The Role of Piloerection in Primate Thermoregulation

In primates, thermoregulation affects behavior and activity levels [Hill et al., 2004; Jablonski et al., 2009], and the requirements of temperature regulation determine dietary caloric requirements and, thus, niche and resource utilization. Primates, like other mammals, regulate body temperature through changes in physiology, morphology, and behavior. Large mammals have smaller surface areas relative to their mass, greater energy reserves as fat, and can have longer coats of hair or fur for insulation. Smaller mammals have larger surface areas relative to their mass, can accommodate only minor energy reserves as stored fat, and cannot grow long coats for insulation. In order to maintain thermal homeostasis, small mammals maintain high rates of activity and food intake; when these conditions cannot be met, they must compensate by hibernation and torpor. Diurnal animals are subject to highly variable temperature regimes as they move in and out of shadows and as the sun traverses the sky. Crepuscular animals experience the highest rates of diurnal thermal variability because the coldest time of the day usually is around dawn and the hottest is in the late afternoon. Nocturnal animals experience a more constant temperature regime. Cathemerality, as an activity pattern, enables the most favorable thermal conditions to be utilized regardless of the solar time or season [Tattersall, 1987; Curtis, 2006].

Animals living in forests, and particularly in low latitude, everwet tropical forests, live under almost constant temperatures [Janzen, 1967]. The climates of the late Cretaceous and early Paleocene, under which many lineages of modern mammals differentiated, were warmer, humid, and less seasonal than those of later epochs [Steuber et al., 2005]. Equable and taxonomically diverse everwet tropical forests of modern aspect existed from equatorial through significantly higher latitudes from the early Paleocene through the early Eocene [Johnson and Ellis, 2002; Archibald et al., 2010], and it is under these conditions that the earliest primates evolved.

Mammals have evolved a wide variety of adaptations to different regimes of environmental temperature, and the nature of these has depended partly on the degree and temporal scale of temperature variation. For small or medium-sized mammals, adaptations to marked diurnal and seasonal temperature fluctuations have involved the evolution of compensatory changes in modes and types of fat storage and pilary insulation, and changes in metabolic rate, and/or activity pattern. Daily or prolonged torpor and hibernation have evolved in many lineages as means of controlling energy expenditure in the face of seasonal changes in climate and food availability. Specific behaviors such as social huddling, nest building, and shelter seeking also appear to have evolved partly or mostly for thermoregulatory purposes.

For most mammals, piloerection is a common method of modifying heat exchange. It is caused by the contraction of the musculi arrectores pilorum or MAP, which are bodies of smooth muscle that extend from the fibrils of the dermis into connective tissue investment of the hair follicle. Innervation to the MAP is supplied by the sympathetic nervous system, and contraction of the muscles is involuntary. Contraction of the MAP elevates the hair and increases the loft of the coat, thereby enhancing its insulative value. In many species, including familiar ones such as dogs, cats, and chimpanzees, raising the hair serves a secondary function of social warning of impending aggression. In mammals such as horses that are well-endowed with MAP, contraction of the muscles raises the effective hair coat depth by 16-32% or by a depth of 0.4-1.4 cm in newborn foals [Cymbaluk, 1994]. The insulating properties of the coat are not determined as much by the intrinsic thermal capacity of the hair itself, but by the layers of motionless air that can be trapped in successive laminations around the body to provide ideal outer insulation [Herrington, 1951]. Piloerection compounds the insulating properties of the hair coat by changing its thickness. It minimizes heat loss primarily through radiation and secondarily through convection, and makes it possible for animals to compensate for moderate cooling of the environment without increasing their metabolism. The ability of piloerection to buffer the need for chemical thermoregulation varies according to body size because heat loss and environmental heat gain are proportional to the square of the surface dimensions of the animal under consideration [Herrington, 1951]. Metabolic heat production scales at a similar rate to the volume at the cubed power of the surface dimensions, while surface heat exchange scales according to the Stefan-Boltzmann law, and is proportional to the fourth power of its absolute temperature [Bligh and Johnson, 1973]. Because of the ‘surface law', piloerection is especially effective in large mammals (>5 kg), but is less so in small mammals because of their higher surface-to-volume ratio and restriction of the length of the coat covering possible.

The MAP are present and piloerection is possible in most mammals, and these states represent the primitive conditions for Mammalia, based on their presence in reptiles, birds, and in most members of most mammalian orders (table 1; online suppl. table 1; see www.karger.com/doi/10.1159/000355007 for all online suppl. material). In reptiles and birds, homologs of the MAP cause erection of scales and feathers. The absence of MAP in mammals is a derived condition that is most often associated with aquatic, fossorial, or aerial lifestyles. The MAP were lost independently in hairless aquatic mammals (Cetacea), and in three lineages of fur-bearing aquatic species where a protective boundary layer of air cannot be maintained next to the body. These lineages are the eared seals, Otariidae, and otters, Lutrinae, within Carnivora, and the beavers, Castoridae, within Rodentia. The MAP were also lost independently in two lineages of burrowing mammals, namely naked mole rats (Batherygidae) within Rodentia and shrews (Soricidae) within Insectivora (Eulipotyphla). The MAP were also mostly or completely lost in the aerial mammals belonging to Chiroptera and Dermoptera. Very strong MAP confer enhanced pilary insulation in most large terrestrial mammals, but robust MAP also serve as shock absorbers to prevent abrasion in some groups (see online suppl. table 1).

The presence of MAP and piloerection is the ancestral condition for Euarchontaglires [comprising Glires (Rodentia and Lagomorpha) and Euarchonta (Dermoptera, Scandentia, and Primates)]. The ancestral condition for Euarchonta is more difficult to interpret because of the divergent character states present in the crown taxa. In the Tupaiidae, MAP are variably present and weak according to one group of researchers [Montagna et al., 1962], but are absent entirely according to another authority [Sokolov, 1983]. In Dermoptera, they are absent [Sokolov, 1983], but - as noted above - this is almost certainly a derived condition associated with an aerial habitus, as it is in Chiroptera. Among Primates, weak MAP are present on the body and stronger MAP on the tail in most groups except for tarsiers and some lorises, where they are entirely lacking. The presence of weak or vestigial MAP on the body and more prominent and functional MAP on the tail is interpreted as the primitive condition for primates. Therefore, based on the distribution of character states among living Euarchonta, the presence of weak MAP on the body was probably the ancestral state for the group.

Thermoregulation in primates has not been studied extensively because of the difficulty of quantifying parameters of the thermal environment in which primates live [Hill et al., 2004] and because field workers have focused primarily on documenting details of locomotor and social behavior, conservation, and ecology. Our interest in primate thermoregulation arose because of our curiosity about the roles played by the integument and behavior in thermoregulation in ring-tailed lemurs, Lemur catta. Our preliminary study revealed an unexpected connection between sunning behaviors in L. catta and anatomical features of the species' skin and hair [Jablonski et al., 2009]. This study represents an extension of that work and is focused specifically on the role played by the MAP in primate thermoregulation.

Our studies of the skin of L. catta [Jablonski et al., 2009] revealed that the torso of the animal is covered with fine, mostly white hair on the venter and coarser, gray and grayish-brown agouti hair on the dorsum. The hairs on the torso are grouped in compound follicles consisting of 6-24 follicles, as noted by Montagna and Yun [1962c] (fig. 1a, b), which exit through a single opening on the surface. The MAP are present on the torso (contra Montagna), but they are sparse and very small relative to the size of the compound follicles with which they are associated. By contrast, the compound follicles of the tail are associated with larger and more abundant MAP (fig. 1c, d). The skin of the ventral torso is black, owing to the high density of melanocytes in the stratum basale of the epidermis and the numerous eumelanin-containing corneocytes in the stratum corneum (fig. 2a). The skin of the dorsal torso is considerably less pigmented (fig. 2b). This is the obverse of the solar exposure pattern.

The sparse distribution and small size of the MAP on the torso of L. catta, especially relative to the large size of the compound hair follicles, suggested that the MAP are probably non-functional. Therefore, they would have a minimal effect on the thermal characteristics of the hair coat. We inferred, further, that the sunning behavior characteristic of the species was related to the functional absence of MAP and the importance of basking in the sun combined with passive stretching of the skin of the ventral torso in order to maximize warming and reduce the need for chemical thermoregulation on cool mornings [Jablonski et al., 2009]. This finding led us to explore the available information on the presence and nature of the MAP in primates and other mammals, and to speculate on the relationships between the MAP, thermoregulation through metabolism, and behavior in primates.

We surveyed the literature on the primate integument in order to gather information on the status of the MAP, and relied primarily on the descriptive histological studies of William Montagna and colleagues. Our survey of the status of the MAP in primates is presented in table 2.

The distribution of the MAP in primates and their close relatives suggests that the probable ancestral condition in early Primates was the same as in basal Euarchonta, i.e. MAP weak or vestigial on the body and present on the tail. Complete loss of MAP occurred independently in tarsiers and some Lorisidae, and is a derived state. In anthropoids, the MAP are present on all of the non-glabrous skin. In many species of larger platyrrhines and catarrhines, they are described as large and ‘strong', especially on the dorsal surface of the trunk. They are generally less prominent in small-bodied callithrichids, but are well developed on parts of the head and distal tail. In figure 2, the distribution of MAP character states in Primates and closely related outgroup taxa was mapped onto a recent and highly supported molecular phylogeny of Primates [Perelman et al., 2011]. Examination of the distribution of MAP character states in figure 3 indicates that most primates retain the probable primitive euarchontan condition of MAP present on body and tail (character states 1, 0), but the size and presumed function of the MAP on the body vary considerably. Dermopterans, tarsiers and some lorisids independently lost MAP entirely (character states 0, 0 in fig. 3), probably for different reasons. Lemurs lost most or all MAP on the body, but retained MAP on the tail (character states 0, 1 in fig. 3).

Primates and many other mammals underwent dramatic evolution and widespread dispersal during the Paleocene-Eocene thermal maximum (PETM) [Lucas, 1998; Storey et al., 2007; Beard, 2008; O'Leary et al., 2013]. Early primates have been found at a wide range of latitudes including far northern sites in Great Britain and Germany, to northern and central China and the northern Great Plains; they were also widespread at equatorial latitudes in Africa [Covert, 2002; Ni et al., 2007]. At the PETM, equable conditions obtained over a wide range of latitudes, and primates lived in hot to warm, humid tropical and paratropical forest biomes. In such environments, animals were probably subjected to very limited ranges of temperature variation (both diurnal and seasonal), and the selection for an increased basal metabolic rate was probably weak [Lovegrove, 2012]. Under these conditions, enhanced pilary insulation for close regulation of the air temperature in the boundary layer close to the skin was probably less important. Mutations leading to the complete or near loss of the MAP probably occurred in stem euarchontans in the late Cretaceous or early Paleocene when temperatures were considerably warmer and daily temperature fluctuations smaller than they are today, even in the tropics. Loss of most MAP may have been under positive selection in order to reduce heat stress, or may have occurred as the result of genetic drift, or in the absence of stabilizing selection. Retention of MAP on the tail in some strepsirhines may have been associated with the importance of the tail in mid-leap maneuvers in some taxa. In galagos and tarsiers, the tail-flick and mid-air rotation of the torso allow the animal to change direction in mid-leap [Crompton and Sellers, 2007; Crompton et al., 2010]. Contraction of the MAP and erection of the hairs on the dorsum of the tail and tail tuft appears to increase air resistance and facilitate braking or a change of direction, working like a sea parachute anchor. Piloerection may also help to resist damage due to abrasion when the tail is being moved during locomotion.

For early primates, the probable absence of functioning MAP except on the tail meant that the roles played by piloerection and pilary insulation in thermoregulation were minimal and that alterations in metabolism and behavior were used instead to maintain thermal homeostasis. Small mammals realize less benefit from piloerection than large mammals, but the absence of any potential for automatic adjustments to the loft of the hair coat in relation to temperature would have meant the evolution of compensatory modes of thermoregulation, including changes in activity patterns, modifications of metabolism, and development of postural and social behaviors favoring conservation of body heat. Many small primates consume diets of high-quality, calorie-dense foods such as insects and gums to maximize metabolic output. In seasonal and energy-limited environments such as Madagascar, lemurs evolved a wide variety of energy-conserving adaptations. Cathemerality buffers animals against the thermal extremes of hot days and cold nights [Tattersall, 1987; Curtis, 2006], and has been suggested as the most likely pattern of activity for the ancestral primate based on short-wave opsin gene sequences [Tan et al., 2005]. For many ancestral primates living in paratropical forests, such as adapiforms living in northern Europe in the early Eocene, cathemerality may have been obligatory due to the vastly varying day lengths at their latitude and the very short period of true darkness (of about 3 h) at midsummer.

Today the most extreme energy-serving adaptations are found in cheirogaleids, and probably not coincidently, in species of the smallest primate, the mouse lemur (Microcebus), where daily torpor is combined with hibernation in the cold and dry season [Schmid and Kappeler, 2005]. In strepsirhines and tarsiers, the limited ability of pilary insulation to buffer animals against environmental temperature fluctuations has led to the evolution of great diversity in activity patterns, metabolic rates, and abilities to enter and maintain torpor. The absence of MAP also helps to account for why many strepsirhines assume postures that conduce to either heating or cooling, depending on environmental temperatures [Muller, 1985; Morland, 1993; Donati and Borgognini-Tarli, 2006; Snodgrass et al., 2007].

In this context, it is instructive to compare the thermoregulatory behaviors of meerkats (Suricata suricatta), of comparable body size (mean weight 800 g [Habicher, 2009]) to many strepsirhines and possessing whole-body MAP. Meerkats live in areas of southern Africa with high diurnal temperature ranges and moderate to high annual temperature ranges. They engage in a range of thermoregulatory behaviors similar to those of L. catta, including sunning and huddling (fig. 4). When cold, meerkats thermoregulate by involuntary piloerection and shivering, and by sunning and huddling; when hot, they pant, lie down in contact with the substrate, and avoid the sun by seeking shade or staying in their burrows [Habicher, 2009]. Meerkats have a 3°C thermal neutral zone but can maintain homeothermy over a much wider range through metabolic and physiological-behavioral adaptations [Habicher, 2009]. When ambient temperatures are below the lower limit of their thermoneutral temperature of approximately 30°C, their body hair is erect regardless of the strength of solar radiation. The animals sunbathe whenever possible. When the temperature is low, sunbathing ceases before the temperature at which piloerection ceases [Habicher, 2009]. The only cold adaptation that enables meerkats to forage on cool to cold and sunless days is piloerection. Piloerection is utilized for up to 86% of the time the animals spend outside [Habicher, 2009]. From this evidence, we infer that the absence of piloerection in small primates appears to have exacted a cost in terms of increased need for chemical thermoregulation and shortened length of the thermal activity window as compared to meerkats and other mammals of comparable size that can piloerect.

Judging from the presence of whole-body MAP in larger Ceboidea, and in Cercopithecoidea, and Hominoidea, MAP were probably reacquired in the ancestor of anthropoids by the later part of the middle Eocene, when the first generally accepted catarrhines are distinguished in the fossil record [Jaeger et al., 2010]. The pattern of MAP over the surface of the body in many Callithrichidae may be instructive in understanding how MAP were reacquired in anthropoids. In Saguinus fuscicollis, MAP are present and better developed on the cheek, scalp, back, and tail; in the tail, the MAP are larger on the distal than on the proximal portion [Perkins, 1966]. In Saguinus oedipus, MAP are present on all body regions except the facial disk (eyebrow, eyelids, lips, and perialar area), external ears, anogenital skin, and suprapubic region; they are largest on the tail [Perkins, 1969a]. Reacquisition of MAP may initially have involved selection for the spread of the epidermal follicle stem cell precursors of the MAP over the dorsal surface of the body via cell migration from the neural tube. It is interesting to speculate as to whether the large-bodied subfossil lemurs also reacquired MAP on the torso as part of an adaptation to increased seasonal temperature fluctuations in the Pleistocene. The thermoregulatory stress on smaller-bodied lemurs at that time must have been severe.

Early primates probably had weak or vestigial MAP on the body, but possessed MAP on the tail. This condition persisted in all lemurs and some lorises, and was probably present in the basal haplorhine. The complete loss of MAP in tarsiers is a derived condition. The well-developed MAP of anthropoids is a shared-derived condition probably related primarily to increased body size.

The lack of functional MAP on the body in basal primates probably did not adversely affect the animals living under continuously warm and equable conditions during the PETM, but after that time would have exacted a considerable metabolic cost for maintaining body temperature and/or in restricting the length of time the animals could be actively foraging. The lack of MAP would not be so important for nocturnal or cathemeral animals or for those restricted to tropical or paratropical forests with relatively narrow temperature fluctuations. After the PETM, it is possible that there was strong selection for high-quality diets, with social, physiological, and morphological adaptations to regimes of more extreme daily and seasonal temperature fluctuations. Despite these changes, primates became extinct in non-tropical latitudes by the late Eocene and earliest Oligocene, and it is inviting to speculate that the effective absence of MAP and the ability to piloerect contributed to their demise. In this connection, it is relevant that all primates, even strepsirhines and tarsiers, are more highly encephalized than other mammals of comparable body mass, and so bear the metabolic cost and thermal sensitivity of a relatively larger brain. The evolution of functional, whole-body MAP can thus be seen as a necessary precursor to the development of large-bodied highly encephalized, diurnal primates.

Our observations on the MAP of L. catta were made on skin samples obtained from 2 cadavers, which came from the Duke Lemur Center (DLC) and from the Lemur Conservation Foundation Myakka City Lemur Reserve (LCF), respectively. We thank these institutions for making these samples available to us. Histological studies of horse and lemur skin were conducted in the facilities of the Department of Anthropology at the Pennsylvania State University at University Park (PSUUP) for the DLC specimens, and in the histopathology laboratory of the Animal Diagnostic Laboratory in the Department of Veterinary and Biomedical Sciences of the PSUUP, for both the DLC and LCF specimens. For generous assistance in conducting tissue processing, slide preparation, and photography of slides we thank Dr. Kazuhito Kawasaki (Department of Anthropology, PSUUP), and Dr. Arthur Hattel and Ms. Roberta Horner (Animal Diagnostic Laboratory, PSUUP). We are grateful to Associate Editor Sarah Elton and to 3 anonymous reviewers for their constructive and insightful comments on the original version of this paper; their suggestions for improvement strengthened the paper.