Negative attitudes are associated with lethal control of rodents in an Indigenous Community from the temperate forest of Mexico

Rodents have been unwanted human commensals for millennia as they decimate crops and spread diseases. However, rodents perform essential ecosystem functions, which need to be known by the public to boost rodent conservation. Furthermore, knowledge about the transmission of diseases via ectoparasites enables safe coexistence between rodents and humans. We aimed to unpack the knowledge, attitudes and practices of indigenous people towards rodents and their ectoparasites. We conducted 108 in-person interviews with farmers and foresters from the indigenous community of Nuevo San Juan, Michoacan, Mexico. Attitudes towards rodents were divided, with 54% of the respondents having a positive attitude and 36% having a negative attitude. Very few respondents could name diseases associated to rodents (44%). Most respondents agreed that rodents have the right to live, enhancing the potential of management focused on the application of preventive measures rather than the lethal control of rodents.

Methods
Study site. The study was conducted within the territory of the NSJIC (Figure 1). The NSJIC people are of Purepecha origin and historically inhabited a large portion of the State of Michoacan in west-central Mexico. Nowadays, they are concentrated in a region known as the Meseta Purepecha in the northwest portion of the state (INPI, 2019). The main town, Nuevo San Juan Parangaricutiro, was founded in 1944, after the population was relocated due to the Paricutin volcano emerging in their previous territory, in 1943 (Trask, 1943). Thus, the people from the NSJIC faced the challenge of adapting to extreme changes in their physical environment and new social contexts (Nolan, 1979). A large proportion of the NSJIC population lives in Nuevo San Juan Parangaricutiro, which has over 16,700 inhabitants (INEGI, 2020). People’s livelihoods in the area primarily revolve around forest use and management (Velázquez et al., 2015) and currently 2,300 landowners are registered for forest management activities, which include forest logging, pine resin extraction, and industrial management of wood (José Campoverde, Research Staff from the Forestry Direction of the NSJIC, personal communication). In the region surrounding NSJIC, avocado orchards have been widely established (Bravo-Espinosa et al., 2014; Fregoso et al., 2001; Latorre-Cárdenas et al., 2023). This “avocado boom” is permeating the NSJIC, where 8,990 ha of avocado orchards were recorded in the municipality in 2022 (SIAP, 2022). However, NSJIC's strong cultural institutions and communal forest management have limited the establishment of new orchards in comparison with other communities in the region (De la Vega-Rivera & Merino-Pérez, 2021).

Our study population included indigenous people who maintain a close relationship with the forests and/or cultivated fields of the NSJIC. We used the snowball method to identify and contact interviewees. This method is a form of convenience sampling, where interviewees provide information about other potential participants for the study (Naderifar et al., 2017). In our case, the interviewer asked each participant about other forest or crop owners within the NSJIC who might be willing to partake in the study.

We designed a questionnaire to capture the knowledge, perceptions, attitudes, and practices of indigenous NSJIC people associated with rodents and their ectoparasites. The questionnaire consisted of five sections: 1) basic information about the respondents (gender, age, occupation, and education level), 2) perceptions and attitudes towards rodents, 3) knowledge about rodents, 4) application of rodent control measures, 5) knowledge, preventive measures, and experiences associated with ectoparasites (Supplementary material 1). We started section 2 by asking open-ended, general questions to avoid skewing the answers of the participants and to capture the diversity of their perceptions. Afterwards, a set of close-ended questions were asked to standardise responses. These were answered on simple three-point scales (Yes, No, I don’t know). Section 3 was composed of eight true/false questions about the basic ecology of rodents. The answers in this section were used to calculate a knowledge score, where only correct answers would increase the value of the score. For sections 2 and 3, given the scarcity of specific studies regarding people’s perceptions and attitudes towards rodents or the knowledge about them, we adapted existing questionnaires designed for other species (Carter et al., 2012; Khamassi Khbou et al., 2020a; Pérez et al., 2021; Thorn et al., 2012). For section 4, we selected relevant questions from existing questionnaires (Burt & Lipman, 2021; Olalekan, 2015; Panti-May et al., 2017). Likewise, for section 5, we adapted questionnaires about attitudes towards ectoparasites and preventive measures associated with them (Abdulkareem et al., 2019; Azrizal-Wahid et al., 2022; Cornall y Wall, 2015; Leibler et al., 2018; Ozioko et al., 2018; Sadeghi Dehkordi et al., 2022; Slunge y Boman, 2018; Torres-Castro et al., 2020). The interviews were conducted in person, by a trained psychologist who self-identifies as a member of the NSJIC (Diana Lucrecia Ruiz Ramírez). The interviews were recorded with the consent of the respondents. The interviews lasted an average of 10 minutes. All standard ethical procedures for conducting interviews were followed. Our study was approved by the Research Committee of the Instituto de Investigaciones Biologicas of the Universidad Veracruzana, Mexico.

To test the internal consistency of this set of questions (items), we obtained the ordinal value of α, which is conceptually equivalent to Cronbach's α, but does not produce negative biases when the responses are presented on ordinal scales with few points (Gadermann et al., 2012; Zumbo et al., 2007). We chose a three-point scale because our respondents struggled to answer the attitude questions on a five-point scale. The value of ordinal α was calculated with the psych package (R Core Team, 2022; Revelle, 2023). The ordinal α of the attitude questionnaire was 0.84 (CI 95% = 0.71 - 0.93), which is considered a robust value (Vaske et al., 2017). This value was obtained after removing five negatively correlated items and adding four more items. We performed a confirmatory factor analysis using the lavaan package (R Core Team, 2022; Rosseel, 2012). The Comparative Fit Index (CFI) was 0.64 and the root mean square error of approximation (RMSEA) was 0.08 (CI 95% = 0.06 - 0.1).

Data analysis. For all the open-ended questions we performed word tokenisation, which implies separating each word in a text. We performed multiple tokenisation processes, joining words with equal meanings together into one category (type token). We used the quanteda R package to calculate token frequencies (Benoit et al., 2018; R Core Team, 2022). We defined our sample size based on the qualitative saturation, which was assessed after the first 55 interviews, and we did not observe new viewpoints to describe rodents after this, which assures the saturation of our sample (Lauret et al., 2020).

The data from the close-ended questions in section 2 were used to obtain a total attitude score for each respondent. The answers to each item were given a punctuation (-1, 0, or 1), depending on the negative, positive, or negative connotation of the answer. All the scores from the answers were summarised into a single total score. We confirmed the normality of the attitude score using a Shapiro-Wilk test (W = 0.98, p = 0.18). We analysed the influence of age, gender, occupation, and knowledge levels (section 3) on the attitude score through a linear model run using the MuMIn R package (Barton, 2018; R Core Team, 2022). We compared attitude scores associated with the application of rodent control measures using a Wilcoxon test.

Regarding the knowledge and experiences related to ectoparasites (section 5), we conducted a descriptive analysis of the frequency of encounters of our respondents with ectoparasites, the preventive or control measures applied against ectoparasites, their perceptions of changes in ectoparasite encounters through time, the species (including domestic or wild species) that carried ectoparasites as well as people’s knowledge about the diseases transmitted by them. The application of preventive measures against ectoparasites was compared concerning the frequency of encounters and knowledge about disease transmission using contingency tables and χ2 tests performed in stats R package (R Core Team, 2022).