Platyrrhine color signals: New horizons to pursue

Abstract Like catarrhines, some platyrrhines show exposed and reddish skin, raising the possibility that reddish signals have evolved convergently. This variation in skin exposure and color combined with sex‐linked polymorphic color vision in platyrrhines presents a unique, and yet underexplored, opportunity to investigate the relative importance of chromatic versus achromatic signals, the influence of color perception on signal evolution, and to understand primate communication broadly. By coding the facial skin exposure and color of 96 platyrrhines, 28 catarrhines, 7 strepsirrhines, 1 tarsiiform, and 13 nonprimates, and by simulating the ancestral character states for these traits, we provide the first analysis of the distribution and evolution of facial skin exposure and color in platyrrhini. We highlight ways in which studying the presence and use of color signals by platyrrhines and other primates will enhance our understanding of the evolution of color signals, and the forces shaping color vision.

multiple mechanisms of sexual selection are acting in concert or antagonistically on the same trait. 23 Additionally, unlike clades like birds, in which all diurnal species are thought to be tetrachromatic, primates exhibit great intraorder diversity in color perception. 24 Primates also exhibit a stunning range of social organizations, and mating systems, including monogamy, polygyny, polyandry, and polygynandry. The latter includes extreme interspecific variation in the distribution of matings and genetic reproduction. 25 Understanding the evolutionary selective pressures acting on color variation of the faces, genitals, hindquarters, and elsewhere, within and between primate species has long been of interest to evolutionary biologists and anthropologists, dating back to Darwin. 1,26 However, with few recent exceptions, [27][28][29] the vast majority of research has focused on one particular taxonomic lineage, the African and Asian monkeys (part of the Parvorder Catarrhini). 30

| SKIN COLOR AND EXPOSURE IN PRIMATES: ROLES IN VISUAL COMMUNICATION
Catarrhine primates are well known to possess large patches of exposed skin on their faces and hindquarters, and to possess skin color that varies in response to fertility, pregnancy, and other reproductive phases, as well as social status. 31 In addition to chroma (the hue and saturation components of color), patterns of luminance variation is a frequent but often underemphasized component of color signaling in catarrhines. 32,33 Chromatic and achromatic components of signals seem to have different mechanisms of inheritance 34 but often covary. 35 For example, red male rhesus macaques (Macaca mulatta) receive more sexual solicitations by more females than pale pink males, suggesting the potential importance of both chroma and luminance. 21,36 A now impressively large body of work has focused on the reddish and other colorful signals of African and Asian primates, and their roles in socio-sexual communication (Table 1).
Facial skin that is distinctly reddish and/or exhibits large luminance contrast relative to surrounding pelage is also present in the other major radiation of haplorhine primates, the monkeys of Mexico, Central and South America (Parvorder Platyrrhini) ( Figure 1). Intriguingly, a wide diversity of social systems occurs in platyrrhines, including polyandry and monogamy, begs investigating the impact of mating systems on signal evolution. However, comparative studies of platyrrhine coloration are almost absent in the literature. The only study investigating intraspecific skin color variation in platyrrhines demonstrates that female genital skin color (luminance and hue) varies across pregnancy and parturition in common marmosets (Callithrix jacchus). 62 It is possible that this color variation is a nonadaptive by-product of hormonal variation with no signaling purpose. 63,64 Alternatively skin color may be part of a suite of sensory cues associated with triggering preparation for male care, 65 or female reproductive suppression. 66 Although speculative, this work calls attention to the potential for intraspecific skin color signals in platyrrhine monkeys and highlights the need and potential for color signal studies within this clade.
In addition to color variation, the extent of exposed skin among the monkeys of the Americas ranges from the most extensive seen in the order (bald uakaris [Cacajao calvus]) to a complete absence of exposed skin (e.g. moustached tamarins [Saguinus mystax]). The increased surface area of exposed skin in African and Asian monkeys has been hypothesized to enhance communication based on skin color signals. 31,67 However, causal explanations are only supported if they are both explicable and predictable-that is, they demonstrate regularity and independence in their evolution in response to similar selective conditions. 68 The platyrrhines provide an independent test of the hypothesis that the amount of exposed facial and genital skin is evolutionarily plastic, and when present, used in socio-sexual communication. To be convergent with catarrhines, we would expect increased facial skin exposure to have evolved independently in platyrrhine species that use colorful skin signals.

| EVOLUTION OF SKIN COLOR AND EXPOSURE IN PLATYRRHINE MONKEYS: A CHARACTER MAPPING APPROACH
To motivate future studies of visual signals and their role in communication among platyrrhine primates, we here provide the first analysis of exposed facial skin, variation, distribution, and evolution in Platyrrhini.
Concordantly, we aim to uncover patterns underlying red, conspicuous skin. Because a dataset of color-calibrated images across the platyrrhines is not yet available, our analysis is color-subjective. We analyzed two to five forward-facing images found in a current encyclopedia of living primate species 69 and from online image libraries (Table S1). We quantified facial skin exposure and color for 96 platyrrhines, and for comparative purposes, 28 catarrhines, 7 strepsirrhines, 1 tarsiiform, and 13 nonprimate outgroups. Coding of skin color and exposure was done independently by L.A.A.M. and G.D. In case of any discrepancy, we recoded the images together to reach consensus. We did not include Homo sapiens in our analyses because of high intraspecific phenotypic variation and numerous derived features.
We created five categories to code patterns of increasing surface area of exposed facial skin: (a) completely exposed skin (cheeks, nose, eyes, forehead [ Figure 2a,b]); (b) exposed skin around the eyes, nose and mouth [ Figure 2c,d]; (c) exposed skin around the nose and eyes, or exposed skin around the nose and mouth [ Figure 2e,f]; (d) exposed skin around eyes [ Figure 2g,h]; and (e) exposed skin around the nostrils [ Figure 2i,j]. We coded for skin color using established categories 29 :   The ancestral character states for skin exposure (continuous trait) were simulated using the Maximum Likelihood method. We tested three models of evolution: Brownian Motion (BM), Ornstein-Uhlenbeck (OU), and Early Burst (EB), before selecting the best-fitting model (OU). Skin exposure evolution was plotted using the phytools contMap function in R. 72 Because the evolutionary history of monkeys in the Americas is still debated, we repeated our analyses across multiple potential phylogenies ( Figures S1 and S2). 73,74 We inferred the evolutionary history of skin color (discrete trait) using a stochastic mapping approach implemented in the R phytools package. 72 The ancestral states at each node were estimated under three basic models: equal rates (ER), all rates different (ARD), and symmetrical transition rate (SYM). The best model fitting (ER) was selected and 1,000 character histories were simulated across the phylogeny using the phytools make.simmap function in R version 3.5.2. 75 Our results indicate that exposure of facial skin is an evolutionarily plastic trait, with shifts from relatively hairy to relatively exposed skin and vice versa being relatively common among platyrrhines ( Figure 4).
The last common ancestor (LCA) of catarrhines and platyrrhines is reconstructed to have predominantly exposed skin, that is, exposed skin around the eyes, mouth, and nose. Many species of platyrrhines possess a lesser amount of exposed skin than catarrhines. Yet, there are radiations of platyrrhines with facial skin that is largely to entirely exposed ( Figure 4). Increases in exposed facial skin appear to have evolved independently in platyrrhines at least five times. We see this trait in the genus Saguinus, and among the Atelidae family, with interesting variation within the genus Cacajao. There is also considerable exposed skin within the genera Cebus and Sapajus, and the callitrichids, including some but not all species of Callithrix, some Saguinus, and all examined members of the genus Leontopithecus Both bald uakaris and red-faced spider monkeys are also among those with the most exposed facial skin, consistent with the hypothesis that the evolution of exposed faces and red signals are correlated. 31 However, it is important to note that while this relationship occurs, many species with completely or predominantly exposed skin do not have red pigmentation, highlighting the presence of considerable variation (e.g., Saguinus bicolor; Figures 4 and 5).
Our results are robust to phylogenetic uncertainty, that is, we come to similar conclusions when the models are run with different phylogenies (Figures S3 and S4). Overall, we demonstrate that there is extensive variation among platyrrhines in the presence of exposed facial skin, and high potential for reddish color signals. We suggest this group offers great promise for studies that mirror the research programs that have been undertaken to examine skin signal evolution in other taxa.
The analyses we present here consider only facial skin, as the diagnostic images of faces are more readily available for many species. However, like many African and Asian primates, the genital skin of at least some Central and South American monkeys is exposed, visible, and of high visual (luminance) contrast to body pelage ( Figure 6). In some species of catarrhines, color is expressed on both the face and genitals, and this color variation is often correlated (e.g., rhesus macaques 35 ; drills 18 ). However, other species only exhibit conspicuous color on the genitals (e.g. vervet monkeys 16 ).
Variation in the conspicuity of genital skin among platyrrhines suggests that there may be convergent evolution on such traits, and that studies of inter-and intra-specific variation in genital color alongside studies of face color are likely to be fruitful. F I G U R E 2 Examples of categories of exposed skin in primates. Completely exposed face (cheeks, nose, eyes, forehead), a, b); exposed skin around the eyes, nose, and mouth, c, d); exposed skin around the nose and eyes; or exposed skin around the nose and mouth, e, f); exposed skin around eyes, g, h); exposed skin around the nostrils, i, j).  (Table 1), provides excellent templates for such studies.
We hypothesize that luminance and/or chromatic variation is involved in socio-sexual communication for at least some species of platyrrhine primates. Specifically, we predict that color variation of facial or genital skin plays a role in communication for species with exposed faces and genitals.
If socio-sexual communication among females along the red-green channel is important, then those taxa are predicted to have greater numbers of opsin alleles and higher instances of trichromacy. By genotyping the OPN1LW opsins of platyrrhine primates with reddish, exposed faces, along with closely related species without red faces or patches, the hypothesis that red and visible facial skin is associated with increased diversity of opsin alleles in platyrrhines can be tested. It is of particular interest to note that among monkeys in the Americas, females may have a private color channel due to the sex-linked nature of color vision variation in platyrrhines. A private channel is a signaling channel that only a subset of potential receivers can see. This has been suggested, for example, for bird UV signaling, which is visible to other birds, but not visible to mammalian predators. 87  foraging ecology. 76 Howler monkeys (Alouatta sp.), whose routine trichromacy has evolved independently from that of catarrhines, and monochromatic owl monkeys (Aotus sp.) may provide unique test cases that generate new insight into primates color vision evolution. 86,90 Finally, we note that it is important to explicitly evaluate female-female signaling, male-male signaling, female-to-male signaling, and male-tofemale signaling separately in platyrrhines, given the sex-linked nature of color vision variation.
Challenges to studying skin color signals in platyrrhines include: (a) difficulties in measuring coloration of small, highly arboreal primates; (b) finding appropriate populations for testing attention to   70 and adapted to include 96 platyrrhines, along with 28 catarrhines, 7 strepsirrhines, 1 tarsiiform, and 13 nonprimate groups to reconstruct ancestral types. The color map represents observed and reconstructed ancestral states for skin exposure ranging from a completely exposed face (brown) to only exposed skin on the nose (green)  (c) relying on DNA from noninvasively collected fecal samples, rather than blood, hair, or tissue, to increase the feasibility of population-level sampling. Fortunately, study of opsin genes from fecal DNA is well established, alleviating need to capture and release wild monkeys. [94][95][96] Careful and comprehensive sampling of wild populations is possible, if time consuming. 97,98 Finally, although we have focused on skin color and intraspecific communication, we would be remiss if we failed to highlight the potential of interspecific pelage and skin color variation to contribute to key evolutionary processes. For example, species boundaries may be enhanced via conspicuous character displacement among callitrichid and titi monkey lineages that have radiated widely in short time periods. [99][100][101] Sensory ecology is at the heart of evolutionary anthropology-the major differences between the haplorhines and the strepsirrhines, and similarly between catarrhines and platyrrhines, are in the ways in which they receive and process information from their environment. A more comprehensive study of primate visual communication, to include understudied major radiations, will enhance our understanding of the evolution of our improved color vision and of social communication, key elements of what it means to be a primate.
F I G U R E 6 Conspicuous scrotum relative to body pelage in mantled howler monkey (Alouatta palliata, a) and bearded saki (Chiropotes satanas, b). Photo credits: Scott Robinson (a) and Allan Hopkins (b) F I G U R E 7 Simulated appearance of a female rhesus macaque (Macaca mulatta; Cayo Santiago) for: a trichromatic observer (a); a dichromatic observer (b). 80 Peak cone sensitivity values on simulations of trichromatic Rhesus macaque vision (S cone = 420 nm; M cone = 530 nm; L cone = 560 nm) and a protanomalous dichromatic type (S cone = 420 nm; M cone = 530 nm). 45 Photo by ADM