Introduction

The cooperation of local communities is important for effective conservation of biodiversity in protected areas (PAs). In conservation circles, a general observation is that positive attitudes toward PAs are likely to engender pro-conservation behaviors (Holmes, 2003; St. John et al., 2012). According to Albarracín, Johnson, Zanna, and Kumkale (2005 p. 4.), “attitude is the psychological tendency of an individual to evaluate an entity (person, place, behavior or thing) with a degree of favour or disfavour.” Attitude plays a cardinal role in predicting human behavior toward the natural environment (Clayton, 2012; St. John, Edwards-Jones, Jones, & Moloney, 2010). Understanding the attitude of people living in communities near PAs is therefore useful, particularly to PAs managers and conservation policy makers (Allendorf, 2006; Kideghesho, Røskaft, & Kaltenborn, 2007; Mamo, 2015). With the right information on attitudes of local communities, PA managers and other conservation authorities can design appropriate management programs and interventions that can win the support of local communities and thus ensure effective biodiversity conservation (St. John et al., 2012).

Conservation attitudes are usually influenced by many factors depending on community needs vis-à-vis the objectives of PAs and the management strategies adopted to achieve those objectives (Bragagnolo, Malhado, Jepson, & Ladle, 2016). Park-community interaction and socioeconomic and demographic factors such as age, gender, household size, education, livelihood activity, and landholding are often used as explanatory variables to determine attitudes toward PAs or conservation in general (Gifford & Sussman, 2012). Age and gender usually interact with other factors, thereby making their unique contribution to attitudes context-specific and difficult to determine (Bragagnolo et al., 2016). The ethnicity, culture, local economy, and religion of a given society determine how gender roles are shared between men and women. For instance, in many African cultures, women and children are assigned the roles of collecting firewood, feeding animals, or fetching water (Ogato, Boon, & Subramani, 2009). Since these activities are based on the natural environment, conservation attitudes of women and children in such settings are likely to be influenced by the availability or scarcity of these resources (Bragagnolo et al., 2016; Gifford & Sussman, 2012). Formal education and conservation awareness are often found to be associated with positive attitudes toward PAs (Allendorf & Yang, 2015; Shrestha & Alavalapati, 2006; Sudarmadi et al., 2001). It is also argued that local residents with formal education are more likely to obey rules protecting PAs than nonliterates (Bragagnolo et al., 2016). However, it is not always the case that education level is positively associated with favorable attitudes. In Kenya, Guthiga (2008) found that educated households were more dissatisfied with the protectionist approach used to manage the Kakamega Forest despite the ecological effectiveness of the approach.

The availability or scarcity of non-timber forest products (NTFPs) to communities fringing PAs does influence attitudes of residents toward protection of natural resources (Kideghesho et al., 2007). Understanding the availability of NTFPs, and the level to which local communities can access those products, can provide important information about the benefits local communities derive from the forest. This helps to establish the level of dependence of local communities on the forest and how conservation regulations can be made and enforced to ensure that community needs and conservation objectives are both met (Schaafsma et al., 2014).

Spatial factors such as “location,” “distance from park to community,” and “landholding” in fringe communities also influence attitudes toward PAs. Spiteri and Nepal (2006) point out that residents in communities closer to park boundary (where conflicts are more likely) often have less positive attitudes than those in more distant settlements. Villages that are located inside PAs or in wildlife corridors may have the tendency to show negative attitude because they have higher chance of negative interaction with wildlife than those in distant areas (Synman, 2014). Landholding by households can be a determinant of attitudes toward PAs since scarcity of land can compel residents to fall back on neighboring PAs for resources. Allendorf (2006) reported land-owning residents to be more likely to show positive attitude toward PAs in Burma than their landless counterparts. Elsewhere in Nepal, Shrestha and Alavalapati (2006) did not find significant association between landholding and PA-attitudes.

Available literature indicates that the influence of cost-and-benefit considerations on PA-attitudes of local communities is stronger than the influence of demographic factors such as age and gender (Acquah et al., 2017; Bragagnolo et al., 2016; Synman, 2014). Local communities usually weigh what they stand to gain and what they stand to lose in any conservation policy or strategy before they form an attitude (Fiallo, & Jacobson, 1995). Weladji and Tchamba (2003) and Mamo (2015) recognize that damage caused by wildlife has been a fundamental source of people-PA conflicts due to lack of compensation and mistrust between victims and PA authorities. While victims are sometimes accused of overestimating destroyed property, local communities often perceive governments and conservation authorities to have more regard for wildlife than human beings (Davis, 2011). PA agencies also blame victims for farming or living too close to park boundaries (Synman, 2014). As a result, most people who experience frequent wildlife destruction of crops or livestock tend to show less positive attitude toward PAs. Employment of locals in PAs can alleviate poverty (Roe, 2008) and strengthen PA-community relationship. This engenders positive attitudes and support toward PAs (Cetas & Yasué, 2016).

This study examined and compared factors that influence attitudes of residents in communities near the Tarangire National Park (Tarangire NP) in Tanzania and the Mole National Park (Mole NP) in Ghana. The Mole and Tarangire NPs were selected for this study in an attempt to compare the PA-attitudes of park-adjacent communities in East and West Africa. Previous studies on this topic have confined their work to comparing local community attitudes in single regions of Africa (Synman, 2014) or in individual countries (Mutanga, Vengesayi, Gandiwa, & Muboko, 2015). The two national parks were created around the same time. Mole was gazetted as a game reserve in 1958 and made a national park in 1971 while Tarangire was gazetted in 1957 and made a national park in 1970. Both parks are located within savannah biomes and thus have similar wildlife resources.

The parks however differ in terms of the local communities bordering them and their level of development for tourism. For instance, the major ethnic groups living around the Mole NP (Gonja, Kamara, Mamprusi, and Chakali) are mostly food-crop farmers who live in nucleated villages and cultivate lands far away from their homesteads. On the other hand, the main ethnic groups living around the Tarangire NP are the Maasai, Iraqw, and Rangi, and the dominant livelihood activities include pastoralism and agro-pastoralist. Their homesteads are scattered and far apart from each other to give space for farmlands. Per their cultures, bushmeat consumption is popular among tribes living near the Mole NP in Ghana but the reverse is true for the tribes in communities near the Tarangire NP.

Community-based conservation interventions practiced in communities around the two parks also differ. In Tanzania, wildlife management areas (WMAs) and support for community initiated projects are two intervention programs that are supposed to be helping park-adjacent communities in terms of development and poverty alleviation (Kiwango, Komakech, Tarimo, & Martz, 2015). In Ghana, the operating community-based conservation strategy is known as community resource management area (CREMA). The CREMA mechanism allows communities to identify a resource and develop it with the support of governmental or nongovernmental organizations for sustainable resource use and poverty reduction (Murray, Agyare, Dearden, & Rollins, 2018).

Tarangire NP has better structures for receiving tourists than the Mole NP and therefore is a more popular tourists’ destination. Different tourism agencies operate over 15 hospitality facilities and safari services in and around the Tarangire NP. However, only few individuals from the neighboring communities are employed by these agencies. The explanation given is that people in the local communities lack expertise needed to work in such agencies. In an effort to involve the neighboring communities in the tourism business, the Tarangire NP offers some skills training and financial assistance to some members of neighboring communities. In communities near the main entry points of the park, selected individuals are given training and financial support to make and sell artifacts to tourists. Others are given horticultural training to produce and supply fruits and vegetables to the hospitality entities in and around the park.

On the other hand, there are only two lodges in the Mole NP. According to operators of the facilities, most of their employees were recruited and trained from the neighboring communities. Again villages near attraction sites in the Mole NP are assisted by the park to develop homestay accommodation facilities to host tourists who are interested in immersing in local culture. Neighboring communities such as Mognori and Laribanga with ecological and cultural landmarks are becoming satellite tourists’ attraction sites that generate some income for the communities (McDowell, 2012). In both national parks, tourists visit neighboring villages to experience traditional setting and culture of the people and this leaves some tourists cash in the villages.

The specific objectives of the study include to:

Methods

Study Areas

Mole NP—Ghana

The Mole NP was established in 1958 as a game reserve and gazetted in 1971 as a national park. The Mole NP is the first and largest (4,577 km²) national park in Ghana. It is located in the northern region of Ghana and straddles four administrative districts: West Gonja Disrict, West Mamprusi District, Sawla-Tuna Kalaba District, and Wa District. There are 33 communities fringing the Mole NP with over 40,000 inhabitants. The park is about 149 km from Tamale, the capital city of the Northern Region and is the most important national park in Ghana in terms of visitor attraction and availability of tourists’ facilities (Mole NP, 2011).

Tarangire NP

Tarangire NP is the sixth largest national park in Tanzania after Ruaha, Serengeti, Katavi, Mikumi, and Mkomazi National Parks. It is about 118 km from Arusha (the regional capital city) and straddles three administrative districts: Babati, Simanjiro, and Monduli. Tarangire NP has an area of 2,850 km² and was designated a game reserve in 1957 and gazzeted a national park in 1970. Tarangire NP is rich in mega fauna, which makes it an attractive tourist’s destination. There are over 58 species of large mammals with the popular ones being zebras (Equus quagga), wildebeests (Connochaetes taurinus), buffalo (Syncerus caffer), and elephants (Loxodonta africana) (Abukari & Mwalyosi, 2018). Tarangire NP is surrounded by 42 local communities, and most of the people in these communities are agro-pastoralists or pure pastoralists. Figure 1 shows the study areas with the administrative districts hosting the Mole and Tarangire NPs.

Figure 1.

Map of the study areas showing the administrative districts hosting the Mole and Tarangire National Parks.

Data collection

Household survey sampling design

A village each was randomly selected from the three and four administrative districts hosting the Tarangire and Mole NPs, respectively. Census reports indicated that there were 2,136 and 2,171 people in villages 10 km from the boundaries of the Tarangire NP and Mole NP, respectively (Ghana Statistical Service, 2014; National Bureau of Statistics, 2013). The sum of the two figures (4,307) was taken as the universal population size and Cochran formula (Cochran, 1963) was used to get a representative sample size from the universal population.

where $n_0$is the sample size, Z² is the abscissa of the normal curve, e is the desired level of precision, p is the estimated proportion of an attribute that is present in the population, and q is 1–p. The value for Z was taken from statistical tables. For this study, confidence level was set at 95%, precission (margin of error) at ±5% and p = .5. The sample size $\left(n_0\right)$ was therefore calculated as illustrated later:

For finite and relatively small populations like the one for this study, Cochran suggests an adjustment of the formular which he calls the finite population correction. The correction is necessary to give a more precise sample size. The adjustment was done using Cochran’s second equation as shown later:

where n = the final sample size, N = the universal population size (4,307), n₀ = initially calculated sample size (385).

The resulting sample size (354) was therefore proportionately shared for villages near the two national parks.

At the individual park level, samples were again distributed proportionately among villages based on the number of households in each village. Households were selected through random sampling by using a random number table.

The questionnaire comprised of questions to measure the dependent variable (attitude toward park) and the predictor variables. The predictor variables were derived from park-community interactions as well as demographic and socioeconomic characteristics (age, gender, education, livelihood activity, landholding, and household size). Park-community interaction variables included (a) distance from community to park boundary (distance), (b) employment by park (employment), (c) knowledge of park rules and regulations (park rules), (d) access to NTFPs, and (e) damage caused by wildlife.

The scale of statements used to measure the dependent variable had two parts. Table 2 shows the statements and percentages of responses given by respondents. Part 1 comprised of statements meant to assess attitudes of respondents toward the national park as an ecological entity, and Part 2 assessed respondents’ perception of the national park as an agent for community development. This was done to avoid generalization of attitudes of respondents towards the parks. Care was taken to counterbalance positively and negatively worded statements so as to have equal number of positive and negative statements on each issue (Ajzen, 2002).

The questions were drawned from responses obtained from elicitation studies conducted in three villages each near the Mole and Tarangire NPs prior to designing the questionnaire. The aim of the elicitation studies was to identify salient issues that arise from interactions between the two national parks and their neighbouring communities (Ajzen & Fishbein, 1980).

A 7-point Likert scale (1 = stongly disagree . . . 7 = strongly agree) was put against each statement for respondents to indicate their level of agreement or disagreement. Statement scores were added together to give overall attitude scores (Aipanjiguly, Jacobson, & Flamm, 2003).

The questionnaire was pretested in October 2016 and was refined and administered in December 2016 in Tanzania. It was translated from English to Kiswahili for use in Tanzania since Kiswahili is the national and most popular language in Tanzania. In Ghana, the questionnaire was pretested in March 2017 and administered in June 2017. In Ghana, the questionnaire was administered in English because all enumerators could fluently speak English and the local languages spoken in the surveyed villages.

Focus group discussion

A check list of questions was prepared to elicit information on residents’ perceptions on: (a) natural resources conservation (e.g., What do you think about protecting some part of the natural environment for future generations?), (b) benefits of the park to the community (e.g., What are the advantages of living next to the national park?), (c) the park’s restriction on access to resource use (e.g., What are the disadvantages of living next to the national park?), (d) park governance and management (e.g., Are you involved in decision-making processes to decide how natural resources in your area should be managed?), and (e) livelihood impacts (e.g., How does the national park affect your livelihood activities?). The discussions were held to collaborate, cross-check, or infer the data collected on the household survey questionnaire. Discussions were held in each of the surveyed villages. Total number of participants were 27 and 32 for Tarangire NP and Mole NP, respectively.

Data analysis

Data analysis was conducted using IBM SPSS version 20 with the level of significance set at p < .05. Reliability of the scale used to measure the dependent variable (Attitude toward park) was checked using Cronbach’s alpha. Cronbach’s alpha values lie between 0 and 1 with values ≥ 0.7 indicating high internal consistency (DeVellis, 2003). Chi-square test and t-test were test was used to determine significant difference between park-community interactions and sociodemographic and socioeconomic variables for the two groups of respondents (Table 1). The Student’s t test was used to find differences in the mean attitude scores between the groups of respondents in Mole and Tarangire NPs. The effects of independent variables on the dependent variable were determined using stepwise regression. The regression process included checking to ensure no violation of the assumptions of normality, linearity, homoscedasticity, and multicollinearity. Respondents’ mean attitude scores were regressed with “Landholding,” “Distance,” “Park rules,” “Employment,” “DCW,” “Household size,” “Age,” “Gender,” “Education,” “Access to NTFPs,” and “Livelihood activity” (crop farming, pastoralism, agro-pastoralism, petty trading, and fishing or hunting).

Results

Background Characteristics of Respondents

We conducted a total of 365 interviews with 181 respondents in Tarangire NP and 184 respondents in Mole NP. More males than females were encountered as household heads in both countries (Mole NP: males = 68.5%, females = 31.5%; Tarangire NP: males = 62.4%, females =37.6%). Table 1 shows details of predictor variables investigated to relate with attitude of respondents in Mole and Tarangire NPs.

Table 1.

Predictor Variables Used to Explain Attitudes of Respondents Toward the Mole and Tarangire NPs.

Table 2.

Statements Used to Measure Attitudes Toward Mole and Tarangire NPs and Percentages of Responses.

About two thirds (70%) of respondents in Tarangire NP and about half (54%) of their counterparts in Ghana indicated their crops or livestock have ever been destroyed by wildlife.

Influence of Park-Community Interactions, Demographic, and Socioeconomic Characteristics on Attitudes

Stepwise multiple regression was performed to predict the overall attitudes of respondents toward the Mole and Tarangire NPs. For respondents in Tarangire NP, significant predictor variables include “distance,” “household size,” “park rules,” “access to NTFPs,” and “livelihood activities.” Only three predictor variables (“distance,” “household size,” and “park rules”) were found to be significant for respondents in Mole NP. Table 3 shows details of the model summary.

Table 3.

Model Summary of Regression Relating Predictor Variables and Attitudes of Respondents Toward the Tarangire and Mole National Parks.

The best models (Model 7 for Tarangire NP and Model 3 for Mole NP) explained 82% and 74% of the variation in respondents’ attitude toward the Tarangire NP (R² = 0.82, F(7, 173) = 114.27, p = < .001) and the Mole NP (R² = 0.74, F(3, 178) = 169.59, p < .001), respectively. Table 4 shows details of Models 7 and 3 for Tarangire and Mole NPs, respectively. Distance from respondents’ location to park boundary was the most important variable predicting attitudes in both parks, accounting for 76% and 69% variation for respondents in Tarangire and Mole NPs, respectively.

Table 4.

Results of Standard Multiple Regression Showing the Effects of Demographic, Socioeconomic Factors, and Park-Community Interaction on Attitudes Toward the Mole and Tarangire National Parks.

Discussion

Our results showed significant difference between the park-attitudes of residents in communities near the Mole NP in Ghana and that of their counterparts in communities near the Tarangire NP in Tanzania. However, residents in communities near both parks showed slightly positive attitudes toward the parks as ecological entities and slightly negative attitudes toward them as agencies for economic advancement. The results demonstrate that while residents in communities near the two parks may not be satisfied with the economic support they get from the parks, they appreciate the noneconomic values of conservation. This is in line with the findings of Allendorf (2006) and Synman (2014) who reported that residents in communities near PAs in developing countries do appreciate the noneconomic values of PAs such as ecosystem services and benefit of future generations. When attitudes of local residents toward the ecological values and socioeconomic values of PAs are combined and measured, it can lead to misconception and misinterpretation of findings where local residents can easily be seen to be negative and confutative to the concept of conservation (Allendorf, 2006).

Difference in community-based natural resources management and use strategies adopted in Ghana and Tanzania could also contribute to the slight difference between attitudes of respondents. The CREMA in Ghana allows the state institution in charge of wildlife management (Wildlife Division of the Forestry Commission) to transfer authority and responsibility for wildlife management to rural communities to manage (Asare, Kyei, & Mason, 2013). The main aim of CREMA is to ensure sustainable use of natural resources and improved livelihoods of local communities. CREMA-owning communities are trained on wildlife monitoring and rehabilitation of degraded vegetation. More sustainable local production systems are adopted and both financial and technical assistance given to community members to boost their livelihood activities (Bosu, 2014; Murray, et al., 2018).

On the other hand, the operating community-based natural resources management mechanism in Tanzania is the WMAs system. WMAs like CREMAs are preferably and mostly established in areas adjacent to large state-owned PAs such as national parks. WMAs are a strategy to sustainably manage and use wildlife resources on village-owned lands in Tanzania. Although local communities are said to be the owners and managers of WMAs, management activities are guided by laws and regulations given by state institutions who also have a share in benefits accruing from the WMAs (Green & Adams, 2015; Kiwango et al., 2015). Park-adjacent communities in Tanzania are also expected to benefit from earnings of national parks through a program known as Support for Community Initiated Projects, sponsored by TANAPA.

It would appear the WMA system in Tanzania may be impacting more positively on attitudes toward PAs than the CREMA system in Ghana as our results show that respondents in Tarangire NP are slightly more positive than their counterparts in Mole NP.

Conservation initiatives under the CREMAs and WMAs have common objectives of ensuring sustainable resource management and utilization (Asare et al., 2013; Kiwango et al., 205); however, their impact on the PA-attitudes of neighboring communities may still be marginal. As our results indicated, respondents in both parks showed almost neutral attitudes toward the parks which may be an indication of lack of interest in the affairs of the parks. The CREMAs and WMAs initiatives benefit only a minute fraction of the population living near the parks, thus limiting their ability to influence change in attitudes in the larger park-adjacent communities.

The Mole and Tarangire NPs both have it as part of their objectives to ensure that local communities benefit from tourism activities. The two parks are employing various strategies to incorporate local communities into the tourism business but so far only very few of the populations living near parks are benefiting from tourism. The contribution of tourism to the improvement of livelihoods in neighboring communities is minute and may not be able to influence attitudes toward the parks.

Implications for Conservation

Our study revealed that residents in communities near the Mole and Tarangire NPs accept the concept of setting land aside to conserve natural resources. However, there is also the perception among neighboring local communities that they do not get equitable share of benefits from the parks. Findings of this study point to the need for livelihood diversification in communities near the Tarangire NP to give residents income-generating activities that are more compatible with wildlife conservation. Apart from the potential to engender positive attitudes and support for the park, alternative livelihood activities will avert future pressures that may result from the increasing livestock population and the area of farmlands in the Tarangire ecosystem (Hariohay & Røskaft, 2015). The Mole NP may have to intensify conservation awareness education and extend it to communities that are more distant in order to change attitudes and behaviors about wildlife use in their neighboring districts. More communities around the Mole NP need to be targeted for CREMA projects so that more local people can receive conservation awareness education (Murray et al., 2018) and may become more committed to sustainable use of forest and wildlife resources.

Damage caused by wildlife and the lack of compensation thereafter engender negative attitudes toward PAs and heighten conservation conflicts. Both Mole and Tarangire NPs need to support local residents to find new ways of curbing wildlife attacks on farms, livestock, and humans. This could be done by supporting communities with financial and technical resources to build barriers that will prevent marauding wild animals from entering farms and villages. For example, the elephant repellant fence technique, which comprises of fabric lased with powdered chili pepper, is found to be effective in repelling the elephant from crop farms (Enukwa, 2017). This technique could be disseminated in communities around the Mole and Tarangire NPs to reduce crop-raiding incidences and promote good relationship between Park and community.