Introduction
Emotions are central to the daily life experiences of humans and to relationships with phenomena in the world ‘out-there’ (LeDoux 1996; Damasio 2000). By extension, emotions are important factors shaping our relationships with wildlife (Manfredo 2008; Jacobs 2009). Emotions influence virtually all other mental processes, including attention (LeDoux 1996), motivation (Damasio 2000), memory formation (Talarico and Rubin 2003) and intention (Frijda 1986). While a consensus definition has not emerged, scholars agree that emotional reactions consist of physiological, expressive and experiential components (Bradley and Lang 2000; Mauss et al. 2005). For instance, a fear response usually includes an increased heartbeat, a scared facial expression, a tendency to freeze or flee and conscious negative feelings such as fear (LeDoux 2013).
Biophobia is the tendency to respond with a negative emotion, such as fear or disgust, towards natural stimuli (slightly amended from Soga et al. 2023). Bats are frequently described as stimuli of biophobia (Kingston 2016; Eklöf and Rydell 2021; Soga et al. 2023)—likely because people find their behaviour and morphology unfamiliar—which often forms the basis of myths and other misperceptions (Eklöf and Rydell 2021). There are at least 1 474 known bat species (Simmons and Cirranello 2024) with 16.5% of 1 335 species assessed by the IUCN considered threatened (IUCN 2024), and bats play important roles in our ecosystems as agents of insect pest suppression, pollinators and seed dispersers (Kunz et al. 2011). However, bats are nocturnal, they rest upside down and, with the exception of Old World fruit bats (family Pteropodidae), and they tend to lack traits that are considered cute or charismatic (e.g., neotenic features such as big eyes and large heads, colourful fur or feathers, large size and majestic bearing; Verbrugge et al. 2013; Estévez et al. 2015; Albert et al. 2018; Jaric et al. 2020). In a direct comparison of 100 mammals, Macdonald et al. (2015) found that species that were smaller and dark (among other characteristics) were less preferred than larger mammals with a single bright colour; for example, the common vampire bat (Desmodus rotundus) was ranked 97th, while species traditionally considered charismatic (especially from the felid family) ranked highly. In addition, while only three of the 1 474 known bat species feed on blood, and those that do are restricted to Central and South America, this habit is often attributed to bats in general, reinforcing negative perceptions and myths. For example, Prokop et al. (2009) found that around 20% of Slovakian students believed that bats feed on blood; probably due to the link between this belief and vampire stories. This belief was even more common (37%) among children aged 10–16 years in Slovakia (Prokop and Tunnicliffe 2008). Further, bats are described as unclean animals in the Bible (Eklöf and Rydell 2021).
Anzeige
Earlier literature also characterised bats as fear-relevant, i.e., highly feared without posing a real threat (as opposed to posing a threat through predatory attacks; e.g., Davey 1994; Ware et al. 1994, Arrindell 2000; Polák et al. 2019). Fear of bats has likely increased in recent years, because several recently emerged infectious viral diseases may have ancestral or recent origins in certain bat populations (e.g., COVID-19, Nipah, Ebola, Olival et al. 2017; Epstein et al. 2020). While links between bats and zoonoses are frequently speculative and confused by inaccurate communication and consequent misunderstanding (López-Baucells et al. 2018; Shapiro et al. 2021), these links have also led to bats being commonly associated with the emotion of disgust. Fear and disgust are the two emotions known to be evoked by animals considered dangerous (Ekman 1999). Feared objects or situations can even elicit a combination of fear and disgust (Coelho et al. 2023). While fear is an adaptive response to an imminent threat of injury or even death (LeDoux 2012) or to other elements, such as sensory experiences (e.g. pain, Coelho et al. 2023), disgust is considered part of the behavioural system that protects humans against pathogen transmission or contamination (Matchett and Davey 1991; Curtis et al. 2011). As already widespread persecution of bats (Straka et al. 2021) increased during the COVID-19 pandemic (Zhao 2019), it is reasonable to hypothesise that biophobia of bats also increased during the pandemic (Soga et al. 2023). In the realm of human-wildlife relationships, other emotions deemed significant include sadness and anger (Ekman 1999). Although not directly linked to animals considered dangerous (Ekman 1999), these emotions remain relevant when it comes to understanding and addressing human-bat interactions (Straka et al. 2020); maybe even more so during the pandemic. Sadness may arise from a sense of empathy and sorrow about the unjust persecution of bats, while anger might stem from feelings of frustration with or resentment towards bats.
Despite the many myths and recent concerns, bats are not universally feared. Fear of bats appears rare or infrequent among people in Michigan, United States (Hoffmaster et al. 2016) and students in Germany (Straka et al. 2020). And while bats were reportedly less well-liked than other wild animals in Norway (Bjerke & Østdahl 2004), positive associations with bats are suggested to be common in Swedish (Nordic) history (Eklöf and Rydell 2021). People living in Kenyan villages near bat caves generally held a greater-than-neutral interest in bats (Musila et al. 2018). Meanwhile, people on Ishigaki Island, Japan, displayed a lack of clear opinion regarding their interest in the Ryukyu flying fox, likely due to either varied sensitivities to the question or limited knowledge about the species (Vincenot et al. 2015). Indeed, a review found that in Asia-Pacific, the vast majority of cultural values associated with bats are positive (Low et al. 2021). The above examples, taken together, illustrate that negative emotions towards bats are contextual (as in the case of vampire bats; López-del-Toro et al. 2009), simply over-emphasised or dependent on which emotions or other psychometric concepts are being compared. For instance, the above-mentioned studies (Davey 1994, Ware et al. 1994; Arrindell et al. 2000; Polák et al. 2019) only measured negative emotions in relation to different animals, including bats—none of these measured positive emotions. Indeed, there is a limited number of published studies that have simultaneously investigated positive and negative bat-related emotions that matter to human-wildlife relationships (but see e.g.Vincenot et al. (2015) and Musila et al. (2018) who measured for instance positive dispositions such as interest besides negative ones towards bats). Consequently, the literature on human emotions towards bats may be influenced by a negativity bias (Buijs & Jacobs, 2020), given the contextual limitations of the studies and the use of different measurement approaches. Clearly, studies focusing on negative bat-related emotions or comparing bats with other animals cannot explore the full range of humans’ emotional relationships with bats. Therefore, we cannot conclude that negative emotions matter more than positive emotions do in shaping human-bat relationships. Practically, an overemphasis on negative bat-related emotions might limit the discovery of potential avenues to understand and raise public support for bat conservation. For example, if positive emotions exist, then instead of working to counter negative emotions, it might be more effective to reinforce or invoke positive emotions in outreach campaigns. Moreover, given that emotions and behaviours may be linked (Strack et al. 2016), gaining a comprehensive understanding of emotions has conservation relevance.
The term ‘emotion’ can refer to a state or a trait (Jacobs et al. 2014). Emotional responses are states, that is, momentary reflections of how one is. The term ‘emotions’ is also used to denote traits, i.e., stable properties reflecting who one is. In the present study, we use emotions in the second sense, labelled as emotional dispositions (Frijda 1986). Emotional dispositions act as mental criteria for judging the emotional relevance of stimuli to the individual (Jacobs et al. 2014). Meaning, if a person claims they fear wolves, this indicates their emotional disposition, but not that they are, in that moment, in a state of fear, which is rare for most people. Compared to emotional responses (which are time and context dependent), emotional dispositions are more amenable to measurement through surveys. Scholars of emotions use two different theoretical approaches—discrete and dimensional—to account for the diversity of emotions. In the discrete approach, emotions are classified as qualitatively different from each other (Izard 2007). Researchers using this perspective examine specific emotions such as fear, disgust or joy. In the dimensional approach, emotions are characterised on a few dimensions, usually valence (positive-negative) and arousal (the level of activation; Russell 2003). Compared to the dimensional approach, the discrete approach better matches how people usually evaluate their own emotions (Jacobs et al. 2014), so we used it for our study.
Our study rationale was informed by the above-mentioned overemphasis on negative emotions towards bats, the limited number of studies of both negative and positive bat-related emotions and the potential influence of the pandemic on emotions towards bats (and the implications for bat conservation). As such, we had three aims, to (1) assess discrete positive and negative emotions towards bats, (2) test for emotional shifts during the pandemic and (3) explore the predictive potential of positive and negative emotions, along with socio-demographics, on behavioural intentions to engage in bat conservation actions.
Anzeige
We surveyed the German public during the pandemic and investigated their self-assessed emotions towards bats and bat conservation-relevant behaviours. We ran the survey twice to examine whether negative emotions occurred and increased during the pandemic. We further assessed the link between emotions and bat-conservation-relevant behavioural intentions. This is specifically relevant to determining which focus is more beneficial to bat conservation: mitigating negative emotions (biophobia) or fostering positive emotions (biophilia).
The present research makes a new contribution to conservation social science by examining negative and positive discrete emotions towards bats sensu Ekman’s (1999) and Izard’s (2007) which we found most useful in this context. An understanding of emotions towards bats at a time when bats have been the focus of (negative) societal attention (due to speculation about the origin of the SARS-CoV-2 virus) requires empirical research. While some studies have considered evaluations of bats during this time (Lu et al. 2021; Ejotre et al. 2022) including how negative emotions towards bats can reduce willingness to cohabit with bats (Lundberg et al. 2021), none has investigated changes in public emotions and their consequences for bat conservation. We asked if COVID-19 exacerbated negative emotions towards bats, as some literature (e.g., Ejotre at al. 2022) suggest it has.
Methods
Survey sample
We launched our first survey approximately ten months after the official declaration of the COVID-19 pandemic (mid December 2020-mid January 2021; hereafter 2021) and the second one twelve months later (January 2022; hereafter 2022). We created our online questionnaire in SoSci survey (https://www.soscisurvey.de/), distributed both iterations on the same online platforms (Survey Circle, Facebook) and sent them to > 500 email addresses from diverse public institutions to reach the broad public (Table S1). We emailed reminders to the institutions and also posted them on social media channels two weeks after the first call. We selected this online convenience sample because it was the most feasible option during the pandemic. The items we analysed for this study were presented as part of a wider survey that aimed to understand the effect of photo stimuli and narratives about bats on people’s emotions and behavioural intentions. In this paper, we focus only on the items relating to self-assessed emotions and behavioural intentions, which appeared at the beginning of the surveys to avoid any potential confounding interactions between questions.
Survey items
We used five-point scales to assess three parameters. First, discrete bat-related emotions deemed relevant to human-wildlife relationships (Ekman 1999; Izard 2007)—three positive ones (joy, interest, compassion) and four negative ones (anger, disgust, fear, sadness). While interest is not always labeled as an emotion in the literature, some studies do consider it to be a basic emotion (Izard 2007; Silvia 2008; Fredrickson 2013). For instance, interest manifests as a distinct, identifiable pattern of facial expressions and heart-rate changes in infants (Izard 2007) and is considered a crucial emotion to practical problems of learning, education and motivation (Silvia 2008). In our case, motivation is relevant in the context of getting involved in bat-conservation. Second, we measured four behavioural intentions to get involved in bat conservation actions: (1) engaging in bat conservation projects, (2) motivating others to do something for bat conservation, (3) donating money to bat conservation projects and (4) learning more about bats (all adapted from Jacobs and Harms 2014). Lastly, we measured socio-demographic variables (Table 1) to compare both sample populations and understand the influence of socio-demographics on emotions and behavioural intentions.
Table 1
Survey items and variable types (original survey found in Fig. S1)
Concept | Item(s) | Variable type |
|---|---|---|
Emotions | When I think about bats, I feel [interest, compassion, joy, fear, anger, disgust, sadness] | Ordinal 1 = completely disagree 2 = disagree 3 = neither disagree nor agree 4 = agree 5 = completely agree |
Behavioural intentions | I want to [engage in bat conservation projects, motivate other people to do something for bat conservation, donate money to bat conservation projects, learn more about bats] | Ordinal 1 = completely disagree 2 = disagree 3 = neither disagree nor agree 4 = agree 5 = completely agree |
Age | Respondents’ age | Ordinal 1 = 18–29 2 = 30–45 3 = 46–55 4 = 56–65 5 = 66 + |
Gender | Respondents’ gender | Categorical 1 = female 2 = male 3 = diverse |
Place of living | Respondents’ current place of living (self-assessed, no definitions of urban, suburban, rural provided) | Categorical 1 = urban 2 = suburban 3 = rural |
Statistical analyses
To assess whether participants in the 2021 and 2022 surveys had similar socio-demographic backgrounds, we used Chi-squared tests using the ‘chisq.test’ function in the ‘MASS’ package (Venables and Ripley 2002) of R 4.2.2 (R Core Team 2022). We specifically compared the distributions of age, gender and place of living between samples (Table 2).
To understand the underlying structure of the seven assessed discrete emotions (Table 1), we performed a factor analysis with a varimax rotation (assuming that the discrete emotions are uncorrelated) using the ‘fa’ function in the R package ‘psych’ (Revelle 2022). In other words, we were interested in whether the discrete emotions that we measured could be condensed into a smaller number of unobserved (latent) emotion factors. Here, the factor analysis revealed that two factors with eigenvalues > 1.0 explained 66% of the variance. Anger, fear and disgust loaded above 0.5 in one component (hereafter, ‘negative emotions‘), explaining 52% of the data, whereas interest, compassion and joy loaded above 0.5 in the second factor (hereafter, ‘positive emotions‘), explaining 48% of the data; Table S2. Sadness loaded below 0.5 in both factors, so we omitted it from further analyses. Using the ‘alpha’ function from the ‘psych’ package, we then estimated the internal consistency for the derived (latent) emotion factors via Cronbach’s alpha reliability coefficient—scores were 0.71 and 0.73 for negative and positive emotions, respectively, i.e., in the “acceptable” range (sensu Cortina 1993). Consequently, we calculated indices for positive and negative emotions derived from average scores from the associated emotions.
We further assessed the internal consistency of the four behavioural intention items (Table 1) using Cronbach’s alpha reliability coefficient values. They were 0.88 for all four items, so we created a composite scale for behavioural intentions from average scores of the associated four items.
After Shapiro-Wilk tests revealed that the distributions of the emotion and behavioural intentions variables were significantly non-normal, we used Wilcoxon rank-sum tests to compare emotions and behavioural intentions within (aim 1) and between (aim 2) samples.
Lastly, we fitted ordinal logistic regression models using the ‘polr’ function in the ‘MASS’ package (Ripley 2002) to understand the effects of the factored positive and negative emotions and of socio-demographic variables on behavioural intentions (aim 3). We tested for interactions between age and place of living and found none. For all statistical analyses, we applied a rejection criterion of α = 0.05.
Results
Data screening and transformation
In total, we received 711 responses (410 in 2021, 301 in 2022). We found no significant differences in socio-demographics between both sample populations (although living place did approach a marginally significant level of difference; Table 1). While this suggests that both sample populations were similar, it is noteworthy that both samples were more skewed towards younger, female and urban-dwelling people compared to the German population (population demographics in Germany in 2020/2021: females∼50.68%, individuals between 20 and 40 years∼24.5% (Federal Statistical Office of Germany, 2024)).
Table 2
Socio-demographic backgrounds of survey respondents in 2021 and 2022
2021 (n = 410) | 2022 (n = 301) | Chi-square (between years) | |
|---|---|---|---|
Gender | χ2 = 0.46, df = 2, P = 0.79 | ||
Female | 244 | 200 | |
Male | 114 | 96 | |
Other | 2 | 3 | |
Age | χ2 = 5.02, df = 4, P = 0.29 | ||
18–29 | 154 | 107 | |
30–45 | 72 | 76 | |
46–55 | 58 | 55 | |
56–65 | 47 | 40 | |
66+ | 30 | 21 | |
Living place | χ2 = 5.25, df = 2, P = 0.07 | ||
Urban | 249 | 227 | |
Suburban | 66 | 51 | |
Rural | 43 | 21 |
Comparison of positive and negative emotions towards bats (aim 1)
Respondents expressed positive emotions more strongly than negative emotions (Table 3). As an illustration, the average score of positive emotions (the index) was ca. 3.5 (on a scale from 1 to 5) in both years. Thus, positive emotions are felt not intensely, but rather moderately. In contrast, the average score for negative emotions was ca. 1.5. This means that respondents hardly express negative emotions towards bats. The differences between positive and negative emotions were large (i.e., two points on a five-point scale) in both years. The observed associations between socio-demographics and emotions include decreasing positive emotions and increasing negative emotions with increasing age (Table S3).
Table 3
Average score and standard deviation (SD) of positive and negative emotions towards bats and behavioural intentions indices in 2021 and 2022 samples. Analyses based on Wilcoxon rank-sum W and P-value marked in grey, significant results in bold
2021 | 2022 | ||
|---|---|---|---|
Emotions | Mean ± SD | Mean ± SD | Between years |
Positive emotion index (average scores from three positive discrete emotions) | 3.51 ± 1.29 | 3.46 ± 1.09 | W = 63,279, P = 0.10 |
Negative emotion index (average scores from three negative discrete emotions) | 1.42 ± 0.80 | 1.48 ± 0.87 | W = 55,782, P = 0.27 |
Positive versus negative emotions | W = 11,505, P < 0.001 | W = 4525, P < 0.001 | |
Behavioural intentions index (average scores from four items) | 2.84 ± 1.11 | 3.20 ± 0.96 | W = 46,457, P < 0.001 |
Shifts in emotions and conservation intentions during the COVID-19 pandemic (aim 2)
We found no evidence of shifts in (estimated at the level of the composite indices) negative (W = 63 279; P = 0.10) or positive (W = 55 782, P = 0.27) emotions between samples (Fig. 1; Table 3) despite observed shifts in discrete emotions (Figure S2). Behavioural intentions were significantly higher in the second sample than in the first one (W = 46 457; p < 0.001).
Fig. 1
Composite indices of positive and negative emotions and behavioural intentions compared between 2021 and 2022. Significant differences from Wilcoxon rank-sum test between years are marked with different letters. Means are marked with an “x”
×
![]()
Predictive potential of emotions and socio-demographics on behavioural intentions (aim 3)
All behavioural intentions were positively predicted by positive but not negative emotions (Table 4). As for socio-demographic variables, age was the only impactful one: participants in the 46 to 55 years old category were more in support of donations and motivating others when compared to the reference group (18–29 years old); respondents aged 66 and above were also more likely to support motivating others than the reference group was. In contrast, the 56–65 age group was less in favour of learning more about bats compared to the reference group.
Table 4
Effect size ± SD of positive (interest, joy and compassion) and negative (fear, disgust and anger) emotions and socio-demographics on behavioural intentions to get involved in bat conservation (four behavioural intentions), based on ordinal-logistic-regression analyses and generalised linear modeling (behavioural intention composite scale). Results that meet the threshold for statistical significance are highlighted in bold (significance levels shown as *** < 0.001, ** < 0.01 and * < 0.05), however it should also be noted that several other results reach a level of near significance but are not highlighted
Learn more Effect ± SD P-value | Engagement Effect ± SD P-value | Donation Effect ± SD P-value | Motivating others Effect ± SD P-value | Behav. intention (comp scale) Effect ± SD P -value | |
|---|---|---|---|---|---|
Emotions | |||||
Positive | 1.38 ± 0.10*** P< 0.001 | 1.16 ± 0.09*** P< 0.001 | 1.06 ± 0.09*** P< 0.001 | 1.31 ± 0.10*** P< 0.001 | 1.50 ± 0.09*** P< 0.001 -0.17 ± 0.12 P = 0.15 |
Negative | -0.15 ± 0.12 P = 0.24 | -0.25 ± 0.13 P = 0.05 | -0.02 ± 0.13 P = 0.88 | -0.21 ± 0.12 P = 0.09 | |
Socio-demographic variables Gender (compared to female) | |||||
Male Other | -0.12 ± 0.16 P = 0.47 0.68 ± 1.00 P = 0.50 | -0.05 ± 0.16 P = 0.74 1.04 ± 0.81 P = 0.20 | -0.04 ± 0.16 P = 0.81 0.37 ± 0.77 P = 0.63 | -0.30 ± 0.16 P = 0.06 -0.44 ± 0.85 P = 0.60 | -0.13 ± 0.15 P = 0.41 0.43 ± 0.75 P = 0.57 |
Age (compared to 18–29) | |||||
30–45 | 0.16 ± 0.20 P = 0.41 | -0.05 ± 0.16 P = 0.78 | 0.36 ± 0.19 P = 0.06 | 0.39 ± 0.20 P = 0.05 | 0.25 ± 0.19 P = 0.18 0.20 ± 0.20 P = 0.32 0.03 ± 0.23 P = 0.88 0.49 ± 0.28 P = 0.08 |
46–55 | -0.35 ± 0.22 P = 0.11 | -0.06 ± 0.21 P = 0.78 | 0.64 ± 0.21* P= 0.03 | 0.46 ± 0.22* P= 0.03 | |
56–65 | -0.67 ± 0.25** P= 0.006 | -0.02 ± 0.24 P = 0.93 | 0.34 ± 0.24 P = 0.15 | 0.39 ± 0.23 P = 0.10 | |
66+ | -0.33 ± 0.30 P = 0.26 | 0.13 ± 0.29 P = 0.65 | 0.58 ± 0.29 P = 0.05 | 1.25 ± 0.29** P= 0.001 | |
Living place (compared to urban) | |||||
Suburban | -0.21 ± 0.20 P = 0.28 | 0.19 ± 0.19 P = 0.32 | -0.09 ± 0.19 P = 0.62 | 0.16 ± 0.19 P = 0.41 | -0.04 ± 0.27 P = 0.89 -0.04 ± 0.23 P = 0.86 |
Rural | -0.24 ± 0.25 P = 0.34 | 0.21 ± 0.24 P = 0.38 | -0.01 ± 0.24 P = 0.97 | 0.20 ± 0.25 P = 0.41 | |
Discussion and conclusion
While the majority of previous studies have suggested that bats are the object of negative emotions and biophobia, they mainly focused on negative emotional responses (e.g., Davey 1994; Ware et al. 1994, Arrindell 2000; Polák et al. 2019) with only a few focusing on positive and negative ones (e.g. Vincenot et al. 2015, Musila et al. 2018). As such, our understanding of the full spectrum of human-bat relationships remains limited. We studied positive and negative bat-related emotions in relation to each other and in response to the COVID-19 pandemic. Our findings suggest that positive emotions matter. First, positive emotions towards bats are felt more intensely than negative emotions (which are hardly felt at all). This is important because positive emotions predict bat-conservation intentions much better than negative emotions do. In addition, in contrast to our expectations, we observed no increase in negative emotions in our study population over the course of the pandemic. Interestingly, we also found that the strength of behavioural intentions increased between samples. Taken together, these results suggest two conclusions: First, at least in this population, bats are not objects of widespread biophobia. Second, even if the moderately positive emotions expressed by this population are not sufficient to constitute love of bats, they still underscore how important it is to measure both positive and negative emotions, particularly given the association between positive emotions and intent to adopt conservation-relevant behaviors.
To fully appreciate our findings, their potential interpretations and their implications for social science and conservation, it is vital to consider the study context. Most of our respondents were young, female and living in urban environments and hence the sample is not representative for the German public. Also, our use of convenience sampling presumably led to some self-selection in the constitution of samples potentially leading to a higher proportion of participants already interested in bats participating in the survey. A repeated-measures design could reduce the potential confounding effects of individual variation, as demonstrated in a study of disgust during the COVID-19 pandemic (Schwambergová et al. 2023). While we are confident in the broad pattern of our results, we also note that these practical constraints could have influenced our observations. However, while generalizing to the broader German public is not possible with our convenience sample, they allow valuable opportunities for testing relationships as in our case the effect of emotions on conservation behavioral intentions.
Although we are unaware of systematic research into associations between demographics and emotions towards wildlife, the literature on wildlife value orientations (basic beliefs that give meaning and direction to fundamental values in the context of human-wildlife interactions) might present a model for informed reasoning about these associations. Two main wildlife value orientations have emerged: domination (wildlife exists for humans to use) and mutualism (an egalitarian orientation translating into willingness to coexist; Jacobs et al. 2018). Research in various countries suggests that mutualism orientation increases among: urbanites, younger people, females and people who grew up in more affluent circumstances (Manfredo et al. 2009; Vaske et al. 2011; Gamborg and Jensen 2016; Jacobs et al. 2022). Because mutualism-oriented individuals are more disposed to expressing positive emotions towards wildlife compared to domination orientated people (Zainal Abidin and Jacobs 2019), we believe our sample may be biased toward positive emotions.
The above-mentioned effect may be magnified by self-selection. Meaning, people who like bats might have been particularly keen to take our survey. In addition, we labeled interest as a positive emotion based on emotion theory and research (Izard 2007; Ortony 2022), even though it does not occur in some well-known lists of basic emotions (e.g., Ekman 1999) and is sometimes treated as a component of attitudes (e.g., scientific attitudes and an interest in the biology of bats; Musila et al. 2018). Besides the theoretical argument to combine interest, joy and compassion into one measure of a latent construct (which we labeled as positive emotions; see methods section), our statistical analyses support doing so (i.e., the factor loading was > 0.5 and the reliability index was 0.73). In other words, our measurements of interest and of joy and compassion behaved the same way. Removing interest from the positive-emotion scale results in a marginal reduction in mean scores (i.e., year one to 3.30 and year two to 3.07) but does not alter the pattern of results or our conclusions.
Another important aspect of context is that European bat diversity is low—all but five species belong to two families: Vespertilionidae and Rhinolophidae (Russo et al. 2016). Furthermore, all European bats are protected by strong conservation laws. Essentially, it is illegal to kill, keep or exploit bats, to damage or destroy their habitats or to disturb them during the breeding, migration and hibernation seasons. In Germany, bat species are “strictly protected” (Federal Nature Conservation Act, BNatSchG), and five species require the designation of “special protection areas” (annex II, Council Directive 92/43/EEC on the Conservation of natural habitats and of wild fauna and flora). Finally, since 1997, the annual European “Bat Night” has been educating people in more than 30 countries about bats. In all these ways, Europe’s context is one of longstanding efforts to protect its remaining bats and bring people closer to them—it might have contributed to a positive image of bats among the public. We did explicitly not ask for negative behavioural intentions in our survey due to the strict protection of bats in Germany. Addressing such sensitive questions would have required an alternative approach (Tourangeau and Ting, 2007). However, as demonstrated by Vincenot et al. (2015) in their study on flying foxes in Japan, negative emotions or a lack of willingness to protect should not be considered equal to tolerance for killing.
Another potentially relevant contextual aspect is public experience with various, economically impactful human-bat conflicts (e.g., vampire bats biting and transmitting rabies to livestock; Reid 2016; or frugivorous bats damaging fruit crops; Aziz et al. 2016), that occur in many regions, especially in the tropics. Human-bat conflicts are much rarer in Germany and mainly involve bats residing in buildings. Germany even has formal services (e.g., Landesbund für Vogelschutz, NABU, BUND) to help people deal with bats in buildings. In countries such as Germany, where bats appear to be viewed positively, or China, where bats hold cultural significance, bat-related emotions may have been uniquely affected by COVID-19 pandemic (Lu et al. 2021), and the global response of conservationists may have reduced bat persecution during the pandemic (Nanni et al. 2022). The underlying factors responsible for such variation (e.g., regulatory measures or other influences) remain unclear. Nevertheless, it would be worth investigating whether and how bat-related behaviours and emotions shift in response to major events within countries that have divergent (i.e., positive vs. negative) views of bats, as such studies could highlight the intersection of cultural context, events and public reactions. It is important to realize though that being considered as attractive or cute by humans or holding significant cultural values does not necessarily predict or guarantee a species’ protection from conservation threats.
For all these reasons, we must be cautious and avoid overgeneralising our findings to represent the emotions held by the German or any other population. Still, we found no association between demographics and conservation intentions and hardly any between demographics and emotional dispositions. The only demographic variable that mattered was age: older people felt less positive and more negative emotions relative to younger people; albeit within a generally young study population (Table S3). While Lundberg et al. (2021) call for further investigation into the role of age in shaping attitudes and emotions towards bats, we also propose a nuanced consideration of cultural context, which may influence what people think and feel about bats (e.g., Low et al. 2021). Therefore, future research should explore the interplay of age, cultural factors and emotional responses to gain a comprehensive understanding of the subject. Also, the differences between positive and negative emotions were large (two points on a five-point scale). As such, we speculate that in other samples and contexts, positive emotions may also prevail over negative ones, but that the differences might be smaller than those we observed. What is needed now is research using representative samples of other populations to assess support for this hypothesis, particularly from regions with greater diversity of species, more frequent human-bat encounters, and weaker protection for bats.
Besides asking whether the relationships among concepts in our study are observable elsewhere, we urge researchers who seek to understand the gamut of human-bat relationships not to be guided by traditional assumptions that bats are the objects of negative appraisals (be they emotions, beliefs, norms or values). Negative appraisals may well prevail in certain contexts, but our study at least provides convincing evidence that this is not always true and adds to indications (e.g., Low et al. 2021) that bats are positively regarded in many societies. Conservation social science is often motivated by the desire to solve conservation problems (Bennett et al. 2017) and this may also be a key motivation for funders of such research (Manfredo 2008). So, the prevailing focus on problematic relationships (e.g., conflicts) that could hamper public support for conservation is understandable (Buijs & Jacobs 2021). Yet, as our findings demonstrate, this focus can easily create a blind spot when it comes to capitalising on positive emotions to foster conservation.
In our view, conservation practitioners should keep modest expectations about opportunities to influence the public’s emotional dispositions towards wildlife for the sake of conservation because opportunities to do so are limited. The mechanisms that underpin emotional dispositions toward wildlife can be diverse and include genetically predisposed and therefore innate tendencies, conditioning and culturally shared narratives (Jacobs 2012). Some of the mechanisms manifest in regions of the brain that are insensitive to linguistic input. Also, because emotional dispositions might be central to one’s identity, they may resist change, especially after one’s formative years. In addition, with the thousands of messages a person receives daily, any one message is highly unlikely to drastically change emotional dispositions.
A more effective avenue to promote bat conservation may be to capitalise on people’s existing emotional dispositions. For example, there is experimental evidence that telling whale-watching tourists stories emphasising positive emotions (as opposed to facts about whales or human responsibility for whales) had a more drastic and widespread effect on their conservation intentions (Jacobs and Harms 2014). Since (in our study), positive emotions predict bat conservation intentions, we recommend testing the impacts of narratives designed to activate these emotions. For example, researchers could ask: What gets people to act meaningfully to conserve bats? Stories that evoke positive emotions, stories that aim to reduce negative emotions, stories that communicate facts or no stories at all? Positive narratives can be expanded to encompass not only the ecosystem services that bats provide, but also their sentience and individuality. Narratives might emphasise the attributes that make bats unique within mammals (e.g., powered flight, sophisticated echolocation for navigation and prey detection) and the ones that point to their kinship with people. Both types of narratives can evoke an appreciation for bats’ complexity and even personal identification (Rogge et al., unpublished data). It is important to recognise that different narratives may resonate with different individuals, allowing for a diverse range of perspectives and connections to be formed. MacFarlane and Rocha (2020) also provide guidelines for communicating about bats out of concern about potential persecution related to the COVID-19 pandemic.
We used the discrete emotion perspective to study emotions toward bats. The two factors that emerged from exploratory factor analysis reflect positive and negative emotions. Yet, this does not imply that our composite indices reproduce valence: the pleasant-unpleasant dimension in the dimensional approach. The concept of valence reflects one continuum, with positive and negative as opposite ends (Russel & Barrett 1999). Our analysis suggests that positive and negative emotions are largely independent dispositions (as a logical consequence of factor analysis). Ancillary analysis revealed that the association between the positive and negative emotion indices (Spearman’s rho) was − 0.39 (P < 0.001). This suggests that most of the covariance is explained by independent variation between these two dimensions. Theoretically, a person can simultaneously be both disgusted by and joyous about bats. After all, bats have many traits, stories about them emphasise many different aspects, and the human mind is complex and able to incorporate different perspectives. Therefore, a disadvantage of the dimensional approach is its inability to capture the simultaneous coexistence of positive and negative emotional dispositions within individuals. So, we recommend using the discrete approach when measuring the potential multidimensionality of emotions (and note that sadness did not load on either factor in our study and thus could be considered a third dimension).
As mentioned, with few exceptions, the literature on human-bat relationships overwhelmingly focuses on negative emotions, e.g., conflicts (see also Straka et al. 2021). We found great utility in using Ekman’s (1999) and Izard’s (2007) measures of discrete emotions. While other studies have explored related emotions and attitudes, comparing emotions towards bats across diverse cultural contexts remains challenging. Therefore, employing consistent emotion measures would facilitate comparative studies on emotions towards bats. Our study demonstrates that, at least in some cases, positive emotions toward bats prevail in both occurrence and influence on conservation intentions. All while reiterating our caution against over-extrapolation to other contexts, we point out that the difference is large enough to question the emphasis on negative appraisals and to suggest that perhaps people love bats more than researchers tend to believe.
Acknowledgements
We thank all survey participants who participated in our two surveys and Ricardo Rocha for helpful suggestions that improved this manuscript.
Declarations
Research ethics
There were no institutional requirements for ethical clearance. However, the survey was undertaken in accordance with the European Union’s General Data Protection Regulation. Participants were given a consent form that ensured their anonymity and provided information about the general aim of the study, the data that will be collected and investigators’ contact information, and assured them that there would be no disadvantage for participants if they withdrew from the study at any stage of their participation. Participants had to provide consent and be at least 18 years to start the survey.
Conflict of interest
The authors declare no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
