Primates, as a primarily arboreal radiation, tend to exhibit a remarkable diversity of locomotor modes and postures, which are fundamental for utilizing the 3-dimensional and complex environment of forest canopies (Prost, 1965; Grand, 1984; Cant, 1992; Bergeson, 1996). Although postural data would provide information about the adaptive significance and evolution of the biological roles of different morphological complexes, postural modes in relation to substrate use have been often neglected in many ecological studies (Garber and Pruetz, 1995; Hunt et al., 1996; Youlatos, 2004).

Locomotor modes and postures constitute major adaptations to exploit econiches and have been strongly diversified in most major primate radiations (Youlatos, 2004; Garber, 2010). Thorough examination of factors at multiple levels, from ecomorphological traits to habitat structure, should help elucidate the ecological pressures that have shaped specific adaptive radiations and, consequently, the evolution of locomotor and postural diversity. In effect, postures, defined as the dynamic maintenance of body stability, are essential in foraging, digestion, social interactions, energy conservation, thermoregulation, and resting/sleeping (Hunt et al., 1996). The latter constitutes the longest and most continuous behavioral phase throughout the 24 h cycle of any primate (Anderson, 1984, 1998, 2000; Fruth and McGrew, 1998; Matsuda et al., 2009; Prasetyo et al., 2009; Wada and Tokida, 1981). Sleeping postures may represent important adaptations for homeostasis and survival and ultimately contribute to fitness (Anderson, 1998, 2000).

Given the lack of data on sleeping postural behavior, it remains largely unexplored which postures primates adopt, which substrates they use, and where they position themselves in tree crowns during sleep. These variables may have important evolutionary implications for survival since selection of stable substrates promotes overall body stability and energy conservation (Rose, 1974), selection of covered areas in the canopy protects from potential predators and extreme weather conditions (Anderson, 1998; Wahungo, 2001), and certain postures contribute to thermoregulation (Anderson, 2000) or digestive processes (Urbani and Bosque, 2007). To test these links, interspecific comparisons are very useful for investigating relevant ecological pressures, associated evolutionary patterns, and taxon-specific constraints and strategies. More particularly, comparisons between closely related taxa, like congeners under different ecological conditions, can be very instructive. In this context, a good example would be howler monkeys (Alouatta), a widely distributed and behaviorally flexible neotropical primate genus (Eisenberg et al., 1972; Kowalewski and Zunino, 2004; Kowalewski, 2007; Di Fiore et al., 2011; Cortés-Ortiz et al., 2015; Kowalewski et al., 2015).

Thus, the present study aims to examine the use of postures, substrate types, sizes and inclinations, and tree crown parts during sleep in mantled howlers (A. palliata), Guianan red howlers (A. macconnelli), and black-and-gold howlers (A. caraya). Alouatta palliata is found in the northern distributional limits of the genus (Central and north-eastern South America), are sexually dimorphic (males 4.5–9.8 kg, females 3.1–7.6 kg), exploit dry to humid forests, are mainly folivorous, and exhibit high levels of intra-sexual tolerance with large groups containing as many as 4–12 adult males (Garber et al., 1999; Di Fiore et al., 2011). Alouatta macconnelli is widespread along the northern parts of South America, display sexual dimorphism (males 7.2–8.0 kg, females 5 kg), exploit mainly humid primary and secondary rain forests, are folivorous–frugivorous, and form small groups that rarely contain more than two adult males (Crockett and Eisenberg, 1987; Crockett and Janson, 2000; Di Fiore et al., 2011). Finally, Alouatta caraya is the species with the southernmost range (northern Argentina), is sexually dimorphic (males 4.0–9.6 kg, females 3.8–5.4 kg) and dichromatic, exploits seasonal forests, is folivorous–frugivorous, and may form small uni-male (1 male, 1–5 females) or multi-male–multi-female groups that contain 4–6 males and 4–8 females (Crocket and Eisenberg, 1987; Neville et al., 1988; Rumiz, 1990; Zunino et al., 2001; Kowalewski and Zunino, 2004; Di Fiore and Campbell, 2010). The analysis of their behavior, substrate, and tree part used will help to evaluate the ways posture, sleeping time, physiology, and digestion are intertwined and to test an ecophysiological hypothesis.

The main goal of this research is to explore the potential ecological and adaptive roles of postures during sleep among different species of howler monkeys. Thus, the following questions are addressed:

Does sleeping postural behavior differ among howler species at various sites?

The sit-in posture was primarily used during sleep, and this pattern was consistent across the three species (Table 1). In terms of substrate inclination, horizontal supports dominated in all three species, with no significant differences across species (Table 1). A similar pattern was detected for substrate size, with all three species using large branches significantly more frequently than medium ones, with no significant differences between them (Table 1). Alouatta palliata and A. caraya showed no significant differences in substrate type use, using almost equally unramified branches and ramified branches (Table 1). Furthermore, the two species appeared to largely utilize branches with bifurcations (A. palliata 75 %, A. caraya 96.2 %) over branches with trifurcations. In contrast, A. macconnelli used exclusively unramified branches (Table 1).

Tree crown vertical and horizontal location of sleeping postures across the three species revealed some common patterns. In terms of vertical location, A. palliata and A. macconnelli largely used the middle parts of tree crowns, while A. caraya used both the upper and middle parts (Table 1). All three howler species used the lower part of the tree crown less often. On the other hand, the different howler species preferred the middle and core sections of the trees to a varying extent and appeared to have used the tree periphery less (Table 1).

The current study is a first attempt to systematically record postural behavior, substrate characteristics, and crown location during sleep in three species of howler monkeys across the distributional range of the genus Alouatta. Previously, Mendel (1976) provided an early attempt to explore the selection of postures in relation to substrates, tree parts, and anatomical correlations, but his study was restricted to mantled howlers (A. palliata). Our comparative study showed that howlers exhibited common sleeping patterns. All three species used the inner parts of tree crowns. Therein, they adopted a sit-in posture on mainly horizontal as well as large, unramified, and ramified branches. Sleep is the major extended behavioral phase of the diel cycle of mammals and appears to play an important role in safety and thermoregulation. The common use of this particular niche (central tree crown with large horizontal branches) by all three species during sleep most likely reflects common adaptive strategies to similar ecological pressures. The central tree crown is a covered zone with reduced access to potential predators and increased protection from extreme weather conditions, and at the same time its abundant large horizontal branches represent steady substrates ideal for postural stability. In this way, they seem to provide the required safety along with mechanical and physiological stability (Anderson, 1984, 1998, 2000).

All three species appeared to have selected parts of the tree crown located in the inner sections, systematically using the periphery and lower part of trees less often. The inner parts of tree crowns usually provide sleeping sites that are relatively inaccessible to potential predators (Zhang, 1995; Anderson, 1998; Ramakrishnan and Coss, 2001). In effect, the central and upper parts of tree crowns provide a natural shelter, reducing visibility to predators, and create a relatively wide buffer zone that may protect animals from approaching predators. In this way, potential howler predators, such as scansorial carnivores (jaguarundis, Herpailurus yagouaroundi, jaguars, Panthera onca), harpy eagles (Harpia harpyja), or humans cannot easily locate prey which is frequently covered in heavy foliage (Peres, 1990; Peetz et al., 1992; Di Fiore, 2002; Kowalewski and Zunino, 2004; Urbani, 2005; Raguet-Schofield, 2008; Urbani et al., 2012). In addition, such locations may also provide protection from extreme weather conditions, as has been suggested for other arboreal monkeys (e.g., Wahungu, 2001; Cui et al., 2006). In our cases, sleeping in central sections of tree crowns may provide cover from heavy and intense rainfalls in French Guiana, from frost during the winter in northern Argentina, and from severe night wind flows in Nicaragua as perceived by the authors in the field sites. However, more detailed data associating these specific factors with canopy part selection are required for further testing.

Apart from a natural hide-out that protects from predators and weather, the inner parts of tree crowns are dominated by relatively large and horizontal branches. Large substrates offer a wide contact area for the body, minimizing any unbalancing forces and the likelihood of toppling over. In effect, all three howler species utilized particularly large substrates, >10 cm in diameter, which is larger than the average width of the pelvis of most howler species (e.g., 7 cm for Alouatta palliata: Leutenegger, 1974). In mantled and black-and-gold howlers, the use of branches with bi- and trifurcations further enlarges the contact area for the ischia and enhances stability and balance. In this way, firm and secure postures are guaranteed over an extended time of inactivity during sleep. In addition, all three species mostly used horizontal or only slightly inclined substrates. Horizontal and low-inclination branches assure that the center of mass be located well within the support area, eliminating any risk of sliding, as would have been the case with steeper inclinations. These features characterize biomechanically strong, firm, and stable substrates which provide the stability and equilibrium that is necessary during long postural modes, such as during sleep (Rose, 1974). In this way, animals maintain a dynamic stability with less energy expenditure, thus contributing to energy conservation. This is essential for mammals, such as howlers, that are mainly folivorous and have adopted an overall energy conservation strategy to likely maximize their intake from a low-quality diet (Milton, 1981). Postural stability and equilibrium were further achieved by the principal use of sit-in, as the dominant sleeping posture by all three howler species. Anderson (1984, 2000) has suggested that postural modes such as crouching, sitting, and squatting during sleep in primates appear to provide enhanced stability. These postures enlarge the contact area between body and substrate and simultaneously lower the center of mass of the body, providing balanced support on the branch (Wells, 1974, in Anderson, 2000).

Additionally, crouched sitting postures during sleep have been functionally related to thermoregulation (Anderson, 1984, 2000). Contact between the extremely flexed hindlimbs, forelimbs, and the lowered head towards the ventral area reduces the exposed surface and consequently heat loss (Anderson, 1984, 2000). During stormy weather, the same posture may further divert rainwater away from the body, keeping it warm (Anderson and McGrew, 1984). Regulating heat loss and maintaining a warm body are necessary during sleep, and this is accomplished by adopting this posture within the inner parts of tree crowns with relatively stable microclimatic conditions. The contribution of this posture to thermoregulation is further suggested when night temperatures drop drastically, like in Argentina, where black-and-gold howlers not only use the ball-like sit-in posture but also huddle with other individuals to increase heat retention, as in many other social mammals (Kowalewski and Zunino, 2004; Martín M. Kowalewski, unpublished data; see also Bicca-Marques and Calegaro-Marques, 1998).

Lastly, adopting this crouched ball-like sit-in posture on stable large horizontal branches during sleep for long time periods may be associated with the prolonged digestion times of howlers (Milton, 1981, 1998; Milton and McBee, 1983; Yumoto et al., 1999). It has been suggested that prolonged time of digestion, and the concomitant extended food retention time in the digestive system, aids in the absorption of nutrients in primates (Clauss et al., 2008). Compared to other New World primates, howlers exhibit particularly high rates of folivory (A. palliata: Illes, 2005; Raguet-Schofield, 2010; A. macconnelli: Julliot and Sabatier, 1993; A. caraya: Bravo and Sallenave, 2003; Kowalewski, 2007) and show comparatively long digestion and ingesta retention time than other neotropical primates, which occurs mainly during the night (Milton, 1981; Espinosa-Gómez et al., 2013, Matsuda et al., 2019). Thus, we suggest that the adoption of the sit-in posture during sleep likely provides the necessary stability and positioning for facilitating advanced digestive processes. Similarly, it seems possible that the digestive system is positioned to facilitate the vertical stratification of food particles for digestion and nutrient absorption in hindgut fermentation, over the prolonged nocturnal inactivity and long daily resting periods as commonly observed in howlers. This contention, then, requires further research that might be tackled through extended comparative studies among neotropical primates with different dietary and digestive patterns and gut passage times (e.g., Lambert, 1998). Thus, this ecophysiological mechanism (as observed for foregut fermenters such as wild sloths: Urbani and Bosque, 2007, and colobines: Matsuda et al., 2017) might eventually further interplay with the thermoregulatory and stability contentions summarized by Anderson (1984, 1998, 2000).

The results of the current research have indicated that independent of forest type, howlers tend to use similar substrate features and locations within trees and adopt similar postures during sleep. This common strategy is very likely associated with biomechanical stability, avoidance of predators and/or extreme weather conditions, thermoregulation, the efficiency of the digestive processes, and ultimately phylogeny. These factors may have played an important role in the selection of certain behavioral and habitat variables during sleep. However, in order to further understand these adaptive interrelations, more studies are required on the postural behavior and microhabitat selection during sleep of other howler species and other diurnal neotropical monkeys, as well as other primates with a strong reliance on folivory (Milton, 1998).