Let us give voice to local farmers: Preferences for farm-based strategies to enhance human–elephant coexistence in Africa

Citation: Montero Botey, M.; Soliño, M.; Perea, R.; Martínez-Jauregui, M. Let Us Give Voice to Local Farmers: Preferences for Farm-Based Strategies to Enhance Human–Elephant Coexistence in Africa. Animals 2022, 12, 1867. https://doi.org/10.3390/ ani12141867 Academic Editors: Bruce Alexander Schulte and Chase LaDue Received: 28 June 2022 Accepted: 19 July 2022 Published: 21 July 2022 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). animals Article Let Us Give Voice to Local Farmers: Preferences for Farm-Based Strategies to Enhance Human–Elephant Coexistence in Africa María Montero Botey 1,* , Mario Soliño 2,3 , Ramón Perea 1and María Martínez-Jauregui 4,5 1Departamento de Sistemas y Recursos Naturales, Universidad Politécnica de Madrid, Avda. Moreras s/n E, 28040 Madrid, Spain; [email protected] 2Institute of Marine Research—CSIC, C/ Eduardo Cabello 6, 36208 Vigo, Spain; [email protected] 3Complutense Institute for International Studies (ICEI), Finca Mas Ferré, Edif. A. Campus de Somosaguas, 28223 Pozuelo de Alarcón, Spain 4Forest Research Centre (INIA-CSIC), Ctra. de La Coruña km. 7.5, 28040 Madrid, Spain; [email protected] 5Sustainable Forest Management Research Institute, University of Valladolid and INIA, Avda. de Madrid 57, 34004 Palencia, Spain *Correspondence: maria.monter[email protected]; Tel.: +34-910671701 Simple Summary: Local communities living on the edge of protected areas often experience negative impacts on their livelihoods due to wildlife. These situations threaten support for long-term conservation of wildlife and wild habitats so a key for conservation sustainability should be based on implementing socially accepted and economically sustainable mitigation practices. For successful design and implementation of mitigation strategies, it is vital to engage local communities and understand their preferences and previous experiences. In this study, we present a choice experiment as a tool to analyze local farmer preferences for the most common farm-based solutions to reduce African elephant crop damage. Results show that there are significant differences among responses triggered by farmers’ previous experience with elephants and socioeconomic situation, with a marked spatial distribution among respondents. This methodology, based on a choice modeling approach considering the differential availability of resources and previous experience with elephants or other wildlife, is highly applicable, with small changes in other areas where wildlife competes with local communities for resources. This approach also represents a suitable instrument for identifying stakeholders’ preferences in each specific context. Abstract: Local communities surrounding wildlife corridors and natural reserves often face challenges related to human–wildlife coexistence. To mitigate the challenges and ensure the long-term conservation of wildlife, it is important to engage local communities in the design of conservation strategies. By conducting 480 face-to-face interviews in 30 villages along and adjacent to the Selous- Niassa Wildlife Corridor (Tanzania), we quantified farmers’ preferences for farm-based measures to mitigate African elephant damage using choice experiments. Results show that farmers considered no action the least preferred option, revealing that they are open to trying different measures. The most preferred management strategy matched with the preferences of wildlife rangers in the area, suggesting low concern about the potential conflicts between stakeholders. However, a latent class model suggests that there are significant differences among responses triggered by farmers’ previous experience with elephants, the intensity of the elephant damage, and the socioeconomic situation of the farmer. Results show a marked spatial distribution among respondents, highlighting the benefits of zone management as conflicts were found to be highly context dependent. Understanding the human dimension of conservation is essential for the successful planification and implementation of conservation strategies. Therefore, the development and broad utilization of methodologies to gather specific context information should be encouraged. Keywords: mitigation measures; choice experiment; human–wildlife conflict; Loxodonta africana; willingness to pay; beehives; chili-oil fences Animals 2022,12, 1867. https://doi.org/10.3390/ani12141867 https://www.mdpi.com/journal/animals Animals 2022,12, 1867 2 of 18 1. Introduction Coexistence between people and wildlife has been long recognized as a global conservation challenge [ 1 , 2 ]. In some cases, coexistence with large-sized wildlife implies impacts on the safety or livelihood of local people. As a result, socio-economic conflicts may arise, confronting local communities negatively affected by the presence of certain species and those who want to promote or protect those species [ 3 ]. Although people and wildlife have co-existed for millennia, wildlife-related conflicts have become more intense and frequent in recent years due to habitat loss and degradation, mainly caused by the expansion and intensification of human activities [ 4 , 5 ]. Africa is a paradigmatic example of increased conflicts related to wildlife due to the charismatic and threatened species involved, the recent growth of its human population [ 6 ], and the strong economic vulnerability of rural areas [7]. Compensation policies, where the government or conservationists pay for the damages occurred due to wildlife, may seem a good strategy to address human–wildlife conflicts [ 8 – 10 ]. However, the conservation of wildlife in Africa is generally encouraged by governments or organizations that are heavily dependent on outside sources of funding. Compensation policies are not advised in areas with limited funds or deficient administrative controls due to possible fraudulent claims and damage of the motivation of local communities to protect their properties from wildlife damage [11,12]. Previous research has shown that management tools to promote human–wildlife coexistence should consider not only the research on technical solutions but the development of shared solutions, where conflicting parties are engaged and cooperate [ 13 ]. This highlights the importance of co-management in addressing human conflicts with wildlife in Africa, where engagement of local communities is necessary for the implementation of successful and economically sustainable mitigation strategies in the long term [14,15]. Empowering farmers to implement simple farm-based but cost-effective measures [ 16 ] could be a particularly successful alternative to mitigate conflicts in African wildlife corridors. In these areas, conservation programs are necessary for the maintenance of wildlife meta-population processes [ 17 ] and connectivity [ 18 ]; however, wildlife shares land and resources with rural communities, triggering important social costs [ 19 ]. Although government and private financial support is frequently scarce, it is already known that affected farmers are more willing to accept changes they have chosen themselves [ 20 , 21 ]. Similarly, the context and experiences farmers have accumulated during their lives have been identified as key factors to engaging farmers in mitigation practices in Asia [ 22 , 23 ]. Therefore, the incorporation of farmers’ preferences for different farm-based measures and their relationship with farmers’ previous experiences is urgently needed for the successful and context-dependent design of wildlife conservation programs. In this study, we used tools from environmental economics to address preferences among farmers in the Selous-Niassa Wildlife Corridor (Tanzania) related to: (i) the specific farmbased measures they consider effective in preventing African elephant ( Loxodonta africana Blumenbach 1797) damage and their willingness to apply them, (ii) the importance of receiving technical advice (conducted by NGOs or the Government) in the implementation of the measures, and (iii) the desirable level of cooperation in their community for this implementation (which was proven to be a key factor in the success or failure of human–elephant conflict mitigation programs in other areas, e.g., [ 16 ]). To avoid false expectations being raised in the local communities, all proposed strategies are supported by science, relatively inexpensive, and applicable by the farmers on their own. In addition, wildlife rangers were informed about these strategies and their preferences were previously analyzed [ 24 ], which will allow us to shed some light on the potential conflicts between rangers and farmers when choosing, planning, and implementing the proposed mitigation measures. Conflicts between rangers and farmers regarding the implementation of mitigation measures influence the success of the measures as wildlife rangers hold a key role in the community awareness and protection of people’s livelihoods from wildlife [ 25 , 26 ]. These conflicts can also undermine trust and cooperation between the parties, influencing the implementation Animals 2022,12, 1867 3 of 18 and success of other conservation activities [ 27 ], as rangers are, in many cases, the most visible actors in conservation to local communities [28]. Finally, and for a better understanding of the local communities’ preferences, including an analysis of the heterogeneous preferences among respondents [ 29 ] and its possible causes, we explored whether there are differences among responses triggered by farmers’ personal previous experience with elephants, either on their own farms or through family, friends’, or neighbors’ experiences (contagious effect of risk perception, [ 30 ]). Moreover, we also explored whether the actual socioeconomic situation of the respondents (measured by the self-reported food insecurity level) influences their preferences for the proposed measures. This exploration is important to identify factors that can influence preferences in other contexts. 2. Materials and Methods 2.1. Farming and Elephant Conservation in the Selous-Niassa Wildlife Corridor The Selous-Niassa Wildlife Corridor (Figure 1) is part of the world’s largest Miombo woodland ecosystems (Selous-Niassa ecosystem) and links Julius Nyerere National Park (established in November 2019 but previously known as Selous Game Reserve) in Tanzania with Niassa National Reserve in Mozambique. The corridor lies within the Tunduru and Namtumbo districts in Ruvuma Region (southern Tanzania), covers traditional elephant movement routes [ 31 ], and harbors a population of 602 ± 258 elephants [ 32 ]. It is located entirely on the land owned by 30 villages. Local people mostly base their economy on subsistence farming, although this is more pronounced in the north part of the corridor. The staple crops grown are maize, rice, and cassava while common cash crops are tobacco, sunflower, cashew nut, sesame, etc. [33,34]. For local communities all around Africa, cohabitation with elephants commonly implies crop losses, damages to infrastructures and water supplies, and, in few cases, injuries or human deaths due to elephants charging at humans [ 35 – 40 ]. These situations disrupt the psychological and physical wellbeing of local communities [ 41 – 43 ] and involve many challenges for elephant conservation [ 44 , 45 ], fueling both legal and illegal retaliation killings of elephants [ 39 , 46 , 47 ] and threatening the maintenance of protected areas in the long term due to increased resistance to conservation [ 39 , 48 ]. In addition, damages have increased in the last century due to the rapid growth of the human population and the colonization of natural areas for its conversion into agriculture land [ 49 , 50 ], spreading all over the African elephant range [51,52]. The current decrease in the elephant populations in the Selous-Niassa ecosystem [ 53 ] and the rise in impacts on humans lives due to frequent human–elephant interaction [ 54 ] make the area a unique place to address large-scale human–elephant coexistence challenges and establish sustainable local initiatives for the mitigation of conflicts related to farming and wildlife conservation. Additionally, Tanzania is an example where the government and local communities are willing to engage in mitigating these types of challenges. This is proved by the “National Human-Wildlife Conflict Management Strategy 2020–2024” [ 54 ] and the fact that rangers commonly work on chasing away elephants from farms and are also involved in citizen science [ 24 ]. In addition, some farmers are already applying some farm-based mitigation measures, such as chili fences, encouraged and supported in the corridor by PAMS (Protected Areas Management Solution) Foundation and WWF (World Wild Fund for Nature). However, in the Selous-Niassa Wildlife corridor, the most common elephant mitigation measures applied by farmers are guarding the crops at night and making noises to chase them away (drumming, clapping, shouting, etc.), which are traditional methods that they have broad knowledge of and do not represent an added cost to their already vulnerable and limited familiar economy. Animals 2022,12, 1867 4 of 18 Animals 2022, 12, 1867 4 of 18 Figure 1. Selous-Niassa Wildlife Corridor map and location of villages where interviews were conducted. 2.2. Data Collection Data was collected by conducting 480 face-to-face interviews in 30 villages along and adjacent to the Selous-Niassa Wildlife Corridor (Figure 1). The sampling unit was the household. Households were chosen randomly, and interviews were restricted to one respondent (above 18 years old) per household. In each village, 16 locals were interviewed, 8 men and 8 women, in equal proportions between people interviewed in the village center and in further farms inside the village land. All interviews were conducted between June and September 2019 in Swahili by five previously trained Tanzanians from the area. The survey was pre-tested in April 2019 on 25 farmers from 3 villages with different intensities of elephant damage to ensure clarity before use and improve the design of the final study. The questionnaire (Supplementary File A) was designed to gather four categories of information: (1) personal data (gender, occupations, food shortage in their household, etc.), (2) previous experience with elephants and elephant crop damage, (3) perception of the effectivity of farm-based elephant mitigation measures using Likert scales (from 1 to 4, where 1 represented strongly disagree, 2 disagree, 3 agree, and 4 strongly agree; don’t know was always available for the respondent), and (4) preferences for mitigation tools and their implementation using a discrete choice experiment [55]. Figure 1. Selous-Niassa Wildlife Corridor map and location of villages where interviews were conducted. 2.2. Data Collection Data was collected by conducting 480 face-to-face interviews in 30 villages along and adjacent to the Selous-Niassa Wildlife Corridor (Figure 1). The sampling unit was the household. Households were chosen randomly, and interviews were restricted to one respondent (above 18 years old) per household. In each village, 16 locals were interviewed, 8 men and 8 women, in equal proportions between people interviewed in the village center and in further farms inside the village land. All interviews were conducted between June and September 2019 in Swahili by five previously trained Tanzanians from the area. The survey was pre-tested in April 2019 on 25 farmers from 3 villages with different intensities of elephant damage to ensure clarity before use and improve the design of the final study. The questionnaire (Supplementary File A) was designed to gather four categories of information: (1) personal data (gender, occupations, food shortage in their household, etc.), (2) previous experience with elephants and elephant crop damage, (3) perception of the effectivity of farm-based elephant mitigation measures using Likert scales (from 1 to 4, where 1 represented strongly disagree, 2 disagree, 3 agree, and 4 strongly agree; don’t know was always available for the respondent), and (4) preferences for mitigation tools and their implementation using a discrete choice experiment [55]. Animals 2022,12, 1867 5 of 18 2.3. Choice Modeling To analyze the local communities’ preferences regarding farm-based management programs, we designed a discrete choice experiment (DCE) composed of four attributes. The DCE is a stated preferences method that involves presenting respondents with various choice cards comprising two or more alternatives (actions, programs, scenarios, etc.) that are described by a set of attributes and different levels. This method is commonly used to obtain comparable measures of preferences across factors and attributes [56,57]. The attributes were equal to those employed in the rangers’ preference exploration in the same study area [ 24 ]. They are: (1) specific farm-based measures that farmers can apply to reduce elephant damage to humans and human means, which include six different scientifically proven effective strategies: (a) chili-oil fences [ 58 – 60 ]; (b) noisemakers [ 61 , 62 ]; (c) beehive fences [ 63 ]; (d) surveillance [ 61 , 64 ]; (e) crop selection [ 34 , 65 – 67 ]; and (f) crop relocation [68]; (2) the level of cooperation in the implementation of different tools, which has been defined as an important key for the success of mitigation measures [ 16 ], defined in a qualitative manner: (a) individual, (b) small groups of neighbors (2–3 households, as represented in Figure 2), and (c) large groups (>10 households, as illustrated in Figure 2) and community levels) [ 69 , 70 ]; (3) the involvement of technical support given by NGOs or the government in the process [ 71 ] considering (a) yes, it is present, and (b) no, it is not, which provides important information about how much farmers trust those institutions; and (4) a monetary attribute to estimate the willingness to pay per household and commonly used to quantify preferences. In this case, we also considered the monetary cost that farmers should assume when implementing the elephant crop damage mitigation program, which was not considered in the rangers’ study performed by Montero-Botey et al. [ 24 ]. The monetary attribute had four levels from 10,000 TZS (~5$) to 40,000 TZS (~20$) and represented the monetary cost per year for a farmer to apply the measure selected in one acre. The levels of cost were established after a discussion in a focus group with members of the community to determine the range of cost that farmers would be willing to invest and could afford as the majority are subsistence farmers. It was also tested in the pilot questionnaire. A more extensive description of the first two attributes is available in Figure 2. Based on the results obtained by the pilot study of 25 farmers in the study area, a D-efficiency criterion to generate efficient designs was considered to identify the lower D-error that minimizes the variances and covariances of the parameter estimates [ 72 ]. We used the Ngene ® 1.2. software [ 73 ] for our experimental design and 48 choice cards were generated. In order to make a feasible choice task, and not overwhelm the respondents with too many choices, a blocking strategy was considered, and twelve choice cards were shown to each individual. Each choice card comprised four alternative programs and an opt-out option that represented a no-intervention alternative to avoid forcing activity choices [ 74 ] (Figure 3). The final data of farmers’ choice was analyzed in two steps. First, for comparison with the wildlife rangers’ preferences reported in Montero-Botey et al. [ 24 ], we estimated a random parameters logit model using the Nlogit ® version 6 software. We assumed that all the attributes are random parameters that are normally distributed and the willingness to pay (WTP) for each attribute level was estimated (see the formulation in Supplementary File B). Secondly, we estimated a latent class model (LCM) with random parameters [ 75 , 76 ] using the Latent GOLD ® version 5.1 software [ 77 ] (see the formulation in Supplementary File B ). This modeling approach is useful for the in-depth analysis of heterogeneous preferences among respondents [ 29 ], possibly associated with previous experience with elephants [ 78 ] and the possible social contagion of risk perception [ 79 ]. For this purpose, we created an artificial variable classifying the farmers directly affected by elephant crop damage; farmers not directly affected by elephant crop damage but whose family, friends, or neighbors have been affected; and farmers not affected without relatives or neighbors affected by crop damage. Based on the results from the latent class model, we carried out a post-hoc descriptive analysis to show the spatial distribution of the classes as zoning management Animals 2022,12, 1867 6 of 18 could improve the achievement of conservation goals [ 80 ]. We also explored the relationship of those classes with food shortage and elephant presence as indicators of vulnerability [ 30 ]. Animals 2022, 12, 1867 6 of 18 a post-hoc descriptive analysis to show the spatial distribution of the classes as zoning management could improve the achievement of conservation goals [80]. We also explored the relationship of those classes with food shortage and elephant presence as indicators of vulnerability [30]. Figure 2. Examples of explanatory cards showed to the interviewees to define the specific farmbased measures that farmers can apply to reduce elephant damage and the level of cooperation in the implementation of those measures. Figure 2. Examples of explanatory cards showed to the interviewees to define the specific farm-based measures that farmers can apply to reduce elephant damage and the level of cooperation in the implementation of those measures. Animals 2022,12, 1867 7 of 18 Animals 2022, 12, 1867 7 of 18 Figure 3. Example of a choice card used in the DCE. 3. Results A total of 241 men and 239 women were interviewed: 95% of them focused on agriculture as their main occupation and 78% were originally from the village where they were interviewed. Elephants were considered the most conflictive wildlife species in the area by 76% of the respondents (see more information in Table S1). Regarding their personal experience with elephants, 75% had seen an elephant, 4 people reported to have been directly charged by elephants, 9% that family members or friends were charged, and 13% that the closest person charged they know about was someone from their village. Regarding elephant crop damage, 55% of them reported that they had been directly affected (average of 4 times in their lifetime), 12% that not them but their family or friends had been affected, and 8.5% that the closest person affected they knew about was someone from the village they live in. Concerning the perceived effectivity of measures to reduce crop damage (Figure 4), noisemakers were considered effective by 52% of respondents (2.48 ± 0.05 in the same Likert scale, from 1 to 4), crop selection by 48% (2.6 ± 0.04), chilioil fences by 47% (2.57 ± 0.05), guarding crops at night by 38% (2.18 ± 0.05), bee-hive fences by 31% (2.53 ± 0.05), and crop translocation by 28% (2.27 ± 0.04). Technical advice was considered effective by 67% (2.93 ± 0.04). Importantly, 34% did not know about the bees as a mitigation measure and 17% and 18% were not sure about the effectivity of crop selection and crop translocation, respectively. Choice experiment results showed that farmers in the Selous-Niassa Wildlife Corridor generally agreed with a farm-based management program to mitigate elephant crop damage. However, 2.5% did not choose any option due to budgetary restrictions (true zeros) and 4.6% (protest responses) refused to choose options in the choice experiment due to other reasons such as, for example, that the mitigation measures should be implemented and paid for by the government and/or the lack of elephants in their area. For the rest of the respondents that made any choice (93%), the option “no action” was chosen in 11.5% of the observations. For the analysis of preferences, we excluded the protest responses (4.6%), and the final sample was composed of 27,420 observations of 457 individuals. Results showed that the alternative specific constant (ASC) was statistically significant (Table 1 and Table S2). Figure 3. Example of a choice card used in the DCE. 3. Results A total of 241 men and 239 women were interviewed: 95% of them focused on agriculture as their main occupation and 78% were originally from the village where they were interviewed. Elephants were considered the most conflictive wildlife species in the area by 76% of the respondents (see more information in Table S1). Regarding their personal experience with elephants, 75% had seen an elephant, 4 people reported to have been directly charged by elephants, 9% that family members or friends were charged, and 13% that the closest person charged they know about was someone from their village. Regarding elephant crop damage, 55% of them reported that they had been directly affected (average of 4 times in their lifetime), 12% that not them but their family or friends had been affected, and 8.5% that the closest person affected they knew about was someone from the village they live in. Concerning the perceived effectivity of measures to reduce crop damage (Figure 4), noisemakers were considered effective by 52% of respondents ( 2.48 ±0.05 in the same Likert scale, from 1 to 4), crop selection by 48% (2.6 ± 0.04), chili-oil fences by 47% (2.57 ± 0.05), guarding crops at night by 38% (2.18 ± 0.05), bee-hive fences by 31% (2.53 ± 0.05), and crop translocation by 28% (2.27 ± 0.04). Technical advice was considered effective by 67% (2.93 ± 0.04). Importantly, 34% did not know about the bees as a mitigation measure and 17% and 18% were not sure about the effectivity of crop selection and crop translocation, respectively. Choice experiment results showed that farmers in the Selous-Niassa Wildlife Corridor generally agreed with a farm-based management program to mitigate elephant crop damage. However, 2.5% did not choose any option due to budgetary restrictions (true zeros) and 4.6% (protest responses) refused to choose options in the choice experiment due to other reasons such as, for example, that the mitigation measures should be implemented and paid for by the government and/or the lack of elephants in their area. For the rest of the respondents that made any choice (93%), the option “no action” was chosen in 11.5% of the observations. For the analysis of preferences, we excluded the protest responses (4.6%), and the final sample was composed of 27,420 observations of 457 individuals. Results showed that the alternative specific constant (ASC) was statistically significant (Tables 1and S2). Animals 2022,12, 1867 8 of 18 Animals 2022, 12, 1867 10 of 18 Figure 4. Farmers´ perception about the effectivity of farm-based mitigation measures to reduce crop damage by elephants. Figure 5. Description of the classes regarding the percentage of respondents that had seen an elephant (blue bar) and the percentage of respondents that had suffered a food shortage in their households (grey bar). The line shows the average duration of the food shortage period (in months). Class 1: Affected and cooperative; Class 2: Not affected and cooperation in small groups; Class 3: Not affected and communal; Class 4: Affected and individualist; Class 5: Not affected whose family, friends, or neighbors have been affected and individualist. Figure 4. Farmers’ perception about the effectivity of farm-based mitigation measures to reduce crop damage by elephants. Table 1. Results of the random parameter logit models (457 face-to-face wildlife rangers and 12 choices per individual; number of observations = 5484 ; Log likelihood function = −6410.09 ; restricted log likelihood = −8826.16 ; McFadden Pseudo R-squared = 0.2737 ; replications for simulated probs. = 500; used Halton sequences in simulations). Coefficient Standard Error Z Prob. |z| > Z * 95% Confidence Interval Random parameters ASC −2.485 *** 0.2080 −11.94 <0.001 (−2.8927, −2.0772) Crop selection 0.288 ** 0.1226 2.35 0.019 (0.0478, 0.5286) Crop translocation −0.38 7 *** 0.1135 −3.41 <0.001 (−0.6092, −0.1645) Noisemakers −0.075 0. 1332 −0.57 0.571 (−0.3365, 0.1858) Chili-oil fences 1.213 *** 0.1131 10.72 <0.001 (0.9908, 1.4343) Bee-hive fences 0.708 *** 0.1257 5.63 <0.001 (0.4617, 0. 9545) Technical support 0.658 *** 0.0748 8.79 <0.001 (0.5111, 0.8044) Cooperation in small groups −0.024 0.0603 −0.39 0. 694 (−0.1419, 0.0944) Cooperation in big groups 0.437 *** 0.0604 7.24 <0.001 (0.3189, 0.5557) BID Cost/year −0.110 *** 0.0081 −13.60 <0.001 (−0.1254, −0.0938) Standard Deviations of random parameters (normally distributed) ASC 3.091 *** 0. 1862 16.60 <0.001 (2.7263, 3.4562) Crop selection 1.741 *** 0.1204 14.46 <0.001 (1.5047, 1.9766) Crop translocation 1.243 *** 0.1115 11.16 <0.001 (1.0248, 1.4617) Noisemakers 1.865 *** 0.1580 11.80 <0.001 (1.5557, 2.1751) Chili-oil fences 1.8465 *** 0.1009 18.30 <0.001 (1.6487, 2.0443) Bee-hive fences 2.070 *** 0.1337 15.48 <0.001 (1.8080, 2.3321) Technical support 1.238 *** 0.0708 17.49 <0.001 (1.0992, 1.3766) Cooperation in small groups 0.218 0.1621 1.34 0.180 (−0.1001, 0.5351) Cooperation in big groups 0.5142 *** 0.0823 6.25 <0.001 (0.3530, 0.6754) BID Cost/year 0.135 *** 0.0070 19.26 <0.001 (0.1217, 0.1492) *** Significance at 1% level; ** Significance at 5% level; * Significance at 10% level. Animals 2022,12, 1867 9 of 18 The results of the random parameters logit model (Table 1) show that regarding the mitigation tools, farmers’ most preferred tool was the use of chili-oil fences, followed by beehive fences, having technical support, promoting cooperation in large groups (community levels), and crop selection. Using noisemakers and surveillance and cooperation in small- and medium-sized groups were not significant, and translocating crops was rejected as it reduces overall farmers’ well-being. The latent class model identified five different classes of behavior among the respondents that explained the mitigation strategies’ choice heterogeneity (Table 2). Combining this information with their experience, food shortage (Figure 5), and the geographical context (Figure 6shows the spatial distribution of every class in the villages of the Selous- Niassa Wildlife Corridor), we characterized and further explained the classes that resulted from the model. The main findings are: (i) 24.1% of the respondents (class 1) were directly affected by elephants, had suffered from a severe food shortage, and were willing to cooperate at the village level; (ii) 23.8% of the respondents (class 2) were not directly affected by elephants, had suffered a moderate food shortage, were not concerned about the economic cost, and were not very demanding on the characteristics of the program; (iii) 21.1% of the respondents (class 3) were not directly affected by elephants, suffered a lower food shortage, were willing to pay much more (almost 6-fold), but they were in favor of a program involving the whole community and technical support; (iv) 18.5% of the respondents (class 4) were directly affected by elephants, had suffered from a severe food shortage, valued technical support but they preferred an individual program, and had a strong negative reaction to crop translocation; (iv) and 12.5% of the respondents (class 5) were not directly affected and were characterized by low cooperation and strong willingness to pay (almost 4 times more than the directly affected classes; Table 2), with little or no apparent value for technical support. Animals 2022, 12, 1867 10 of 18 Figure 4. Farmers´ perception about the effectivity of farm-based mitigation measures to reduce crop damage by elephants. Figure 5. Description of the classes regarding the percentage of respondents that had seen an elephant (blue bar) and the percentage of respondents that had suffered a food shortage in their households (grey bar). The line shows the average duration of the food shortage period (in months). Class 1: Affected and cooperative; Class 2: Not affected and cooperation in small groups; Class 3: Not affected and communal; Class 4: Affected and individualist; Class 5: Not affected whose family, friends, or neighbors have been affected and individualist. Figure 5. Description of the classes regarding the percentage of respondents that had seen an elephant (blue bar) and the percentage of respondents that had suffered a food shortage in their households (grey bar). The line shows the average duration of the food shortage period (in months). Class 1: Affected and cooperative; Class 2: Not affected and cooperation in small groups; Class 3: Not affected and communal; Class 4: Affected and individualist; Class 5: Not affected whose family, friends, or neighbors have been affected and individualist. Animals 2022,12, 1867 16 of 18 35. Thouless, C.R. Conflict between humans and elephants on private land in northern Kenya. Oryx 1994,28, 119–127. [CrossRef] 36. Naughton-Treves, L. Predicting Patterns of Crop Damage by Wildlife around Kibale National Park, Uganda. Conserv. Biol. 1998,12, 156–168. [CrossRef] 37. Naughton, L.; Rose, R.; Treves, A. The Social Dimensions of Human-Elephant Conflict in Africa: A Literature Review and Case Studies from Uganda and Cameroon; A Report to the African Elephant Specialist Group, Human-Elephant Conflict Task Force; IUCN: Glands, Switzerland, 1999. 38. Sitienei, A.J.; Jiwen, G.; Ngene, S.M. Assessing the cost of living with elephants (Loxodonta africana) in areas adjacent to Meru National Park, Kenya. Eur. J. Wildl. Res. 2014,60, 323–330. [CrossRef] 39. Mariki, S.B.; Svarstad, H.; Benjaminsen, T.A. Elephants over the cliff: Explaining wildlife killings in Tanzania. Land Use Policy 2015,44, 19–30. [CrossRef] 40. Nsonsi, F.; Heymans, J.C.; Diamouangana, J.; Mavinga, F.B.; Breuer, T. Perceived human-elephant conflict and its impact for elephant conservation in northern Congo. Afr. J. Ecol. 2017,56, 208–215. [CrossRef] 41. Hill, C.M. Conflict of interest between people and baboons: Crop raiding in Uganda. Int. J. Primatol. 2000 ,21, 299–315. [CrossRef] 42. Ogra, M.V. Human–wildlife conflict and gender in protected area borderlands: A case study of costs, perceptions, and vulnerabilities from Uttarakhand (Uttaranchal), India. Geoforum 2008,39, 1408–1422. [CrossRef] 43. Barua, M.; Bhagwat, S.A.; Jadhav, S. The hidden dimensions of human-willdife conflict: Health impacts, opportunity and transaction costs. Biol. Conserv. 2013,157, 309–316. [CrossRef] 44. Sitati, N.W.; Walpole, M.J.; Smith, R.J.; Leader-Williams, N. Predicting spatial aspects of human-elephant conflict. J. Appl. Ecol. 2003,40, 667–677. [CrossRef] 45. Songhurst, A.; Coulson, T. Exploring the effects of spatial autocorrelation when identifying key drivers of wildlife crop-raiding. Ecol. Evol. 2014,4, 582–593. [CrossRef] 46. TAWIRI. Tanzania Elephant Management Plan 2010-2015; TAWIRI: Arusha, Tanzania, 2010. 47. Graham, M.D.; Adams, W.M.; Kahiro, G.N. Mobile phone communication in effective human elephant-conflict management in Laikipa County, Kenya. Oryx 2012,46, 137–144. [CrossRef] 48. Norgrove, L.; Hilme, D. Parking resistance and resisting the park: ’weapons of the weak’, Confronting conservation at Mount Elgon, Uganda. Dev. Chang. 2006,37, 1093–1116. [CrossRef] 49. Hoare, R.E. Determinants of human-elephant conflict in a land-use mosaic. J. Appl. Ecol. 1999,36, 689–700. [CrossRef] 50. Pozo, R.A.; McCulloch, G.; Stronza, A.; Coulson, T.; Songhurst, A. Determining baselines for human-elephant conflict: A matter of time. PLoS ONE 2017,12, e0178840. [CrossRef] [PubMed] 51. Walpole, M.J.; Linkie, M. Mitigating Human-Elephant Conflict: Case Studies from, Africa and Asia; Fauna and Flora International: Cambridge, UK, 2008. 52. Lamarque, F.; Anderson, J.; Fergusson, R.; Lagrange, M.; Osei-Owusu, Y.; Bakker, L. Human-Wildlife Conflict in Africa: Causes, Consequences and Management Strategies (No. 157); Food and Agriculture Organization of the United Nations (FAO): Rome, Italy, 2009. 53. Chase, M.J.; Schlossberg, S.; Griffin, C.R.; Bouché, P.J.C.; Djene, S.W.; Elkan, P.W.; Ferreira, S.; Grossman, F.; Kohi, E.M.; Landen, K.; et al. Continent-wide survey reveals massive decline in African savannah elephants. PeerJ 2016 ,4, e2354. [CrossRef] [PubMed] 54. MNRT. National Human-Wildlife Conflict Management Strategy 2020–2024; Ministry of Natural Resources and Tourism, United Republic of Tanzania: Arusha, Tanzania, 2020. 55. Louviere, J.J.; Hensher, D.A.; Swait, J.D. Stated Choice Methods: Analysis and Applications; Cambridge University Press: Cambridge, UK, 2000. 56. Bartkowski, B.; Lienhoop, N.; Hansjürgens, B. Capturing the complexity of biodiversity: A critical review of economic valuation studies of biological diversity. Ecol. Econ. 2015,113, 1–14. [CrossRef] 57. Czajkowski, M.; Hanley, N. Using labels to investigate scope effects in stated preference methods. Environ. Resour. Econ. 2009,44, 521–535. [CrossRef] 58. Parker, G.E.; Osborn, F.V.; Hoare, R.E.; Niskanen, L.S. Human-Elephant Conflict Mitigation: A Training Course for Community-Based Approaches in Africa. Participant’s Manual; Elephant Pepper Development Trust, Livingstone, Zambia and IUCN/SSC AfESG: Nairobi, Kenya, 2007. 59. Graham, M.; Ochieng, T. Uptake and performance of farm-based measures for reducing crop raiding by elephants Loxodonta africana among smallholder farms in Laikipia District, Kenya. Oryx 2008,42, 76–82. [CrossRef] 60. Chang’a, A.; Souza de, N.; Muya, J.; Keyyu, J.; Mwakatobe, A.; Malugu, L.; Ndossi, H.P.; Konuche, J.; Omondi, R.; Mpinge, A.; et al. Scaling-up the use of chili fences for reducing human-elephant conflict across landscapes in Tazania. Trop. Conserv. Sci. 2016,9, 921–930. [CrossRef] 61. Hoare, R. African elephants and humans in conflict: The outlook for co-existence. Oryx 2000,34, 34–38. [CrossRef] 62. King, L.E.; Douglas-Hamilton, I.; Vollrath, F. African elephants run from the sound of disturbed bees. Curr. Biol. 2007,17, R832–R833. [CrossRef] 63. King, L.E.; Lawrence, A.; Douglas-Hamilton, I.; Vollrath, F. Beehive fence deters crop-raiding elephants. Afr. J. Ecol. 2009,47, 131–137. [CrossRef] 64. Sitati, N.W.; Walpole, M.J.; Leader-Williams, N. Factors affecting susceptibility of farms to crop raiding by African elephants: Using a predictive model to mitigate conflict. J. Appl. Ecol. 2005,42, 1175–1182. [CrossRef] Animals 2022,12, 1867 17 of 18 65. Chiyo, P.I.; Cochrane, E.P.; Naughton, L.; Basuta, G.I. Temporal patterns of crop raiding by elephants: A response to changes in forage quality or crop availability? Afr. J. Ecol. 2005,43, 48–55. [CrossRef] 66. Parker, G.E.; Osborn, F.V. Investigating the potential for chilli Capsicum spp. to reduce human-wildlife conflict in Zimbabwe. Oryx 2006,40, 343–346. [CrossRef] 67. Gross, E.M.; McRobb, R.; Gross, J. Cultivating alternative crops reduces crop losses due to African elephants. J. Pest Sci. 2016,89, 497–506. [CrossRef] 68. Inogwabini, B.I.; Mbende, L.; Bakanza, A.; Bokika, J.C. Crop damage done by elephants in Malebo Region, Democratic Republic of Congo. Pachyderm 2013,54, 59–65. 69. Bouma, J.; Bulte, E.; van Soest, D. Trust and cooperation: Social capital and community resource management. J. Environ. Econ. Manag. 2008,56, 155–166. [CrossRef] 70. Austin, J. Farmers’ Perceptions of ECAN’s Proposed, “Good Practice Discharge Allowance” in the Waimakariri Sub Region of Environment Canterbury’s (ECAN) District of New Zealand. Ph.D. Thesis, Lincoln University, Lincoln, New Zealand, 2014. 71. Espinosa-Goded, M.; Barreiro-Hurlé, J.; Ruto, E. What do farmers want from agri-environmental scheme design? A choice experiment approach. J. Agric. Econ. 2010,61, 259–273. [CrossRef] 72. Olsen, S.B.; Meyerhoff, J. Will the alphabet soup of design criteria affect discrete choice experiment results? Eur. Rev. Agric. Econ. 2017,44, 309–336. [CrossRef] 73. Ngene, C. 1.2 User Manual & Reference Guide; ChoiceMetrics Pty Ltd.: Sydney, Australia, 2018. 74. Hanley, N.; Mourato, S.; Wright, R.E. Choice modelling approaches: A superior alternative for environmental valuatioin? J. Econ. Surv. 2002,15, 435–462. [CrossRef] 75. Soliño, M.; Farizo, B.A. Personal traits underlying environmental preferences: A discrete choice experiment. PLoS ONE 2014,9, e89603. [CrossRef] 76. Soliño, M.; Oviedo, J.L.; Caparrós, A. Are forest landowners ready for woody energy crops? Preferences for afforestation programs in Southern Spain. Energy Econ. 2018,73, 239–247. [CrossRef] 77. Vermunt, J.K.; Magidson, J. Upgrade Manual for Latent Gold Choice 5.1: Basic, Advanced, and Syntax; Statistical Innovations Inc.: Belmont, MA, USA, 2016. 78. Kansky, R.; Knight, A.T. Key factors driving attitudes towards large mammals in conflict with humans. Biol. Conserv. 2014,179, 93–105. [CrossRef] 79. Muter, B.A.; Gore, M.L.; Riley, S.J. Social contagion of risk perceptions in environmental management networks. Risk Anal. 2013,33, 1489–1499. [CrossRef] [PubMed] 80. Linnell, J.D.; Nilsen, E.B.; Lande, U.S.; Herfindal, I.; Odden, J.; Skogen, K.; Andersen, R.; Breitenmoser, U. Zoning as a means of mitigating conflicts with large carnivores: Principles and reality. In People and Wildlife, Conflict or Co-Existence? Woodroffe, R., Thirgood, S., Rabinowitz, A., Eds.; Cambridge University Press: Cambridge, UK, 2005; p. 162. 81. Sitati, N.W.; Walpole, M.J. Assessing farm-based measures for mitigating human-elephant conflict in Transmara District, Kenya. Oryx 2006,40, 279–286. [CrossRef] 82. Ballantyne, P. Ownership and partnership: Keys to sustaining ICT-enabled development activities. IICD Res. Brief 2003,8, 1–8. 83. Weeks, J.; Anderson, D.; Cramer, C.; Geda, A.; Hailu, D.; Muhereza, G.; Rizzo, M.; Ronge, E.; Stein, H. Supporting Ownership: Swedish Development Cooperation with Kenya, Tanzania and Uganda; Sida Evaluation, Sida: Stockholm, Sweden, 2002; Volume 2. 84. Hedges, S.; Gunaryadi, D. Reducing human-elephant conflict: Do chillies help deter elephants from entering crop fields? Oryx 2010,44, 139–146. [CrossRef] 85. Baishya, H.K.; Dey, S.; Sarmah, A.; Sharma, A.; Gogoi, S.; Aziz, T.; Ghose, D.; Williams, A.C. Use of chilli fences to deter Asian elephants—A pilot study. Gajah 2012,36, 11–13. 86. Xiang, P.; Zhang, H.; Geng, L.; Zhou, K.; Wu, Y. Individualist–collectivist differences in climate change inaction: The role of perceived intractability. Front. Psychol. 2019,10, 187. [CrossRef] 87. Shaffer, L.J.; Khadka, K.K.; Van Den Hoek, J.; Naithani, K.J. Human-elephant conflict: A review of current management strategies and future directions. Front. Ecol. Evol. 2019,6, 235. [CrossRef] 88. Branco, P.S.; Merkle, J.A.; Pringle, R.M.; King, L.; Tindall, T.; Stalmans, M.; Long, R.A. An experimental test of community-based strategies for mitigating human–wildlife conflict around protected areas. Conserv. Lett. 2020,13, e12679. [CrossRef] 89. Matsika, T.A.; Adjetay, J.A.; Obopile, M.; Songhurst, A.C.; McCulloch, G.; Stronza, A. Alternative crops as a mitigation measure for elephant crop raiding in the eastern Okavango Panhandle. Pachyderm 2020,61, 140–152. 90. Rajamma, G. Changing from subsistence to cash cropping: Sakaramma’s story. Gend. Dev. 1993,1, 19–21. [CrossRef] [PubMed] 91. Baiphethi, M.N.; Jacobs, P.T. The contribution of subsistence farming to food security in South Africa. Agrekon 2009 ,48, 459–482. [CrossRef] 92. Gunaryadi, D.; Hedges, S. Community-based human–elephant conflict mitigation: The value of an evidence-based approach in promoting the uptake of effective methods. PLoS ONE 2017,12, e0173742. [CrossRef] [PubMed] 93. Thuppil, V.; Coss, R.G. Playback of felid growls mitigates crop-raiding by elephants Elephas maximus in southern India. Oryx 2016,50, 329–335. [CrossRef] 94. Stokke, S.; Du Toit, J.T. Sexual segregation in habitat use by elephants in Chobe National Park, Botswana. Afr. J. Ecol. 2002,40, 360–371. [CrossRef] Animals 2022,12, 1867 18 of 18 95. Smith, R.J.; Kasiki, S.M. A Spatial Analysis of Human–Elephant Conflict in the Tsavo Ecosystem, Kenya. AfESG Report; IUCN/SSC: Gland, Switzerland, 2000. 96. Haile, G.G.; Tang, Q.; Sun, S.; Huang, Z.; Zhang, X.; Liu, X. Droughts in East Africa: Causes, impacts and resilience. Earth Sci. Rev. 2019,193, 146–161. [CrossRef] 97. Kubo, T.; Shoji, Y. Spatial tradeoffs between residents’ preferences for brown bear conservation and the mitigation of human–bear conflicts. Biol. Conserv. 2014,176, 126–132. [CrossRef] 98. Treves, A.; Karanth, K.U. Human-carnivore conflict and perspectives on carnivore management worldwide. Conserv. Biol. 2003,17, 1491–1499. [CrossRef] 99. Van Eeden, L.M.; Crowther, M.S.; Dickman, C.R.; Macdonald, D.W.; Ripple, W.J.; Ritchie, E.G.; Newsome, T.M. Managing conflict between large carnivores and livestock. Conserv. Biol. 2017,32, 26–34. [CrossRef] 100. König, H.J.; Kiffner, C.; Kramer-Schadt, S.; Fürst, C.; Keuling, O.; Ford, A.T. Human-wildlife coexistence in a changing world. Conserv. Biol. 2020,34, 786–794. [CrossRef] 101. Røskaft, E.; Bjerke, T.; Kaltenborn, B.; Linnell, J.D.; Andersen, R. Patterns of self-reported fear towards large carnivores among the Norwegian public. Evol. Hum. Behav. 2003,24, 184–198. [CrossRef] 102. Kleiven, J.; Bjerke, T.; Kaltenborn, B.P. Factors influencing the social acceptability of large carnivore behaviours. Biodivers. Conserv. 2004,13, 1647–1658. [CrossRef] 103. Thornton, C.; Quinn, M.S. Risk perceptions and public attitudes towards cougars in the southern foothills of Alberta. Hum. Dimens. Wildl. 2010,15, 359–372. [CrossRef] 104. Tierney, K.J.; Lindell, M.K.; Perry, R.W. Facing the unexpected: Disaster preparedness and response in the United States. Disaster Prev. Manag. 2002,11, 222. [CrossRef] 105. Kirschenbaum, A. Preparing for the inevitable: Environmental risk perceptions and disaster preparedness. Int. J. Mass Emergencies Disasters 2005,23, 97–127. 106. Grothmann, T.; Reusswig, F. People at risk of flooding: Why some residents take precautionary action while others do not. Nat. Hazards 2006,38, 101–120. [CrossRef] 107. Winter, G.; Fried, J.S. Homeowner perspectives on fire hazard, responsibility, and management strategies at the wildland-urban interface. Soc. Nat. Resour. 2000,13, 33–49. [CrossRef] 108. McGee, T.K.; McFarlane, B.L.; Varghese, J. An examination of the influence of hazard experience on wildfire risk perceptions and adoption of mitigation measures. Soc. Nat. Resour. 2019,22, 308–323. [CrossRef] 109. Tiller, L.N.; Oniba, E.; Opira, G.; Brennan, E.J.; King, L.E.; Ndombi, V.; Wanjala, D.; Robertson, M.R. “Smelly” Elephant Repellent: Assessing the Efficacy of a novel olfactory approach to mitigating elephant crop raiding in Uganda and Kenya. Diversity 2022,14, 509. [CrossRef] 110. Treves, A.; Wallace, R.B.; Naughton-Treves, L.; Morales, A. Co-managing human-wildlife conflicts: A review. Hum. Dimens. Wildl. 2006,11, 383–396. [CrossRef] 111. Bennett, N.J.; Roth, R.; Klain, S.C.; Chan, K.; Christie, P.; Clark, D.A.; Cullman, G.; Curran, D.; Durbin, T.J.; Epstein, G.; et al. Conservation social science: Understanding and integrating human dimensions to improve conservation. Biol. Conserv. 2017,205, 93–108. [CrossRef]