Adapting the facial action coding system for chimpanzees (Pan troglodytes) to bonobos (Pan paniscus): the ChimpFACS extension for bonobos

AU1+2—Brow Raiser

The raising of the browridge is brought about by the contraction of the frontalis muscle in chimpanzees, which is also the case here. The description of appearance changes for AU1+2 in the ChimpFACS can be applied to bonobos without major modifications: the brows are pulled up in their entirety, the wrinkles on the forehead deepen and the skin under the browridge becomes visible. The human FACS AU1—Inner Brow Raiser and AU2—Outer Brow Raiser do not seem to act independently in chimpanzees nor bonobos. In all of the AU1+2 observed in bonobos, the inner and outer sections of the brow were raised together (e.g., Figs. S5–S6, Videos S1–S6).

AU1+2 may be more difficult to detect in some bonobos than in chimpanzees, as the browridge is less prominent, so that less skin will be moving over it. In addition, like in many other AUs, the black skin colour of bonobos may render it more difficult to see the changes in lines and wrinkles.

AU41—Glabella Lowerer

Humans can produce AU4—Brow Lowerer by the contraction of three different muscles: procerus, depressor supercilii, and corrugator supercilii. The procerus pulls the medial end of the eyebrow downward, the depressor supercilii pulls the eyebrow and the skin above the orbit downward, and the corrugator supercilii pulls the eyebrows towards the midline and downwards. No such independent movement was described for chimpanzees in the ChimpFACS, and it was not seen either in bonobos in the initial ChimpFACS adaptation study (Kuchenbuch, 2010). In both works, AU4 was only observed with AU9—Nose Wrinkler. However, this single movement of lowering the brows has been described in the literature in bonobos (often with an emotional label). De Waal (1988) and Gaspar (2006) both use the word “frown” for certain appearances of bonobo faces. One photograph in de Waal’s article (Fig. 1A in De Waal, 1988) shows an individual with the inner brow area appearing to be drawn inward and downward, but there is no neutral picture to compare this appearance to, which makes it difficult to identify an AU4 with certainty. The face of one of the individuals observed in Kuchenbuch’s (2010) study has an appearance similar to human frowning in its neutral state, but no individual showed a singular movement in accord with the minimum criterion for an AU4, i.e., downwards movement of the brow region. In the non-FACS literature, this “frowning” is generally described as a medial and downward movement of the inner brow area, just like it is observed in humans. Using the human FACS approach, this would likely include at the very least the lowering of the medial part of the browridge caused by the procerus muscle, i.e., an AU41—Glabella Lowerer. However, as explained above, caution is needed to not code a false indicator due to permanent features of an individual’s face in a neutral state. AU41—Glabella Lowerer has also been identified in another primate taxa, the Macaca genus and thus included in the MaqFACS and its extensions (Parr et al., 2010; Julle-Danière et al., 2015; Clark et al., 2020; Correia-Caeiro, Holmes & Miyabe-Nishiwaki, 2021).

In the larger dataset with bonobos used in the current work, we clearly identified several examples of an independent movement in which the browridge was lowered. However, it is unclear if corrugation is widely present (only observed in one individual once). Hence, despite the three muscles producing AU4 in humans being present in bonobos (Diogo, Molnar & Wood, 2017; Diogo et al., 2017), and in bonobos these muscles producing some of the same appearance changes seen in humans, without corrugation present as one of the appearance changes of this movement, AU4—Brow Lowerer cannot be coded. Corrugation of the brows, i.e., brows coming closer together and creating wrinkles on the glabella, has not yet been found in any other AnimalFACS, and thus seems to be unique to humans. Therefore, in bonobos, AU41—Glabella Lowerer is coded for downwards browridge movements, which we describe next with the corresponding sections A–D (e.g., Videos S7–S10).

The glabella area (i.e., area between the brows) differs slightly between humans and bonobos. While in humans the glabella is a flat area of smooth skin between the brows, in bonobos the glabella area usually has permanent wrinkles, and may sit deeper than the rest of the browridge or the salient arches of the browridge. In humans, the glabella contrasts with the surrounding areas due to being a (usually) hairless area between two strips of hair, while in bonobos either the whole browridge may be covered in hair or may be fully naked (see Fig. S4 for examples of variation in this feature), which may make identifying AU41 more challenging in bonobos.

A. Proposed muscular basis: procerus.

B. Appearance changes:

1. The browridge moves downwards, with either the browridge moving down or the skin of the browridge sliding down and inwards.

2. The eye aperture may be narrowed.

3. In more intense movements, the underbrow region may become less visible, the eye cover fold may disappear from view, and the root of the nose may also be covered.

4. If hair is present above or on the browridge, it will move downwards.

5. In very intense movements, the root of the nose might be completely covered with the glabella.

6. In some individuals, wrinkles may appear or may be deepened on the glabella and root of the nose.

C. Minimum criteria: downwards movement of the browridge, glabella, or brow arches.

D. Subtle differences between AUs: even though AU1+2 and AU41 act in opposite directions, in humans they can be coded simultaneously, impacting each other’s appearance changes and creating new appearance changes. In contrast, in bonobos they are mutually exclusive, i.e., they cannot be coded simultaneously, and hence do not share any appearance changes. In bonobos, either an AU1+2 is acting whenever the browridge goes upwards from neutral, or an AU41 is acting whenever the browridge goes downwards from neutral. However, these two movements are sometimes observed in succession, in which AU1+2 is immediately followed by AU41 (or vice-versa), sometimes without a clear temporal return to neutral (e.g., Videos S11–S13). In addition, the return of AU41 to neutral might be difficult to distinguish from a weak AU1+2, and the return of AU1+2 to neutral might be confused with a weak AU41. However, the release of AU1+2 should not be coded as AU41, and vice-versa. To define when to code one or the other, comparison with the neutral browridge for each individual may be necessary, as well as frame-by-frame analysis of the succession of movements (Fig. S7). In still frames, weak occurrences of this movement might be hard to code in absence of a neutral picture of the same individual.

AU5—Upper Lid Raiser

The action of this AU opens the eyes widely by raising the upper lid beyond the neutral open position. The raising of the upper lid is brought about by the contraction of the levator palpebrae superioris muscle in humans and chimpanzees, but this muscle was not described for bonobos as the extra-ocular musculature was not examined in the recent dissections (Diogo et al., 2017). In some species, the orbicularis oculi muscle may also act to further increase the opening of the eye.

In any case, this AU was reported to be present in chimpanzees and bonobos, with the same basic appearance in both species (Kuchenbuch, 2010). The eye aperture is widened, exposing more of the upper part of the eyeball, while part of the upper eyelid itself is concealed under the browridge. In some bonobos, as the browridge may be smaller, the eyes are less overshadowed than in chimpanzees, so that the upper eyelid may be more visible. This makes this AU slightly easier to discern in individuals with smaller browridges. If AU1+2 is acting, this may also expose more of the eye region and facilitate the identification of AU5 (Fig. S8). The influence of head angle and eye movements on the identification of AU5 is as important in bonobos as in chimpanzees; changes in head angle can lead the observer to erroneously assume that the upper eyelid has been raised, and the same can happen if the individual is looking up. Figure S9 illustrates these effects, showing a combination of AU5, looking up, and an upward head movement.

However, both in the ChimpFACS and the initial work done in bonobos, AU5 was not observed in isolation, and all examples are accompanied by either AU41 or eye/head movements. Similarly, we did not identify this movement in our current sample as an independent movement (e.g., temporally distinct from other AUs). Hence, this is another AU that may be hard to code depending on individuals or video conditions.

AU6—Cheek Raiser and AU7—Lid Tightener

In chimpanzees, AU6—Cheek Raiser is produced by the contraction of the orbicularis oculi (pars orbitalis) muscle, and it has the same muscular basis in bonobos. This muscle action pulls the skin from around the eyes towards the centre of the eye. Narrowing of the eye aperture, as described for chimpanzees, is also seen in bonobos. The medial section of the browridge can be seen to be pulled downwards, which also happens in stronger actions in chimpanzees. In some movements, beyond the lowering of the brows, the skin of the browridge and frontal region is seen sliding over the browridge and downwards towards the eyes (e.g., Videos S14–S15). AU6 can be easily identified in bonobos from these cues alone. However, it has less conspicuous appearance changes on the corners of the eyes than in chimpanzees: certain wrinkles lying laterally of the eye region seem to be shallower in bonobos than in chimpanzees, which makes it harder to see any deepening of these or increase of bagging beneath the eyes, as described in the ChimpFACS. The black skin colour may also make these appearance changes harder to see. The possibility to use wrinkles and bagging as hints for the action of AU6 is therefore more limited than in chimpanzees. Like in chimpanzees, this action in bonobos seems to be more spread around the whole eye area, in a concentric movement (e.g., Video S16).

It is important to note that we could not find an example of AU6 in isolation in bonobos, but only in conjunction with other AUs, such as AU45/43, AU9 (see sections below for description of these AUs, Fig. S10), or AU41. This was the case for both the initial work in applying FACS to bonobos (Kuchenbuch, 2010) and in the current larger sample. However, there is a slight temporal difference in AU6 and the other AUs, and thus we consider it as an independent AU for bonobos, similar to what was determined for chimpanzees. In humans, likely due to anatomical differences (e.g., fat deposits on the zygomatic bone that form the human-like cheeks), AU6 can be coded as an independent action.

In bonobos, we also detected another movement that often accompanies AU6: AU7 - Lid Tightener, in which the lower eyelid is raised or bulged (Fig. S11). In humans AU7 can be coded as an independent AU and is produced by the other portion of the orbicularis oculi (pars palpebralis) muscle. However, in chimpanzees this action was not described, as the authors thought several factors could influence the reliability of detecting this movement (Vick et al., 2007). In bonobos, AU7 may likewise be hard to detect reliably due to the difficulty in clearly visualising the eyelid, and was only observed twice independently and as a very subtle movement in the analysis of our whole video dataset (Videos S17–S18). Hence, we recommend to code this movement only in optimal conditions of visibility of the eye area.

As both AU6—Cheek Raiser and AU7—Lid Tightener seem to be slightly different in bonobos compared to chimpanzees, we include the appearance changes for both of these below.

A. Proposed muscular basis for AU6—Cheek Raiser: orbicularis oculi (pars orbitalis).

B. Appearance changes for AU6—Cheek Raiser:

1. The skin around the eye is pulled towards the centre of the eye, including skin from browridge, IOT (infra-orbital triangle, Figs. S1–S2), nose and temples.

2. Skin wrinkles around the eye deepen, and new wrinkles may form around the eye.

3. The skin of the IOT is pulled towards the eye, but may not decrease in size.

4. Deepens the infraorbital furrow.

5. Even if eye closure is not present, it usually narrows eye aperture.

6. It was only observed in conjunction with other AUs, such as AU41, AU43/45, and AU9 (see below for further AUs descriptions).

7. When eye closure or AU7 are present, it can make the eyelids appear compressed and bulging.

C. Minimum criteria for AU6—Cheek Raiser: concentric movement of skin globally around the eyes being pulled inwards toward the centre of the eye.

D. Subtle differences between AUs for AU6—Cheek Raiser: as AU6 was not observed alone, this makes it harder to distinguish from other AUs. Nonetheless, there has to be at least a temporal distinction with other AUs appearance changes. AU6 may be confused with AU41, AU9, or AU43/45. However, with AU41 and AU9, movement will not be global, but with AU41 will be seen coming from the browridge downwards only, and with AU9 there will be only movement coming from the nose and IOT area upwards. With AU43/45 alone, there is some movement around the eye, but its area is less extensive and less globalised.

A. Proposed muscular basis for AU7—Lid Tightener: orbicularis oculi (pars palpebralis).

B. Appearance changes for AU7—Lid Tightener:

1. The eyelid is tightened and moves upwards covering more of the eyeball. There might be some movement also towards the corner of the eye.

2. It was only observed in the lower eyelid.

3. It narrows eye aperture.

4. It may change shape slightly from a curved U line to a more straight or arched line.

5. The skin right underneath below the eyelid may move upwards, but not beyond the IOT.

6. Lower eyelid furrow may become more apparent.

C. Minimum criteria for AU7—Lid Tightener: lower eyelid is pushed upwards covering more of the eyeball.

D. Subtle differences between AUs for AU7—Lid Tightener: AU7 may be confused with AU6, as in AU7 there is also movement in the IOT and eyelid. However, in AU7, this movement is not concentric, and more just a push of the eyelid upwards or towards the inner corner of the eyes, whilst in AU6 the movement is concentric towards the centre of the eye. Furthermore, we observed AU6 in both eyelids, whilst we only observed AU7 in the lower eyelid.

AU9—Nose Wrinkler

In humans, the levator labii superioris alaeque nasi muscle wrinkles the nose, producing AU9—Nose Wrinkler, but in chimpanzees, this AU may recruit an additional muscle, the procerus muscle. Like in humans, in chimpanzees, this AU pulls the nostrils upwards and medially, causing wrinkling at the root of the nose, while the medial portion of the browridge is pulled downwards. The lateral portion of the nasolabial furrow is deepened in chimpanzees.

Both muscles are present in bonobos (Miller, 1952; Diogo, Molnar & Wood, 2017; Diogo et al., 2017). The AU9 has similar appearance changes in this species (Fig. S12, Videos S20–S22), but because the bonobo nose is broader than in chimpanzees, the medial movement of the nostrils is less pronounced. The deepening of the nasolabial furrow described in chimpanzees is more difficult to discern in bonobos, and the same can be said of the wrinkling at the root of the nose. Therefore, this AU is easier to recognise in bonobos by the slight upward movement of the nose, which can be seen more clearly in profile than in frontal view. In addition, it seems that smaller movements produce more appearance changes around the nostril area, pulling the nostrils upwards, whilst in more intense movements the nostril, side of nose, upper lip, and glabella may have visible changes.

AU10—Upper Lip Raiser

AU10—Upper Lip Raiser refers to the upper lip being pulled upwards. This action may reveal parts of the upper teeth and gums. The muscle responsible for this movement in chimpanzees is the levator labii superioris muscle. This AU looks very similar in bonobos; a noticeable shortening of the distance between the nose and the upper lip is the best cue for recognition in bonobos as it is in chimpanzees. However, the upper lip seems to thicken less than in chimpanzees, although this may be due to many occurrences of AU10 being accompanied by AU12—Lip Corner Puller, which stretches the lips horizontally. The lighter lip colouration in some bonobos creates a higher contrast area, which may facilitate the identification of this movement (e.g., Figs. S13–S15, Videos S23–S26).

AU12—Lip Corner Puller

This AU pulls the corners of the lips upwards and laterally towards the ears (e.g., Figs. S16–S17, Videos S27–S29). The underlying muscle action is the contraction of the zygomaticus major muscle in chimpanzees. The mouth is elongated in the horizontal plane; in chimpanzees, the stretching of the top lip caused by this AU makes the vertical wrinkles between mouth and nose less visible. This appearance change can also be seen in bonobos. However, the appearance change that describes the deepening of the semi-circular mouth corner furrows into wrinkles mentioned in the ChimpFACS is not as straightforward in bonobos to detect. In many bonobos, the hair comes down and medially further than in many chimpanzees. The area where wrinkling is said to occur in chimpanzees when AU12 is acting is often covered by hair in bonobos, so that wrinkling in the mouth corner can then not be used as a reliable cue to recognise the action of AU12.

However, the bright lips make the elongation of the mouth into a crescent shape easier to see in bonobos, which compensates for when the lip corners are not visible. As in chimpanzees, a slight upward curving of the mouth corners is often present in the neutral face (AU0) of bonobos (Fig. S18). The degree of this curvature is individually different. The presence of AU12 alone is therefore often hard to discern. Thus, it is especially important for the correct coding of AU12 to consider the AU0 of the individual or for the actual movement of the lip corners to be detected.

AU16—Lower Lip Depressor

This AU pulls the lower lip towards the mental region. The inner lip area, the lower teeth and gums may be revealed, the lips are always parted (AU25), and the lower lip may bulge (but not fall anteriorly). The responsible muscle is the depressor labii inferioris muscle in chimpanzees, and this muscle is also present in bonobos (Miller, 1952; Diogo, Molnar & Wood, 2017; Diogo et al., 2017). We found no differences in AU16 between bonobos and chimpanzees, hence the ChimpFACS appearance changes can be used without any modification for bonobos (e.g., Figs. S19–S20, Videos S30–S32).

AU160—Lower Lip Relax

The Lower Lip Relax - AU160 describes the strong relaxation of the lower lip to the point that it falls forward and hangs loose. This AU is not observed in humans, due to the small size and low flexibility of the human lower lip. It was however described for chimpanzees (Vick et al., 2007), orangutans (Correia-Caeiro et al., 2013), and gorillas (Correia-Caeiro et al., 2025), since these species have a very large, thick, and highly mobile lower lip. In bonobos, AU160 is also observed, and it presents similar appearance changes to the other apes (e.g., Fig. S21, Videos S33–S35).

AU17—Chin Raiser

In humans, the main appearance change of this AU is an upward movement of the chin and the lower lip, causing the lower lip to protrude. In humans and chimpanzees, it is caused by the contraction of the mentalis muscle. This muscle is not mentioned in Miller’s work on bonobo facial musculature (Miller, 1952), but it was identified in the more recent dissections (Diogo et al., 2017). In bonobos, there is a very obvious presence of AU17 illustrated by the several examples included in Kuchenbuch’s work (2010) and here included. The description of this AU in chimpanzees can be applied to bonobos without major modifications (e.g., Figs. S22–S23, Videos S36–S39). This AU is also clearly visible in bonobos because of the contrasting bright lip colour, which makes the lip and the mental area appearance changes much more visible than in chimpanzees.

AU18—Lip Pucker

The action of this AU in humans pushes the lips slightly forwards, the mouth opening is de-elongated becoming smaller and rounder, and wrinkles appear on the upper and lower lip (Ekman, Friesen & Hager, 2002a). AU18 is produced by the incisivii labii superioris and inferioris muscles to produce the typical “kiss/pucker” mouth shape in humans (Ekman, Friesen & Hager, 2002a; Hur, 2018). In the ChimpFACS, the presence of these muscles and the AU are noted as unclear. The published dissections do not seem to agree regarding the nomenclature, description, and inclusion of these muscles in NHP (Diogo et al., 2017), but it is unclear if this is because it was not dissected or it could not be found, or due to some technical issues. However, a recent work targeting these muscles was published with one chimpanzee sampled (Iwanaga et al., 2021), in which this muscle was clearly identified, at least in the lower lip, but with a different proposed function than the one proposed in the human FACS (Ekman, Friesen & Hager, 2002a). Hence, more studies on these muscles are needed to clarify their presence and function in human and NHP.

Due to the yet tentative muscular basis in bonobos, Kuchenbuch (2010) included only a suggestion for an AU18, consisting of several appearance changes: a clearly visible elongation of the lips with a pointed centre, but with the lips pulled flat in the vertical plane, and without outwards flaring. These appearance changes, albeit with slight differences, were suggested to be like the human FACS AU18 - Lip Pucker (Ekman, Friesen & Hager, 2002a). However, when examining these picture and video examples from the work of Kuchenbuch (2010) (Fig. S24), there seems to be more similar appearance changes from a weak AU22—Lip Funneler (see below for description), without the extensive outwards flaring of the lips (but still there is a slight flare in both videos), than an AU18, according to the human FACS. However, both videos also contain an AU18 (see below for still frames of AU18). The AU18 in the human FACS does not require lip protrusion (just mouth de-elongation and pursing of the lips), whilst the AU22 in the human FACS always requires the lips to be projected forwards (Ekman, Friesen & Hager, 2002a). The human FACS concedes that AU22 can happen without AU25, if AU17 is acting to push the lower lip upwards towards the upper lip and thus keeping the lips together. Furthermore, in other ape species, such as orangutans (Correia-Caeiro et al., 2013), similar movements to these ones observed in Kuchenbuch’s work were classified as AU22 in the OrangFACS, in which the projected lips can take two forms: funnelled or flattened. In the OrangFACS, AU18 was also included, but with the appearance changes of pursing the lips medially and the appearance of wrinkles in both lips, with none or minimal protrusion of the lips. As such, here we complement both movements (AU18 and AU22) with further video and picture examples, and opted for following the appearance changes of ChimpFACS and OrangFACS for AU18 and AU22.

In the current work, we identified several instances of AU18 in bonobos (e.g., Fig. S25, Videos S40–S42), and we further add that AU18 can be produced with or without wrinkling, and can be coded whenever the lip corners are drawn medially, even if no puckering is observed. As this movement was not described in the ChimpFACS, we include below its detailed description for bonobos.

A. Proposed muscular basis: orbicularis oris and incisivii labii.

B. Appearance changes:

1. The lip corners move towards the mouth midline or the lips are pushed towards the mouth midline.

2. Some medial bulging of the lips may be observed, but there is no outwards flaring of the lips.

3. Wrinkling may be observed or the existent wrinkles become more conspicuous, particularly on the upper lip.

4. The mouth area may appear to become narrower as it appears compressed medially. In frontal view, the lip corners become visible.

5. The lip corners may appear to be slightly pointing upwards due to the movement of the upper lip being squeezed towards the mouth midline.

6. If the mouth is open, fewer teeth may become visible. The sharp angle of the lip corners becomes rounder.

7. Skin and hair accompany the lip corner movement towards the medial area of the mouth.

8. It can be observed as a unilateral movement on one side of the mouth (AU18L or AU18R, Videos S40). It can also occur only on one of the lips (top lip: AU18T or bottom lip: AU18B, Videos S41).

9. In low intensity movements, only the lip corners move slightly forward, without wrinkling or mouth shape change. Hence, in low intensity movements, AU18 may not be visible from a full frontal view if the lip corners are not visible (e.g.,  Videos S42).

C. Minimum criteria: the lip corners are pushed medially or the lips are drawn together medially.

D. Subtle differences between AUs: AU18 and AU12 are mutually exclusive movements as they move the lip corners in opposing directions. However, AU18 might be confused with AU12 returning to neutral and vice-versa. Therefore, identifying the neutral position of the lip corners in a particular individual is important to distinguish these two movements and the release of the opposing movement. In addition, with AU18 the lip corners may appear to be slightly pointing upwards (Fig. S25), which is a false indicator for AU12. However, in AU18 the lip corners move forward, not towards the ears as it happens in AU12.

AU18 can also be confused with AU22, since both movements may present some degree of lip protrusion and wrinkles on both lips. However, the main difference is the shape of the lips, where in AU18 the protrusion is due to the bunching up of the lips in the medial area of the mouth, whilst in AU22 the protrusion is due to one or both lips flaring out and taking a funnel or flattened shape.

Caution is needed when coding AU18 in bonobos, as the neutral position of the lip corners is hard to determine in many individuals. Hence, as a minimum criterion to code this AU, movement of the lip corners or the lip itself moving medially must be observed.

AU22—Lip Funneler

In humans, chimpanzees, and bonobos, AU22—Lip Funneler is produced by the orbicularis oris muscle to project the lips medially forwards and outwards in a typical funnelled or flattened shape. In chimpanzees and bonobos, due to the larger size of the lips in comparison to humans, AU22 is a more conspicuous movement and can have more variation of how the lips are positioned during this movement. In bonobos and chimpanzees, this AU can result in circular parting of the lips in the middle, resulting in the exposure of the inner lip area and often the separation of the lips (which needs the additional code AU25 - Lips Parted). The mental region is pulled slightly upwards. The vertical lines on the upper lip deepen during this action in both species. Pronounced actions of this AU are well visible in bonobos from a frontal view because of the bright lips: a flesh-coloured circle appears when the lips are funnelled; but it is easy to detect movement from a side view as well, as the lips are projected forwards (Figs. S26–S29, Videos S43–S48).

In the ChimpFACS, action of AU22 for the top lip only was described, but in our sample of bonobo videos, funnelling of the upper lip alone was not observed. Instead, we observed different intensities of AU22 in each lip, so it is possible that bonobos also have the ability of producing AU22 in only one lip (coded as AU22T for the top lip or AU22B for the bottom lip). In addition, we observed a higher mobility and flexibility of the lower lip in bonobos, beyond the funnel/flattened shape described for a typical AU22 described for humans or chimpanzees. In humans the lower lip is short, and other than the AU22, does not have a lot of flexibility to curl outwards or extend extensively. In bonobos, the lower lip is much larger than in humans, being often used to hold and manipulate objects or food. This differentiated lower lip morphology allows the production of movements that extend and elongate the lower lip outwards, which we also include as appearance changes for AU22 (e.g., Fig. S29, Video S48).

AU24—Lip Presser

This AU presses both lips together, causing lip bulging, particularly in the upper lip, and can only be coded if both lips are together. This action is produced by the orbicularis oris muscle in both chimpanzees and bonobos. The appearance changes of the AU are the same in chimpanzees and bonobos, except there seems to be more bulging and less wrinkling on the lips in bonobos than in chimpanzees (e.g., Figs. S30–S31, Videos S49–S51). The visibility in bonobos may be slightly better than in chimpanzees because of the contrasting lip colour.

AU25—Lips Part

Action Unit 25 parts the lips. Part of the teeth, gums and the inner mucosal lip area may be revealed. These appearance changes were found to be the same in bonobos as in chimpanzees (e.g., Videos S52–S53). Other AUs often are acting to part the lips (e.g., AU10 and AU16); AU25 is then coded additionally.

AU26—Jaw Drop

This AU lowers the mandible by relaxation of the jaw musculature. The lips may part, and if they do, a small space between the upper and lower teeth is visible. The appearance changes in bonobos are the same as in chimpanzees, and the same cues can be used for identification (e.g., Fig. S32–S33, Videos S52–S53).

AU27—Mouth Stretch

AU27 actively pulls down the mandible, which results in a wide mouth opening. The mouth is elongated vertically. Visually, the only difference between AU26 and AU27 is the degree of separation of the teeth, where in AU26 there is a small space and in AU27 a large space between the teeth. Although AU27 describes a larger degree of mouth opening in relation to AU26, AU27 is mutually exclusive to AU26, as different muscles are involved. Due to the degree of mouth opening always separating the lips in AU27, it is coded as AU25+AU27. In addition, although this is mostly not visible, there is more muscular tension around the mouth and lips which are being stretched, hence the skin might appear smoother around the lips and lip corners. However, this only becomes visible in high intensity AU27. High intensity AU26 and low intensity AU27 may be hard to distinguish and only by continued visualisation of several examples can this discrimination be achieved by coders. All these appearance changes apply equally to chimpanzees and bonobos (e.g., Figs. S34–S351, Videos S54–S55).

AU28—Lips Suck

In chimpanzees and bonobos, AU28—Lips Suck have similar appearance changes, with the main difference being the brighter colour of the lips in bonobos, which makes this movement more conspicuous (e.g., Figs. S36–S37, Videos S56–S58). This movement in bonobos seems often to accompany AU24—Lip Presser, and this AU24+AU28 combination is described as possible in humans, but it was not reported for chimpanzees. However, the subtle differences reported in the ChimpFACS between AU24 and AU28 also apply to bonobos, with bulging of the upper lip being the minimum criterion to code AU24. If both lip insertion into mouth and bulging are present, AU24+AU28 should be coded together.

AU38—Nostril Dilator and AU39—Nostril Compressor

In humans the nostrils are widened with AU38—Nostril Dilator, or constricted with AU39—Nostril Compressor, by the action of the nasalis muscle. In the ChimpFACS this movement is stated as not observed and it was not observed either in Kuchenbuch’s work (2010), but we found these in the current video sample. Bonobos present a much larger and more developed nasal region than humans and to some extent chimpanzees, and a more connected nose to the upper lip. These differences in facial morphology affect the appearance changes to code AUs. These movements can be subtle movements or very quick movements, and hence harder to detect than other AUs.

A. Proposed muscular basis: nasalis.

B. Appearance changes:

1. In AU38, the nostril increases in size (e.g., Video S59), and in AU39 decreases in size (e.g., Video S60). In AU38/AU39 the nostrils usually change shape or at least the nostril wings will present some movement.

2. The skin next to the nose and on the nose shield as a whole might move, enlarging, contracting, projecting forward, or flattening, accompanying each of the respective movements.

C. Minimum criteria: the nostrils are widened in AU38 and narrowed in AU39.

D. Subtle differences between AUs: caution regarding appearance change 2 is needed, as movement of the skin around the nose might be due to other AUs pulling the skin globally in the mouth region (e.g., AU10, AU16). Hence, AU38/AU39 are coded only if the minimum criterion for each AU is present, i.e., if the movement is clearly originating in the nostrils or nasal shield. Additionally, these nose AUs may be difficult to detect depending on the angle and distance of the individual from the camera. Changes in head position might appear to change the nostril shape. Therefore, if there are head movements that can change the observed nostril shape or the nasal shield shape, the nose AUs should be coded only when movement is detected in the nostrils or the nasal shield.

In addition, AU38/AU39 might be hard to code in a side view if the nostrils are not fully visible.

AU38 is mutually exclusive to AU39 as these are opposite movements.