow the seeds and defecate them later at a distance from the parent plant. Angiosperms have become very successful in part because of this elaborate coevolutionary system [Go?mez & Verdu?, 2012; Janson, 1983; Stiles, 1989; van der Pijl, 1982]. The evolution of modern birds and mammals is directly related to that of angiosperms chronologically and there are good reasons to believe the relationship was a driving force in the evolution of adaptive features of all groups involved in the coevolutionary relationship; in this context, planteating vertebrates and dispersers have had a powerful influence on angiosperm evolution and vice versa [Friis et al., 1987, and see references in Sussman, 1991]. Although the general outline of evolutionary events was initially traced between angiosperms and a number of animals, mainly insects and birds, this was later applied to primates [Sussman, 1991, 2003a; Sussman & Raven, 1978]. Although they surely arose somewhat earlier [Bell et al., 2010], the oldest known angiosperm fossils are from the early Cretaceous, approximately 120–130 Ma [Benton, 2010; Friis et al., 2011; Morley, 2000; Sun et al., 2011]. The earliest flowering plants were pollinated by insects, produced small seeds, and were wind or water dispersed with few or no specializations for animal dispersal [Crane et al., 1995; Friis & Crepet, 1987; Wing & Tiffney, 1987]. These angiosperms were small shrubs and herbs located mainly in unstable environments. By the late Cretaceous, many groups of angiospermous trees had appeared, with angiosperms becoming dominant about 90 Ma [Friis et al., 1987; Niklas et al., 1980; Tiffney, 1984]. During the Cretaceous, major coevolutionary events occurred between flowering plants and the animals with which they were interacting. For example, birds and mammals dispersed large fruits and seeds, the fruits often providing special attractions to their dispersers [Wing and Tiffney, 1987]. In the latter parts of the Cretaceous Period, from 90–65 Ma, angiosperms became dominant in most forests throughout the world. With the appearance of specialized fruits, seeds that had earlier mainly been dispersed by wind and water began to be dispersed by animals, predominantly vertebrates [Tiffney, 1981, 1984; Wing & Tiffney, 1987]. The total diversity of angiosperms continued to increase across the Am. J. Primatol. Cretaceous–Tertiary boundary and into the early Tertiary [Friis et al., 1987; Niklas et al., 1980]. This coincided with the extinction of the dinosaurs and with major evolutionary radiations of mammals and birds. These radiations included the origin and diversification of the plesiadapiforms in Europe and North America during the Paleocene. A similar diversification of marsupials occurred at the same time in South America [Clemens, 1968; Clemens et al., 1979]. The plesiadapiform radiation included the invasion of arboreal habitats by some lineages [Collinson & Hooker, 1987; Kay & Cartmill, 1977; Kay et al., 1990; Szalay & Dagosto, 1980]. Thus, the evolution of angiosperms between 80 and 130 Ma “was a key event in the diversification of mammals and birds” [Meredith et al., 2011 p 523]. Arboreality seems to have arisen in eutherian mammals concurrent with this angiosperm diversification of the Late Crecateous–Early Paleocene [Goswami et al., 2011]. A number of these mammals have dental morphology that suggests at least a partial switch to plant foods. This new feeding niche was the small branch milieu of the newly radiating flowering plants, which offered an array of previously unexploited resources, for example, flowers, fruits, flower and leaf buds, gums, nectars, and also the insects attracted to these resources. It cannot be understated that angiosperm products necessarily will have insects associated with them; from the point of view of an early primate foraging in terminal branches, angiosperm products, and insects are found conveniently together with great reliability [Rasmussen, 1990, 2001]. The radiation of angiosperms during the Cretaceous and early Tertiary included a systematic increase in fruit and seed sizes and this correlated with an increasing number of mammal-dispersed taxa [Tiffney, 2004]. By the Late Paleocene large seeds with large nutritious reserves became common. Thus, fruits and seeds with clear adaptations for mammal dispersal (thick walls, attractive flesh) were relatively common by the end of the Paleocene. Coevolutionary interactions between flowering plants and animals appear to have multiplied in late Paleocene–Eocene time [Tiffney, 2004]. At that time, much of the land from the Equator to the poles particularly after the Early Eocene Climatic Optimum was occupied by highly diverse subtropical to warm temperate rich forests [Smith et al., 2006, Townsend et al., 2010]. In these forests and in those that evolved from them later, there was a great diversity of large fruits and seeds, doubtless accompanied by an increased importance of seed dispersal by mammals and birds. As the Tertiary Period proceeded, development of sharper temperature gradients from Equator to poles led to the appearance of many modern ecosystems, including tundra and taiga near the poles in the Neogene and earlier that Primate Origins Again / 101 of evergreen tropical rainforests near the Equator [Graham, 2011]. During this same period, the earliest euprimates appeared and diversified [Rasmussen, 2007; Rose et al., 2011]. The appearance of modern primates coincides with the first record of fruit-eating birds [Olson, 1985; Tiffney, 1984]. Bats also appear and diversify in the early Eocene [Gunnell & Simmons, 2005, Habersetzer & Storch, 1987]. Although some early bats were insectivorous and even had evolved specialized echolocation [Gunnell et al., 2003], the morphological characters of bat-dispersed fruit and, likely, fruit bats also evolved during this time period. (The dentition of one early fruit bat was misidentified as a prosimian primate given their common angiosperm diet and inflicted with the name Propotto!; [Simpson, 1967]). Many other mammalian orders, including Rodentia, also appear in the fossil record for the first time [Tiffney, 1984, 2004]. Thus, the establishment of biological interactions between angiosperms and their pollinators and dispersers is reflected in the rapid appearance of modern families and genera in the Eocene. This further coincides with the maximum northwards geographic extent of tropical and subtropical ecosystems, including arborescent angiosperms, as evident by to the occurrence of tropical groups far in the north, such as Eocene tapirs, crocodilian, and banana-like plant lineages within the Arctic Circle [see Graham, 2011, p 345–346]. Indeed, the sweep of tropical ecosystems northwards probably carried with it many tropical groups, like primates, where they first appear to us in the fossil record in northern areas that today are temperate [Smith et al., 2006]. It was also a time of high angiosperm biodiversity [Wilf et al., 2003, 2005]. The evolution of modern primates parallels that of other omnivorous mammals, of plant-eating birds, and of many of the features of modern angiosperms. Viewed in this light, it appears likely that many of these organisms were linked by coevolutionary relationships. At present, frugivorous birds, bats, and primates are the most important seed dispersers in the tropics [Chapman, 1995; Go?mez & Verdu?, 2012; Russo & Chapman, 2011; Stiles, 1989; Terborgh, 1992]. Recent research has illuminated this coevolutionary relationship. Go?mez & Verdu? [2012] combined phylogentic, neontological, and paleontological data to show that a facultative mutualistic plant– animal interaction emerging from frugivory and seed dispersal contributed to the diversification of primates. Compiling data from 381 extant and 556 extinct primates, they found that mutualistic extant primates had higher speciation rates, lower extinction rates, and higher diversification rates than nonmutualistic ones. Similar results were found among fossil primates, with mutualistic ones having higher geographic durations and smaller per capita rates of extinction than nonmutualistic ones. Go?mez and Verdu? also found that both extant and extinct mutualistic primates have significantly larger geographic ranges, which promotes diversification by reducing extinction rates and increasing geographic speciation. They concluded that “these outcomes together strongly suggest that the establishment of a facultative mutualism with plants has greatly benefited primate evolution and fueled its taxonomic diversification” [Go?mez & Verdu?, 2012, p 567]. The evolution of modern primates, therefore, as well as that of fruit bats and fruit-eating birds, may be directly related to the evolution of improved means of exploiting, or we might say synergizing with, the fruits and seeds of flowering plants. TESTS OF THE ANGIOSPERM EVOLUTION HYPOTHESIS How would we know which of the above hypotheses were correct? Is there a way to test the alternatives? Cartmill [1992] suggested a way that this could be done. If the earliest primates had grasping feet and blunt teeth adapted for eating fruit, but also possessed small, divergent orbits like those of plesiadapiforms, the hypothesis that their features evolved in connection with their feeding on terminal branches of trees would be supported. In, 2002, Bloch and Boyer discovered a remarkably well-preserved, late Paleocene (56-million-year-old) skeleton of a carpolestid plesiadapiform (Carpolestes simpsoni) in Wyoming. The completeness of the fossil of this tiny animal is a testament to the collecting and fossil preparation skills of the researchers. This fossil was much more complete than those of plesiadapiforms described earlier, and includes tiny hand and foot bones, as well as a skull. It allows us to test the proposed theories concerning the evolution of modern primates. Carpolestes lacks primate visual specializations such as convergent orbits and a postorbital bar. However, it has a divergent and opposable big toe or hallux, with a nail that was adapted for strong grasping of small-diameter supports similar to that of the euprimates. It also has correlated grasping specializations in the hand. In intricate detail, the foot bones of Carpolestes resemble those of primates. Carpolestes appears to have had finger and toe proportions supporting the hypothesis that it moved among slender branches. In addition, the morphology of the molar teeth is consistent with a fruit-rich diet. Bloch & Boyer [2002, p 1609] concluded: “The fossil find presented here is consistent with the hypothesis that early euprimates evolved grasping first and convergent orbits later and inconsistent with the visual predation hypothesis. [Carpolestes] simpsoni had feet with strikingly euprimate-like grasping, low-crowned molar teeth adapted for eating fruit, and small divergent orbits.” This sequence of Am. J. Primatol. 102 / Sussman et al. acquisitions is that predicted by the angiosperm exploitation theory [see discussion by Bloch & Boyer, 2003; Kirk et al., 2003]. Another way to test theories of primate origins relies on the comparative study of early primate analogs alive today, using them as surrogates for the fossil record. The marsupials are an incredibly diverse monophyletic group of mammals, containing creatures as diverse as kangaroos, opossums, wombats, koalas, and many more. Among these diverse marsupials are some that are remarkably primatelike [Rasmussen & Sussman, 2007]. Those with these features have grasping hands and feet, relatively high degrees of stereoscopic vision, relatively larger brains than their kin, and lower reproductive rates than most other marsupials. The evolution of these traits in just a few marsupial lineages suggests that they may have undergone selective pressures that parallel the evolutionary forces that shaped the euprimates and seems to have resulted in similar features in both groups. The primate-like marsupials include the woolly opossum (Caluromys) of South America, and the phalangeroid group of Australia (“phalanx” means finger, and the group is named after their primatelike fingers), which includes among other forms the pygmy possums (Cercartetus). Studies of the primate-like woolly opossums and pygmy possums in the wild have revealed many details of their behavior and ecology. These animals are specialists on angiosperm fruits and flowers and on the insects associated with them in the terminal branches of rainforest trees [Rasmussen, 1990; Rasmussen & Sussman, 2007; see also Dominy et al., 2001; Melin et al., 2012; Muchlinski & Perry, 2011]. The fact that two groups not directly related to primates have acquired primate-like traits as adaptations to feeding among the terminal branches of trees and shrubs on fruits, flowers, and insects provides persuasive support for the angiosperm theory of primate origins [Rasmussen & Sussman, 2007]. It is also interesting to note that the phalangeroid marsupials such as the pygmy possum evolved in Australia, which was not colonized by primates until the relatively recent arrival of human beings, and the woolly opossum evolved in South America, which primates appear to have colonized only about 30 Ma [Takai et al., 2000]. In other words, primate-like marsupials apparently evolved in the terminal branches of angiosperms only in places that primates had not already reached. The study of parallelisms between primates and marsupials (or other primate-like mammals) is not simply a matter of pointing out a similarity in one aspect of behavior observed on occasion. Rather, the intent is to evaluate the diversity within a phylogenetically distant group in which one lineage or clade is primate-like both in morphology and behavioral ecology while other members of the group are not. Am. J. Primatol. What about orbital convergence? If the first primates of modern aspect were not visually oriented predators, what might have been the advantage of forward-facing eyes? Recently, Changizi & Shimojo [2008] and Changizi [2009] have proposed the “X-ray vision” hypothesis for the evolution of primate visual adaptations. They propose that the degree of binocular convergence is selected to maximize how much of its environment a mammal can see. Using data from 319 species and 17 orders of mammals [derived from Heesy, 2003], they found that, in “noncluttered” or “nonleafy” environments, mammals can see the most of their surroundings with panoramic, laterally directed eyes. On the other hand, in cluttered environments such as dense forest canopy, mammals can see more of their environment when their eyes face forward, because binocularity has the power of “seeing through” the leaf clutter. Using sophisticated calculations, Changizi [2009] estimated that two forward facing eyes can see two to eight times as far in a highly cluttered environment than can a lone eye. Furthermore, Changizi predicted that, if the hypothesis that forward-facing eyes are an adaptation for leafy environments is correct, there should be a relationship between body size and forward-facing eyes because for small animals, in dense leafy habitats, the leaves would be too large for them to benefit greatly from X-ray vision. On the other hand, larger “leafy-loving” animals would be able to better “see through” leaves and thus benefit more from X-ray vision, and should have progressively more forwardfacing eyes. Using the same data set of 319 mammals, Changizi [2009] found that among animals living in leafy environments, large animals tend to have eyes that face farther forward than small animals, regardless of whether they are predators or not, and regardless of their activity cycles. Changizi [2009, p 96] stated: “It may be, then, that one finds some predators with forward-facing eyes not because they need the three-dimensional stereo vision we get from our binocular field to catch prey, but because of their leafy-loving nature.” Changizi & Shimojo [2009], further argued that the “optical blurring” hypothesis of Allman [1977] does not sufficiently explain why the advantages proposed are worth the cost of the loss of vision in the periphery. They point out that the “conventional view” of nocturnal visually oriented predation [Howland, 2009] consists of an unparsimonious layered mix of multiple subhypotheses, which makes it difficult to theorize about and test. In contrast, their ‘‘X-ray’’ hypothesis is simple, relying only on the desideratum that animals should evolve to ‘‘see the most’’ with their two eyes. Changizi & Shimozo [2008, p 764] summarized by stating: “The results we have seen do not tend to support the conclusion that predators have the greatest convergence . . . (and also that) convergence is low for small predators but increases with body Primate Origins Again / 103 size, something that the visual predation hypothesis cannot explain . . . .The central idea behind (the angiosperm hypothesis) is . . . that there was “diffuse coevolutionary” interaction between flowering plants and animals . . . so that as flowering plants evolved, . . . some animals moved into the terminal branch niches . . . .Our hypothesis fits well with Sussman’s hypothesis for primate origins.” SUMMARY Primates is a monophyletic order of mammals containing considerable diversity among extant and fossil forms, which derived from a common ancestor, one most likely similar to the newly discovered late Paleocene fossil C. simpsoni. Members of this order share a number of morphological traits, especially of the locomotor anatomy, skull morphology, dentition, and reproductive biology, which distinguish them from other mammals. These traits indicate that the adaptive shift accompanying the appearance of the primates was the occupation of new locomotor and feeding niches made available by the coevolving flowering plants of tropical rain forests. Although bats and birds can reach the relatively slender terminal branches of large rain forest trees by flying to them, primates need their grasping appendages to obtain the same advantage and need forward-facing eyes to see through the clutter of the dense forest canopy. In fact, apart from a few primate-like marsupials, primates are the only major taxonomic groups of nonflying vertebrates to exploit the terminal branch niche of the tropical forest regularly. ACKNOWLEDGMENTS We acknowledge the excellent suggestions of one anonymous reviewer, Paul Garber, and Alan Graham. Their suggestions greatly improved this manuscript. This research adhered to the American Society of Primatologists principles for the ethical treatment of primates. REFERENCES Adkins RM, Honeycutt RL. 1991. Molecular phylogeny of the superorder Archonta. Proc Nat Acad Sci, USA 88:10317– 10321. Allman J. 1977. Evolution of the visual system in the early primates. Prog Psychobiol Physiol Psychol 7:1–53. Ankel-Simons F. 2007. Primate anatomy. 3rd edition. San Diego: Elsevier Inc. Ankel-Simons F, Rasmussen DT. 2008. Diurnality, nocturnality, and the evolution of primate visual systems. Yearb Phys Anthropol 47:100–117. Armelagos GJ. 1997. Smith, (Sir) Grafton Elliot. In: Spencer F, editor. History of physical anthropology, Volume 2. New York: Garland Publishing. p 955–956. Atsalis S. 2008. A natural history of the brown mouse lemur. New York: Prentice Hall. Beard KC. 1990. Gliding behavior and palaeoecology of the allged primate family Paromomyidae (Mammalia, Dermoptera). Nature 345:340–341. Beard KC. 1993. Origin and evolution of gliding in early Cenozoic Dermoptera [Mammalia, Primatomorpha]. In: MacPhee RDE, editor. Primates and their relatives in phylogenetic perspective. New York: Plenum Press. p 63–90. Bell CD, Soltis DE, Soltis PS. 2010. The age and diversification of angiosperms re-visited. Am J Botany 97:1296– 1303. Benton MJ. 2010. The origins of modern biodiversity on land. Phil Trans R Soc 365:3667–3679. Bloch JI, Boyer DM. 2002. Grasping primate origins. Science 298:1606–1610. Bloch JI, Boyer DM. 2003. Response to comment on “Grasping primate origins.” Science 300:741. Bloch JI, Silcox MT. 2001. New basicrania of Paleocene-Eocene Ignacius: re-evaluation of the plesiadapiform-dermopteran link. Am J Phys Anthropol 116:184–198. Bloch JI, Silcox MT, Boyer, DM, Sargis EJ. 2007. New Paleocene skeletons and the relationship of plesiadapiforms to crown-clade primates. Proc Natl Acad Sci 104:1169– 1164. Burrows AM, Nash LT, editors. 2010. The evolution of exudativory in primates. New York: Springer. Butler PM. 2010. Neogene Insectivora. In: Werdelin L, Sanders WJ, editors. Cenozoic mammals of Africa. Berkeley: University of California Press. p 573–580. Cartmill M. 1972. Arboreal adaptations and the origin of the order Primates. In: Tuttle R, editor. The functional and evolutionary biology of primates. Chicago: Aldine Press. p 97– 122. Cartmill M. 1974. Rethinking primate origins. Science 184:436–443. Cartmill M. 1992. New views on primate origins. Evol Anthropol 1:105–111. Changizi M. 2009. The vision revolution. Dallas: BenBella Books. Changizi MA, Shimojo S. 2008. “X-ray vision” and the evolution of forward-facing eyes. J Theoret Biol 254:756– 767. Changizi MA, Shimojo S. 2009. Response to H.C. Howland, “Orbital orientation is not visual orientation.” J Theoret Biol 257:524–525. Chapman CA. 1995. Primate seed dispersal: coevolution and conservation implications. Evol Anthropol 4:74–82. Charles-Dominique P. 1977. Ecology and behavior of nocturnal primates. New York: Columbia University Press. Charles-Dominique P, Cooper HM, Hladik A, Hladik CM, Pages E, Pariente GE, Petter-Rousseaux A, Schilling A. 1980. Nocturnal Malagasy primates. New York: Academic Press. Charles-Dominique P, Martin RD. 1970. Evolution of lorises and lemurs. Nature 227:257–260. Clemens WA. 1968. Origins and early evolution of marsupials. Evolution 22:1–18. Clemens WA. 1974. Purgatorius, an early paromomyid primate [Mammalia]. Science 184:903–906. Clemens WA. Lellegraven JA, Linsday EH, Simpson GG. 1979. Where, when and wha . . . a survey of known Mesozoic mammal distribution. In: Lillegraven JA, Kielan-Jaworowska Z, Clemens WA, editors. Mesozoic mammals. Berkeley; University of California Press. p 7–58. Collinson ME, Hooker JJ. 1987. Vegetational and mammalian faunal changes in the early tertiary of southern England. In: Friis EM, Chaloner WG, Crane PR, editors. The origins of angiosperms and their biological consequences. Cambridge: Cambridge University Press. p 259–304. Conroy GC. 1990. Primate evolution, New York: Norton. Crane PR, Friis EM, Pederson KR. 1995. The origin and early diversification of angiosperms. Nature 374:27–33. Am. J. Primatol. 104 / Sussman et al. Crompton RH. 1995. “Visual predation.” Habitat structure and the ancestral primate niche. In: Alterman L, Doyle GA, Izard MK, editors. Creatures of the dark: the nocturnal prosimian. New York: Plenum. p 11–30. Digby L, Ferarri S, Saltzman W. 2011. Callitrichines. In: Campbell C, Fuentes CA, MacKinnon K, Panger M, Bearder S, editors. Primates in perspective. New York: Oxford University Press. p 85–106. Dominy N, Lucas PW, Osorio D, Yamashita N. 2001. The sensory ecology of primate food perception. Evol Anthropol 10:171–186. Doyle GA. 1974. Behavior of prosimians. In: Schrier AM, Stolnitz F, editors. Behavior of nonhuman primates. New York: Academic Press. p 155–353. Dunn RH, Rasmussen DT. 2009. Skeletal morphology of a new genus of Eocene insectivore [Mammalia, Erinaceomorpha] from Utah. J Mammol 90:321–331. Fleagle JG. 1999. Primate adaptation and evolution. New York: Academic Press. Friis EM, Chaloner WG, Crane PR, editors. 1987. The origins of angiosperms and their biological consequences. Cambridge: Cambridge University Press. Friis EM, Crane PR, Pedersen KR. 2011. The early flowers and angiosperm evolution. New York: Cambridge University Press. Friis EM, Crepet WL. 1987. Time and appearance of floral features. In: Friis EM, Chaloner WG, Crane PR, editors. The origins of angiosperms and their biological consequences. Cambridge: Cambridge University Press. p 259– 304. Garber PA. 1993. Feeding ecology and behaviour of the genus Saguinus. In: Rylands AB, editor. Marmosets and tamarins, Oxford: Oxford University Press. p 275–294. Garber PA, Sussman RW. 1984. Ecological distinctions in sympatric species of Saguinus and Sciureus. Am J Phys Anthropol 65:135–146. Gingerich PD. 1976. Cranial anatomy and evolution of early Tertiary Plesiadapidae [Mammalia, Primates]. Papers on paleontology, University of Michigan Papers on Paleontology 15:1–140. Gingerich PD. 1986. Early Eocene Cantius torresi: the oldest primate of modern aspect from North America. Nature 319:319–321. Goswami A, Upchurch P, Boyer DM, Seiffert ER, Verma O, Gheerbrant E, Flynn JJ. 2011. A radiation of arboreal basal eutherian mammals beginning in the Late Cretaceous of India. Proc Nat Acad Sci USA 108:16333–16338. Go?mez JM, Verdu? M. 2012. Mutualism with plants drives primate diversification. Syst Biol 61:567–577. Graham A. 2011. The age and diversification of terrestrial New World ecosystems through Cretaceous and Cenozoic time. Am J Botany 98:336–351. Gregory WK. 1910. The orders of mammals. Bull Am Museum Nat Hist 27:1–524. Gunnell, GF. 1989. Evolutionary history of Microsyopodia and the relationship between Plesiadiformes and Primates. Univ Michigan Papers Paleontol 27:1–154. Gunnell GF, Simmons NB. 2005. Fossil evidence and the origin of bats. J Mammal Evol 12:209–246. Gunnell GF, Jacobs BF, Herendeen PS, et al., 2003. Oldest placental mammal from Sub-Sarahan Africa: Eocene microbat from Tanzania: evidence for early evolution of sophisticated echolocation. Palaeontographica 5:2224–2226. Habersetzer J, Storch G. 1987. Klassification and funtionelle Flugelmorphologie palaongener Fledermause [Mammalia, Chiroptera]. Courier Forschungsinstitut Senckenberg 91:11–150. Hamrick MW. 1998. Functional and adaptive significance of primate pads and claws: evidence from New World anthropoids. Am J Phys Anthropol 106:113–127. Hamrick MW. 2001. Development and evolution of the mam- Am. J. Primatol. malian limb: adaptive diversification of nails, hooves, and claws. Evol Dev 3:355–63. Harding RSO. 1981. An order of omnivores: nonhuman primates diets in the wild. In: Harding RSO, Teleki G, editors. Omnivorous primates. New York: Columbia University Press. p 191–214. Heesy CP. 2003. The evolution of orbit orientation in mammals and the function of the primate postorbital bar. [Ph.D. thesis]. Stony Brook, NY: Stony Brook University. Hoffstetter R. 1977. Phyloge?nie des primates. Confrontation des resultats obtenus par les diverses voies d’approche du probleme. Bulletins and Me?moires Socie?te? d’Anthropologie de Paris t.4, se?rie XIII:327–346. Howells WW. 1947. Mankind so far. Garden City: Doubleday. Howland HC. 2009. Letter to the editor: Orbital orientation is not visual orientation: a comment on “X-Ray Vision and the evolution of forward-facing eyes” by M.A. Changizi and S. Shimojo. J Theoret Biol 257:522–523. Janson CH. 1983. Adaptation of fruit morphology to dispersal agents in a neotropical forest. Science 219:187–189. Jones FW. 1916. Arboreal man. London: Edward Arnold. Kay RF. 1984. On the use of anatomical characters to infer foraging behavior in extinct primates. In: Rodman PS, Cant JGH, editors. Adaptations for foraging in nonhuman primates. New York: Columbia University Press. p 21–53. Kay RF, Cartmill M. 1974. Skull of Palaechthon nacimienti. Nature 252:37–38. Kay RF, Cartmill M. 1977. Cranial morphology and adaptations of Palaecthon nacimienti and other Parmomyidae [Plesiadapoidea? Primates], with description of new genus and species. J Hum Evol 6:19–53. Kay RF, Thorington RW, Houde P. 1990. Eocene plesiadapiform shows affinities with flying lemurs not primates. Nature 345:342–344. Kirk EC, Cartmill M, Kay RF. 2003. Comment on “Grasping primate origins.” Science 300:741. Krause DW. 1991. Were paromomyids gliders? Maybe, maybe not. J Hum Evol 21:177–188. Le Gros Clark WE. 1963. The antecedents of man. New York: Harperand Row. Lemelin P. 1996. The evolution of manual prehensility in primates: a comparative study of prosimians and didelphid marsupials [Ph.D. thesis]. Stony Brook, NY: Stony Brook University. Lemelin P. 1999. Morphological correlates of substrate use in didelphid marsupials: implications for primate origins. J Zool 247:165–175. MacPhee RDE., Cartmill M, Gingerich PD. 1983. New Palaeogene primate basicrania and the definition of the order Primates. Nature 301:509–511. Martin R. 2005. Tree-kangaroos of Australia and New Guinea. Australia: Collingwood Press. Martin RD. 1972. A preliminary field-study of the lesser mouse lemur [Microcebus murinus J.F. Miller 1777]. Tierpsychologie 9:43–89. Martin RD. 1986. Primates: a definition. In: Wood BA, Martin LB, Andrews P, editors. Major topics in primate and human evolution. Cambridge: Cambridge University Press. p 1–31. Martin RD. 1990. Primate origins and evolution: a phylogenetic reconstruction. Princeton: Princeton University Press. McKenna MC. 1966. Paleontology and the origins of primates. Folia Primatologica 4:1–25. Melin AD, Moritz, GL, Fosbury AE, Kawamura S, Dominy NJ. 2012. Why aye-ayes see blue. Am J Primatol 74:185–192. Meredith RW, Janec?ka JE, Gatesy J, et al. 2011. Impacts of the cretaceous terrestrial revolution and KPg extinction on mammal diversification. Science 334:521–524. Montgomery GG, Sunquist ME. 1978. Habitat selection and use by two- and three-toed sloths. In: Montgomery GG, editor. The ecology of arboreal folivores. Washington D.C.: Smithsonian Institution Press. p 329–359. Primate Origins Again / 105 Morley RJ. 2000. The origin and evolution of tropical rainforests. Chichester: John Wiley and Sons. Muchlinski MN, Perry JMG. 2011. Anatomical correlates to nectar feeding among the strepsirrhines of Madagascar: implications for interpreting the fossil record. Anat Res Int doi:10.1155/2011/378431. Nash LT. 1986. Dietary, behavioral, and morphological aspects of gumnivory in primates. Yearb Phys Anthropol 29:113–137. Neimitz C. 1979. Outline of the behavior of Tarsius bancanus. In: Doyle GA, Martin RD, editors. The study of prosimian behavior. New York: Academic Press. p 631–660. Nekaris KAI. 2005. Foraging behavior of the slender loris [Loris lydekkerianus]: implications for theories of primate origins. J Hum Evol 49:289–300. Nekaris KAI, Bearder SK. 2007. The lorisiform primates of Asia and mainland Africa: diversity shrouded in darkness”. In: Campbell C, Fuentes A, MacKinnen K, Panger M, Bearder SK, editors. Primates in perspective. New York: Oxford University Press. Nekaris KAI, Rasmussen DT. 2003. Diet and feeding behavior of Mysore slender loris. Int J Primatol 24:33–46. Nekaris KAI, Starr CR, Collins RL, Wilson A. 2010. Comparative ecology of exudate feeding by lorises [Nycticebus, Loris] and pottos [Perodicticus, Arctocebus]. In: Burrows AM, Nash LT, editors. Evolution of exudativory in primates. New York: Springer. p 155–168. Ni X, Wang W, Hu Y, Li C. 2004. A euprimate skill from the early Eocene of China. Nature 427:65–68. Niklas KJ, Tiffney BH, Knoll AH. 1980. Apparent changes in the diversity of fossil plants: a preliminary assessment. In: Hecht MK, Steere WC, Wallace B, editors. Evolutionary biology, Vol. 12. New York: Plenum. p 1–89. Olson SL. 1985. The fossil record of birds. In: Farner D, King J, Parkes JK, editors. Avion biology, Vol. 2. New York: Academic Press. p 79–237. Oxnard CE, Crompton RH, Lieberman SS. 1990. Animal lifestyles and anatomies: the case of the prosimian primates. Seattle: University of Washington Press. Pariente G. 1979. The role of vision in prosimian behavior. In: Doyle GA, Martin RD, editors. The study of prosimian behavior. New York: Academic Press. p 411–459. Pettigrew JD. 1986. Flying primates? Megabats have the advanced pathway from eye to midbrain. Science 231:1304– 1306. van der Pijl L. 1982. Principles of dispersal in higher plants. Berlin: Springer-Verlag. Rasmussen DT. 1990. Primate origins: lessons from a Neotropical marsupial. Am J Primatol 22:263–277. Rasmussen DT. 2001. Primate origins. In: Hartwig WC, editor. The Primate Fossil Record. London: Cambridge University Press. p 5–10. Rasmussen DT. 2007. Fossil record of the primates from the Paleocene to the Miocene. In: Henke W, Tattersall I, editors. Handbook of paleo-anthropology, Berlin: Springer-Verlag. p 889–890. Rasmussen DT, Nekaris KAI. 1998. The evolutionary history of the lorisiform primates. Folia Primatologica 69(Supp. 1):250–287. Rasmussen DT, Sussman RW. 2007. Parallelisms among primate and possums. In: Ravosa M, Dagosto M, editors. Primate origins: adaptations and evolution. New York: Springer. p 775–803. Rehg JA. 2006. Seasonal variation in polyspecific associations among Callimico goeldii, Saguinus labiatus, and S. fuscicollis in Acre, Brazil. Int J Primatol 27:1399–1428. Rose KD. 2006. The beginnings of the age of mammals. Baltimore: Johns Hopkins University Press. Rose KD, Chester SGB, Dunn RH, Boyer DM, Bloch JI. 2011. New fossils of the oldest North American euprimate Teilhardina brandti [Omomyidae] from the Paleocene– Eocene thermal maximum. Am J Phys Anthropol 146:281– 305. Russo SE, Chapman CA. 2011. Primate seed dispersal: linking behavioral ecology with forest community structure. In: Campbell CJ, Fuentes AF, MacKinnon KC, Bearder S, Stumpf RM, editors. Primates in perspective. 2nd edition. Oxford: Oxford University Press. p 523– 534. Seiffert ER. 2010. Paleogene “Insectivores”. In: Werdelin L, Sanders WJ, editors. Cenozoic mammals of Africa. Berkeley: University of California Press. p 253–260. Silcox MT, Gunnell GG. 2008. Plesiadapiformes. In: Janis CM, Gunnell GF, Uhen MD, editors. Evolution of tertiary mammals of North America. Cambridge: Cambridge University Press. p 207–238. Silcox MT, Sargis EJ, Bloch JI, Boyer DM. 2007. Primate origins and supraordinal relationships: morphological evidence. In: Henke W, Tattersall I, editors. Handbook of paleo-anthropology. Berlin: Springer-Verlag. p 832– 860. Simpson GG. 1945. The principles of classification and a classification of mammals. Bull Am Museum Nat Hist 85:i-xvi, 1–350. Simpson GG. 1955. The meaning of evolution. Denver: Mentor Books. Simpson GG. 1961. Principles of animal taxonomy. New York: Columbia University Press. Simpson GG. 1967. The tertiary lorisiform primates of Africa. Bull Mus Comp Zool 136:39–62. Smith GE. 1912. The evolution of man. Smithsonian Institution Annual Report. Washington DC: Smithsonian Institution. Smith T, Rose KD, Gingerich PD. 2006. Rapid Asia-EuropeNorth America geographic dispersal of earliest Eocene primate Teilhardina during the Paleocene thermal maximum. Proc Nat Acad Sci 103:11233–11277. Steiper ME, Seiffert ER. 2012. Evidence for a convergent slowdown in primate molecular rates and its implications for the timing of early primate evolution. ProcNat Acad Sci 109:6006–6011. Stiles EW. 1989. Fruits, seeds, and dispersal agents. In: Abrahamson WG, editor. Plant-animal interactions. New York: Mcgraw-Hill. p 87–122. Strait SG. 1993. Molar morphology and food texture among small-bodied insectivorous mammals. J Mammol 74:391– 402. Sun G, Dilcher DL, Wang H, Chen Z. 2011. A eudicot from The Early Cretaceous of China. Nature 471:625–628. Sussman RW. 1991. Primate origins and the evolution of angiosperms. Am J Primatol 23:209–223. Sussman RW. 2003a. Primate ecology and social structure: volume 1, lorises, lemurs, and tarsiers. New York: Prentice Hall. Sussman RW. 2003b. Primate ecology and social structure: volume 2, new world primates. New York: Prentice Hall. Sussman RW, Raven PH. 1978. Pollination by lemurs and marsupials: an archaic coevolutionary system. Science 200:731–736. Szalay FS. 1968. The beginnings of primates. Evolution 22:19– 36. Szalay FS. 1972. Paleobiology of the earliest primates. In: Tuttle R, editor, The functional and evolutionary biology of primates. Chicago: Aldine. p 3–35. Szalay FS, Dagosto M. 1980. Locomotor adaptations as reflected on the humerous of Paleogene primates. Folia Primatologica 4:1–45. Takai M, Anaya F, Shigehara N, Setoguchi T. 2000. New fossil materials of the earliest new world monkey, Branisella boliviana, and the problem of platyrrhine origins. Am J Phys Anthropol 111:263–281. Am. J. Primatol. 106 / Sussman et al. Terborgh J. 1992. Diversity and the tropical rain forest. New York: Scientific American Library. Tiffney BH. 1981. Diversity and major events in the evolution of land plants. In: Niklas KJ, editor. Paleobotany, paleoecology, and evolution, Volume II. New York: Praeger. p 193–230. Tiffney BH. 1984. Seed size, dispersal syndromes, and the rise of the angiosperms: evidence and hypothesis. Ann Missouri Bot Gard 71:551–576. Tiffney BH. 2004. Vertebrate dispersal of seed plants through time. Annu Rev Ecol Syst 35:1–29. Townsend KE, Rasmussen DT, Murphey PC, Evanoff E. 2010. Middle Eocene habitat shifts in the North American Am. J. Primatol. western interior: a case study. Palaeogeogr Palaeoclimatol Palaeoecol 297:144–158. Wilf P, Johnson KR, Rube?n N, et al. 2005. Eocene plant diversity at Laguna del Hunco and R??o Pichileufu?, Patagonia, Argentina. Am Natural 165:634–650. Wilf P, Rube?n N, Cu?neo NR, et al. 2003. High plant diversity in Eocene South America: evidence from Patagonia. Science 300:122–125. Wing SL, Tiffney BH. 1987. Interactions of angiosperms and herbivorous tetrapods through time. In: Friis EM, Chaloner WG, Crane PR, editors. The origins of angiosperms and their biological consequences. Cambridge: Cambridge University Press. p 203–224. Rethinking Primate Origins The characteristic primate traits cannot be explained simply as adaptations to arboreal life. Matt Cartmill The author is an assistant professor in the departments of anatomy and anthropology at Duke University, Durham, North Carolina 27710. 436 explanation in question; and that such objections must be raised systematically if we wish to arrive at adequate explanations of historical processes. These assumptions underlie the following reassessment of what has been called the arboreal theory of primate evolution. The Arboreal Theory and Its Background The Linnean concept of the order Primates, which included the bats and colugos, was still current as late as 1870 (4). In 1873, Darwin's antagonist Mivart proposed ordinal boundaries which excluded these animals, but which (unlike the taxonomies then advocated by Milne-Edwards, Grandidier, and Gervais) included the prosimians as a suborder of Primates (5). Mivart also proposed a list of traits that distinguished prosimians and anthropoids from other placental mammals. These traits included a complete bony ring around the eye, a well-developed occipital lobe of the cerebral cortex, and a grasping hind foot with an opposable, clawless first toe. In the second decade of the 20th century, G. E. Smith and his pupil, F. W. Jones, put forth the first systematic attempts at explaining these and other characteristic primate traits in terms of natural selection. Smith, a comparative neuroanatomist, was principally concerned with explaining the distinctive features of primate brains. He proposed (6) that the remote ancestors of the primates were shrewlike terrestrial creatures that entered upon an arboreal way of life. In the complex networks of tree branches through which these early primates moved and foraged, the olfactory and tactile receptors in the snout did not provide adequate guidance; snuffling blindlly along Downloaded from http://science.sciencemag.org/ on May 9, 2021 If you asked a student of human evolution to explain why human beings, unlike other mammals, walk around on only two legs, you would be 'baffled and unhappy if he answered, "Because in man's ancestral lineage, individuals who could not run away from predators left fewer offspring." You would be justified in retorting that the same remarks apply equally to thousands of other species of mammals, yet none of these have developed upright bipedal locomotion. The purported explanation, you would properly conclude, may be a true proposition, but is worthless as an explanation. An explanation is-a hypothesis of a complex sort. Ordinarily, to explain one fact in terms of another requires that there be an a posteriori rule which allows us to deduce the first from the second, and which warrants testable expectations other than the one in question (1). We reject the foregoing "explanation" of human bipedality because we sense that its explanatory force depends on the lawlike generalization, "Natural selection favors bipedal locomotion in any mammal species that has predators," and that this generalization is false. Yet some evolutionary biologists and philosophers of science (2) have argued that evolutionary explanations do not involve any such generalizations, and hence are not subject to refutation by counterexamples. In this view, we have no grounds for dismissing the "explanation" with which I began; the objection that the same remarks apply to species which have remained quadrupedal is beside the point. I have suggested elsewhere (3) that this and similar objections are very much to the point; that, when valid, they demonstrate the inadequacy of the in hopes of scenting something edible, as most living insectivores do, was no longer a viable foraging pattern. Accordingly, vision gradually replaced olfaction as the dominant sense. In correlation with *this, the hand assumed the tactile and grasping functions primitively served by the mouth and lips; eye-hand coordination replaced nosemouth coordination. Arboreal life alsto required more precise and rapid motor responses. Thus, Smith was able to account for the primates' reduced olfactory centers and elaborated visual, tactile, motor, and association cortex in terms of the selection pressures exerted by the arboreal environment. Jones's reinterpretation of these ideas (7) reflects his professional interest in the anatomy of the hand and foot. Jones proposed that the arboreal habit led to a functional differentiation of the limbs. While the foot remained a relatively passive organ of support and propulsion, the hand, used by the primate ancestors for reaching out and grasping new supports when climbing about in trees, became specialized for prehension-and therefore preadapted to take over the mouth's functions of manipulation and food-gathering. As the snout lost importance as a sensory and manipulative organ, it dwindled in size; and the eyes were perforce drawn together toward the middle of the flattening face. The progressive specialization of the hind limb for support and propulsion led to a more upright posture, with correlated changes in the axial skeleton, gut, and reproductive organs. For Jones, most of the things that distinguish human beings from typical quadrupedal mammals were originally adaptations to living in trees, The arboreal theory was open to th obvious objection that most arborea mammals-opossums, tree shrews, paln civets, squirrels, and so on-lack the short face, close-set eyes, reduced olfactory apparatus, and large brains that arboreal life supposedly favored Jones tried to account for these counter examples. Accepting Matthew's thesis (8) that primitive mammals had beer arboreal creatures with opposable thumbs and first toes, Jones proposec that the absence of primate-like trait, in other arboreal lineages resulted fron a period of adaptation in each lineag to terrestrial locomotion. During thi period, the thumb and first toe becam4 reduced, the primitive reptilian flexibil ity of the forelimb was lost, and th4 primitive flat nails were replaced b! claws. These changes blocked the speSCIENCE, VOL. 184 cialization of the forelimbs for prehension. Accordingly, in nonprimate mammals that had reentered the trees, the primate evolutionary trends did not materialize. Stated thus baldly, Jones's thesis is obviously inconsistent. His treatment of the evolution of the brain, which he borrows from Smith, presupposes that primitive mammals were small-eyed terrestrial beasts that nosed their way through the world, guided by specialized olfactory and tactile receptors in the snout; but when the evolution of the limbs is in question, he assumes that arboreality is primitive and that early mammals were neither terrestrial nor typically quadrupedal. The late W. E. Le Gros Clark's reformulation of the arboreal theory, which more skilfully conceals this inconsistency, has been almost univer- sally accepted by other students of primate evolution. Much of Le Gros Clark's primatological work centered around the now-discredited (9) proposition that the tree shrews (Tupaiidae) are persistently primitive lemuroids that have somehow failed to develop the perfected adaptations to arboreal life seen in the other extant primates. Le Gros Clark believed that primitive Insectivora were tree-climbing beasts Downloaded from http://science.sciencemag.org/ on May 9, 2021 44 26 APRI 197 437 m+; g! Fig. 1. The Carolina gray squiirrel, Sciuru-is car-oliniensis, (A) hanging from wire grid, showing nonopposable first digits; (B) climbing thin sloping support; (C) descending uinderneath thin sloping SuIpport; (D) (squilrel shown by arrow) leaping across gap in the canopy, about 20 m above the gr-ound; (E) clinging to vertical cinder block wsall; and (F) foraging in terminal branches of a wiAllow oak (Qiuer-cls phellos), hanging bipedally. 26 APRIL 1974 437 The Comparative Evidence If progressive adaptation to living in trees transformed a treeshrew-like ancestor into a higher primate, then primate-like traits must be better adaptations to arboreal locomotion and foraging than are their antecedents. This expectation is not borne out by studies of arboreal nonprimates. The diurnal tree squirrels (Sciurinae) provide the most striking counterexample. The eyes of squirrels face laterally, the two visual fields having only about a 600 arc of overlap (11); the olfactory apparatus is not reduced by comparison with terrestrial rodents (12); all the digits (except the diminutive thumb) bear claws, which are sharper and more recurved than those of terrestrial sciurids (13); and the marginal digits of the hand and foot are not opposable or even very divergent (Fig. 1). Yet squirrels are highly successful axboreal mammals, and seem to have little difficulty in accomplishing the arboreal activities in which primates might be expected to excel. Despite their laterally directed eyes (and presumed lack of stereoscopy), squirrels of several genera may leap from 13 to 17 body lengths from tree to tree (Fig. 1 D) (14), which compares favorably with the 20 body lengths reported for the saltatory lemu438 roid Propithecus verrauxi (15). Although squirrel hands and feet are not adapted for grasping, squirrels easily walk atop or underneath narrow, sloping supports, and can forage for long periods in slender terminal branches hanging by their clawed hind feet (Fig. 1, A to C, F). Clearly, successful arboreal existence is possible withou-t primaelk adpttions. A partisan of Le Gros Clark's form of the arboreal theory might still postulate that tree squirrels are under selection pressure which favors their developing primate-like morphology, but have not undergone a long enough period of adaptation to arboreal life for them to have converged markedly with primates. Accepting this, we would still expect that arboreal squirrels would differ in primate-like ways from terrestrial sciurids, at least to a slight extent. We would have similar expectations about arboreal members of other nonprimate families. The facts do not bear out these expectations. Virtually the only features of the hands and feet which systematically distinguish arboreal from terrestrial squirrels are the longer fourth digits and generally larger carpal pads of the former; the arboreal genera show no tendency toward enlargement of the thumb, reduction of claws, or development of a wide or deep cleft,etween the first and second digits (16). Orbital convergence in all sciurids is slight, and is actually greater in the more terrestrial species (Fig. 2E), although the optic axes of ground squirrels' eyes are not more convergent than those of tree squirrels'. Since small mammals have relatively large eyes, orbital-margin convergence in most mammals varies inversely with size, other things being equal (3). For a given skull length, this convergence is somewhat greater in higher primates than in lemurs (17). When convergence is plotted against skull length for several families of arboreal mammals and the lemuriform and haplorhine regressions are traced on the plot (Fig. 2), it is evident that arboreality (or saltatory arboreal locomotion, in wholly arboreal taxa) does not correlate with proximity to the primate regressions. The slowmoving lorises have, for their size, more convergent orbits than the saltatory galagos (Fig. 2A). Among feloid carnivores (Fig. 2B), the terrestrial Felis bengalensis approaches the primate regressions most closely. Both arboreal and terrestrial procyonids (Fig. 2D) fit a regression parallel to those of the primates, from which the semiarboreal coatimundi is widely displaced away from the primate lines. Certain primate-like specializations of the visual pathways of the brain may perhaps represent adaptations to arboreal life per se. Diamond and his coworkers (18, 19) have found that the common tree shrew and the Carolina gray squirrel resemble Galago senegalensis in having little or no overlap between the projection from the retina to the occipital visual cortex (relayed via the lateral geniculate) and a significant visual projection to the temporal cortex from the superior colliculus (via the pulvinar). This is not the case in the cat, in which these areas overlap widely and the temporal cortex is given over to projections from the medial geniculate. Since arboreality is about the only thing that tree shrews, squirrels, and galagos have in common, the suggestion that this represents a specifically arboreal adaptation (18) may be correct. However, its adaptive significance is obscure. The expectation that "any mammalian line that relies heavily on visual cues" will develop a visual temporal lob- (19) is clearly unwarranted; cats rely heavily on visual cues, and in fact show several primate-like features of the visual system that are absent or unknown in squirrels and tree shrews-for example, parallel optic axes, substantial ipsilateral radiations of each optic nerve, and the presence of "binocular depth cells" in the striate cortex (20, 21). These features are all functionally related to stereoscopic depth perception. Since most of the projection from the retina to the lateral geniculate body seems to correspond to the binocular portion of the visual field (11, 22), the relative de-emphasis of the older tectopulvinar system in cats can even be described, from a different perspective (20), as a special similarity to higher primates. The comparative evidence, then, does not suyport the idea that the selection pressures of arboreal life favor the replacement of tree shrew-like morphology by primate-like morphology. In many respects, the first sort of morphology is actually of superior adaptive value. Clawed fingers and toes are superior adaptations for locomotion on nonhorizontal surfaces with large radii of curvature-including vertical walls (Fig. 1E) as well as tree trunks (23). Like marmosets (24), squirrels tend to avoid very thin branches in normal arboreal locomotion, but can walk on them easily enough, relying on the largely SCIENCE. VOL. 184 Downloaded from http://science.sciencemag.org/ on May 9, 2021 with clawed, nonprehensile hands and feet, small eyes and brains, and elaborate olfactory apparatus. The unspecialized, squirrel-like climbing habit of tree shrews (and ancestral primates) is invoked by Le Gros Clark to explain their incipiently primate-like morphology; tree shrews have a complete bony ring around the orbit, a relatively extensive visual cortex, a highly differentiated retina, some simplification of the olfactory apparatus, and a few minor grasping adaptations of the joints and muscles of the hind foot. More perfect arboreal adaptations, of the sort seen in lemurs, involve the replacement of sharp claws by flattened nails overlying enlarged friction pads, the divergence and enlargement of the first toe and thumrb to produce effective grasping organs, and the approximation of the two eyes toward the center of the face. This last change, in Le Gros Clark's view, had a positive selective advantage for acrobatic arboreal mammals; it produced a wide overlap of the two visual fields, allowing stereoscopic estimation of distance in jumping from branch to branch (l0). passive grip of the proximal volar pads when the support is horizontal and (unlike marmosets) gripping with opposed hands and opposed feet when the support is sloping (Fig. 1B). Primate-like approximation of the orbits increases visual field overlap, 'but decreases parallax, reducing the distance over which visual field disparities can provide distance cues. In a leaping arboreal animal, selection should act against the extreme orbital approximation seen in tarsiers and higher primates. This expectation is 'borne out by a comparison of lorises with galagos; the slowmoving Loris and Nycticebus have more convergent and closely approximated orbits than the saltatory galagos (25), whose wider interorbital space allows stereoscopic ranging over greater dis- Were Primitive Mammals Arboreal? Jones's version of the arboreal theory holds, not that the primate characteristics will be selected for in any arboreal mammal lineage, but that they all result from the primates' unique preservation of the grasping hands and mobile forelimbs supposedly found in the arboreal ancestors of the Mammalia. This conception of what early mammals were like can be traced to several sources. Huxley (26) and Dollo (27) proposed that the last common ancestor of the living marsupials had a grasping hind foot, but they thought this represented an arboreal specialization and that early mammals were terrestrial. Matthew (8), following Cope (28), reinterpreted this trait as a primitive retention, and suggested that Eocene and Paleocene placental mammals (and early ungulates in particular) also showed features indicating derivation from an arboreal ancestor. Most of the supposedly arboreal features identified or inferred for the ancestral mammals by Matthew and his inheritors (8, 29, 30) can be shown (17, 31 ) to be either chimerical or irrelevant to arboreality. Others represent specializations fixed at various points along the reptilian lineage leading to mam26 APRIL 1974 mals (such as the loss of all but two phalanges in the thumb and first toe, the "anomalous" arrangement of the thumb's extrinsic muscles, and the appearance of a tuber calcanei). Some are mere amphibian retentions (for example, persistence of the clavicle) that were lost in later mammalian lineages that developed cursorial specializations. Most of those who have believed that primitive mammals were lemur-like arboreal animals have also thought that terrestrial habits select for cursorial locomotion and thus for simplification and stabilization of the limlbs; that "the final stage of this process is exemplified in the horse" (7); and that primates could therefore not be descended from ancestors that had long been terrestrial. However, the fact that placental ancestors could not have been very much like horses does not imply that they were very much like lemurs. The same suite of primitive retentions seen in the primates is also seen in many terrestrial Insectivora. Most extant insectivores manifest no ungulate-like trends toward simplifying the limb skeleton-apart from a general but not universal tendency toward distal tibiofibular fusion, which can also occur in arboreal primates (Tarsius) and marsupials (Marmosa) (32). Cursorial specializations are adaptations for rapid visually directed pursuit of prey or rapid and prolonged flight from predators, and are best de- veloped in large mammals inhabiting open country. They would have had little or no selective advantage for the small, shrewlike mammals of the Mesozoic, and their absence does not imply arboreality. In support of Matthew's hypothesis, Lewis (33) points out that in reptiles the peroneal muscles arising from the fibula insert on the fifth metatarsal, but in mammals part of this musculature forms a peroneus longus muscle, whose tendon runs across the sole to insert on the first metatarsal. Lewis suggests that peroneus longus originally acted to adduct a divergent first toe in arboreal grasping. However, in extant mammals with rudimentary first toes, the peroneus longus typically persists, shifting its attachment one toe over to the base of the second metatarsal. This demonstrates that it has some important function unrelated to adduction of the first toe. An alternative explanation of its original adaptive value is that it acted to evert the foot against resistance. If the earliest mammals walked with their feet pointing somewhat sideways, as echidnas do (34), eversion would have added propulsive thrust at the end of the stance phase, and would have worked more efficiently if part of the everting musculature exerted its force through an attachment at the anterior (preaxial) edge of the foot. Intermediate stages in the shift of this attach439 Downloaded from http://science.sciencemag.org/ on May 9, 2021 tances. Evidently, the close-set eyes and grasping extremities typical of extant primates are adaptations to some activity other than simply running about in the trees; arboreal life per se cannot be expected to transform a primitive tree shrew-like primate into a lemur. Le Gros Clark's version of the arboreal theory is not adequate. Fig. 2. Five bivariate plots of species mean values of skull lengths (prosthion to inion, centimeters) and orbital convergence (dihedral angle between orbital and midsagittal planes, degrees): (A) lorisiform prosimians, (B) feloid carnivores, (C) didelphids (dashed line) and diprotodont marsupials, (D) procyonid carnivores and (E) sciurids. White symbols represent terrestrial animals (such as Moniodelphis) or slow-moving arboreal forms (such as Phalanger); stippled symbols represent semiarboreal animals (such as Didelphis); stars represent predominantly carnivorous animals (such as Monodelphis). In each plot, the diagonal lines represent the least-squares regression of convergence on skull length for Madagascar lemurs (upper line) and haplorhine primates (tarsiers and anthropoids: lower line). [Data from (17)] Fig. 3. Upper left (above) and lower right (below) molar teeth of (A) the Cretaceous opossum AIphadon wilsoni, (B) the mid-Paleocene plesiadapoid Palenochtha minor, (C) the Late Cretaceous ungulate Protungulatum donnae, and (D) the early rodent Paramys copei. In the 'I A B C D latter three, the stylar shelf (vertical arrows, above) and trigonid (horizontal arrows, below) are reduced by comparison with the more primitive condition seen in Aiphladon. is made for allometry, insectivorous diprotodonts also have more convergent orbits than other marsupials (see Fig. 2C). The Visual Predation Hypothesis If primate traits cannot be interpreted either as the products of a primitive arboreality retained only in primates, or as specializations necessarily selected for in any lineage of arboreal mammals, then neither form of the arboreal theory can explain why primates differ from squirrels or opossums, and an alternative set of explanations is 440 needed. One recently proposed alternative (3, 23) has been induced from a survey of the distribution of primatelike traits in other taxa. Grasping hind feet with a divergent first toe are characteristic of marsupials, chameleons, and certain arboreal mice and rats. Their adaptive significance varies. In at least some climbing mice, the grasping hallux is an adaptation to locomotion on the large siliceous stems of bamboos (39), on which claw grip is useless. In chameleons, grasping extremities represent a predatory adaptation, permitting prolonged and stealthy locomotion on slender terminal branches in pursuit of insects, which these specialized lizards stalk in the dense marginal undergrowth and lower canopy of tropical forests (40). The notion that ancestral marsupials had a grasping hallux remains generally accepted. In the smaller South American opossums like Marmosa robinsoni, this trait correlates with a chameleonlike way of life involving visually directed predation on insects "in the intricate interlacing of vine and branch that characterizes the second growth which abounds around the edges of clearings" (41). Insects, which these small didelphids require for adequate nutrition (42), are seized either in the hands or the mouth, bitten, and eaten held in one or both hands (41, 43, 44). The occasional use of the hands by didelphids in seizing prey becomes the most frequent pattern in small bushfrequenting Australian marsupials, including diprotodonts like Cercartetus as well as polyprotodonts like Antechintus (43, 45). Cercartettus and related small insect-eating diprotodonts like Burrainys differ from other arboreal marsupials and resemble primates in having muchreduced claws (46). When allowance SCIENCE, VOL. 184 Downloaded from http://science.sciencemag.org/ on May 9, 2021 ment across the sole would yield progressively more efficient eversion, whereas, if its original function had been to adduct the first toe, selectively advantageous intermediate stages would not be possible. In short, there is no reason to believe that the Triassic ancestors of the MZammalia had clawless, grasping extreniities, as Jones's version of the arboreal theory requires. The point may be settled by forthcoming studies of the virtually complete skeleton of the Triassic mammal Megazostrodon (35). There is in any event ample evidence to show that late cynodont reptiles and their mammalian descendants progressively developed a more elaborate olfactory apparatus than is found in other reptilian lineages (36), and that the earliest mammals had relatively small and degenerate eyes, in which the sauropsidan mechanisms of accommodation and nictitation had been lost (37). These facts suggest that the earliest mammals were shrewlike terrestrial creatures, guided largely by olfactory and tactile stimuli. This does not mean that early mammals were incapable of climbing branches that presented themselves as supports or obstacles; as Jenkins (38) points out, any small mammal needs this ability in a forest community. These comparisons suggest that the close-set eyes, grasping extremities, and reduced claws characteristic of most post-Paleocene primates may originally have been adaptations to a way of life like that of Cercartetus or Burramys, which forage for fruit and insects in the shrub layer of Australian forests and heaths. By this interpretation, visual convergence and correlated neurological specializations are predatory adaptations, comparable to the similar specializations seen in cats and owls, and allowing the predator in each case to gauge its victim's distance accurately without having to move its head. The grasping feet characteristic of primates allow insectivorous prosimians like the smaller cheirogaleines and lorisiforms to move cautiously up to insect prey and hold securely onto narrow supports when using both hands to catch the prey. Although claws are advantageous in most arboreal locomotor situations, they are actually a hindrance for a bush-dwelling animal that grasps slender twigs by opposition of preaxial and postaxial digits, and has little occasion to climb on larger supports (23). Olfactory regression has not been characteristic of most arboreal mammals. The slight simplification of the olfactory apparatus seen in strepsirhine prosimians, and the marked regression found in haplorhines (tarsiers and higher primates), are necessary results of the approximation of the medial walls of the two orbits; since the optic nerve leaves the base of the skull and the orbital openings lie in the dermal bones of the skull roof, the olfactory connections between braincase and snout must necessarily be constricted if the orbital cones draw closer together. This effect is evident in a comparison of small felids with canids: in the former, the interorbital space is generally narrower, and the olfactory bulbs are correspondingly smaller and have constricted con- similar changes are seen in the earliest rodents and ungulates (Fig. 3), Szalay (50, 51) has proposed that the differentiation of the Primates from the Insectivora involved an adaptive shift from an insectivorous diet to a predominantly herbivorous one. If true, this vitiates the visual-predation hypothesis. Szalay's thesis has recently been challenged by Simons (52), who suggests that, in at least four of the six families of early Tertiary mammals usually assigned to the order Primates, the earliest representatives have molars functionally similar to those of the carnivorous prosimian Tarsius. Although it has been said that the carnivorous diet of Tarsius could not be inferred from the morphology of its dentition (51), my colleague R. F. Kay has recently developed a multivariate bio- metric statistic which is over 90 percent accurate in "predicting" the dietary habits of the extant primates, includ- ing Tarsius. Despite the reduction of the stylar shelf in extant prosimians, at least some of them have recognizable dental adaptations for masticating prey; other shearing mechanisms have replaced the primitive shear of trigonid against paracrista and metacrista (53). The application of Kay's procedure to early primate dentitions will permit us to test certain aspects of the visualpredation theory. The plesiadapoids of the Paleocene (Plesiadapidae, Paromomyidae, Carpolestidae) are assigned by paleontologists to the order Primates, although they show none of the diagnostic primate traits listed by Mivart (5). WVhere known, plesiadapoid orbits are small Downloaded from http://science.sciencemag.org/ on May 9, 2021 nections with the olfactory fossa (47). In Tarsius, the close approximation of the huge eyeballs reduces the interorbital volume (filled, in typical mammals, by olfactory scrolls of the ethmoid) to a single plate of compact bone, the interorbital septum, over the top of which a few olfactory fibers arch to reach a much-reduced nasal fossa (3, 48, 49). Small ceboids and cercopithecoids resemble Tarsius in these respects. Other lineages of visually directed predators have achieved comparable degrees of visual field overlap without pronounced olfactory constriction; in marsupials (cover photograph), optic convergence is produced by the coexistence of a low frontal region with a broad and high zygomatic arch (3), while in lorises the eyeballs come together around and outside the olfactory connections, which reach the nasal fossa between the optic nerves (3, 49). The unique arrangement seen in the smaller extant haplorhine primates probably reflects derivation from a bigeyed Eocene prosimian like Pseudoloris (which appears to have had a Tarsiuslike interorbital septum); it does not represent perfected adaptation to arboreal life. Marsupial lineages which have evidently been arboreal since the Cretaceous have undergone no olfactory regression; arboieal life per se does not encourage loss of olfactory acuity. Most of the distinctive primate characteristics can thus be explained as convergences with chameleons and small bush-dwelling marsupials (in the hands and feet) or with cats (in the visual apparatus). This implies that the last common ancestor of the extant primates, like many extant prosimians (for example, Tarsius, Microcebus, Loris, Arctocebus, and the smaller galagines), subsisted to an important extent on insects and other prey, which were visually located and manually captured in the insect-rich canopy and undergrowth of tropical forests. 1I I \ TO KNGAROS \ I is I PETAURUS The Fossil Record Like any other evolutionary explanation, the visual-predation theory must be tested against the relevant paleontological data. Here it encounters difficulties. However we choose to define the order Primates, its early representatives differ from the earliest placentals in several features of the molar teeth, including reduction of the stylar shelf and associated cristae and decrease in the size and height of the trigonid. Since 26 APRIL 1974 ---- herbivores insectivores and mixed feeders - Fig. 4. (Above) Representatives of the plesiadapoid radiation (left to right: Plesiadapis tricuspidens, Carpodaptes aulacodon, Palaechthon alticuspis, Phenacolemur jepseni). (Below) Possibly comparable extant representatives of the phalangeroid marsupial radiation: phylogenetic relationships after Kirsch (58). The morphological shift at (i), which established the dental tr...