“Are there caimans here?” Influence of multiple drivers on local ecological knowledge about an apex predator

Local Ecological Knowledge (LEK) regarding apex predators, such as crocodilians, is crucial for understanding human–wildlife dynamics and supporting conservation strategies. This knowledge is not uniformly acquired but is predicted by key variables. Direct contact through activities such as fishing provides experiential understanding of species ecology. Age generally facilitates knowledge accumulation, although this relationship is not strictly linear and may be affected by knowledge erosion in older individuals. LEK is also shaped by the interplay between formal education, which can augment understanding, and culturally embedded knowledge transmission modes, in which vertical, intergenerational learning coexists with the growing influence of mass media. We analyzed how emotion, contact, age, education level, and knowledge transmission mode influence LEK concerning the biology and ecology of the broad-snouted caiman (Caiman latirostris). The study was conducted around the Tapacurá Reservoir in northeastern Brazil, which sustains a structured population of this species. We interviewed 61 adults (19–79 years old) from local communities, predominantly fishermen and farmers, ensuring the inclusion of respondents with frequent direct contact with caimans and their environment. The mean LEK score was 5.1 ± 2.9 (scale from 0 to 10.5). Age and education were the main factors associated with LEK: young people (<30 years old) and individuals with higher education had higher LEK, while contact and emotion did not significantly influence it. Negative emotions (72%) were related to cultural and educational factors, with no direct link to local incidents. Television was the main transmission route (43%), followed by horizontal transmission (33%), with the latter occurring especially among older adults. Young people showed greater influence from mass media, while traditional modes, such as vertical transmission, had a greater impact on older adults. The observed average LEK can be attributed to urbanization and the low dependence of the community on resources associated with caimans. These findings show that cultural changes, urbanization, and modernization can shape LEK, highlighting the importance of adapting to new transmission dynamics to ensure conservation. Participatory strategies, such as environmental education programs, are essential to value LEK, include it in the school curriculum, promote its integration with protected areas, and strengthen the cultural link with local biodiversity.

The human–caiman relationship dates back to the historical coexistence of these groups (Barboza et al., 2021) and can be both harmonious and conflicting (Barboza et al., 2021; Cavalier et al., 2022; Ojeda-Adame et al., 2023). In this context, the broad-snouted caiman (Caiman latirostris) emerges as a medium-sized crocodilian species with a low risk of accidents involving humans, but no less susceptible to conflicts than other species, particularly given its history with fishers (Verdade et al., 2010; Mascarenhas-Júnior et al., 2018). Like other caimans, females build their nests on land during the nesting period, lay eggs, and usually provide parental care against predators (Campos and Mourão, 1995; Rodrigues et al., 2021). It is during this period that local residents in certain areas supplement their diet with caiman eggs (Torralvo et al., 2017). Knowledge about these animals can be constructed and shaped through lived experience and under the strong influence of cultural, physiological, and psychological aspects, among others (Alves et al., 2012; Santoro et al., 2018). In this context, local ecological knowledge (LEK) emerges as a set of “knowledge, practices and beliefs developed by adaptive processes and transmitted through generations, through cultural transmission, primarily oral, in the relationships between living beings and their environment” (Berkes, 1999; Berkes et al., 2000). LEK can have positive or negative relationships with several variables, such as emotion (Castillo-Huitrón et al., 2020), age (Knoop et al., 2020), education (Reyes-García et al., 2016), gender (Hanazaki et al., 2000; Torres-Avilez et al., 2016), mode of knowledge transmission (Henrich and Gil-White, 2001), social relationships, length of residence (Iniesta-Arandia et al., 2014), homophily (Bond and Gaoue, 2020), socioeconomic profile (Voeks and Leony, 2004; Voeks, 2007), and human nutritional status (Reyes-García et al., 2008), among others.

To date, studies on LEK related to crocodilians are generally descriptive and do not evaluate hypotheses (Zamudio et al., 2013; Barboza et al., 2021). Balaguera-Reina and González-Maya (2010) and Valdelomar et al. (2012) provide in-depth results on reproductive aspects of Caiman crocodilus fuscus and Crocodylus acutus, while Zamudio et al. (2011) report LEK without relating it to knowledge predictors, and Ojeda-Adame et al. (2023) primarily offer a descriptive, emic perspective on the LEK of Crocodylus acutus. Pooley (2016) and Cavalier et al. (2022) provide historical accounts of interaction and coexistence between humans and crocodilians, but these remain incipient and scarce. Nonetheless, these studies do not explore hypotheses regarding variables that influence LEK of biological and ecological aspects of crocodilians, with an emphasis on reproduction. To fill this gap and understand how specific predictors affect LEK, we address the following question: how do predictive LEK variables influence LEK about the broad-snouted caiman? To this end, the present study focused on biological and ecological aspects, with an emphasis on reproduction due to its marked presence in the terrestrial environment during the nesting period, which is of great importance in the caiman life cycle. In this study, five variables were considered due to their influence on LEK (Reyes-García et al., 2016; Castillo-Huitrón et al., 2020; Zhang et al., 2020; Costa et al., 2023): emotion, contact, age, education, and knowledge transmission mode.

Emotions are human mental states or expressions closely related to individual experience and physiological and psychological reactions, producing feelings that are incorporated into culture (Frijda and Mesquita, 1998; Castillo-Huitrón et al., 2020). Human emotions toward wild animals can be associated with different perspectives, including importance, affection, and conflict (Castillo-Huitrón et al., 2020). For example, large predators can elicit positive feelings (e.g., happiness, admiration, affection, or respect) or negative ones (e.g., anger, fear, disgust, or sadness). These feelings may also be directly related to other factors, such as animal appearance (e.g., coloration, skin texture, body shape, exposure of teeth, presence or absence of fur, life stage—juvenile or adult); activity patterns and stakeholder perspective (e.g., a hare perceived as charismatic by a city dweller but as a pest by a farmer); potential risks to human health and safety; historical construction of the relationship (e.g., large predators, venomous animals, competition for fish with caimans, or slaughter of caimans for livestock consumption); cultural contexts (e.g., religious influence on negative attribution to snakes); and evolutionary contexts (Prokop and Fančovičová, 2010, ProkopFančovičová, 2013; Castillo-Huitrón et al., 2020; Barboza et al., 2021). Zajonc (1980) states that both positive and negative intense feelings and events promote learning, thus increasing the cognitive repertoire. Relationships between humans and crocodilians in Brazil are generally classified as negative due to hunting, incidents with riverside dwellers, and conflicts in fishing and livestock activities (Barboza et al., 2013, Barboza et al., 2021; Lourenço-de-Moraes et al., 2023). When these relationships are reported as negative events in the news and emphasized by mass media and folklore (Notaro and Grilli, 2022), the emotions pervading the general public, particularly local populations, are expected to be predominantly negative.

Contact consists of personal experience with the “study object” and involves various aspects of the relationship, including interaction, intensity, frequency, and sensory modalities involved. This can be explored in greater detail through the type and level of experience (e.g., fishing, hunting, extractivism, mystical or religious rituals), location (e.g., city, countryside, sea), length of residence, and occasional experiences related to tourism and environmental education (Walter, 2009; Cheng and Wu, 2015; Torres Junior et al., 2016; Alves and Albuquerque, 2017; Sampaio et al., 2018; Joa et al., 2018; Sabinot and Lescureux, 2019; Cebrián-Piqueras et al., 2020). Individuals who have direct contact through proximity, interaction, conflict, or direct use of animal resources (such as hunters and fishers) tend to have considerable knowledge about animal ecology and behavior, as capture success is intrinsically related to ecological knowledge (Røskaft et al., 2003; Rosa et al., 2005; Nordi et al., 2009; Alves, 2012; Alves and Souto, 2015).

Like contact, age plays an important role in LEK (Souto and Ticktin, 2012). Studies show that LEK is often directly proportional to age (Bortolotto et al., 2015; Geng et al., 2016); however, this positive relationship is not universal, as differences between age groups may arise due to learning dynamics, knowledge erosion, or memory loss in older individuals (Almeida et al., 2012; Ferreira Júnior et al., 2013; Torres-Avilez et al., 2016). Age also allows for different perceptions among informants regarding population information and behavioral changes of the target species, or even due to the reference points each generation previously experienced (Shifting Baselines Syndrome—Pauly, 1995).

Another important variable is the level of formal education. In addition to knowledge acquired through personal experience or orally from family members and close social networks, schools play a role in providing information and developing skills (Reyes-García et al., 2016; Zhang et al., 2020). LEK about a primate species has been shown to be positively associated with education level, indicating that higher levels of formal education are linked to greater LEK, which in turn is related to more favorable conservation outcomes (Zhang et al., 2020). However, the relationship between education and LEK is complex; in multicultural countries such as Brazil, regional cultural particularities should be incorporated into school curriculum, since, according to Reyes-García et al. (2010), the traditional format of schools promotes a loss in LEK, leaving aside these particularities.

Within the context of learning, knowledge transmission emerges as a key predictor of LEK (Mesoudi, 2019). Individual knowledge construction occurs primarily through direct experience, while transmission occurs through interaction between individuals, also referred to as social learning (Heyes, 1994; Mesoudi et al., 2006; Soldati, 2015). Knowledge transmission processes can be related to content, context, and mode, including number of participants (one-to-one, many-to-one) and direction (horizontal, vertical, oblique) (Eerkens et al., 2014). Horizontal, vertical and oblique transmission are classic modes of cultural transmission that describe how knowledge moves within and between generations. Vertical transmission occurs when knowledge is passed from parents or primary caregivers to children. Oblique transmission involves adults of the parental generation who are not the learners’ parents (e.g., teachers, elders, or community leaders). Horizontal transmission refers to knowledge sharing among individuals of the same generation, such as peers or friends. Studying knowledge transmission is increasingly challenging due to the presence of multiple non-local mass transmission channels, such as television, the internet, books, radio, schools, associations, and cooperatives, which operate alongside traditional mechanisms (Leonti, 2011; Reyes-García et al., 2016). In addition, different age groups may be influenced differently by specific transmission modes, as proposed in multistage learning models (Reyes-García et al., 2016).

Given that LEK is an important ally for conservation (Hallwass et al., 2020), it can help elucidate people’s perceptions of large predators. However, to date, there are no studies assessing the aforementioned variables in relation to LEK about crocodilians. We hypothesize that emotion, contact, age, education level, and knowledge transmission mode directly influence LEK about reproductive and behavioral biology of the broad-snouted caiman. Specifically, we expect that individuals with extreme positive (e.g., charisma) or negative (e.g., fear) emotions will exhibit higher LEK, as they may observe caimans more closely, better identify their presence and behavior, and thus acquire greater knowledge. We also expect that individuals engaged in river- and forest-related activities, with greater contact with caimans, longer residence time, older age, and higher education levels will have higher LEK. Finally, we anticipate that knowledge transmission will influence LEK primarily through vertical transmission from parents, and secondarily through mass media, such as the internet and television, due to increased exposure to these technologies.

Study area

The study area comprised the surroundings of the Tapacurá reservoir, built in the 1970s, located in two municipalities (São Lourenço da Mata and Vitória de Santo Antão) in the state of Pernambuco, Brazil (8°03 ’ S and 35°10 ’ W) (Almeida and Oliveira, 2009) (Figure 1). The region is characterized by semideciduous forests with a tropical climate and an annual average rainfall of 1,300 mm (Moura et al., 2012). Although both municipalities are predominantly rural, one of them encompasses part of the metropolitan region of the state capital, located approximately 60 km from the reservoir (de Azevedo et al., 2024).

The area includes private properties where various agricultural activities, artisanal fishing, and extensive cattle and goat farming are performed, as well as remnants of the Atlantic Forest that comprise three state protected areas (Moura, 2018). Agricultural activity includes short-cycle crops (beans, corn, pineapple, and pumpkin) and fruit growing (acerola, Malpighia emarginata), while fishing mainly occurs using gillnets to capture the main commercial species (Plagioscion squamosissimus, Oreochromus niloticus, and Astyanax spp.) (Mascarenhas-Junior et al., 2024). The study area comprises the localities of Ato Grande, Campo Alegre, Barro, Oiteiro, Engenho Poço Sagrado, Santo Antônio PA, São Bento, and Veneza.

Participant selection

It is common to use calculations considering the total population of the study area to determine sample size (Albuquerque et al., 2014; de Souza et al., 2022). However, because households surrounding the reservoir are located in rural areas spanning different municipalities, no disaggregated population data are available. Therefore, we conducted a frame-type sampling to obtain a more accurate estimate of the study population. This consisted of a visual survey using satellite images of residences within a radius of 1.5 km from the reservoir shoreline, followed by interviews with residents of these residences (Shannon et al., 2012; Escamilla et al., 2014) (Figure 1).

We obtained an estimate of 337 buildings within the study area. Many of the identified structures were not residences but summer homes, garages, stables, farm support houses, churches, businesses, schools, or houses that were unoccupied during the data collection period. Residences were visited to recruit participants, excluding those that were inaccessible by river due to the massive presence of aquatic macrophytes. Eligibility criteria included residing within the established radius of 1.5 km from the Tapacurá reservoir shoreline or immediately adjacent to this limit and being 18 years of age or older (adults according to Brazilian law). Interviews were conducted individually, with only the participant and the interviewer present. Exclusion criteria included children, adolescents, and individuals not residing in the area during the study period.

Procedure

Interviews were conducted between March and May 2022 by one of the authors, with assistance from a student (C.A.S.). Access to households was primarily via river through the reservoir using a small boat, with support from a local resident (R.S.), followed by land access on foot. Due to this resident’s knowledge of the area and the local community, he assisted in recruiting participants and establishing a relationship of trust with interviewees. Participants were approached personally by the researchers, mainly at their homes. During the first visit, residents were invited to participate in the interview. If they were unable to participate at that time, a second visit was scheduled.

Data were collected through semistructured interviews (Costa et al., 2023; Silva et al., 2019) investigating socioeconomic factors such as age, education, and contact (direct experience with caimans through occupation and length of residence), as well as participants’ knowledge of the topic, taking into account the source of knowledge (personal experience, family, friends, reports, school, books, television, internet, etc.).

The questionnaire consisted of 15 questions and subquestions (Supplementary Materials 1, 2), organized into the following sections: (i) expression of basic feelings through two questions, one open-ended and one with predetermined categories (negative, indifferent, and positive) (Ekman and Cordaro, 2011; Matas, 2018); (ii) contact with caimans, assessed through time spent living in the area and occupation-related activities associated with caimans and their environment (e.g., agriculture, public service, fishing, hunting, medicinal use, collection of firewood, timber, or honey); (iii) age; (iv) education; and (v) knowledge transmission, assessed whenever the interviewee referenced a source of LEK. Finally, we obtained information about the LEK (response variable) by asking questions related to nesting and parental care of caimans in the reservoir, always letting the dialogue flow as the interviewee answered.

Statistical analysis

We created an index based on interview responses to assess LEK level (Nazario and Bittencourt, 2003), assigning each response a value according to the respondent’s self-reported certainty (0 = no answer/does not know (e.g. “ I don’t really know, I’ve never seen a caiman’s nest ”); 0.5 = uncertain answer (e.g. “ I reckon they lay eggs like turtles do, but I’ve never seen it myself ”); 1 = certain answer (e.g. “ The female stays by the nest, leaves to eat, and comes right back ”) (Cebrián-Piqueras et al., 2020; Costa et al., 2023). This scoring system is a good proxy for LEK (Mikołajczak et al., 2021), as it reflects whether participants were fully confident (e.g., based on direct experience), hesitant (partial or indirect knowledge), or lacked knowledge entirely. Certainty level was self-reported, reflecting whether participants had directly witnessed the phenomenon or obtained information through other sources (e.g., other people, internet, television, books, or school, etc.). Finally, we calculated the scores for each interviewee by adding the value assigned to each response based on the maximum score value (11), arriving at the index. We classified the LEK index as low (0-4.5), medium (5-8) and high (8.5-11).

Variables such as age, education level, and length of residence were categorized to facilitate interpretation and comparison with similar studies using standardized groupings (Costa et al., 2023; Zhang et al., 2020). Categorization also enabled the application of nonparametric tests when normality assumptions were violated. Categories were defined based on established sociodemographic benchmarks, such as legal adulthood, elderly age classification, completion of compulsory education, and retirement age. Emotion was categorized as negative, indifferent, or positive (Riaz et al., 2018). Following Costa et al. (2023), contact was categorized as yes or no for activities related to the caiman’s environment; length of residence as 0–10 years, 11–20 years, or more than 20 years; age as young (<30 years), middle-aged (30–60 years), or older (>60 years); and education level as no education, incomplete or complete elementary education, or complete secondary education (high school).

Normality was assessed using the Shapiro–Wilk test, and homogeneity using the Bartlett test (Table 1). Differences in means for independent variables were analyzed using Student’s t-test when normality assumptions were met, and the Mann–Whitney test otherwise. Variables with more than two categories were analyzed using ANOVA followed by Tukey’s post hoc test when homogeneity assumption was met. When the assumption was not met, the data were assessed using the Kruskal–Wallis test followed by Fisher’s LSD post hoc test (when the assumption was violated). We used multinomial logistic regression to perform univariate and multivariate analysis to provide better interpretation of the results due to the richness of the elements provided. The magnitude of these associations was estimated with Odds Ratio (OR) using 95% confidence intervals. Statistical significance was set at 5%. Analyses were conducted using R Core Team (2023).

Because the emotion variable comprised two questions, the closed question was analyzed as described above, while responses to the open question were represented using a word cloud (www.wordclouds.com). The results are presented visually, distinguishing the words used with the highest and lowest frequency of citation (font size) within a scale of positive and negative emotions (in colors) (Than et al., 2022; Ostrovski et al., 2021; Wyndham and Park, 2022).

Only 42% of interviewees answered questions related to knowledge transmission, limiting statistical analysis for this variable. Therefore, citation frequencies for each transmission mode were calculated and results were presented graphically (Salali, 2017).

The sampling design, statistical methodology, and analysis were developed in collaboration with a professional statistician (co-author G.T.N. Diniz), ensuring methodological rigor. Interpretation considered not only statistical significance but also effect size and test power (Cumming, 2014; Halsey et al., 2015; Wasserstein and Lazar, 2016). Values such as p = 0.058 (education) and p = 0.081 (age) were interpreted as indicative trends due to their proximity to the conventional alpha threshold and elevated odds ratios. Given the relatively small sample size (n = 61), the risk of Type II error was acknowledged. Accordingly, these results were interpreted cautiously, avoiding definitive claims while recognizing their potential explanatory value while avoiding definitive claims. This approach aligns with a comprehensive inferential analysis framework that prioritizes effect estimates and confidence intervals over rigid significant/non-significant dichotomies (Amrhein et al., 2019).

A total of 61 adults were interviewed, comprising 48% (n = 29) men and 52% (n = 32) women, with ages ranging from 19 to 79 years. The number of years interviewees had lived in the study area ranged from 3 to 71 years, and their education level was mainly incomplete elementary school (67%; n = 41). The following economic activities were mentioned: agriculture (29%; n = 18), fishing (28%; n = 17), retirement (16%; n = 10), housewife (10%; n = 6), and others (16%; n = 10), such as caretaker, small business, beekeeping, cowboy, security guard, and service provision. Approximately 50% of interviewees performed two of the economic activities mentioned.

We calculated LEK within the previously defined ranking, with a variation from 0 to 10.5 and a mean of 5.1 ± 2.9. Age and education were the only variables that showed a relationship with LEK (Table 1; Figure 2). The LEK of young people (<30 years) was significantly higher than that of middle-aged individuals (30–60 years) (p = 0.042). Young people also tended to have higher LEK (p = 0.081) compared to older adults (>60 years) (Table 1; Figure 2A).

Participants with higher education levels (6.73 ± 2.51; high school) presented higher LEK (p = 0.058) compared with those with elementary education (4.65 ± 2.81; Table 1; Figure 2B). However, emotion (p = 0.396), contact through occupation (p = 0.781), and length of residence (p = 0.449) were not related to LEK (p > 0.05) (Table 1). Regarding contact, few people reported having eaten caiman meat (5%), only one interviewee (1.6%) reported hunting “on special occasions,” and four (6.6%) mentioned knowing that caiman fat could be used in traditional medicine to treat general pain, earache, and skin inflammation.

Age and education were mutually dependent (p = 0.013); the highest concentration of participants with the highest education level (high school) were young people (61.5%) (Table 2). Among young people, there was a gradual decrease in the proportion of interviewees from the high school level to the no-education level (0%). Older adults showed an inverse trend, with a gradual increase in the proportion of participants from the highest level (high school; 0%) to the lowest (no education; 29%). Middle-aged individuals showed the highest concentration in the lower classes, elementary (54%) and no education (71%).

Participants with a high school education were 11.6 times more likely to have high LEK compared with those with intermediate education (OR = 11.6/p value = 0.067. Figure 2. Supplementary Material 3). Young people were 10.6 times more likely to have average LEK compared with older adults (OR = 10.6/p-value = 0.056. Figure 2. Supplementary Material 3; multinomial logistic regression). Other associations observed in the univariate analysis were no longer present after adjusting for education and age in the multivariate model (p < 0.05) (Supplementary Material 3).

The most frequently answered LEK questions concerned the presence of caimans in the locality (95% of respondents), whether caimans reproduce by laying eggs (75%), and caiman vocalization (63%). In contrast, information about predators of caiman hatchlings was least frequently expressed with high self-reported certainty, reported by only 8% of respondents.

In response to the question, “Do you think humans interfere with nest building or hatchling care? How?”, just over half of respondents (55%) stated that humans do interfere, while roughly one third (32%) perceived no interference and a smaller proportion (13%) did not provide an answer. Among those who acknowledged interference, responses emphasized direct disturbance of reproduction and parental care, particularly through fishing activities (e.g., use of nets and noisy and physically invasive techniques), deliberate persecution and killing of animals, removal of eggs, and broader environmental degradation, such as deforestation and pollution. In contrast, negative responses often framed the environment as sufficiently large to dilute human impacts or justified human use of resources in the caimans’ area, revealing tensions between livelihood needs and perceived interference.

Although positive or negative feelings did not influence LEK (p = 0.396), representations of emotional valence from the open question were mostly clustered in negative feelings (72%), compared with positive and neutral feelings (28%) (Figure 3).

Knowledge transmission was most frequently cited through mass communication (television; 43%), followed by horizontal transmission (33%), which occurs among individuals of the same generation (Figure 4). Under horizontal transmission, participants reported, for example, hearing from friends that the caiman found in the area was the broad-snouted caiman, that it vocalized, and that it built nests to lay eggs. Notably, transmission citations differed by age group: horizontal transmission was mostly reported among older adults, whereas television was cited mostly among young people (Figure 5).

Herein, we sought to assess how the variables emotion, contact, age, education level, and knowledge transmission mode directly influence LEK regarding the biological and ecological aspects of the broad-snouted caiman. We found that LEK among residents around the Tapacurá reservoir was of an intermediate level and was mainly related to age and education level. Surprisingly, the emotion and contact variables did not show any relationship with LEK, although negative feelings were more prominent than positive ones. LEK transmission regarding information about caimans occurred most frequently via television, followed by horizontal transmission.

LEK among participants revealed an interesting relationship with age, in which young people exhibited higher LEK. Studies suggest that differences in LEK distribution across age groups may result from learning dynamics, knowledge erosion, or memory loss in older individuals (Almeida et al., 2012; Ferreira Júnior et al., 2013; Torres-Avilez et al., 2016). As with age, education level was shown to be an important factor in LEK construction. The dependence identified between age and education, whereby young people have the highest education levels, appears to reflect educational expansion across generations, as demonstrated by shifts in educational attainment over time. However, it is worth noting that formal schooling can also be a degrading factor in LEK when it does not consider elements of the culture and environment of each region (Reyes-García et al., 2010, Reyes-García et al., 2016). In contrast to such contexts, our results indicate that higher levels of formal education were associated with greater LEK, suggesting that schooling may represent an important pathway for transmitting cultural and environmental knowledge.

Crocodilians are often associated with negative interactions, eliciting feelings such as disgust, fear, and danger (Barboza et al., 2021). However, emotions and perceptions are not static and may vary over time. Notably, crocodilians ranked 15th in a survey of 69 countries assessing animal charisma, surpassing dolphins, koalas, and bears (Albert et al., 2018). Although the emotion variable was not related to LEK in our study, we observed a predominance of negative representations (72%). These feelings may be partly explained by biophobia, often described as the opposite of biophilia and broadly defined as an aversion shaped by animal appearance and perceived threat (Wilson, 1984). Biophilia can also be “artificially” constructed over time; as an example, whales were negatively judged as sea monsters for years, however they currently have an absolutely charismatic status (Bearzi et al., 2010; Mazzoldi et al., 2019). Despite the biophobia observed here, there are no records of incidents involving the broad-snouted caiman in its natural environment in Brazil (Barboza et al., 2021; Mascarenhas-Junior et al., 2021) or in the study area, suggesting that direct conflict history is unlikely to be the primary driver of negative perceptions. Negative feelings can also have other origins (Røskaft et al., 2003), such as lower education levels which can contribute to fear of wild animals (Soga et al., 2023), previous unpleasant experiences or experiences incorporated into culture as mechanisms of mass transmission (Mesoudi, 2016) and also urbanization and/or modernization processes (Júnior et al., 2016). Importantly, emotions interpreted as “negative,” such as fear, can also serve adaptive functions by acting as defense mechanisms against potentially dangerous animals (Ekman and Cordaro, 2011).

Given the importance of fishing in the study area, we expected that occupation type and length of residence would influence LEK (Nordi et al., 2009; Alves, 2012; Gandolfo and Hanazaki, 2014; Alves and Souto, 2015; Medeiros et al., 2016; Sampaio et al., 2018). As the area is predominantly rural, it was also expected that residents would exhibit less fear of wild animals due to greater contact and interaction (Røskaft et al., 2003; Castillo-Huitrón et al., 2020). However, the absence of an emotion effect on LEK may be linked to the lack of a contact effect, both in terms of occupation and length of residence length. Since the study region includes a reservoir that was created in the 1970s, our results showed that many residents come from other locations. Therefore, due to the low contact levels, especially from the low number of responses about the consumption of caiman meat consumption or medicinal use—suggests that caimans currently have little direct nutritional, medicinal, or economic importance to the local community.

The role of urbanization and local modernization in the study area is further illustrated by patterns observed in other systems, such as the inverse relationship between access to public services and local medicinal knowledge (Albuquerque and Alves, 2016; Júnior et al., 2016), which parallels the LEK patterns identified here. Although our methodology did not explicitly assess urbanization or modernization processes, their influence can be inferred from indicators such as infrastructure improvements (e.g., paved roads and electrification), proximity to urban centers, internet availability, building upgrades from traditional to modern construction, adoption of non-traditional livelihood strategies, and demographic trends showing increasing urban and decreasing rural populations in the municipalities studied (IBGE, 2023).

Narratives reported by participants portray humans as important sources of stress, injury, and mortality for caimans and emphasize invasion of animal space and territorial disruption as key elements shaping local understandings of human–caiman interactions. These responses indicate that LEK includes a clear perception of human activities as potential threats to caiman reproductive success, while simultaneously reflecting ambivalence and negotiation around coexistence within a shared landscape.

Knowledge transmission was diverse, with an emphasis on mass communication via television, followed by horizontal transmission; both modes are considered rapid forms of cultural transmission (Mesoudi, 2011). The distribution of transmission modes across age groups highlights the importance of generational analyses for understanding transmission dynamics. It is noteworthy that knowledge transmission, which was presumably an important factor that promoted greater LEK among young people, mainly occurred via television, revealing that this is still a relevant method of mass communication. Evaluating the transmission by internet is not an easy task because the means of transmission (such as smart TVs) are already interconnected, making it difficult. Our findings differ from those of Soldati et al. (2015), who identified vertical transmission via parents as the dominant mode. According to Okui et al. (2021), the content to be propagated depends on its ideal transmission mode; thus, there are themes which are not naturally transmitted through traditional channels. This pattern is reflected in our results, where different age groups were influenced by distinct transmission mechanisms (Reyes-García et al., 2016). Processes that reduce direct contact with the environment, particularly when combined with mass media exposure, may contribute to LEK decline (Ostrovski et al., 2021). Thus, higher LEK found at younger ages demonstrates the role of the knowledge transmission mode’s influence in society, as well as the importance of school (Zhang et al., 2020).

The interview-based data collection approach used in this study may have favored explicit forms of knowledge transmission (e.g., direct instruction), as such methods can underestimate implicit transmission pathways that require extended immersive observation (e.g., observational learning or imitation) (Heyes, 1994; Hewlett et al., 2011). Future studies could incorporate ethnographic approaches or systematic behavioral observations to better capture these processes. Despite the influence of education level on LEK, the method used to assess knowledge transmission suggests that school played a relatively minor role, alongside the internet, oblique transmission, and vertical transmission, among the six mechanisms identified in the Tapacurá reservoir area. As previously noted, the region exhibits considerable urbanization, which may contribute to local evasion due to discontinuation of the parents’ professions by young people (Hanazaki et al., 2000), or even due to the decline in resources (Merétika et al., 2010), and when associated with low education, can justify the average LEK value observed in the location. This discontinuity in the parents’ professions may be one of the reason for the low oblique and vertical transmission.

We evaluated a set of variables to explain their effects on LEK regarding the broad-snouted caiman in communities surrounding the Tapacurá reservoir. However, we recognize that several other factors exhibit complex, often non-linear relationships that can influence LEK in different ways, including synergistically. Our study provides a portrait that deepens understanding of the relationship between local communities and the broad-snouted caiman. We suggest that urbanization and modernization processes, negative emotions, and the low importance of direct use by the community contribute to the observed intermediate level of local ecological knowledge about caimans, which is positively influenced by age and education level. In this context, we also identified that knowledge transmission through mass media can play a relevant role, especially among younger people. Thus, we emphasize that this transmission mode may have ecological implications (Ferreira Júnior et al., 2022).

Processes of “acculturation” or cultural transformation that involve drastic changes require careful consideration to avoid undesirable cultural loss within communities, even when such processes occur naturally. Therefore, we suggest that future studies incorporate temporal perspectives and additional elements potentially related to knowledge transmission, particularly aspects associated with urbanization and modernization.

Taken together, our findings show that local ecological knowledge about Caiman latirostris can be reshaped by urbanization, schooling, mass media, and changes in knowledge transmission. This study contributes a hypothesis-driven understanding of how social and demographic factors influence human–caiman relationships in the Tapacurá reservoir region. From a conservation perspective, our results indicate that LEK can affect detection, tolerance, and potential conflict, with direct implications for the design of conservation and management strategies for C. latirostris. For local communities, our findings highlight that education, media, and shifting transmission pathways can both weaken traditional connections and create opportunities to reframe caimans as valuable components of local biodiversity, thereby informing participatory actions and environmental education programs that support both human well-being and conservation.

Understanding the social and emotional factors underlying complex human–animal relationships facilitates the development of effective conservation strategies. Efforts to promote the effectiveness of protected areas animals are used must ensure the continued quality of life of the people who live in these places and conservation of these areas, and this should be best achieved through strategies with participatory methodologies (Than et al., 2022). One such strategy is to expand LEK related to local natural resources by inserting and reinforcing environmental education themes in school curricula, as well as strengthening connections between local communities and protected areas by involving both adults and children in research and extension activities.