The study was conducted in the southern Mayan Forest, within the Greater Lacandona Ecosystem in Chiapas, Mexico (Fig. S1). The region is dominated by tropical rainforest (Ibarra-Manríquez and Martínez-Ramos, 2002), but jaguar habitat is increasingly threatened by cattle ranching expansion. Jaguars in this area use both protected and ejido lands (see Fig. S1), resulting in human-jaguar conflicts with livestock ranchers in these interface areas (de la Torre et al., 2017).

Established in 2016 in response to increasing human-jaguar conflict in the Greater Lacandona Ecosystem, the “Jaguares de la Selva Maya” program aims to ensure long-term jaguar conservation by addressing key threats through collaborative initiatives with Natural Protected Area administrators and relevant agencies. From 2021 to 2024, the program implemented activities to promote human-jaguar coexistence which included: 1) Capacity building for livestock owners; 2) Jaguar predation deterrents; 3) Community-based monitoring of jaguars and their prey; 4) Environmental education; and (5) Informative talks (Fig. 1). A detailed description of the human–jaguar conflict mitigation program is provided in Appendix S1.

Fig. 1.

Flow chart illustrating the methodology followed in this study, which includes four main stages: (I) conflict level assessment; (II) intervention actions; and (III) evaluation. Intervention actions comprised: (1) capacity building; (2) predation deterrents; (3) community monitoring; (4) environmental education; and (5) informative talks. The evaluation was based on four aspects: (1) number of predation cases before and after the intervention; (2) cost–benefit analysis of predation deterrents; (3) jaguar occurrence in the intervened ranches; and (4) changes in jaguar acceptance before and after the intervention.

To monitor the effectiveness of deterrents, the team conducted bimonthly visits to cattle owners to record new predation events and assess the functionality of the predation deterrents which included electric fences and night enclosures (Fig. 1; and see a detailed description of the implementation of jaguar predation deterrents in Appendix S2). During these visits, data on livestock numbers, fencing design, new predation incidents, intervention functionality and costs, and adoption of new livestock management practices post-training were collected (Appendix S2).

Jaguar and other predator presence on intervened ranches was assessed using 19 camera-trap stations across 13 ranches and adjacent areas, and through terrestrial transects, to verify species occurrence after project implementation (see more details in Appendix S3). Camera traps were placed opportunistically with the assistance of trained community members, with the aim of maximizing detection probability.

Baseline data on ranchers' perceptions of jaguars were collected through 100 interviews, prior to intervention actions, from March to May 2022 across eight ejidos (a communal land tenure system in which land is collectively owned by a legally recognized community; individual members hold rights to use and manage specific parcels, while major decisions regarding land use and governance are made collectively) (Fig. 1). A follow-up evaluation in September 2023, after the intervention actions, assessed changes in human-jaguar interactions and ranchers' perceptions by comparing responses from the same participants to specific questions from the initial survey. The questionnaire examined human-jaguar coexistence, conflict economic impacts, jaguar knowledge, psychological factors, livestock management, and wildlife valuation. The baseline survey was conducted using snowball sampling (80% men, 20% women), followed by purposive sampling in the second evaluation to assess changes in ranchers’ acceptance of jaguars and management actions after the intervention. In accordance with ethical research standards, all participants provided written informed consent prior to the start of the interviews.

To assess conflict levels (dispute, underlying conflict, deep-rooted conflict), we analyzed 34 questions from the initial rancher questionnaire, evaluating five key categories as proposed by Zimmermann et al., 2020: (i) jaguar perceptions, (ii) human-jaguar conflict situation, (iii) history of mitigation attempts, (iv) willingness to find solutions, and (v) perceptions of other stakeholders (Fig. 2; Table S1). Conflict level was determined based on responses within these categories using predefined criteria (Zimmermann et al., 2020).

Fig. 2.

Summary of the level of conflict evaluation. The percentage represents the level of agreement of respondents with the questions related to each of the five main topics. Shades closer to green indicate higher levels of agreement, whereas shades closer to red indicate lower levels of agreement.

The success of jaguar predation deterrents was evaluated by 1) monitoring predation cases pre- and post-intervention; 2) cost-benefit analysis; and 3) jaguar presence on intervened ranches. Changes in predation rates were analyzed using a generalized linear mixed model with a negative binomial distribution (de la Torre et al., 2021). The number of predated animals (normalized per month) was the response variable, mitigation treatment (before/after) the fixed effect, and ranch the random effect. A likelihood ratio test assessed statistical significance (see Appendix S2). This analysis was performed in R using “MASS” package (R Core Development Team, 2019; Venables and Ripley, 2002).

The cost-benefit ratio of electric fences and nighttime enclosures (10-year lifespan) was assessed by comparing material costs to the economic benefit of livestock protection (based on historical data; de la Torre et al., 2021). A ratio > 1.0 indicated economic benefit. Jaguar presence and detections on ranches measured intervention effectiveness and landowner tolerance (see Appendix S3).

To evaluate jaguar acceptability, we used the Wildlife Acceptance Capacity (WAC) and Acceptability of Management Actions (AMA), which are normative belief metrics commonly applied as proxies for tolerance (Brenner and Metcalf, 2020) and predictors of behavioral intentions toward wildlife (Bruskotter et al., 2015). Changes in WAC and AMA before and after the intervention were assessed using ordinal logistic regression (proportional odds model). Responses were coded as an ordered categorical variable ranging from least positive (e.g., kill the jaguar) to most positive (e.g., do nothing). The predictor variable was time (before vs. after the intervention), and odds ratios (OR) with 95% confidence intervals (CI) were calculated to estimate the likelihood of selecting more positive categories after the intervention. This analysis and the figures were conducted in R using “MASS” and “ggplot2” packages (R Core Development Team, 2019; Venables and Ripley, 2002; Wickham, 2016).

Conflict analysis reveals a “dispute” level of human-jaguar conflict, with mostly positive or neutral perceptions of jaguars (Fig. 2). Notably, 52% reported positive experiences, 69% had no negative encounters, and 42% enjoyed coexisting with jaguars. A strong majority, 78%, would protect an injured jaguar, and 73% empathize with them during a predation event. Most also disagreed with feeling no pity for jaguars in trouble. However, responses revealed a mix of situational concern, fear, and respect towards jaguars. Fear was noted when hearing a roar (52%) or finding tracks (37%). Concern stemmed from tracks in cattle pastures (28%) and reports of livestock attacks (33%). Anger was felt by 41% if their livestock were attacked. Reactions to the killing of a problematic jaguar included sadness (22%), happiness (20%), anger (16%), and displeasure (9%).

Regarding the conflict situation, economic losses due to jaguar predation were a major concern, reported by 61% of ranchers in the last five years. While 38% considered the presence of jaguars somewhat costly and 30% costly for the community, 66% reported jaguars visiting their pastures at least once, though most (72%) did not perceive an increasing trend of jaguars approaching communities. Opinions on the frequency of human-jaguar interactions varied, with 27% believing they increased in the last five years. Notably, 52% did not consider jaguars dangerous to humans despite predation concerns.

Past conflict mitigation efforts were viewed neutrally, with 93% considering jaguar killing as a mitigation strategy. Other methods like electric fences (40%), hazing (52%), moving cattle (51%), and the Livestock Insurance Fund (46%) were seen as potentially effective. However, 86% had not reported predation to the Livestock Insurance Fund, and those who did were largely dissatisfied. While recognizing killing as a potential strategy, 46% found it ineffective. Ranchers were open to assistance and willing to modify practices, but 75% lacked information on attack prevention, mainly learning from NGOs (18%) or neighbors (5%). Some invested in prevention (30%) or management adjustments (33%) after attacks. Most ranchers (61%) admitted limited jaguar knowledge, relying on personal experience (55%), yet 81% expressed interest in learning more and improving coexistence.

Monitoring of pilot ranches revealed limited predation incidents after the intervention (Table 1). Analysis showed a significant decrease in predation rates post-intervention (χ2 = 4.61, P < 0.05), indicating the effectiveness of electric fences and nighttime enclosures. In all cases, predation events recorded after intervention occurred either outside the area protected by the electric fence or because the animals had not been secured within the nighttime enclosures. The average material cost per ranch was USD 943 (range: 656–1259), with an estimated average annual saving of USD 1534 (range: 176–4320) due to avoided predation. The average cost-benefit ratio was highly favorable at 16.7 (range: 1.4–65.9). Furthermore, eight pilot ranches implemented practices from training workshops, including: (1) farm management plans, (2) better animal nutrition (silage, tree forage, supplements), (3) pasture division with electric fencing, (4) deworming and vaccination schedules, and (5) protein banks.

Jaguars were documented in 8 of the 13 intervened ranches (61%) during follow-up visits and camera-trap surveys. In two ranches, detections were based on tracks along transects, while in six ranches jaguars were recorded by camera traps. In total, with the camera traps we obtained 11 independent records, yielding a relative abundance index of 2.08 jaguars per 1000 camera-trap days. A total of 7 jaguars (4 females and 3 males) were identified in the videos obtained by camera traps post-intervention, though the actual count might be 5 females and 4 males, considering that some individuals were only photographed from one side.

Following the intervention, the evaluation showed a significant increase in the WAC of among ranchers (Fig. 3; Table S2). A larger proportion believed that jaguars' presence was beneficial (t = −4.46, p < 0.0001), and ranchers were more willing to accept them near cattle pastures (t = −3.74, p < 0.001). Remarkably, fewer ranchers would choose to kill a jaguar after a single predation incident (t = −3.12, p < 0.001), with more preferring to do nothing. When faced with repeated attacks, more individuals opted for inaction or hazing the jaguar. Conversely, the choice to kill the jaguar showed a significant decrease (t = −3.06, p < 0.01).

Fig. 3.

Stacked bar charts showing respondents’ answers before and after the intervention for questions related to Wildlife Acceptance Capacity (a, b, c, d) and Acceptability of Management Actions (e, f, g, h, i). The probability represents the level of respondents’ acceptance, where colors closer to green indicate more positive responses toward jaguar acceptance capacity or acceptability of management actions, and colors closer to red indicate less positive responses. It is noteworthy that after the intervention, responses were generally more positive regarding both jaguar acceptance and acceptability of management actions.

Furthermore, after the intervention, we documented a significant increase in the AMA. Ranchers' perceptions of strategy effectiveness also shifted. The perceived effectiveness of electric fences increased significantly (t = 5.05, p < 0.0001). Similarly, there was a significant increase in the perceived effectiveness of hazing jaguars (t = −2.63, p < 0.01), moving livestock (t = −4.30, p < 0.0001), and the Livestock Insurance Fund (t = −2.95, p < 0.01). Aligning with this, a significantly larger number of ranchers considered killing jaguars an ineffective strategy (t = −3.92, p < 0.01).