Prehensile tails capable of suspending the entire body weight of the animal evolved twice among extant primates, both times within the platyrrhine (New World monkey) infraorder (Napier,1976; Rosenberger,1983) (Fig. 1). The prehensile tail found among the members of the Atelinae subfamily of platyrrhines (including the modern forms Alouatta, Ateles, Brachyteles, and Lagothrix) (Fig. 2) represents one of these evolutionary events; the second occurred somewhere in the evolutionary history of the genus Cebus (subfamily Cebinae) (Napier,1976; Rosenberger,1983).1
Primate prehensile tails are distinct morphologically from nonprehensile tails in a number of ways. Caudal vertebrae are stronger and more rigid, and have more expanded muscle attachment sites, than those of nonprehensile tails (German,1982; Organ,2010; Organ and Lemelin, 2011). Prehensile tails have more extensive ventral (flexor) musculature with extrinsic tendons that cross fewer joint segments than those of nonprehensile tails (Lemelin,1995). Lateral tail musculature is also better developed in prehensile tails (Organ et al.,2009). These different morphologies are clearly adaptive to the different mechanical demands placed upon prehensile and nonprehensile tails, and can be attributed to different tail-use behaviors observed among taxa (Organ,2010). Atelines regularly use full or hindlimb-assisted tail suspension during bouts of locomotion and feeding (Jenkins et al.,1978; Bergeson, 1996; Turnquist et al.,1999; Schmitt et al.,2005). Cebus uses its tail primarily during feeding bouts with the tail wrapped around a substrate to form a tripod with its hindlimbs (Bergeson, 1996; Garber and Rehg,1999; Garber,2011).
In light of these behavioral differences, it is important to note the ways in which the ateline prehensile tail differs from that of Cebus. First, the ateline prehensile tail is relatively longer than the prehensile tail of Cebus (Rosenberger,1983; Organ,2007; Garber,2011), and accounts for a slightly higher percentage of total body mass in atelines (Grand,1977). Second, the dorsal (extensor) musculature of ateline prehensile tails is diminutive. In Cebus, the dorsal musculature is well developed, as it is in nonprehensile tails (Lemelin,1995). Third and perhaps most obvious is the difference in morphology of the skin of the distal tail. Atelines have a glabrous friction pad on the ventrodistal surface of the tail, marked by papillary ridges and flexure lines similar to palm of the hand and sole of the foot (Hill,1962) (Fig. 3). This friction pad is communally referred to as a “volar pad.” However, the term volar refers specifically to the skin of the palm of the hand and sole of the foot. Although the friction pad found in many New World monkey tails does converge morphologically with the palm of the hand and the sole of the foot, it is more accurate to call this structure a “friction pad,” which removes any potential confusion with the volar surfaces of the hand and foot.
The ateline friction pad is important for maintaining contact with the substrate while grasping (Biegert,1963; Niemitz,1990; Lemelin,1995; Organ,2007). The friction pad is hypothesized to be more sensitive than other hairless skin because its skin may contain a higher density of specialized mechanoreceptors [e.g., Meissner's corpuscles] that are sensitive to light touch and can assist in texture discrimination (Biegert,1961; Hill,1962). Meissner's corpuscles are found in the glabrous skin of primate hands, feet, and lips (Winkelmann, 1964; Hoffman et al.,2004).
The prehensile tail of Cebus is completely covered in hair and devoid of a friction pad. Cebus mirrors its nonprehensile-tailed sister taxon Saimiri in its external skin morphology. Because Meissner's corpuscles are not commonly associated with hairy skin, it is generally assumed, although not clearly documented, that Cebus does not possess these specialized touch receptors in its tail (e.g., Bergeson, 1996; Garber and Rehg,1999; Organ,2007,2010; Organ et al.,2009; Garber,2011). A large percentage of extant mammals with a prehensile tail lack a specialized friction pad and maintain a fully haired tail (e.g., Potos flavus—kinkajou) (Nowak,1991; Bergeson, 1996; Organ,2007). The presence of fur in the skin does not preclude the presence of other specialized mechanoreceptors (e.g., Merkel's cells) and nerve endings that could give the prehensile tail of Cebus—and many other prehensile-tailed mammals that lack a glabrous tail skin—sufficient tactile sensitivity without Meissner's corpuscles. The purpose of this study is to document the types of somatosensory receptors in the tail skin of ateline and cebine platyrrhines, to understand how the prehensile tail of atelines and Cebus function in similar ecological roles (Garber and Rehg,1999).
The somatosensory system is the sensory system of bodily perception (Smith, 2000). The types of stimulus or modalities associated with the somatosensory system include mechanoreception, proprioception, nociception, and thermoreception. Mechanoreception is discriminative touch such as pressure, stretch, tension, and vibration. Proprioception is the sense of static position and movement of muscles, tendons, and joints. Nociception senses pain while thermoreception senses temperature (Saladin,2011). Each of these modalities is associated with a specific type of stimulus energy, receptor class, and receptor cell(s).
The somatosensory system conveys two major classes of sensation: epicritic sensation and protopathic sensation. Epicritic sensations detect gentle contact of the skin and source localization, vibrations (amplitude and frequency), textures (spatial details that allow for two close points to be discriminated), and shapes. Protopathic senses detect pain, temperature, and some crude touch (Markl,1977; Kalman and Csillag,2005). Epicritic senses use mechanoreceptors to communicate discriminant touch and locomotor sensations to the brain (Kandel et al.,2000; Kalman and Csillag,2005). Epicritic mechanoreceptors are the peripheral endings of one or more afferent neurons, and are the mediators between an organism's environment and the somatosensory cortex of the brain. They are present in skin throughout the body and are sensitive to stimuli that stretch, compress, twist, or distort their cell membranes (Martini and Timmons,1997).
Merkel discs, Meissner's corpuscles, Pacinian corpuscles, and Ruffini corpuscles are the four epicritic mechanoreceptors of the skin that are pertinent to this study (Fig. 4). Merkel discs are flattened unencapsulated nerve endings adjacent to specialized tactile Merkel cells. Merkel discs are found in the epidermis of both hairy and glabrous skin and are sensitive to sharp points, edges, and curves, with remarkably high spatial resolution (less than 0.5 mm) (Kandel et al.,2000; Johnson,2001; Kalman and Csillag,2005). Meissner's corpuscles are tall, ovoid, encapsulated mechanoreceptors that consist of two or three afferent nerve endings meandering through a mass of Schwann cells (Paré et al.,2001; Saladin,2011). Meissner's corpuscles detect light touch, movement, and vibration, and they are localized to the dermal papillae of sensitive glabrous skin. They are insensitive to static skin deformation and are four times more sensitive to dynamic skin deformation than Merkel discs but with poorer spatial resolution (Kandel et al.,2000; Johnson,2001; Kalman and Csillag,2005). Ruffini corpuscles are flattened, elongated encapsulated mechanoreceptors containing a few afferent nerve endings and are located in the dermis and subcutaneous tissue of hairless and haired skin (Saladin,2011). Ruffini corpuscles are sensitive to tensile stresses, and thus detect heavy touch, pressure, and stretching of skin (Kandel et al.,2000; Johnson,2001; Kalman and Csillag,2005; Saladin,2011). Pacinian corpuscles are large (1–2 mm) mechanoreceptors consisting of a single afferent dendrite within an encapsulated lamellar mass of Schwann cells and (mostly) fibroblasts (Saladin,2011). Pacinian corpuscles are found in the hypodermis of hairless skin, and are sensitive to deep pressure, stretch and vibration (Saladin,2011). In fact, Pacinian corpuscles are thought to be the most sensitive of all cutaneous mechanoreceptors due to their ability to detect extremely low frequency vibrations of less than 50 Hz (Kandel et al.,2000; Johnson,2001; Kalman and Csillag,2005).
The four epicritic mechanoreceptors identified above can be further classified according to their responsiveness to sensory adaptation—when sensitivity to a stimulus is reduced after repeated presentation—as rapid adapting or slow adapting. Rapid adapting receptors, such as Meissner's corpuscles and Pacinian corpuscles, signal the onset and offset of skin indentations but do not detect continuous pressure and/or touch. Continuous pressure and/or touch is transmitted to the brain by slowly adapting receptors such as Merkel cells/discs and Ruffini corpuscles (Coleman et al.,2001; Johnson,2001). Each mechanoreceptor type serves a very distinct function, and no one type of receptor can be viewed as more important or superior to another type. Accordingly, the nature and extent of tail function and positional behavior may require only a subset of receptor types and not a full complement.
Empirical data are lacking in the literature with regard to the types of mechanoreceptors present in the skin of platyrrhine prehensile tails. Therefore, this study aims to fill this void by documenting the types of epicritic mechanoreceptors found in the ateline and cebine prehensile tails. The glabrous skin of the ateline ventrodistal prehensile tail is expected to contain all four epicritic mechanoreceptors, which would enable a high degree of tactile sensitivity in the region. The hairy distal tail of Cebus and Saimiri is expected only to contain Merkel cells/discs and Ruffini corpuscles.
MATERIALS AND METHODS
The tail skin of two ateline monkeys, Lagothrix (n = 18) and Ateles (n = 7), and two cebine monkeys, Cebus (n = 3) and Saimiri (n = 8), was examined histologically. The cebines differ from one another and the two atelines in their use of the tail in locomotor and manipulatory behaviors. Tail samples were obtained from both fresh frozen and alcohol preserved captive specimens. Subadult and adult individuals were used in this study. No animals were euthanized for the purpose of this study.
Sample Preparation and Imaging
Atelines have a mound-shaped formation of mesenchymal tissue (a pad) elevating thick glabrous skin, which possesses distinct dermal ridges. The friction pad is located at the distal end of the tail on the ventral glabrous surface (Fig. 3). In atelines, tissue was sampled from the proximal, middle, and distal segments of the friction pad. Cebines do not have glabrous skin on the ventral surface of their tails. To systematically sample and compare equivalent tail segments from primates with and without a friction pad, a friction pad length to tail length ratio was derived. On average, the friction pad represents approximately one-third of total tail length and is located distally on the tail. For the cebines that do not have a friction pad, total tail length was measured and then divided by three to derive a ratio similar to the above. The distal third of the tail was then further subdivided into proximal, middle, and distal segments, and hairy skin samples were obtained from each segment.
Once tail samples were harvested, the samples were soaked in a 10%-buffered formalin solution for 1 week. Tissue samples were then processed using a graded alcohol and xylene dehydration protocol following Smith et al. (2007). Following dehydration, tissues were infiltrated and embedded in paraffin. Fifty transverse 8-μm serial sections were cut for the proximal, middle, and distal segments of each individual. Forty serial sections for each segment were mounted and stained with hematoxylin and eosin. The remaining 10 serial sections were stained using Masson's trichrome (American MasterTech: Protocol #KTMTR). Masson's trichrome is a three-color staining protocol that allows an observer to better distinguish cells from surrounding connective tissues. Trichrome stained sections were used to evaluate and confirm results obtained from hematoxylin and eosin stained sections.
Each series of sections was examined using a Zeiss Axio Lab light microscope mounted with a ProGres digital camera. Each serial section was examined for the presence or absence of each of the four epicritic mechanoreceptors described above: Merkel discs, Meissner's corpuscles, Ruffini corpuscles, and Pacinian corpuscles. Mechanorecepting free nerve endings were present in all tissues sampled.
Meissner's corpuscles were the most superficial encapsulated mechanoreceptors. Meissner's corpuscles were located in the apices of dermal papillae and were observed only in Ateles and Lagothrix (Figs. 5 and 6). The capsule of the Meissner's corpuscle only covers the most distal portion of the receptor (Fig. 6). The inner core of the Meissner's corpuscles contained horizontally lamellated modified Schwann cells. It was not possible to count axons associated with each Meissner's corpuscle.
Pacinian corpuscles are large encapsulated receptors (Fig. 7). Pacinian corpuscles were observed deep to the subcutaneous connective tissue of the glabrous tail skin of atelines. These mechanoreceptors were easily seen at very low magnification (e.g., 4×). The Pacinian corpuscles of atelines were characterized by an outer capsule of 20–35 layers of perineural cells, a capsular space, and an inner core of lamellated modified Schwann cells surrounding an axon terminal (Fig. 7). Although Pacinian corpuscle density was not calculated, over 50% of total histological sections for Ateles had Pacinian corpuscles. The length range of 10 Pacinian corpuscles sampled in this study was 0.9–2.2 mm, thus potentially visible with the naked eye.
Ruffini endings were found in the dermis of Saimiri and Cebus as well as the glabrous skin of the atelines. Over 150 serial sections were sampled for each individual. Within these sections, atelines were documented to have one to two Ruffini corpuscles compared to 20–30 in Saimiri and 5–10 in Cebus. Ruffini endings lie deeper than the observed Meissner's corpuscles but are more superficial than Pacinian corpuscles. Ruffini corpuscles are encapsulated receptors, and the capsule space is fluid filled. A single, larger, myelinated axon enters the core of the receptor, where it gives rise to numerous small branches (Fig. 8).
Merkel cells and their associated discs were the most difficult to individually identify. They are located in the dermis. In atelines, these cells were located at the base of papillary ridges. They are best observed after immunohistochemical staining. Merkel cells are clustered in groups of five to eight. Merkel cells interdigitate with keratinocytes in the surrounding basal epidermis. The disc is formed by unmyelinated axons (Fig. 9). Merkel cells were positively identified in the glabrous skin of the atelines and hairy skin of Cebus but were not documented in the hairy skin of Saimiri.
The prehensile tail of atelines and Cebus evolved independently (Napier,1976; Rosenberger,1983). Even though the musculoskeletal structure of the tail is similar in many respects among atelines and Cebus (Lemelin,1995; Organ,2007,2010; Organ et al.,2009; Organ and Lemelin, 2011), the morphology of the tail integument is different. Ateline monkeys possess a friction pad on the ventrodistal surface of the tail, while the ventrodistal skin of the tail in Cebus is completely haired. The ateline friction pad is thought to serve two functions. First, the dermal ridges of the glabrous skin increase friction between the tail and the substrate for tail suspensory behaviors (Niemitz,1990). Second, the glabrous skin of the ateline tail contains Meissner's corpuscles that are sensitive to light touch. Meissner's corpuscles are the mechanoreceptors responsible for relaying light touch sensory information from the tail to the somatosensory cortex of the brain (Biegert,1961,1963). The fully haired prehensile tail of Cebus, however, has not been studied extensively and/or histologically. It was assumed that tactile sensitivity of the tail in Cebus is dramatically lower than in atelines because Meissner's corpuscles are not commonly associated with hairy skin (Bergeson, 1996; Garber and Rehg,1999; Organ,2007; Garber,2011).
The purpose of this study was to document the types of epicritic mechanoreceptors of the ateline friction pad and hairy skin of the cebine tail, focusing specifically on four mechanoreceptor types: Meissner's corpuscles, Pacinian corpuscles, Ruffini corpuscles, and Merkel discs. Meissner's corpuscles and Pacinian corpuscles are classified as rapid adapting receptors, in that they signal onset and offset of skin indentations (touch) but do not relay sensations of continuous pressure/touch. Rapid adapting receptors are typically associated with glabrous skin (Saladin,2011). Slow adapting receptors, such as Ruffini corpuscles and Merkel discs, are important for relaying continuous pressure/touch sensations, and are found in both hairless and hairy skin (Saladin,2011). Ateline friction pads were hypothesized to contain all four epicritic mechanoreceptors providing this skin with a high degree of tactile sensitivity through slow and rapid adapting receptors. Conversely, the tail skin of Cebus (being completely haired) was expected to contain only the slow adapting receptors. Results of this study confirm the presence of all four epicritic mechanoreceptors in the glabrous friction pad of the ateline prehensile tail. These data represent the first direct published evidence of the presence of mechanoreceptors in the ateline friction pad (but see Biegert,1961,1963; Hill,1962, for indirect reference), and demonstrate the ateline friction pad is capable of sensing light and heavy touch through both rapid and slow adapting mechanoreceptors. This study also documents the presence of Ruffini corpuscles in the ventrodistal tail skin of Cebus. Ruffini corpuscles are slow adapting receptors sensitive to heavy touch, pressure, and stretching of the skin.
All of these data together suggest that atelines have a higher degree of tactile sensation in their ventral tails than Cebus. This is not a surprising result, as atelines are known to have relatively expanded caudal sensorimotor cortices of the brain compared to Cebus (Falk,1980). Moreover, atelines and Cebus use their tails in different ways. Cebus often uses its tail as an anchor along with bipedal crouching postures during feeding bouts (Bergeson, 1996; Garber and Rehg,1999). Atelines also suspend from their tails during feeding but are often observed using their tails during locomotor activities as well (Jenkins et al.,1978; Bergeson, 1996; Turnquist et al.,1999; Schmitt et al.,2005). However, the ateline prehensile tail is clearly capable of much finer resolution in touch sensitivity than the Cebus prehensile tail. This begs the question of the importance of rapid- and slow-adapting mechanoreceptors during postural and locomotor activities. Because only slow-adapting mechanoreceptors are found in the hairy skin of the Cebus prehensile tail (and also in nonprehensile-tailed Saimiri)—while both slow- and rapid-adapting mechanoreceptors are found in the ateline friction pad—we predict that slow-adapting mechanoreceptors are important for postural behaviors but not for locomotor behaviors. Rapid-adapting mechanoreceptors are found in the ateline friction pad because their tails are also used during locomotion.
Given the subtle differences in anatomy among the prehensile tails of atelines and Cebus (i.e., musculoskeletal and integumentary differences), there appears to be very strong support for parallel evolution of prehensile tails among platyrrhines (e.g., Napier,1976; Rosenberger,1983; Rosenberger and Strier,1989; Lemelin,1995; Organ,2007,2010; Organ and Lemelin, 2011). Less certain are the underlying pressures that would have selected for prehensile tail evolution in these two separate subfamilies. Meldrum (1998) suggests that the prehensile tails of platyrrhines evolved because of their important functions during hindlimb suspensory behaviors. In particular, he argues that tail twining and tail bracing behaviors in conjunction with hindlimb suspension (behaviors observed in some nonprehensile-tailed platyrrhine primates like Pithecia and Chiropotes) would have created strong selective pressures for stronger and more robust tails with prehensile capabilities, “especially (among) ancestral atelines and Cebus” (Meldrum,1998:154). Garber (2011), however, suggests that because of drastic differences in Cebus and ateline behavior and ecology, prehensile tails likely evolved in these lineages because of alternative selective pressures. The data presented here demonstrate differences in total mechanoreceptivity among the tails of atelines and cebines regardless of prehensile ability. Therefore, while slow-adapting touch sensitivity is likely important during postural behaviors, the rapid-adapting sensitivity of the ateline tail may have specialized in response to different selective pressures related to fine touch—that is, use of the tail during locomotor as well as postural behaviors.
This study did not quantify mechanoreceptor densities, which is important for understanding variation in mechanoreceptivity across taxa. Nevertheless, the abundance of Ruffini corpuscles in each histologic section appears to vary across taxa, where the atelines have fewest, Saimiri has least, and Cebus has an intermediate number. This qualitative assessment will allow us to make explicit predictions about mechanoreceptor densities in correlation to tail-use behavior for future study.
The authors thank A. Rosenberger for inviting them to contribute to this special issue on New World monkeys. The authors also thank G. Smith, T. Smith, and an anonymous reviewer for comments that greatly improved the quality of this work. For access to platyrrhine skin samples and/or cadavers, the authors are grateful to C. Ross, M. Teaford, K. Bishop, D. Wescott, R. Burns, and the staff of the Louisville Zoo. For references, the authors thank K. MacKinnon and S. Segal. Various aspects of this research were formed from discussions with N. Dominy and C. Lawson. The authors extend special thanks to C. Sethi for her assistance with dissection and sampling and S. Paesani and L. Jutras for their assistance with histological processing. For A.H. Organ.
Platyrrhine phylogeny at the family level is questioned, as differences result from using either morphologic or molecular data sets to reconstruct relationships (e.g., Ford, 1986; Kay, 1990; Rosenberger, 1992; Schneider et al., 1993; Steiper and Ruvolo, 2003; Schrago, 2007). There is, however, partial agreement in identifying five distinct subfamilies (disregarding family ranks) (see Schneider and Rosenberger, 1996). In the present study, a classification scheme based on subfamilies is used.