Predicting ranchers’ intention to kill jaguars: Case studies in Amazonia and Pantanal

https://doi.org/10.1016/j.biocon.2012.01.002Get rights and content

Abstract

The killing of jaguars by ranchers in Amazonia and the Pantanal is a major threat to the species. We used the Theory of Planned Behavior to examine the role of ranchers’ perceptions, norms, attitudes and intentions concerning jaguar killing, in determining their jaguar-killing behavior. We also investigated the influence of: (1) descriptive norm and social identity on ranchers’ intention to kill jaguars on their properties; and (2) the effect of perceptions of jaguar impact on human livelihoods (livestock and human safety), and of property size, on the variables that influence intention to kill. Results based on interviews with 268 cattle ranchers indicated that the impact of jaguars on livestock is not the only predictor of a rancher’s intention to kill jaguars. Fear, personal and social motivations, and internal and external barriers (e.g. lack of skills and force of law, respectively) to killing jaguars can also influence jaguar killing. The relative importance of these factors in determining intention to kill varies with region and affluence. We recommend ways of deterring jaguar killing behavior through communication interventions. In addition to the economic and legal incentives that have already been considered by conservationists, effective strategies to protect jaguars on privately owned land will need to address the social and psychological factors that determine the killing of jaguars by ranchers. Conservationists need to find and support ways to make jaguar killing not only unprofitable and illegal, but also socially and personally unacceptable.

Introduction

The most urgent issue in jaguar (Panthera onca) conservation is the killing of jaguars by humans (Zeller, 2007). Because jaguars are renowned for preying on livestock (Hoogesteijn, 2000), one might assume that people kill jaguars in an effort to reduce economic losses associated with livestock predation. Conservationists have approached jaguar killing within the framework of human–wildlife conflicts (Rabinowitz, 2005) and to date research and conservation efforts have focused on the ecological (Azevedo and Murray, 2007, Cavalcanti, 2008, Michalski et al., 2006) and economic (Silveira et al., 2006) dimensions of the conflict. Little has been done to understand the link between jaguars killing livestock and people killing jaguars. The killing of jaguars may not be strictly retaliatory and might have motivations besides the economic. If we are to curb the killing of jaguars, first we have to understand the underlying causes of this behavior. We used an approach that went beyond the usual human–wildlife conflict framework to examine the reasons – both related and not related to conflict – for ranchers killing jaguars in Amazonia and Pantanal.

The few studies that have attempted to address human–jaguar conflict from the human side have assessed attitudes towards jaguars among people that were directly involved in conflict with jaguars over livestock (Conforti and Azevedo, 2003, Palmeira and Barrella, 2007, Santos et al., 2008, Zimmermann et al., 2005). Unfortunately, the relationship between attitudes and behaviors may not always be strong or direct. Moreover, by limiting the study of attitudes to the context of human–wildlife conflict, factors involved in the decision to kill a certain species, but not directly related to the impact of that species on human livelihood, may be overlooked. Theoretical frameworks have been developed by social scientists to predict human behavior from attitudes (Ajzen, 1985, Fazio, 1986), and these could provide a useful structure for studying the killing of jaguars. Key aspects of such an approach are the proper measurement of attitudes and inclusion of additional explanatory variables. In this study, we develop a framework based on the Theory of Planned Behavior (TPB) (Ajzen, 1985) to explore the relationships between landowners’ perceptions of jaguar impact on human livelihood and their own jaguar killing behavior on the Amazon deforestation frontier and in northern Pantanal. We incorporate factors not related directly to the impact of jaguars on human livelihood, such as social motives for killing jaguars and perceived barriers to doing so.

The TPB proposes that human behaviors are governed not only by personal attitudes, but also by social pressures and perceived control over one’s own behavior. According to the TPB, the most proximal determinant of a person’s behavior is their intention to engage in that behavior (Fig. 1). In turn, behavioral intentions are influenced by three main factors: attitudes towards the behavior, subjective norms and perceived behavioral control (PBC). Attitude scores reflect an individual’s overall positive or negative evaluation of performing the behavior. Subjective norms represent an individual’s perception of whether people important to them would approve of them performing the behavior. Perceived behavioral control reflects the extent to which an individual perceives the behavior to be under their volitional control. Thus, according to TPB, people who have positive attitudes towards killing jaguars, think that there is normative support for killing jaguars, and perceive that they can easily kill jaguars (or pay someone else to do it for them), should have strong intentions to kill jaguars. In addition, to the extent that PBC is a proxy for the actual control (accurately accounts for both the internal factors [e.g. knowledge, skills, courage] and external factors [e.g. legal barriers, money, equipment, help from others]) needed to perform the behavior, it may also have a direct impact on behavior.

While the great majority of attitudinal studies conducted in wildlife or conservation research have assessed attitudes about the species in question (Bruskotter et al., 2007, Kaczensky et al., 2004, Lindsey et al., 2005), the TPB recognizes that attitudes will not predict behavior unless they are measured with corresponding levels of specificity: attitudes about objects (such as jaguars) will not necessarily predict behaviors (such as killing jaguars). In order for attitudes to predict behavior, the attitude and behavior must correspond on four levels of specificity: action, target, context and time. In this study, we could have had four different attitude objects: (1) killing (action), (2) killing a jaguar (action and target), (3) killing a jaguar on own property (action, target and context), and (4) killing a jaguar on own property in the near future, or the next jaguar that appears (action, target, context and time). Questions about attitude objects 1 and 2 should predict general intentions to kill jaguars in the future or indices of killing jaguars in the past, but would be poor predictors of the behavior described in attitude objects 3 and 4. Questions about attitude objects 3 and 4 are specific and would be expected to predict the behaviors described in the statements (Manfredo, 2008). We therefore assessed attitudes (and other TPB variables) related to the specific rancher’s behavior of “killing (action) the next (time) jaguar (target) that appears on his property (context)”.

A multitude of variables can be related to or influence attitude, subjective norm and PBC: age, gender, education, socioeconomic status, personality, emotions, general attitudes, religion, knowledge, past experience and so forth (Fig. 1). The TPB recognizes the potential significance of such background factors. Although reviews and meta-analyses have demonstrated broad support for the basic TPB (e.g., Ajzen and Cote, 2008, Armitage and Conner, 2001), it is acknowledged that for some behaviors and contexts, the inclusion of other variables may increase the predictive utility of the model. Descriptive norms and social identity are examples of such variables. Descriptive norms reflect an individual’s perception of whether other people perform the behavior in question (Cialdini et al., 1990). Descriptive norms describe what is typical or normal, and motivate action by indicating what is likely to be effective, adaptive and appropriate action (White et al., 2009). For instance, a rancher who believes that all his neighbors kill jaguars will feel motivated to do the same. Social identity is that component of an individual’s concept of himself that is derived from his knowledge of group membership, and the value and emotion attached to that membership (Tajfel, 1981). According to the social identity theory, people define and evaluate themselves in terms of distinct social categories [e.g. rancher, pantaneiro (i.e. native to the Pantanal)]. By allocating himself a particular social identity an individual is encouraged to accentuate both the similarities between himself and other group members, and the differences between himself and people outside the group (Fielding et al., 2009). A social identity approach assumes that if a certain behavior, for example killing a jaguar, is centrally linked to a social identity, then that behavior will be influenced by the norms of that social group rather than by the expectations and desires of generalized others.

Perceptions of conflict with jaguars (more specifically, perception of jaguar impact on livestock and on human safety) were expected to affect intention: people who perceive that their livestock, or safety, is threatened by jaguars are more likely to intend to kill jaguars (Hypothesis 1). However, we hypothesized that the killing of jaguars by ranchers is not strictly retaliatory, and that by taking into account factors that are not related directly to livestock depredation, or threat to human safety (i.e. subjective norm and PBC), the TPB would offer a more predictive model of intention to kill jaguars (Hypothesis 2).

During the preliminary stages of this study we gathered anecdotal evidence of a social dimension to jaguar killing, especially in the Pantanal surveyed site. Evidence included photographs of hunted jaguars and jaguar paw trophies displayed prominently in ranchers’ living rooms, boastful stories told by ranchers about their bravery in hunting jaguars, and repeated mention that killing jaguars is something that everybody does, or have done for generations, as an important element of the local tradition. Therefore, including descriptive norm and group identity in the model was expected to increase significantly the predictive utility of the TPB model (Hypothesis 3): ranchers who believe that other ranchers kill jaguars and identify themselves with those ranchers will have stronger intentions to kill jaguars themselves.

Attitude towards a particular behavior is measured using experiential items (i.e. how it feels to perform the behavior e.g. unpleasant–pleasant) and instrumental items (i.e. whether the behavior achieves something e.g. detrimental–beneficial and worthless–useful) (Ajzen and Fishbein, 1980). Retaliatory killing of jaguars is instrumental in nature: it is expected to bring the benefit of decreased livestock loss or increased human safety. Therefore, we predicted that attitudes towards killing jaguars would be affected by perceptions of jaguar impact on livestock and on human safety. Indeed, in the TPB, background factors are assumed to influence intentions and behavior indirectly by affecting attitude, norms and PBC: we predicted therefore that perceptions of conflict with jaguars would influence jaguar persecution indirectly by affecting attitude towards the killing of jaguars (Hypothesis 4).

Commercial, sport, and recreational hunting are prohibited in Brazil. Nonetheless, in remote areas of rural Amazonia and Pantanal people do not expect to be penalized for breaking the law. Expectation of law enforcement and perceived risk associated with breaking the law are probably lower among large, affluent, and consequently, more influential ranchers. Furthermore, jaguars are more likely to be present on larger properties and their landowners more likely to have the resources for killing jaguars. In summary, the larger the property the more volitional control its owner should have over killing jaguars. Therefore, we hypothesized that perceived behavioral control increases with property size (Hypothesis 5).

While hosting relatively large populations of both cattle and jaguars (Cavalcanti et al., 2010), the two study areas differ distinctly in some socioeconomic aspects. The Amazon frontier was colonized recently by people from different parts of Brazil, including areas where jaguars had long been extirpated. Small family-owned farms coexist with large commercial cattle ranches on the Amazon frontier. In contrast, Northern Pantanal is home to a few traditional families that own large ranches and have been raising cattle in the region for generations. In communities where residents come from different parts of a country, for example, in frontier areas, there may be no consensus on what are acceptable or unacceptable behaviors towards wildlife. In other words, descriptive norms are unclear. Even if there is some agreement, many people may not care about the opinions of others, and so the subjective norm and group identity are less likely to influence behavior. In contrast, where traditional communities are intact and there is collective thinking and articulation about what constitutes acceptable practice, subjective norms and group identity may be sufficiently powerful to maintain practices that are illegal but socially encouraged. Therefore, we hypothesized that the relative importance of the social determinants of jaguar killing and perceived control over this behavior would differ between the Amazon deforestation frontier and northern Pantanal (Hypothesis 6).

Section snippets

Study areas and participants

The Amazonia survey was carried out in the district of Alta Floresta, in the north of the State of Mato Grosso. Alta Floresta was founded in 1976 and colonized by migrant farmers from other parts of Brazil. Alta Floresta has approximately 49 000 urban inhabitants, and overall the district rears over 748 000 head of cattle (Brazilian Institute of Geography and Statistics, 2007). Cattle depredation by jaguars is regarded as a severe problem in Alta Floresta (Michalski et al., 2006), and persecution

Characteristics of sample

On average, respondents in the Pantanal (n = 48) were older (mean = 54.8 years, SD = 11.8, range = 27–77) than those in Amazonia (n = 220, mean 42.4 years, SD = 13.1, range = 17–82). In the Pantanal, all respondents were male, whereas 26.7% of Amazonian respondents were female. This sample properly reflects relevant characteristics of the study populations: young landowners are more common in the small family-owned farms of the Amazon agricultural frontier and women are rare in the large, remote ranches of the

Discussion and conclusions

This study shows that the impact of jaguars on livestock is not the only predictor of a cattle rancher’s intention to kill jaguars. Intention to kill jaguars is also determined by fear, personal and social motivations, and internal and external barriers to killing jaguars. The relative importance of these factors in determining intention varies with region and affluence. Conservation measures that aim to decrease jaguar persecution by ranchers will be more effective if these factors are taken

Acknowledgements

We gratefully acknowledge the advice of P. Johnson and the comments of S.E. Baker and A.J. Dickman. This study was possible thanks to support from Cristalino Ecological Foundation, Anglo American Brazil, Instituto HSBC Solidariedade, O Boticário Foundation, Rainforest Concern, Cleveland Metroparks Zoo, Woodland Park Zoo, Chester Zoo, Fauna & Flora International, the generosity of K. Duncan, and assistance from C. Vicenti and M.O. Petretto.

References (41)

  • Brazilian Institute of Geography and Statistics, 2007. Pesquisa nacional por amostra de domicílios: Volume Brasil 2007....
  • Cavalcanti, S.M., 2008. Predator–Prey Relationships and Spatial Ecology of Jaguars in the Southern Pantanal, Brazil:...
  • S.M. Cavalcanti et al.

    Jaguars, livestock and people: reality and perceptions behind the conflicts in Brazil

  • R.B. Cialdini et al.

    A focus theory of normative conduct: recycling the concept of norms to reduce littering in public places

    J. Pers. Soc. Psychol.

    (1990)
  • S. Clayton et al.

    Conservation Psychology: Understanding and Promoting Human Care for Nature

    (2009)
  • Dickman, A.J., Marchini, S., Manfredo, M., in press. The importance of the human dimension in addressing conflict with...
  • R.H. Fazio

    How do attitudes guide behavior?

  • K.S. Fielding et al.

    Integrating social identity theory and the theory of planned behavior to explain decision to engage in sustainable agricultural practices

    Br. J. Soc. Psychol.

    (2009)
  • Google Inc., 2009. Google Earth (Version 5.1.3533.1731) [Software]....
  • A. Herda-Rapp et al.

    Mad About Wildlife

    (2005)
  • Cited by (177)

    View all citing articles on Scopus
    View full text