Closely related species living in sympatry face a potential challenge in discriminating between conspecifics and heterospecifics. Such decision-making has important selective outcomes, particularly in behaviors such as mate choice, with individuals choosing heterospecific mates often incurring substantial fitness costs (Coyne and Orr, 2004; Gröning and Hochkirch, 2008). One mechanism for avoiding the costs of interacting with heterospecifics is the use of species-specific signals that structure behavioral interactions between species. For instance, mating signals and associated mating preferences that differ between sympatric heterospecifics can function to maintain reproductive isolation across species boundaries (Andersson, 1994). Such signals are predicted to be salient and distinctive (Dawkins and Guilford, 1997), with sympatric species under selective pressure to diversify. A pattern in which signal distinctiveness increases with degree of sympatry, known as character displacement (Brown and Wilson, 1956; Pfennig and Pfennig, 2009), has been observed in a wide variety of animal groups (insects: Higgie et al., 2000; Lukhtanov et al., 2005; Stanger-Hall and Lloyd, 2015; birds: McNaught and Owens, 2002; gastropods: Kameda et al., 2009; mammals: Allen et al., 2014; anurans: Pfennig and Rice, 2014; Gordon et al., 2017). Importantly, signals that function to promote behavioral interactions with conspecifics should elicit increased interest from conspecifics compared to heterospecifics (Mendelson and Shaw, 2012).

Species in evolutionarily young animal radiations may be at particular risk of hybridization and other costly interactions with heterospecifics due to behavioral similarities and a lack of post-mating barriers to reproduction (Mallet, 2005). One such radiation is the guenons (tribe Cercopithecini), a group of African primates consisting of 25–38 recognized species (Grubb et al., 2003; Groves, 2005; Lo Bianco et al., 2017) that diverged from papionin primates around 11.5 million years ago (Tosi et al., 2005; Guschanski et al., 2013). Guenons exhibit high degrees of sympatry and often form polyspecific groups in which multiple species travel and forage together (Jaffe and Isbell, 2011). Many guenons therefore interact with heterospecifics that share general patterns of morphology (e.g. overall body size/shape) and behavior (e.g. activity patterns). In such circumstances, discriminating between con- and heterospecifics may be particularly important, especially in a mating context. Hybridization between sympatric guenon species has been observed in captive and wild settings but is rare in most natural circumstances (Detwiler et al., 2005), suggesting the existence of barriers to heterospecific mating within mixed-species groups.

Guenons are among the most colorful and visually patterned groups of primates, with many species exhibiting extraordinary and unique face markings (Kingdon, 1980; Kingdon, 1988; Kingdon, 1997; Bradley and Mundy, 2008; Allen et al., 2014) that cannot be reliably computationally classified by sex across all guenon species (Kingdon, 1997; Allen and Higham, 2015). Kingdon hypothesized that guenons use their divergent facial appearances to distinguish between species and select appropriate mates (Kingdon, 1980; Kingdon, 1988; Kingdon, 1997). This young and impressively diverse primate radiation represents a fascinating test case of how visual signals are involved in species radiations and mixed-species interactions (Pfennig and Pfennig, 2009; Pfennig and Pfennig, 2010; Pfennig, 2016; Stuart et al., 2017). Recent empirical work has begun to generate evidence for a key role for face patterns in guenon phenotypic and species diversification. Images of guenon faces can be reliably classified by species using computer algorithms (Allen et al., 2014; Allen and Higham, 2015), demonstrating that guenon faces exhibit species-specific patterns that are potentially informative and biologically relevant to receivers. Guenon face patterns also exhibit character displacement, with facial distinctiveness between species increasing with degree of geographic range overlap across the group (Allen et al., 2014). Moreover, facial components common across species (nose spots and eyebrow patches) alone can be used to computationally classify species even when segmented from the rest of the face (Allen and Higham, 2015). This suggests that guenon faces may be somewhat modular, with particular face regions being potentially informative regarding species identity. Which face regions are most important, and the extent to which such regions vary across species, remains an open question that is of key significance to understanding how complex signals involved in species discrimination evolve. Critically, it is unknown whether variation across guenon species in purported species discrimination signals is perceived and acted on by con- and heterospecific receivers.

Here, we use a machine learning approach to identify guenon face regions that are most important for correct species classification by a computer. These results objectively identify the signal components most likely to be useful to guenon receivers. We use them to determine which signal properties to systematically investigate in behavioral experiments with guenon observers. The machine-learning stage is critical, as many experiments that investigate behavioral responses to complex signals select manipulations based on the perceptions of investigators, which introduces anthropocentric bias (Patricelli and Hebets, 2016). Using the guenon face image database produced by Allen et al. (2014), we couple eigenface decomposition of guenon faces (Turk and Pentland, 1991) with a novel occlude-reclassify scheme in which we systematically block each part of the face and reclassify the image. This allows us to document the spatial distribution of species-typical information across guenon faces by identifying which face regions, when obscured, cause the break-down of correct species classification. Eigenface decomposition was originally developed for individual face discrimination in humans (Turk and Pentland, 1991); feature detection based on eigenfaces is also applicable to other types of discrimination tasks involving complex animal signals (Weeks et al., 1999; Shiau et al., 2012; Perera and Collins, 2015) and has been used previously to quantify guenon facial variation (Allen et al., 2014). The type of perceptual face space generated by eigenface decomposition has correlates in mammalian visual processing systems (Leopold et al., 2001; Cowen et al., 2014; Chang and Tsao, 2017), lending biological credibility to this approach.

We ran occlude-reclassify analyses using both the entire set of fifteen guenon species as well as two subsets, which included species with geographic ranges that overlap those of two guenon species. The full analysis including all species allows us to identify facial traits that differentiate each species from others in the tribe, and these results are therefore broadly applicable across species. Additional subset analyses represent a more biologically realistic discrimination task for the two species that were the focus of experiments; here, the key results are for the two species that form the basis for these subsets.

After identifying the face regions that cause a break-down in classification, and thus those that should be important for correct species identification, we then presented captive putty nosed monkeys (Cercopithecus nictitans) and mona monkeys (C. mona) with images of con- and heterospecific faces in a behavioral discrimination task measuring the looking time toward each stimulus. Ours is the first direct measure of guenon responses to con- and heterospecific faces, which is crucial for clarifying the biological relevance of guenon face patterns and for validating previous correlational results. Looking time experiments are commonly used in studies of human infants and non-human animals to gauge visual attention and interest, and in so doing to infer underlying perceptual and cognitive abilities and biases (reviewed in Winters et al., 2015). Our looking time experiments are not designed to assess the neurological mechanisms underlying object detection or classification, but rather to assess the level of interest elicited by different classes of stimuli and to identify the facial features that contribute to this interest. Differences in looking time between classes of stimuli can be difficult to interpret due to various and often unpredictable novelty and familiarity effects (Winters et al., 2015), however primates reliably exhibit a visual bias (i.e. greater looking time) toward images of conspecifics compared to those of heterospecifics (Fujita, 1987; Demaria and Thierry, 1988; Dufour et al., 2006; Méary et al., 2014; Rakotonirina et al., 2018). Our experimental trials involve the simultaneous presentation of paired con- and heterospecific faces, with trial stimuli varying in facial pattern components across experimental conditions. In this approach, a statistical difference in looking time between classes of stimuli is a behavioral indicator of an ability to discriminate between these classes as well as an indicator of relative interest in the two classes. Experimental stimuli are designed to vary in the presence or absence of nose spots for putty nosed monkeys and eyebrow patches for mona monkeys, on the basis that each of these features is within the region of the face identified by our machine learning approach as being critical for that species. In our first experimental condition, subjects are presented with the face of a conspecific and that of a heterospecific which does not display the relevant focal face trait. This condition represents a baseline in which there is no conflation between species and species-typical face traits, and confirms that the general pattern of visual biases for conspecific faces holds for the individuals in the two guenon species tested here. It also represents a biologically realistic comparison, as guenons are typically surrounded by heterospecifics with faces dissimilar to their own (due to character displacement, Allen et al., 2014). A conspecific bias in this condition would suggest that subjects find this class of stimuli more interesting, and may translate into behavioral differences in interactions with conspecifics relative to heterospecifics. In our subsequent experimental conditions, we present stimuli in which species and focal face traits are conflated: in condition two, we present a conspecific face paired with a heterospecific face that shares the facial trait; in condition three, we present a conspecific face modified to remove the facial trait paired with a heterospecific face that shares the facial trait (for example stimuli, see Figure 1). These latter trials therefore ask the question: how is a conspecific bias affected by manipulations of which facial pattern components are available? The stimuli in these trials do not mimic natural scenarios, but including artificial stimuli such as modified conspecific faces allows us to evaluate the role of these face traits in directing guenon visual attention. We present results analyzing visual biases in the first condition, as well as a unified analysis of visual biases across all experimental conditions to identify the overall drivers of looking time across conditions. This approach allows us to assess generalized species biases in degree of interest as well as the extent to which particular face regions influence these biases.

Figure 1 with 1 supplement see all

Download asset Open asset

We predicted variability across species in the face regions identified by our occlude-reclassify procedure, but made no predictions regarding which regions in particular would be essential for each species. In looking time experiments, we predicted that putty nosed and mona monkeys would exhibit visual biases toward face images of conspecifics, and that these biases would be influenced by species-typical facial characteristics identified as important for correct species classification. Such a pattern of results is consistent with a role for species discrimination signals in facilitating interspecific interactions, such as maintaining reproductive isolation via mate choice, and in generating and maintaining phenotypic variation in one of the most speciose and diverse primate radiations. Species discrimination requires differentiating between conspecifics and all heterospecifics at the very least, however it may be possible that animals are also able to discriminate between different heterospecific species. Ultimately, by examining how aspects of highly complex signals encode species identity and influence receiver biases, this research increases our understanding of how selection for species identity signaling generates phenotypic diversity.

Our machine classifier results identified certain face regions as being critical to correct species classification across the guenons, and experiments undertaken for two target species provided results that were consistent with the results of the classifier. This convergence of results using disparate methods reinforces the validity of both, and ties computationally derived results directly to guenon perception, demonstrating the utility of machine learning for identifying biologically relevant signal components. To our knowledge, ours is the first analysis to use machine classification combined with the systematic occlusion of image regions to characterize the relevant signaling information encoded in an animal’s appearance. This approach, based on research in the field of computer vision designed to assess the contribution of image contents to object classification (Zeiler and Fergus, 2014), is useful for objectively quantifying the relative roles of different signal components with respect to overall signal function. In closely-related sympatric species, selection against mating or other behavioral interactions with heterospecifics is often associated with the evolution of species-typical traits used to maintain reproductive and behavioral isolation. The guenons, a recent and diverse radiation that exhibit mixed species groups in which hybridization is rarely observed, exemplify this phenomenon. By showing how species classification is dependent on different aspects of face patterning and that this links with looking time toward con- and heterospecifics, our analyses support a role for guenon face patterns in species discrimination, and identify specific face regions potentially critical for this function. This parsing of critical signal components is essential for understanding the phenotypic evolution of complex signals and identifying relevant axes of signal variation for additional analyses.

Our occlude-reclassify analysis identified face regions critical to correct species classification by a machine classifier in all guenon species included in our study. Critical regions differed in both location and spread across the face, suggesting variation in potential use across species. For some guenons, reliance on a single facial characteristic may be sufficient for species discrimination. The best example of this in our data set is the putty nosed monkey, where our machine classifier relied exclusively on the white nose spot to classify this species. That is, occlusion of any other region resulted in correct classification, but when the nose spot was occluded classification failed. This result is reinforced by our experiments, in which putty nosed monkey visual attention was driven wholly by the presence of nose spots. Putty nosed monkeys exhibited a conspecific bias when presented with natural con- and heterospecific faces, as is typical in primates, however including stimuli depicting heterospecifics with nose spots and conspecifics without nose spots completely obscured this conspecific bias. This combination of results illustrates the importance of nose spots in this species. It is worth noting that putty nosed monkey nose spots are the most straightforward facial trait documented in our analysis (i.e. putty nosed monkeys were only misclassified when the nose spot was occluded and occluding the nose spot led to a high rate of misclassification) and that putty nosed monkeys are the only species in our analyses for which the classifier identified only one critical face region; the relative simplicity of the face and related visual biases in this species is likely exceptional. On the whole, species discrimination signals in a large radiation with varying patterns of sympatry are expected to be complex and multidimensional, and it is likely that only some species can exhibit single-trait-based signals and visual biases without the system breaking down. This is supported by our results showing that for all other guenon species our classifier identified multiple face regions that were critical for species discrimination.

Not all guenons exhibited critical face regions restricted to isolated facial traits, and our machine classifier sometimes relied on disparate face regions. In our data set, the mona monkey is a good example of such a species. As in putty nosed monkeys, our experiments with mona monkeys supported these computational results. Mona monkeys exhibited a conspecific bias across all trials, regardless of single trait manipulations, as well as an additional bias based on the presence of eyebrow patches. Thus, eyebrow patches alone do not appear to be the sole driver of visual attention in mona monkeys. We predict that additional manipulation of other face regions would be necessary to redirect their visual attention. Nonetheless, that mona monkey attention is still influenced by this species-typical trait shows that it is important but not essential, a result predicted by our computational analyses. It is unclear why mona monkeys would look longer at stimuli without eyebrow patches; however, it is possible that utilization of the whole face causes increased attention to incongruency (e.g. conspecifics without eyebrow patches or heterospecifics with them). Another possibility is that this manipulation resulted in odd-looking faces that drew subjects’ attention, although the absence of eyebrow patches in most guenon species would suggest that these features are not necessary for general face processing. Our results suggest that in mona monkeys, species discrimination may be based on broader face information, and the perceptual processes involved in assessing potential mates could be similar to generalized holistic face processing mechanisms observed in other primates (Dahl et al., 2009). Complex multicomponent signals also have a greater potential for signal redundancy (Partan and Marler, 1999; Hebets and Papaj, 2005; Higham and Hebets, 2013), although our results from mona monkeys showing that a single component of a complex signal does not drive attention on its own suggest that redundant face regions are more likely to represent enhancement (i.e. an increased response to the combined signal compared to each component) than equivalence (i.e. the same response to each component or the combined signal) of alternative regions (Partan and Marler, 1999).

Our results suggest that guenons, while united by a general pattern of facial diversification and the probable use of faces in mate choice, may vary across species in the specific traits and processes that are involved in discriminating between conspecifics and heterospecifics. Our pattern of experimental results for putty nosed monkey nose spots and mona monkey eyebrow patches is interesting because we know that both traits are sufficiently informative to allow discrimination between species that share these features (Allen and Higham, 2015), yet they influence attention differently in the two species. This disparity highlights the importance of testing receiver perception directly, and also indicates the importance of future experimental work on live guenons in additional species. The fact that our experimental results with the two species of guenons tested in the present study line up with predictions generated by our occlude-reclassify analysis nonetheless implies that these computationally derived results may be biologically valid across the radiation. While the observed differences in looking time were small, the magnitude of these differences is similar to those reported in looking time studies of humans (Shimojo et al., 2003) and other non-human primates (Dubuc et al., 2014; Dubuc et al., 2016), in which phenotypes that elicit increased visual gaze have also been associated with reported judgements of attractiveness and behavioral indicators of mating interest, respectively. This suggests that these differences are biologically meaningful and may influence guenon decision making. Here, our results strictly show that visual attention in guenons is influenced by variation in face pattern components; the extent to which these attentional biases translate to other behavioral contexts awaits additional investigation.

Interestingly, we found no sex differences in visual biases for either species, suggesting that selective pressures on species discrimination signaling and visual preference traits are similar between sexes. Most guenons live in polygynous mating systems (one male, multiple females) and males exhibit minimal paternal care. This could theoretically lead to the evolution of increased choosiness and discriminative capabilities in females, if interspecies interactions such as hybridization are more costly to them. However, many guenon groups experience seasonal influxes of males during the mating season; these extra-group males tend to be attractive to females, and the group-male is often unable to competitively exclude them (Cords, 2012; Cords, 2004). In such circumstances, group-males that opt to protect reproductive access to heterospecifics would incur substantial fitness costs, and males in mixed-species groups (as are typical for most guenons: Jaffe and Isbell, 2011) could gain fitness advantages by conceding reproductive access to heterospecific females while preferentially mate guarding conspecifics. As such, we predict that engagement with heterospecific partners is costly in both sexes. Field observations documenting the roles of males and females in initiating and preventing interactions between species would help to evaluate this prediction and contextualize our results.

In guenons, an observed lack of hybrids in most polyspecific groups (Detwiler et al., 2005) is notable given that hybridization is known to be possible between many guenon species (Kingdon, 1980; Kingdon, 1997; Detwiler et al., 2005) and indicates the existence of pre-mating barriers to reproduction. Increased eye gaze is associated with increased mating interest in humans (Shimojo et al., 2003) and non-human primates (Dubuc et al., 2014; Dubuc et al., 2016), suggesting that our experimental results could be indicative of an increased interest in mating with conspecifics based on their face patterns in our putty nosed and mona monkey subjects. Combined with previous work (Kingdon, 1980; Kingdon, 1988; Kingdon, 1997; Allen et al., 2014; Allen and Higham, 2015), our results support the hypothesis that guenon face patterns play a role in mate choice and reproductive isolation in this group. However, it remains possible that the selection pressure for species discrimination traits in guenons arises partially or entirely from other functions where behavioral coordination or avoidance between species is advantageous, such as in foraging decisions (Kingdon, 1997; Jaffe and Isbell, 2011). Careful field observations would be needed to distinguish between such possibilities.

Our occlude-reclassify approach is a novel method for identifying the distribution of information in complex signals and can be used for any question that can be conceptualized as a discrimination problem and analyzed using machine classification. This method therefore has broad utility within sensory ecology and could help to better understand the link between form and function in the complex signals that are common in many animal groups. The objectivity of the approach is important, as it allows researchers to intelligently targetspecific signal components for further analysis without reference to their own perceptions of their salience. This is particularly important when studying species with sensory and perceptual systems very different from our own (Endler, 1990; Bennett et al., 1994). Where possible, combining this approach with a biologically realistic encoding scheme, such as classification within a perceptual face space based on eigenface scores (Leopold et al., 2001; Cowen et al., 2014; Chang and Tsao, 2017) as used here, increases the biological validity of results.

Our research broadens our understanding of how morphology and social decision-making can interact to structure interactions between species living in sympatry. In guenons, facial features like white nose spots are highly salient, attention-grabbing, and distinctive, and our combined results suggest that these may be important for species discrimination. Guenon behavioral repertoires, such as nose-to-nose touching observed in wild putty nosed monkeys (SW, personal observation) and red-tailed monkeys (Estes, 1991), further reflect the importance and biological relevance of these traits. Primates preferentially attend to facial information (Tsao et al., 2008; Kano and Tomonaga, 2009), making face patterns particularly suited to influencing behavior and decision-making in con- and heterospecifics. Interestingly, multiple other primate groups are also characterized by rapid speciation, high degrees of sympatry, and diverse facial appearances (e.g. callitrichids, Opazo et al., 2006; gibbons, Carbone et al., 2014), and comparative analyses have linked face pattern complexity to number of sympatric congenerics in primates (Santana et al., 2012; Santana et al., 2013), suggesting that selection for species discrimination may be a key driver of diversity in this group. Our work linking signal appearance to visual attention in a recent and diverse primate radiation highlights how such evolutionary processes can be important in generating animal phenotypes.

All project data and code are available in Zenodo repositories under the following DOIs. The guenon face image database is available in repository https://doi.org/10.5281/zenodo.3572780, and code implementing the occlude-reclassify analysis is available in repository https://doi.org/10.5281/zenodo.3574512. Experimental looking time data and original videos are available in repositories https://doi.org/10.5281/zenodo.3572791 (Cercopithecus mona) and https://doi.org/10.5281/zenodo.3572789 (Cercopithecus nictitans), and code implementing looking time analyses is available in repository https://doi.org/10.5281/zenodo.3574529.

1. 1. Sandra Winters
   2. William L. Allen
   3. James P. Higham

   (2019) Zenodo

   The structure of species discrimination signals across a primate radiation: guenon image database.

   https://doi.org/10.5281/zenodo.3572780
2. 1. Sandra Winters
   2. William L. Allen
   3. James P. Higham

   (2019) Zenodo

   The structure of species discrimination signals across a primate radiation: Cercopithecus mona looking time trials.

   https://doi.org/10.5281/zenodo.3572791
3. 1. Sandra Winters
   2. William L. Allen
   3. James P. Higham

   (2019) Zenodo

   The structure of species discrimination signals across a primate radiation: Cercopithecus nictitans looking time trials.

   https://doi.org/10.5281/zenodo.3572789

1. 1. Allen WL
   2. Stevens M
   3. Higham JP

   (2014) Character displacement of cercopithecini primate visual signals

   Nature Communications 5:4266.

   https://doi.org/10.1038/ncomms5266

   • Google Scholar
2. 1. Allen WL
   2. Higham JP

   (2015) Assessing the potential information content of multicomponent visual signals: a machine learning approach

   Proceedings of the Royal Society B: Biological Sciences 282:20142284.

   https://doi.org/10.1098/rspb.2014.2284

   • Google Scholar
3. Book

   1. Andersson M

   (1994)

   Sexual Selection

   Princeton: Princeton University Press.

   • Google Scholar
4. 1. Bates D
   2. Mächler M
   3. Bolker B
   4. Walker S

   (2015) Fitting linear Mixed-Effects models using lme4

   Journal of Statistical Software 67:1–48.

   https://doi.org/10.18637/jss.v067.i01

   • Google Scholar
5. 1. Bennett ATD
   2. Cuthill IC
   3. Norris KJ

   (1994) Sexual selection and the mismeasure of color

   The American Naturalist 144:848–860.

   https://doi.org/10.1086/285711

   • Google Scholar
6. 1. Bowmaker JK
   2. Astell S
   3. Hunt DM
   4. Mollon JD

   (1991)

   Photosensitive and photostable pigments in the retinae of old world monkeys

   The Journal of Experimental Biology 156:1–19.

   • PubMed
   • Google Scholar
7. 1. Bradley BJ
   2. Mundy NI

   (2008) The primate palette: the evolution of primate coloration

   Evolutionary Anthropology: Issues, News, and Reviews 17:97–111.

   https://doi.org/10.1002/evan.20164

   • Google Scholar
8. 1. Brown WL
   2. Wilson EO

   (1956) Character displacement

   Systematic Zoology 5:49.

   https://doi.org/10.2307/2411924

   • Google Scholar
9. 1. Carbone L
   2. Harris RA
   3. Gnerre S
   4. Veeramah KR
   5. Lorente-Galdos B
   6. Huddleston J
   7. Meyer TJ
   8. Herrero J
   9. Roos C
   10. Aken B
   11. Anaclerio F
   12. Archidiacono N
   13. Baker C
   14. Barrell D
   15. Batzer MA
   16. Beal K
   17. Blancher A
   18. Bohrson CL
   19. Brameier M
   20. Campbell MS
   21. Capozzi O
   22. Casola C
   23. Chiatante G
   24. Cree A
   25. Damert A
   26. de Jong PJ
   27. Dumas L
   28. Fernandez-Callejo M
   29. Flicek P
   30. Fuchs NV
   31. Gut I
   32. Gut M
   33. Hahn MW
   34. Hernandez-Rodriguez J
   35. Hillier LW
   36. Hubley R
   37. Ianc B
   38. Izsvák Z
   39. Jablonski NG
   40. Johnstone LM
   41. Karimpour-Fard A
   42. Konkel MK
   43. Kostka D
   44. Lazar NH
   45. Lee SL
   46. Lewis LR
   47. Liu Y
   48. Locke DP
   49. Mallick S
   50. Mendez FL
   51. Muffato M
   52. Nazareth LV
   53. Nevonen KA
   54. O'Bleness M
   55. Ochis C
   56. Odom DT
   57. Pollard KS
   58. Quilez J
   59. Reich D
   60. Rocchi M
   61. Schumann GG
   62. Searle S
   63. Sikela JM
   64. Skollar G
   65. Smit A
   66. Sonmez K
   67. ten Hallers B
   68. Terhune E
   69. Thomas GW
   70. Ullmer B
   71. Ventura M
   72. Walker JA
   73. Wall JD
   74. Walter L
   75. Ward MC
   76. Wheelan SJ
   77. Whelan CW
   78. White S
   79. Wilhelm LJ
   80. Woerner AE
   81. Yandell M
   82. Zhu B
   83. Hammer MF
   84. Marques-Bonet T
   85. Eichler EE
   86. Fulton L
   87. Fronick C
   88. Muzny DM
   89. Warren WC
   90. Worley KC
   91. Rogers J
   92. Wilson RK
   93. Gibbs RA

   (2014) Gibbon genome and the fast karyotype evolution of small apes

   Nature 513:195–201.

   https://doi.org/10.1038/nature13679

   • PubMed
   • Google Scholar
10. 1. Chang L
    2. Tsao DY

    (2017) The code for facial identity in the primate brain

    Cell 169:1013–1028.

    https://doi.org/10.1016/j.cell.2017.05.011

    • PubMed
    • Google Scholar
11. 1. Chiarelli B

    (1961)

    Ibridologia e sistematica in primati

    I. Raccolta Di Dati. Atti Assoc Genet Ital 6:213–220.

    • Google Scholar
12. Book

    1. Cords M

    (2004)

    When are there influxes in blue monkey groups?

    In: Glenn ME, Cords M, editors. The Guenons: Diversity and Adaptation in African Monkeys. Kluwer Academic/Plenum Publishers. pp. 189–201.

    • Google Scholar
13. Book

    1. Cords M

    (2012)

    The behavior, ecology, and social evolution of Cercopithecine monkeys

    In: Mitani JC, Palombit RA, Silk JB, editors. The Evolution of Primate Societies. Chicago: University of Chicago Press. pp. 91–112.

    • Google Scholar
14. 1. Cowen AS
    2. Chun MM
    3. Kuhl BA

    (2014) Neural portraits of perception: reconstructing face images from evoked brain activity

    NeuroImage 94:12–22.

    https://doi.org/10.1016/j.neuroimage.2014.03.018

    • PubMed
    • Google Scholar
15. Book

    1. Coyne JA
    2. Orr HA

    (2004)

    Speciation

    Sunderland, MA: Sinauer Associates, Inc.

    • Google Scholar
16. 1. Crete F
    2. Dolmiere T
    3. Ladret P
    4. Nicolas M

    (2007) The blur effect: perception and estimation with a new no-reference perceptual blur metricElectronic imaging

    International Society for Optics and Photonics 6492:64920I.

    https://doi.org/10.1117/12.702790

    • Google Scholar
17. 1. Dahl CD
    2. Wallraven C
    3. Bülthoff HH
    4. Logothetis NK

    (2009) Humans and macaques employ similar face-processing strategies

    Current Biology 19:509–513.

    https://doi.org/10.1016/j.cub.2009.01.061

    • PubMed
    • Google Scholar
18. Book

    1. Dawkins MS
    2. Guilford T

    (1997)

    Conspicuousness and diversity in animal signals

    In: Owings DH, Beecher MD, Thompson NS, editors. Communication. Boston, MA: Springer. pp. 55–75.

    • Google Scholar
19. 1. de Jong YA
    2. Butynski TM

    (2010) Three Sykes's Monkey Cercopithecus mitis × vervet monkey Chlorocebus pygerythrus hybrids in Kenya

    Primate Conservation 25:43–56.

    https://doi.org/10.1896/052.025.0109

    • Google Scholar
20. 1. Demaria C
    2. Thierry B

    (1988) Responses to animal stimulus photographs in stumptailed macaques (Macaca arctoides)

    Primates 29:237–244.

    https://doi.org/10.1007/BF02381125

    • Google Scholar
21. 1. Detwiler KM
    2. Burrell AS
    3. Jolly CJ

    (2005) Conservation implications of hybridization in african cercopithecine monkeys

    International Journal of Primatology 26:661–684.

    https://doi.org/10.1007/s10764-005-4372-0

    • Google Scholar
22. 1. Dubuc C
    2. Allen WL
    3. Maestripieri D
    4. Higham JP

    (2014) Is male rhesus macaque red color ornamentation attractive to females?

    Behavioral Ecology and Sociobiology 68:1215–1224.

    https://doi.org/10.1007/s00265-014-1732-9

    • PubMed
    • Google Scholar
23. 1. Dubuc C
    2. Allen WL
    3. Cascio J
    4. Lee DS
    5. Maestripieri D
    6. Petersdorf M
    7. Winters S
    8. Higham JP

    (2016) Who cares? experimental attention biases provide new insights into a mammalian sexual signal

    Behavioral Ecology 27:68–74.

    https://doi.org/10.1093/beheco/arv117

    • PubMed
    • Google Scholar
24. 1. Dufour V
    2. Pascalis O
    3. Petit O

    (2006) Face processing limitation to own species in primates: a comparative study in Brown capuchins, tonkean macaques and humans

    Behavioural Processes 73:107–113.

    https://doi.org/10.1016/j.beproc.2006.04.006

    • PubMed
    • Google Scholar
25. 1. Endler JA

    (1990) On the measurement and classification of colour in studies of animal colour patterns

    Biological Journal of the Linnean Society 41:315–352.

    https://doi.org/10.1111/j.1095-8312.1990.tb00839.x

    • Google Scholar
26. Book

    1. Estes R

    (1991)

    The Behavior Guide to African Mammals: Including Hoofed Mammals, Carnivores, Primates

    Los Angeles: University of California Press.

    • Google Scholar
27. 1. Fleishman LJ
    2. Endler JA

    (2000) Some comments on visual perception and the use of video playback in animal behavior studies

    Acta Ethologica 3:15–27.

    https://doi.org/10.1007/s102110000025

    • Google Scholar
28. Book

    1. Fox J
    2. Weisberg S

    (2011)

    An {R} Companion to Applied Regression (Second Edition)

    Thousand Oaks: Sage.

    • Google Scholar
29. 1. Fujita K

    (1987) Species recognition by five macaque monkeys

    Primates 28:353–366.

    https://doi.org/10.1007/BF02381018

    • Google Scholar
30. Book

    1. Gelman A
    2. Hill J

    (2007)

    Data Analysis Using Regression and Multilevel Hierarchical Models

    New York: Cambridge University Press.

    • Google Scholar
31. 1. Gordon NM
    2. Ralph MZ
    3. Stratman KD

    (2017) Rapid character displacement of different call parameters in closely related treefrogs (Hyla cinerea and H. gratiosa)

    Behavioral Ecology and Sociobiology 71:112.

    https://doi.org/10.1007/s00265-017-2341-1

    • Google Scholar
32. 1. Gröning J
    2. Hochkirch A

    (2008) Reproductive interference between animal species

    The Quarterly Review of Biology 83:257–282.

    https://doi.org/10.1086/590510

    • PubMed
    • Google Scholar
33. Book

    1. Groves C

    (2005)

    Order Primates

    In: Wilson D, Reader D, editors. Mammals Species of the World: A Taxonomic and Geographic Reference. Baltimore: Johns Hopkins University Press. pp. 111–184.

    • Google Scholar
34. 1. Grubb P
    2. Butynski TM
    3. Oates JF
    4. Bearder SK
    5. Disotell TR
    6. Groves CP
    7. Struhsaker TT

    (2003) Assessment of the diversity of african primates

    International Journal of Primatology 24:1301–1357.

    https://doi.org/10.1023/B:IJOP.0000005994.86792.b9

    • Google Scholar
35. 1. Guschanski K
    2. Krause J
    3. Sawyer S
    4. Valente LM
    5. Bailey S
    6. Finstermeier K
    7. Sabin R
    8. Gilissen E
    9. Sonet G
    10. Nagy ZT
    11. Lenglet G
    12. Mayer F
    13. Savolainen V

    (2013) Next-generation museomics disentangles one of the largest primate radiations

    Systematic Biology 62:539–554.

    https://doi.org/10.1093/sysbio/syt018

    • PubMed
    • Google Scholar
36. 1. Hebets EA
    2. Papaj DR

    (2005) Complex signal function: developing a framework of testable hypotheses

    Behavioral Ecology and Sociobiology 57:197–214.

    https://doi.org/10.1007/s00265-004-0865-7

    • Google Scholar
37. 1. Higgie M
    2. Chenoweth S
    3. Blows MW

    (2000) Natural selection and the reinforcement of mate recognition

    Science 290:519–521.

    https://doi.org/10.1126/science.290.5491.519

    • PubMed
    • Google Scholar
38. 1. Higham JP
    2. Hughes KD
    3. Brent LJN
    4. Dubuc C
    5. Engelhardt A
    6. Heistermann M
    7. Maestriperi D
    8. Santos LR
    9. Stevens M

    (2011) Familiarity affects the assessment of female facial signals of fertility by free-ranging male rhesus macaques

    Proceedings of the Royal Society B: Biological Sciences 278:3452–3458.

    https://doi.org/10.1098/rspb.2011.0052

    • Google Scholar
39. 1. Higham JP
    2. Hebets EA

    (2013) An introduction to multimodal communication

    Behavioral Ecology and Sociobiology 67:1381–1388.

    https://doi.org/10.1007/s00265-013-1590-x

    • Google Scholar
40. Book

    1. Hosmer DW
    2. Lemeshow S
    3. Sturdivant RX

    (2013)

    Applied Logistic Regression (Third edition)

    Hoboken, New Jersey: Wiley series in probability and statistics.

    • Google Scholar
41. Book

    1. Jaffe KE
    2. Isbell LA

    (2011)

    The guenons: polyspecific associations in socioecological perspective

    In: Campbell CJ, Fuentes A, MacKinnon KC, Bearder SK, Stumpf RM, editors. Primates in Perspective. New York, NY: Oxford University Press. pp. 277–300.

    • Google Scholar
42. 1. Kameda Y
    2. Kawakita A
    3. Kato M

    (2009) Reproductive character displacement in genital morphology in Satsuma land snails

    The American Naturalist 173:689–697.

    https://doi.org/10.1086/597607

    • PubMed
    • Google Scholar
43. 1. Kano F
    2. Tomonaga M

    (2009) How chimpanzees look at pictures: a comparative eye-tracking study

    Proceedings of the Royal Society B: Biological Sciences 276:1949–1955.

    https://doi.org/10.1098/rspb.2008.1811

    • Google Scholar
44. 1. Kingdon JS

    (1980) The role of visual signals and face patterns in african forest monkeys (guenons) of the genus Cercopithecus

    The Transactions of the Zoological Society of London 35:425–475.

    https://doi.org/10.1111/j.1096-3642.1980.tb00062.x

    • Google Scholar
45. Book

    1. Kingdon J

    (1988)

    What are face patterns and do they contribute to reproductive isolation in guenons?

    In: Gautier-Hion A, Bourlière F, Gauthier JP, Kingdon J, editors. A Primate Radiation: Evolutionary Biology of the African Guenons. Cambridge, UK: Cambridge University Press. pp. 227–245.

    • Google Scholar
46. Book

    1. Kingdon J

    (1997)

    The Kingdon Guide to African Mammals

    London: Academic Press.

    • Google Scholar
47. 1. Leopold DA
    2. O'Toole AJ
    3. Vetter T
    4. Blanz V

    (2001) Prototype-referenced shape encoding revealed by high-level aftereffects

    Nature Neuroscience 4:89–94.

    https://doi.org/10.1038/82947

    • PubMed
    • Google Scholar
48. 1. Lo Bianco S
    2. Masters JC
    3. Sineo L

    (2017) The evolution of the cercopithecini: a (post)modern synthesis

    Evolutionary Anthropology: Issues, News, and Reviews 26:336–349.

    https://doi.org/10.1002/evan.21567

    • PubMed
    • Google Scholar
49. 1. Lukhtanov VA
    2. Kandul NP
    3. Plotkin JB
    4. Dantchenko AV
    5. Haig D
    6. Pierce NE

    (2005) Reinforcement of pre-zygotic isolation and karyotype evolution in Agrodiaetus butterflies

    Nature 436:385–389.

    https://doi.org/10.1038/nature03704

    • PubMed
    • Google Scholar
50. 1. Mallet J

    (2005) Hybridization as an invasion of the genome

    Trends in Ecology & Evolution 20:229–237.

    https://doi.org/10.1016/j.tree.2005.02.010

    • PubMed
    • Google Scholar
51. 1. McNaught MK
    2. Owens IPF

    (2002) Interspecific variation in plumage colour among birds: species recognition or light environment?

    Journal of Evolutionary Biology 15:505–514.

    https://doi.org/10.1046/j.1420-9101.2002.00431.x

    • Google Scholar
52. 1. Méary D
    2. Li Z
    3. Li W
    4. Guo K
    5. Pascalis O

    (2014) Seeing two faces together: preference formation in humans and rhesus macaques

    Animal Cognition 17:1107–1119.

    https://doi.org/10.1007/s10071-014-0742-3

    • PubMed
    • Google Scholar
53. 1. Mendelson TC
    2. Shaw KL

    (2012) The (mis)concept of species recognition

    Trends in Ecology & Evolution 27:421–427.

    https://doi.org/10.1016/j.tree.2012.04.001

    • PubMed
    • Google Scholar
54. 1. Opazo JC
    2. Wildman DE
    3. Prychitko T
    4. Johnson RM
    5. Goodman M

    (2006) Phylogenetic relationships and divergence times among new world monkeys (Platyrrhini, primates)

    Molecular Phylogenetics and Evolution 40:274–280.

    https://doi.org/10.1016/j.ympev.2005.11.015

    • PubMed
    • Google Scholar
55. 1. Partan S
    2. Marler P

    (1999) Communication goes multimodal

    Science 283:1272–1273.

    https://doi.org/10.1126/science.283.5406.1272

    • PubMed
    • Google Scholar
56. 1. Patricelli GL
    2. Hebets EA

    (2016) New dimensions in animal communication: the case for complexity

    Current Opinion in Behavioral Sciences 12:80–89.

    https://doi.org/10.1016/j.cobeha.2016.09.011

    • Google Scholar
57. 1. Perera TA
    2. Collins J

    (2015) Image based species recognition system

    IEEE pp. 195–199.

    https://doi.org/10.1109/ICSensT.2015.7438391

    • Google Scholar
58. 1. Pfennig KS

    (2016) Reinforcement as an initiator of population divergence and speciation

    Current Zoology 62:145–154.

    https://doi.org/10.1093/cz/zow033

    • PubMed
    • Google Scholar
59. 1. Pfennig KS
    2. Pfennig DW

    (2009) Character displacement: ecological and reproductive responses to a common evolutionary problem

    The Quarterly Review of Biology 84:253–276.

    https://doi.org/10.1086/605079

    • PubMed
    • Google Scholar
60. 1. Pfennig DW
    2. Pfennig KS

    (2010) Character displacement and the origins of diversity

    The American Naturalist 176:S26–S44.

    https://doi.org/10.1086/657056

    • PubMed
    • Google Scholar
61. 1. Pfennig KS
    2. Rice AM

    (2014) Reinforcement generates reproductive isolation between neighbouring conspecific populations of spadefoot toads

    Proceedings of the Royal Society B: Biological Sciences 281:20140949.

    https://doi.org/10.1098/rspb.2014.0949

    • Google Scholar
62. Software

    1. R Development Core Team

    (2018) R: a language and environment for statistical computing, version v3.5.2

    R Foundation for Statistical Computing, Vienna, Austria.

    https://www.r-project.org/
63. 1. Rakotonirina H
    2. Kappeler PM
    3. Fichtel C

    (2018) The role of facial pattern variation for species recognition in red-fronted Lemurs (Eulemur rufifrons)

    BMC Evolutionary Biology 18:19.

    https://doi.org/10.1186/s12862-018-1126-0

    • PubMed
    • Google Scholar
64. 1. Santana SE
    2. Lynch Alfaro J
    3. Alfaro ME

    (2012) Adaptive evolution of facial colour patterns in neotropical primates

    Proceedings of the Royal Society B: Biological Sciences 279:2204–2211.

    https://doi.org/10.1098/rspb.2011.2326

    • Google Scholar
65. 1. Santana SE
    2. Alfaro JL
    3. Noonan A
    4. Alfaro ME

    (2013) Adaptive response to sociality and ecology drives the diversification of facial colour patterns in catarrhines

    Nature Communications 4:e2765.

    https://doi.org/10.1038/ncomms3765

    • Google Scholar
66. Conference

    1. Shiau Y-H
    2. Lin S-I
    3. Chen Y-H
    4. Lo S-W CC-C

    (2012)

    Fish observation, detection recognition and verification in the real worldProceedings of the International Conference on Image Processing, Computer Vision, and Pattern Recognition (IPCV)

    The Steering Committee of the World Congress in Computer Science, Computer Engineering And Applied Computing (WorldComp).

    • Google Scholar
67. 1. Shimojo S
    2. Simion C
    3. Shimojo E
    4. Scheier C

    (2003) Gaze Bias both reflects and influences preference

    Nature Neuroscience 6:1317–1322.

    https://doi.org/10.1038/nn1150

    • PubMed
    • Google Scholar
68. 1. Stanger-Hall KF
    2. Lloyd JE

    (2015) Flash signal evolution in Photinus fireflies: character displacement and signal exploitation in a visual communication system

    Evolution 69:666–682.

    https://doi.org/10.1111/evo.12606

    • Google Scholar
69. 1. Stevens M
    2. Parraga CA
    3. Cuthill IC
    4. Partridge JC
    5. Troscianko TS

    (2007) Using digital photography to study animal coloration

    Biological Journal of the Linnean Society 90:211–237.

    https://doi.org/10.1111/j.1095-8312.2007.00725.x

    • Google Scholar
70. 1. Stuart YE
    2. Inkpen SA
    3. Hopkins R
    4. Bolnick DI

    (2017) Character displacement is a pattern: so, what causes it?

    Biological Journal of the Linnean Society 121:711–715.

    https://doi.org/10.1093/biolinnean/blx013

    • Google Scholar
71. Software

    1. The GIMP Development Team

    (2014)

    GNU Image Manipulation Program, version  Version 2.8.14

    GNU Image Manipulation Program.
72. Software

    1. The MathWorks, Inc

    (2019)

    MATLAB and Image Processing Toolbox release 2019a

    The MathWorks, Inc, Natick, USA.
73. 1. Tosi AJ
    2. Detwiler KM
    3. Disotell TR

    (2005) X-chromosomal window into the evolutionary history of the guenons (Primates: cercopithecini)

    Molecular Phylogenetics and Evolution 36:58–66.

    https://doi.org/10.1016/j.ympev.2005.01.009

    • PubMed
    • Google Scholar
74. 1. Tsao DY
    2. Schweers N
    3. Moeller S
    4. Freiwald WA

    (2008) Patches of face-selective cortex in the macaque frontal lobe

    Nature Neuroscience 11:877–879.

    https://doi.org/10.1038/nn.2158

    • Google Scholar
75. 1. Turk M
    2. Pentland A

    (1991) Eigenfaces for recognition

    Journal of Cognitive Neuroscience 3:71–86.

    https://doi.org/10.1162/jocn.1991.3.1.71

    • PubMed
    • Google Scholar
76. 1. Weeks PJD
    2. O'Neill MA
    3. Gaston KJ
    4. Gauld ID

    (1999) Species–identification of wasps using principal component associative memories

    Image and Vision Computing 17:861–866.

    https://doi.org/10.1016/S0262-8856(98)00161-9

    • Google Scholar
77. 1. Westland S
    2. Ripamonti C

    (2004)

    Computational Colour Science Using MATLAB

    Computational Colour Science Using MATLAB, Hoboken, NJ.

    • Google Scholar
78. 1. Winters S
    2. Dubuc C
    3. Higham JP

    (2015) Perspectives: the looking time experimental paradigm in studies of animal visual perception and cognition

    Ethology 121:625–640.

    https://doi.org/10.1111/eth.12378

    • Google Scholar
79. Book

    1. Zeiler MD
    2. Fergus R

    (2014)

    Visualizing and understanding convolutional networks

    In: Fleet D, Pajdla T, Schiele B, Tuytelaars T, editors. Computer Vision – ECCV 2014. Springer International Publishing. pp. 818–833.

    • Google Scholar
80. Book

    1. Zuur AF

    (2009)

    Mixed Effects Models and Extensions in Ecology with R

    New York; London: Springer.

    • Google Scholar

Author details

1. Sandra Winters

   1. Department of Anthropology, New York University, New York, United States
   2. New York Consortium in Evolutionary Primatology, New York, United States

   Present address

   School of Biological Sciences, University of Bristol, Bristol, United Kingdom

   Contribution

   Conceptualization, Data curation, Software, Formal analysis, Funding acquisition, Validation, Investigation, Visualization, Methodology, Project administration

   For correspondence

   sandra.winters@bristol.ac.uk

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-9561-5950

2. William L Allen

   1. Department of Anthropology, New York University, New York, United States
   2. New York Consortium in Evolutionary Primatology, New York, United States
   3. Department of Biosciences, Swansea University, Wales, United Kingdom

   Contribution

   Resources, Methodology

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-2654-0438

3. James P Higham

   1. Department of Anthropology, New York University, New York, United States
   2. New York Consortium in Evolutionary Primatology, New York, United States

   Contribution

   Conceptualization, Resources, Supervision, Funding acquisition, Methodology, Project administration

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-1133-2030

• 1,307

  views

• 110

  downloads

• 22

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.