The Effect of Animal Bipedal Posture on Perceived Cuteness, Fear, and Willingness to Protect Them

Human–animal relationships have long been the subject of controversy because they are shaped by several cultural, inter-individual, and evolutionary factors. Understanding these relationships, however, is important to optimize conservation efforts. There is agreement that perceived similarity between animals and humans is associated with more positive attitudes. The human–animal similarity could be caused by phylogenetic closeness. We hypothesized that the bipedal posture of an animal may be perceived as a cue of phenotypic similarity with humans. We examined this topic by comparing perceived cuteness, fear, and willingness to protect animals differing in body posture, body size, and phylogenetic closeness with humans on a sample of N = 349 Slovak participants. We found that the bipedal posture enhanced perceived cuteness, but this effect was most pronounced in small-bodied animals, particularly those with direct eye contact. Phylogenetically close and small-bodied species (e.g., small mammals) received greater conservation support than phylogenetically distant species (e.g., invertebrates). However, anthropomorphic-looking animals received greater conservation support, suggesting that pictures of animals that more closely resemble humans can be used in conservation campaigns.


INTRODUCTION
We are not seeing animals as animals, but merely as reflections of ourselves, and if the mirror distorts too badly, we either bend it into shape or discard it. Morris (1969, p. 200) The complexity of human-animal relationships is the subject of controversy (Mullin, 1999) because animal roles in human lives are multidimensional (Kellert, 1980(Kellert, , 1983(Kellert, , 1993Serpell, 2004). Throughout our evolutionary history, animals have been predators of our ancestors (Hart and Sussman, 2008), important sources of food (Ungar and Teaford, 2002), and vectors of zoonotic diseases (Kruse et al., 2004). Since the beginning of the Neolithic era, many animals have been treated as companions (Podberscek et al., 2005), with others being agricultural pests (Ordish, 1976). The enormous diversity of animal shapes, sizes, and behaviors heavily contributes to human evaluations of them (Serpell, 2004). Human preferences are of interest to conservationists because the low popularity of an animal, regardless of why it occurs, erodes public support for its conservation (Kellert, 1985;Houston et al., 2010;MacFarlane and Rocha, 2020).
Human activities contribute to the global loss of biodiversity comparable with the five previous mass extinctions of Earth's history (Barnosky et al., 2011;Dirzo et al., 2014). More than 400 vertebrates became extinct since 1500 (IUCN, 2020). Although invertebrates received much less attention than vertebrates, their extinction rates seem to be similar. For instance, insect abundance in protected natural areas in Germany dropped by 75% over the last 27 years (Hallmann et al., 2017), and roughly 50% of butterfly species (Lepidoptera) were extirpated in Singapore since 1854 (Theng et al., 2020). The reduction of biodiversity can negatively influence ecosystem functioning and human wellbeing by means of disease spread, climate change, and decreased farm productivity (Newbold et al., 2019;van der Plas, 2019). Because all animals play important roles in ecosystems, it is necessary to investigate the factors underlying their preferences by the general public in order to make conservation more effective (Frynta et al., 2013).
Preferences for some animals over others are noteworthy because the perceived attractiveness of an animal species can be an important determinant of conservation support (Gunnthorsdottir, 2001;Tisdell et al., 2006;Martín-López et al., 2007;Knight, 2008;Wang et al., 2018;Curtin and Papworth, 2020). For instance, more popular species at the Paris Zoological Park can receive 46 times the funds of less popular species (Colléony et al., 2017).
Animals that are cognitively, behaviorally, physically, or phylogenetically similar to humans are perceived by humans in a more positive light than distant or dissimilar animals (Plous, 1993;Allen et al., 2002;Knight et al., 2003;Serpell, 2004). Humans express increased empathy and compassion with decreasing phylogenetic distance from humans (Miralles et al., 2019). It has been shown that our perception of "the others" may depend on the group to which we perceive that "the others" belong, that is, the more we perceive the members of the other person/group in need as being similar to us, the more we value their welfare (e.g., Batson, 2011). Małecki et al. (2020) found that the subjective perception of the level of kinship between humans and animals depicted in a narration had a significant influence on the improvement of attitudes toward animal welfare. However, the objective measure of the phylogenetic distance was not related to any changes in attitudes toward animal welfare.
Another largely overlooked cue of species similarity with humans is bipedal posture. Morris (1969) argues that the popularity of penguins, primates, bears, pandas, or dogs is significantly influenced by their ability to stand vertically because animals in bipedal postures resemble humans more than in quadrupedal postures. As far as we are aware, no research has investigated the degree to which the bipedal posture of an animal influences human perceptions and attitudes. We postulate that perceived cuteness, fear of animals, and willingness to protect animals would be worthwhile to investigate because these variables are related to the psychological aspects of animal conservation (Prokop and Fančovičová, 2013;Castillo-Huitrón et al., 2020).
We hypothesized that animals with bipedal posture would be perceived as cuter (and thus more likely to induce human willingness to protect them) than animals with quadrupedal posture. We also hypothesized that large animals would be perceived as more dangerous than small animals (Prokop et al., 2010;Staňková et al., 2021). Thus, the perceived cuteness of small animals in bipedal posture should be greater than the perceived cuteness of large animals in bipedal posture. Furthermore, we hypothesized that phylogenetically distant animals would be perceived as less cute than phylogenetically close animals (Miralles et al., 2019). Finally, both bipedal posture and phylogenetical closeness should be associated with a greater willingness to protect animals.

Participants
The research was carried out during the summer semester of 2020. Roughly 1,500 freshman students at a university in Slovakia were asked to voluntarily participate in the research. Furthermore, the participants were recruited through social networks. A total of 349 participants (272 females) comprised the final sample (age range = 18-67 years, mean = 25, SE = 0.45; Table 1).

Measurement of Cuteness, Perceived Fear, and Willingness to Protect Animals
The participants were asked to rate 52 pictures of 26 animal species for cuteness, perceived fear, and willingness to protect these animals on a seven-point scale (e.g., 1 = not at all cute, 7 = extremely cute). The ratings showed excellent reliabilities (Cronbach α = 0.97, 0.97, and 0.98, respectively).

Measurement of Phylogenetic Distance From Humans
The phylogenetic divergence time from humans (in millions of years) was obtained for each species from timetree.org (Kumar et al., 2017).

Measurement of Direct Eye Contact
Animals on pictures were scored binomially (yes or no) according to direct or indirect eye contact. This procedure was used because it was not possible to obtain pictures of all animals with a standard gaze. This variable could be important because direct eye contact is a signal of potential threat (Emery, 2000).

Measurement of Anthropomorphism
Animals which looked like humans by their posture or activity (e.g., a basilisk running on the surface of water) were binomially classified as anthropomorphic or not. Two of us separately and independently scored the pictures for anthropomorphism. In the few cases where our scorings differed, we discussed the pictures of animals until we agreed on the score to be awarded. Anthropomorphism, in this sense, was not always associated with bipedal posture (e.g., standing Komodo dragon), and we therefore considered it as an additional independent variable.

Visual Stimuli
Colorful pictures, with their original background, of 14 vertebrate and 12 invertebrate species which at least occasionally use bipedal posture were downloaded from Google. The species were further selected according to their phylogenetical similarity and body size ( Table 2). We used two pictures for each species: one in bipedal posture and one in quadrupedal posture. Throughout the paper, we used the term quadrupedal also for invertebrates with more than four legs. For invertebrates, pictures of bipedal animals were searched using common English or Latin names together with the word "bipedal" or "standing." Birds were completely omitted because they are invariably bipedal.

Statistical Analyses
The cuteness, fear, and protection scores were defined as dependent variables in a cumulative link mixed model (CLMM) in R software 3.6.3 (R Core Team, 2019) by applying the clmm function in an ordinal package (Christensen, 2019). Bipedal posture (bipedal or quadrupedal), anthropomorphic (yes or no), eye contact (yes or no), body size (large or small), and gender of the participant were categorical predictors, whereas phylogeny was a continuous predictor (fixed effects). The identity of the participants and the species of animal in the pictures were used as a grouping factor (random effects) to deal with correlations within the participant ratings. To select the particular factor, stepwise selection was used and evaluated by the Akaike information criterion (AIC). The syntax of the final CLMM models are as follows: Cute.score ∼ Phylogeny + Anthropomorphy + Eye.contact + Biped + Gender + Size + Phylogeny:Gender + Phylogeny: Biped + Anthropomorphy:Eye.contact + Size:Gender + Size: Phylogeny + Eye.contact:Size + Anthropomorphy:Size + Size: Protection ∼ Phylogeny + Anthropomorphy + Biped + Size + Size:Biped + Anthropomorphy:Phylogeny + (1| ID respondent) + (1| ID.species).

RESULTS
The best CLMM models for cuteness and fear ratings contain six fixed effects (phylogeny, anthropomorphy, bipedal posture, size, eye contact, and gender) and numerous interactive effects (eight for cuteness and five for fear; Tables 3, 4). In the case of the protection rating model, four fixed factors and two interactions were included ( Table 5). The final CLMM model values of log likelihood ratio, AIC, and marginal, conditional, and pseudo r-squared are shown in Table 6. The likelihood ratio test showed the random effects, which have been included in all the CLMM models, to be significant (p < 0.001).

Bipedal Posture
Animals with bipedal posture were rated as significantly less cute than animals with quadrupedal posture (Figure 1A). However, the combination of bipedal posture with other variables increased the perceived cuteness to a greater degree-for example, bipedals that are small in size were considered to be significantly cuter than larger ones (Figure 2A). Additionally, high cuteness scores were specifically associated with animals in bipedal posture that are phylogenetically closer to humans (e.g., bipedal mammals; Table 3).
Animals with bipedal posture received higher fear scores than animals with quadrupedal posture (Figure 1B). Interactive effects revealed that quadrupedal animals with direct eye contact were considered more menacing than other combinations of these factors ( Table 4). Higher fear scores were associated with animals in bipedal posture that are phylogenetically close to humans, such as bipedal mammals ( Table 4).
When considering the factor of bipedalism regarding willingness to protect, people were generally more willing to protect quadrupedal animals than animals with bipedal posture (Figure 1C). Animal size moderated this effect, such that willingness to protect was strongest toward small animals in bipedal posture (Figure 2B).

Phylogeny
The factor of phylogeny was very strong when assessing all dependent variables (cuteness, fear, and willingness to protect; Tables 3-5). Animals phylogenetically closer to humans (e.g., mammals) were perceived as cuter than phylogenetically distant animals (e.g., insects; Figure 3A). Small animals phylogenetically distant from human (e.g., small insects) received lower cuteness   scores ( Figure 3B). When rated by women, phylogenetically distant animals received significantly lower cuteness scores relative to the ratings provided by men (Table 3).
Phylogenetically close animals were associated with higher fear scores when compared to phylogenetically distant animals ( Figure 3A). The interaction terms showed that the fear scores were highest when considering animals phylogenetically close to humans and displaying a bipedal posture (e.g., mammals in bipedal posture; Table 4). Phylogeny significantly affected the willingness to protect animals. People were willing to protect animals phylogenetically close to humans, such as mammals (Figure 3A), especially when it comes to small mammals ( Table 5). The lowest protection scores were for animals that are phylogenetically distant from humans but looked anthropomorphic (e.g., insect with raised front legs; Table 5).

Anthropomorphism
Animals resembling humans in terms of posture or activity (e.g., a bipedally running reptile or standing gorilla) were perceived as cute ( Figure 4A). Moreover, small animals resembling humans were evaluated as cuter than large anthropomorphic animals ( Table 3). High cuteness scores were assigned to small anthropomorphic-looking animals with direct eye contact ( Table 3).
In general, the respondents were not afraid of animals that looked anthropomorphic (lower fear scores; Figure 4B) and were willing to protect them (anthropomorphism increases the  willingness to protect such animals; Figure 4C), but when considering the interaction with phylogeny, the willingness to protect decreased toward anthropomorphically looking animals that are phylogenetically less related to humans (e.g., insect with raised front legs; Table 5).

Size
Cuteness scores strongly depended on whether the animal is large or small and are biased in favor of small creatures (Figure 5A). In general, small animals that are phylogenetically closely related to humans were perceived as cuter than distant animals (Figure 3B), while small animals resembling humans with direct eye contact were rated cuter than large animals ( Table 3). Women had a tendency to give higher cuteness ratings when considering small animals ( Figure 6A).
As we expected, size modified the fear score as well. Small animals were rated as less menacing when compared to larger ones (Figures 5B, 6B).
Our results showed that people were willing to protect small animals ( Figure 5C) and that small quadrupedals were evaluated with slightly higher protection scores than small bipedals ( Figure 2B).

Eye Contact
When considering animals with eye contact, they were evaluated as less cute in comparison with no eye contact (Figure 7A), even FIGURE 3 | Estimated probability of ratings for cuteness, fear and willingness to protection according to Phylogeny (A), interaction effects of Phylogeny:Size for Cuteness (B). Error areas indicate 95% confidence interval; numbers 1-7 refer to rating scores. though small anthropomorphic animals with eye contact were rated as cute (Table 3).
With regard to fear, eye contact increased the fear scores ( Figure 7B), especially together with quadrupedalism ( Table 4).
Eye contact was not statistically significant for willingness to protect as an individual factor nor for interactions ( Table 5).

Gender
Even though gender was not statistically significant as an individual factor for the cuteness evaluation, the interactions with other factors were significant ( Table 3). As we have mentioned, animals that are phylogenetically distant to humans (e.g., insects) were rated as less cute than phylogenetically close animals (e.g., mammals; Figure 3A), but the cuteness scores decreased more when they were evaluated by women ( Figure 8A). Small animals were assigned higher cuteness scores when rated by women compared to the ratings of men ( Figure 6A).
For fear evaluations, gender was statistically significant as an individual factor, and it also interacted with several other variables. Women had a tendency to give higher fear scores relative to men (Figure 8B). The most fearful evaluations were phylogenetically close animals rated by women, and the least fearful evaluations were phylogenetically distant animals rated by men (Table 4).  Gender was not statistically significant for willingness to protection as an individual factor nor for interactions.

Relationship Between Cuteness, Fear, and Willingness to Protect Animals
Total cuteness scores positively correlated with willingness to protect animals and negatively correlated with perceived fear (Pearson r = 0.50 and -0.37, both P < 0.001, N = 349, respectively). Willingness to protect animals negatively correlated with perceived fear (Pearson r = -0.14, P = 0.01, N = 349).

DISCUSSION
As far as we are aware, this is the first study which systematically investigated whether the bipedal posture of animals influences their perception by humans. We found that bipedal posture significantly influenced all three investigated domains of animal perception: perceived cuteness, fear, and willingness to protect them. Moreover, these associations were influenced by the size of the animal, phylogenetical distance from humans, direct eye contact, and perceived anthropomorphism.
We hypothesized that animals with bipedal posture would be perceived as cuter than animals with quadrupedal posture. In line with this hypothesis originally proposed by Morris (1969), bipedalism enhanced the perceived cuteness under certain contexts-for example, the cuteness of small animals with bipedal posture was enhanced more than the cuteness of large species with bipedal posture. With regard to fear, large animals, particularly carnivores, elicited stronger fear judgments than small animals (Prokop et al., 2010;Staňková et al., 2021). In addition, large, but not small, animals with bipedal posture induced greater fear. These results can be explained from an evolutionary perspective, as large carnivores were common  predators of our ancestors in Africa (Treves and Palmqvist, 2007). Women are more frequently victims of predatory attacks than men (Treves and Naughton-Treves, 1999), which can explain their greater fear of large-bodied animals. Moreover, predation pressure on humans continues, although on a smaller scale than in our evolutionary past (e.g., Treves and Naughton-Treves, 1999;Löe and Røskaft, 2004;Packer et al., 2005).
Phylogenetically distant animals were perceived as less cute than phylogenetically closer animals. This finding is consistent with the different methodological approaches previously documented (the capacity for an animal to feel pain: Plous, 1993;empathy toward animals: Miralles et al., 2019). The participants also indicated lesser willingness to protect phylogenetically distant species (e.g., invertebrates) than phylogenetically closer species (e.g., mammals), which suggests that our innate preference for cute animals (baby scheme, Borgi et al., 2014) and/or the ability to empathize with animals (Miralles et al., 2019) enhances our interest in the conservation of animals similar to us (Samples et al., 1986;Plous, 1993;DeKay and McClelland, 1996).
Although it is well known that people have negative attitudes toward invertebrates (e.g., Kellert, 1993;Schlegel and Rupf, 2010;Fukano and Soga, 2021), it is hard to establish powerful strategies regarding how these attitudes can be improved (Cardoso et al., 2011). We did not observe a positive trend toward willingness to protect invertebrates if they look more anthropomorphic (e.g., an ant with raised front legs). It seems that bipedal posture itself does not fully enhance anthropomorphism (Morris, 1969), though it makes a moderate contribution. We suggest that this result can be used in conservation campaigns, where certain unpopular invertebrates could be used as flagship species if they are photographed in a natural way. Fukano and Soga (2021), for instance, showed that when people saw photographs of insects with an indoor background, perceived disgust and danger for insects were higher compared to pictures with an outdoor background.
Contrary to our hypothesis, the participants were generally more willing to protect small quadrupedal animals compared with animals with bipedal posture. Small species could enhance perceived cuteness and, consequently, our interest in the conservation of these animals. However, human preferences and conservation efforts are directed toward rather large-bodied, charismatic mammals (Kontoleon and Swanson, 2002;Clucas et al., 2008;Sitas et al., 2009;Smith et al., 2012;Veríssimo et al., 2018). Albert et al. (2018) found that charismatic animals are beautiful/cute, dangerous/impressive, or rare/endangered. We used only two species (gorilla and bear) belonging to the list of charismatic animals of Albert et al. (2018). Thus, it seems that human willingness to protect is directed toward small, cute species over large-bodied but non-charismatic animals. It is still not clear, however, why bipedal posture did not contribute to willingness to protect animals. We speculate that the bipedal posture of certain large-bodied species could be perceived as too uncommon (e.g., capybara) and/or threatening, thus reducing perceived conservation needs. Perhaps the bipedal posture of small, cute species could enhance willingness to play rather than willingness to protect them. To support this idea, small animals in anthropomorphic posture with direct eye contact were rated as cute, and eye contact provides a foundation for communication and social interaction (Kleinke, 1986;Senju and Johnson, 2009).

LIMITATION
Since the perception of cuteness, fear, and willingness to protect an animal only according to a picture is a very complex process, we chose ratings (CLMM model) over simple pairwise comparisons. The main reason behind this methodological strategy is that the ratings for two items tell us not only which item is preferred but also the degree to which it is preferable, so ratings can be more informative than pairwise comparisons. Additionally, if we took into account the expected complex information from pairwise comparisons, we would need many more respondents. Heterogeneity and overparameterization of the model need to be solved in multifactorial dimensions.

CONCLUSION
Our research on how looking similar to humans influences the perception of animals showed that bipedal posture contributes to the perceived cuteness of (particularly small) animals. Cuteness is associated with willingness to protect animals, but bipedal posture itself did not enhance the conservation support for animals. It seems that the bipedal posture of large-bodied animals could be perceived as threatening or perhaps non-typical. Small bipedal animals, particularly those with direct eye contact, could initiate social interactions different from conservation needs. However, if phylogenetically distant and less cute animals look more anthropomorphic, they receive lesser intentions of willingness to protect. We did not experimentally manipulate perceived anthropomorphism; thus, more research in this field should be done before a definitive conclusion can be made.

DATA AVAILABILITY STATEMENT
The data analyzed in this study is subject to the following licenses/restrictions: Data are available from authors upon request. Requests to access these datasets should be directed to corresponding author.

ETHICS STATEMENT
Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
PP, MZa, and PF designed the study. MZi analyzed the data. AP commented on the manuscript and improved the English. PP, MZa, and PF collected the data. All authors contributed equally in writing the manuscript.