Beyond a single perspective to conservation relationships: exploring factors influencing protected area staff and local community relationships in Zimbabwe

ABSTRACT

We explored the relationships between protected area (PA) staff and adjacent communities in and around four PAs in Zimbabwe. A total of 938 local people and 133 PA staff participated in the survey conducted between July 2013 and February 2014. Our results showed that communities generally perceived a negative relationship with PA staff, while PA staff generally perceived a positive relationship with local communities. Only benefit-sharing had a different effect on PA staff–community relationship for the PA staff and community samples. In contrast, communication, perceptions (both communities’ and PA staff’s), community involvement in conservation, history of PA creation, and problems caused by PA existence to the communities or by communities to PAs had the same effect on PA staff–community relationship from both perspectives. We recommend that both conservation agencies and communities should pay more attention to factors that influence PA staff–community relationships so as to nurture positive relationships between PA staff and local communities.

KEYWORDS:

• Conservation
• factors
• perceptions
• tourism
• community

Edited by Christian Albert

1. Introduction

Most state protected areas (PAs) were initially inhabited by or used by people who were displaced when these PAs were established (Borrini-Feyerabend et al. 2006; Mombeshora & Le Bel 2009; Muboko & Murindagomo 2014). Many people largely depend on the resources found within PAs for their survival (Lockwood & Kothari 2006). As a result, people whose livelihoods primarily involve the direct exploitation of local natural resources often come into conflict with PA’s management, for example, where issues such as illegal resource harvesting, habitat encroachment and destruction, and human–wildlife conflict are involved (Gandiwa et al. 2013a; Muboko et al. 2014; Matema & Andersson 2015). These conflicts continue to influence local communities’ perceptions of wildlife, PAs, PA staff, and tourism among other factors (Kideghesho et al. 2007; Triguero-Mas et al. 2009).

Community involvement and support for the conservation of natural resources has been suggested as a prerequisite for the long-term sustainability of PAs (Fiallo & Jacobson 1995; Tessema et al. 2010 Andrade & Rhodes 2012). Thus, it is widely postulated that PAs cannot coexist in the long term with communities that are hostile to them (West & Brechin 1991; Mcneely 1995; Pimbert & Pretty 1995; Holmes 2013), despite some authors, for example, Brockington (2004) and Stern (2008), arguing that local community support is not necessarily crucial for the survival of PAs. Accordingly, there is growing scientific literature on PA staff–community relationships as being of vital importance to wildlife conservation (Borrini-Feyerabend et al. 2002; Berkes 2004; Brockington 2004; Buscher & Whande 2007; Hausser et al. 2009; Tessema et al. 2010).

Earlier studies highlighted various factors that influence PA staff–community relationships which this present study builds on. Grunig and Huang (2000) identified trust, relationship commitment, control mutuality, and relationship satisfaction as the most important indicators of successful relationships. Other studies have identified various factors that influence PA staff–community relationships such as history of PAs creation (Choudhury 2004, Mombeshora & Le Bel 2009), benefit-sharing (Allendorf 2010; Tessema et al. 2010), problems faced by the communities from the PAs such as loss of crops and livestock, and safety to human lives (Kideghesho et al. 2007; Harihar et al. 2014), communication between PA staff and communities (Ormsby & Kaplin 2005; Bruyere et al. 2009), community involvement in conservation (Fiallo & Jacobson 1995; Tessema et al. 2010), and community attitudes and perceptions (Triguero-Mas et al. 2009; Allendorf 2010). Recently, Mutanga et al. (2015a) observed other factors influencing PA staff–community relationships, that is, PA staff perceptions on communities and problems faced by PAs from the communities such as illegal resource harvesting and veld fires. However, relationships between PA staff and communities have largely been evaluated taking into consideration the communities’ perspectives with PA staff’s perspectives having been largely ignored (Bruner et al. 2001). Thus, there is a limited understanding of conservation relationships between PA staff and local communities from these two perspectives. This study attempts to fill this gap by incorporating PA staff perspectives of the factors that influence their relationship with the community. PA staff–community relationship refers to the interactions between PA staff and local communities based on interdependence, and where the behaviour of each affects the other (Mutanga et al. 2015a).

Understanding both sides of a relationship can contribute highly to enhancing biodiversity conservation considering that PA staff and local communities are interdependent and their perceptions of each other can positively or negatively affect natural resource conservation. Since perceptions are regarded as attitude-forming processes (Allendorf et al. 2012), negative perceptions by the local communities could imply that conservation problems such as illegal resource harvesting and habitat encroachment could remain a challenge, while negative perceptions by PA staff could also imply that they will put little effort to maintain positive relationships with local communities.

Previous studies on PA staff–community relationship have mainly focussed on single PAs as study sites (e.g., Adams & Infield 2003; McCleave et al. 2006; Tomic´Evic´ et al. 2011; Allendorf et al. 2012) and on the single perspectives, that is, local community perspectives on their relationships with PAs (Kappelle 2001; McCleave et al. 2006; Allendorf 2010; Nagendra et al. 2010). Here, we contribute to scientific knowledge on conservation relationships through focussing on multiple PAs with varying management regimes and also examine the conservations relationships from two perspectives, that is, those of PA staff and local communities. The present study explores PA staff–community relationships in four conservation areas in Zimbabwe covering both state and private PAs, and adjacent communities with and without the Communal Areas Management Programme for Indigenous Resources (CAMPFIRE). The objectives of this study were twofold: (i) to assess how communities and PA staff view their relationship and (ii) to determine factors influencing PA staff–community relationships from the perspectives of both PA staff and communities.

2. Materials and methods

2.1. Study area and study sites

Zimbabwe was chosen as a case study due of its known history of wildlife conservation and the land reform programme whose effects on wildlife conservation has been widely reported (Gandiwa et al. 2014a). Stratified sampling design (Hair et al. 2006) was employed to divide PAs into state and privately owned PAs and adjacent communities into with and without CAMPFIRE. CAMPFIRE utilises wildlife and other natural resources, and promotes devolution of rights to manage, use, dispose of, and benefit from natural resources to rural institutions and improved governance and livelihoods (Martin 1986). CAMPFIRE is based on the principle that if communities receive economic benefits from wildlife, they will change their attitudes, hence effectively conserve and manage the natural resources (Murombedzi 2001).

Four study sites located in different districts of Zimbabwe were selected purposively to give a broad view of PA staff and community views on conservation relationships in Zimbabwe, that is, Umfurudzi Park, Gonarezhou National Park, Matusadona National Park and Cawston Ranch (Figure 1). The wide geographical distribution and varying management regimes provided a good opportunity for a detailed assessment of PA staff–community relationships.

Figure 1. Location of the four study sites in Zimbabwe (see Table 1 for details).

All the sampled villages surrounding a PA are referred to as a community in this study, hence the four communities (i.e., Umfurudzi, Gonarezhou, Matusadona, and Cawston Ranch). All sampled villages surrounding a PA live within a defined spatial boundary, are socially bound by a common cultural identity, and are assumed to have common socio-economic and cultural interest in the resources of the neighbouring PA (see Barrow & Murphree 2001). Moreover, grouping the villages together into a single community allowed for easy comparison among the study sites. Table 1 outlines the general characteristics of the four study sites, that is, PAs and their surrounding communities.

Table 1. General characteristics of the study sites: four PAs and adjacent communities, Zimbabwe.

Attributes	Study site
	Umfurudzi	Gonarezhou	Matusadona	Cawston Ranch
Status	Safari Area	National Park	National Park	Safari area
Ownership	Government	Government	Government	Private
Management	Public–private partnership	Public–private partnership	Public	Private
Year established	1981	1930 as a Game reserve, upgraded to a National Park in 1975	1963 as a Game reserve, upgraded to a National Park in 1975	1988
Size (km^2)	760	5053	1400	128
Animal species	Mixed species of both carnivores and herbivores	Wide variety of both large carnivores and herbivores	Wide range of carnivores and herbivores	Mainly small herbivores
Tourism facilities and attractions	Chalets, tented camps, camp sites, caves with paintings, mountains	Tented camps, campsites, waterfalls, cliffs and natural water pans	Lodges, campsites	Bush camps
Forms of tourism	Photographic, sport fishing	Photographic, sport fishing	Photographic, sport fishing	Trophy hunting
Study areas (as depicted in Figure 1)	1-Sanye, 2-Mufurudzi 1, and 3-Mufurudzi 2	1-Chizvirizvi, 2-Mupinga, 3-Chitsa, 4-Mutandahwe, and 5-Mahenye	1-Nebiri, 2-Musambakaruma 2, and 3-Musambakaruma 1	1-Ward 10 and 2-Ward 9
CBNRM projects	None	CAMPFIRE	CAMPFIRE	None
Sources of community livelihoods	Small-scale subsistence and cash crop farming Small-scale livestock production Gold panning	Small-scale substance and cash crop farming Small-scale livestock production	Small scale subsistence and cash crop farming Very little livestock production due to tsetse fly prevalence	Small-scale subsistence and cash crop farming Small scale livestock production
Local languages	Shona	Shangani	Tonga, Shona	Ndebele
Potential conflicts between PAs and communities
Community benefits from PAs	No benefits	Mainly CAMPFIRE benefits	Employment benefits CAMPFIRE benefits	A number of material benefits, for example, game meat, employment
Human–wildlife conflict	Loss of crops and livestock	Loss of crops and livestock	Minimal crop and livestock destruction	Loss of crops
Compensation for losses from wildlife	No compensation	No compensation	No compensation	No compensation
Community involvement in decision-making	None	Limited involvement only in CAMPFIRE management	Limited involvement only in CAMPFIRE management	None

Source: Gandiwa et al. (2012); Muboko et al. (2014); Muposhi et al. (2014); Muposhi et al. (2015); Mutanga et al. (2015b).

CBNRM: Community-Based Natural Resource Management; CAMPFIRE: Communal Areas Management Programme for Indigenous Resources. Shona people identify themselves by clans such as Karanga, Korekore, Ndau, Manyika or Zezuru which are not captured in this Table.

2.2. Data collection

We used systematic sampling to select households within communities adjacent to the four selected PAs. The communities comprised wards divided into villages and then households within each village setup; hence, the household was used as the sampling unit. A ward is made up of six or seven villages (Madzudzo 1997; Gandiwa et al. 2013a). Systematic sampling was used because samples are easier to draw and execute. Moreover, a systematic sample spread the members selected for measurement more evenly across the entire population, thus, is more precise and representative of the population (Thompson 2012). As part of data collection, we used maps of the PAs showing the adjacent villages to choose transects through the communities that would allow us to cover all the study villages (Messer & Townsley 2003). Sampled households were restricted to within 10 km of the PA boundary due to the likelihood of increased local people–PA interaction (Gandiwa et al. 2014b). On entering a village, we randomly marked the first household as the starting point, after which a random direction from that household was selected and then every third household close to the transect was interviewed. Convenience sampling was used to select PA staff respondents and questionnaires were distributed to all the PA staff that happened to be on duty during the data collection period so as to get as many respondents as possible.

Close-ended questionnaires were used as these are comparatively easy to administer and manage, especially considering large sample sizes. Moreover, close-ended questionnaires are quick and easy to code and interpret, and therefore are amenable to rapid statistical analysis (Leung 2001) despite their limitations in gathering in-depth and detailed information (Barribeau et al. 2005). However, in our case we were more interested in determining respondents’ perceptions using predetermined specific indicators (factors) informed by the literature review, hence the suitability of close-ended questionnaires. The questionnaires were revised after a pilot test with 38 community members from Magazi village adjacent to Umfurudzi Park to remove ambiguities and misunderstandings. Since Magazi village had already been involved in the research, it was not included in the final sample for fear that those who participated in the pilot study could influence the later behaviour of research subjects (Haralambos 2008). The measures used for the pilot study and actual data collection were therefore slightly different.

The questionnaires consisted of three major sections, that is: (i) factors that influence PA staff–community relationships in which eight factors (namely, history of PA creation, benefit-sharing, problems caused by PA existence to adjacent communities, communication between PA staff and communities, community involvement in conservation, community perceptions on tourism, conservation, and PA staff), were used for the community questionnaire. The PA staff questionnaire, however, had six factors because the other two factors: perceptions on conservation and perceptions on tourism were removed since conservation and tourism are part of the PA responsibility. (ii) Community/PA staff perceptions of their relationship with each other, were assessed using Grunig and Huang (2000)’s relationship measurement scale and slightly modified to apply to PA staff–community relationships; and (iii) respondents’ demographics. Respondents were asked to indicate how much they agreed with the given statements on a 7-point Likert scale ranging from ‘strongly disagree/very less extent’ to ‘strongly agree/very great extent’ (Malhotra & Peterson 2006). The 7-point Likert scale (with 1–3 representing a negative perception of the relationship; 4 representing a neutral perception; and 5–7 representing a positive perception) was used to expand response options available to respondents (Colman et al. 1997) and enable respondents to make better discrimination (Fornell et al. 1996).

A total of 1000 questionnaires were distributed to the study communities and 938 were returned (response rate = 93.8%) (i.e., Umfurudzi 74, Gonarezhou 278, Matusadona 281; Cawston Ranch 305). As for the PA staff, a total of 180 questionnaires were distributed and 133 were returned (response rate = 73.9%) (i.e., Umfurudzi 22, Gonarezhou 37, Matusadona 28; Cawston Ranch 46; see Appendix 1 for details on demographic profiles). For the community sample, a questionnaire was given to a household head or an adult with at least 18 years who could read and write present at the target households at the time of survey. For those who could not read and write, an interview was conducted by research assistants with the aid of the close-ended questionnaire. For the PA sample, a questionnaire was given to every staff member on duty during survey period. Data were collected between July 2013 and February 2014. Completion of each questionnaire or interview took 20–30 minutes. Questionnaires were administered with the help of local research assistants with at least 4 years of secondary education who were trained and received instructions about the objectives of the study, and how to collect data. We obtained permission and/or informed consent to conduct and participate in the survey from property holders, district authorities, traditional leaders, and respondents.

2.3. Data analysis

Collected data were grouped into two sets, that is, for communities and PA staff (see Appendices 2 and 3 for descriptive statistics for community and PA data, respectively). Frequencies were used to summarise responses on community–PA relationship. The eight factors from the community perspectives and six factors from PA staff perspectives on PA staff–community relationships were analysed using the ordinal logistic regression. Ordinal logistic regression predicts an ordinal dependent variable given one or more independent variables (Fullerton 2009). Ordinal variables are categorical variables with ordered categories, for example, Likert items, among other ways of ranking categories (Ananth & Kleinbaum 1997; Lall et al. 2002). In our case, we had one ordinal dependent variable, ‘PA staff–community relationship’, with seven ordered categories: ‘1 = Strongly Disagree’, ‘2 = Disagree’, ‘3 = Somewhat Disagree’, ‘4 = Neither Agree nor Disagree’, ‘5 = Somewhat Agree’, ‘6 = Agree’ and ‘7 = Strongly Agree’. This meant that we could not use the binary choice model for the analysis of the data but multinomial or ordered choice models that allow for more than two dependent variables (Ezebilo et al. 2013). However, the multinomial model is often used for modelling unordered dependent variables, while the ordered choice model is more suitable for ordered dependent variables. As with other types of regression, ordinal regression can also use interactions between independent variables to predict the dependent variable (Fullerton 2009).

Prior to running the data in ordinal logistic regression, two tests, that is, (i) a multicollinearity test and (ii) full likelihood ratio test to evaluate the proportional odds were run and the results of these tests confirmed suitability of the ordinal regression model. Multicollinearity in this study was acceptable as indicated by tolerance levels ranging from 0.39 to 0.88 as well as variance inflation factor values between 1 and 3 (see Mertler and Vannatta (2002) and De Vaus (2002) for comparisons). The odds ratios were used to indicate the change in odds resulting from a unit change in the predictor (independent) variable (Field 2009). Predictors greater than 1 indicate that as the predictor increases, the odds of a positive PA staff–community relationship occurring increases, and predictors less than 1 indicates that as the predictor increases, the odds of a positive PA staff–community relationship occurring decreases. All the independent variables were statistically significant at p < 0.001 suggesting that for our models, the proportional odds assumption appears to have held (Bruin 2006). All analyses were conducted using the Statistical Package for the Social Sciences Version 21 (SPSS, Chicago, IL).

3. Results

3.1. Community and PA staff perceptions of their relationship

Approximately, 62.5% (n = 586) of the community respondents perceived the relationship they had with the PAs to be negative, 15.0% (n = 141) perceived a neutral relationship with the PA staff, while about 22.5% (n = 211) perceived a positive relationship with PA staff. As for the PA sample, about 48.9% (n = 65) of the PA staff rated their relationship with the communities positively, while 30.8% (n = 41) perceived a neutral relationship with the communities, whereas about 20% (n = 27) rated their relationship with the communities to be negative (Figure 2).

Figure 2. Community and PA staff perceptions of the relationship they have with each other.

About 54.7% (n = 513) of the community sample reported a very low level of trust in PA staff, 72.0% (n = 675) attested to the fact that PA staff and the communities did not agree on their power to influence, and 65.8% (n = 617) reported as not being satisfied with their current relationship with the PA staff. About 52.6% (n = 428) of the community respondents, however, indicated that they were committed to maintaining a good relationship with PA staff. As for the PA sample, about 56.4% (n = 75) reported as being satisfied with their current relationship with the communities and 64.6% (n = 86) indicated that they were committed to maintaining a good relationship with the communities. Moreover, 48.1% (n = 64) of the PA staff indicate to have some trust in the local community. The lowest relationship indicator was control mutuality with approximately 45.9% (n = 61) indicating that the PA staff and the communities agreed on their power to influence (Figure 2).

3.2. Factors influencing PA staff–community relationships

3.2.1. Community perspectives

The ordinal regression model from the community sample explained a significant amount of the original variability [χ²(8) = 915.76, p < 0.001; R² = 0.77]. Our results showed that an improvement in six of the eight tested factors was associated with an increase in the odds of having a positive PA staff–community relationship (Table 2). These are, communication with an odds ratio of 1.58 (95% CI, 1.41–1.77), Wald χ²(1) = 61.97, p < 0.001; community perceptions of tourism with an odds ratio of 1.7 (95% CI, 1.5–1.93), Wald χ²(1) = 66.96, p < 0.001; community perceptions of conservation with an odds ratio of 1.75 (95% CI, 1.58–1.94), Wald χ²(1) = 117.49, p < 0.001; community perceptions of PA staff with an odds ratio of 1.66 (95% CI, 1.45– 1.9), Wald χ²(1) = 53.35, p < 0.001; benefit-sharing with an odds ratio of 1.37 (95% CI, 1.18–1.58), Wald χ²(1) = 17.68, p < 0.001; and community involvement in conservation with an odds ratio of 1.28 (95% CI, 1.11–1.47), Wald χ²(1) = 11.68, p < 0.001. Contrastingly, an increase in the effects of the history of PA creation was associated with a decrease in the odds of having a positive PA staff–community relationship, with an odds ratio of 0.74 (95% CI, 0.67–0.82), Wald χ²(1) = 32.75, p < 0.001. Although problems caused by PA existence to adjacent communities had an odds ratio of 0.92 (95% CI, 0.83–1.02), whether the problems increased did not significantly affect PA staff–community relationship, Wald χ²(1) = 2.8, p > 0.05.

Table 2. Ordinal logistic regression results explaining the influence of eight factors on PA staff–community relationships from the community data sample.

		95% Confidence intervals for odds ratios
Model	Coefficient estimate (Std. error)	Lower	Odds	Upper
PA–community relationships are influenced by:
History of PA creation	−0.28 (0.05)*	0.67	0.74	0.82
Communication	0.46 (0.06)*	1.41	1.58	1.77
Perceptions of tourism	0.53 (0.07)*	1.50	1.70	1.93
Perceptions of conservation	0.56 (0.05)*	1.58	1.75	1.94
Perceptions of PA staff	0.51 (0.07)*	1.45	1.66	1.90
Problems caused by PA existence to adjacent communities	−0.08 (0.05)^a	0.83	0.92	1.02
Benefit-sharing	0.31 (0.07)*	1.18	1.37	1.58
Community involvement	0.25 (0.07)*	1.11	1.28	1.47
R^2 = 0.77
Model x^2(8) = 915.76*

*p < 0.001, ^aNon-significant (ns, p > 0.05).

3.2.2. PA staff perspectives

The ordinal regression model from the PA staff sample explained a significant amount of the original variability [χ²(6) = 55.32, p < 0.001; R² = 0.50]. Our findings showed that an improvement in three of the six tested factors was associated with an increase in the odds of having a positive PA staff–community relationship (Table 3). These are, communication with an odds ratio of 1.81 (95% CI, 1.17–2.79), Wald χ²(1) = 7.2, p < 0.01; PA staff perceptions of communities with an odds ratio of 1.82 (95% CI, 1.15–2.88), Wald χ²(1) = 6.59, p < 0.05; and community involvement in conservation with an odds ratio of 2.02 (95% CI, 1.13–3.63), Wald χ²(1) = 5.55, p < 0.05. Contrastingly, as in the community sample, an increase in the effects of the history of PA creation was associated with a decrease in the odds of having a positive PA staff–community relationship, with an odds ratio of 0.65 (95% CI, 0.72–0.84), Wald χ²(1) = 6.25, p < 0.05. Although problems caused by communities to PAs had an odds ratio of 0.64 (95% CI, 0.73–1.78), whether the problems increased did not significantly affect PA staff–community relationship, Wald χ²(1) = 0.18, p > 0.05. Similarly, although benefit-sharing had an odds ratios of 1.89 (95% CI, 0.52–1.53), whether benefit-sharing was improved did not significantly affect PA staff–community relationship, Wald χ²(1) = 0.35, p > 0.05.

Table 3. Ordinal logistic regression results explaining the influence of six factors on PA staff–community relationships from the PA staff data sample.

		95% Confidence intervals for odds ratios
Model	Coefficient estimate (Std. error)	Lower	Odds	Upper
PA–community relationships are influenced by:
History of PA creation	−0.44 (0.17)***	0.72	0.65	0.84
Communication	0.59 (0.22)**	1.17	1.81	2.79
Perceptions of communities	0.60 (0.23)***	1.15	1.82	2.88
Problems caused by communities to PAs	−0.13 (0.23)^a	0.73	0.64	1.78
Benefit-sharing	0.17 (0.28)^a	0.52	1.89	1.53
Community involvement	0.70 (0.30)***	1.13	2.02	3.63
R^2 = 0.50
Model x^2(6) = 55.32*

*p < 0.001, **p < 0.01, ***p < 0.05, ^aNon-significant (ns, p > 0.05).

4. Discussion

Our results showed differences in perceptions of PA staff and adjacent communities concerning their relationships. There were noticeable difference in their levels of trust for each other, their perceptions on the degree of power that they have to influence one another, their satisfaction levels with each other, and their levels of commitment to each other. These differences can be attributed to different values and understanding between PA staff and communities, especially on conservation issues and their importance. The local residents are often ignorant of many environmental issues (Fischhoff 1985). Locke (1975) suggests that because of different levels of understanding among stakeholders, good arguments could sometimes lead to human misunderstandings.

History of PA creation was significant in influencing PA staff–community relationship from both the communities and PA staff perspectives. Consistent with earlier studies, the impacts of forced removal during the establishment of PAs, for example, prohibition of access to resources in the PAs such as bush meat, grazing areas, and firewood led to problems between PAs and the communities often leading to increased illegal resource harvesting, habitat encroachment, and destruction (Graham et al. 2005; Fischer et al. 2011; Gandiwa et al. 2011). Since history cannot be changed, it would help if the communities benefited in a way that would not make them feel alienated. This could be done through compensating them either in monetary terms or through some land rights. Alternatively, the benefit-sharing schemes could be improved, for example, improving the CAMPFIRE programme. Currently, the CAMPFIRE programme is striving mostly on migratory animals which mean if there are no healthy wild animal populations in PAs, there will be less revenue accruing to communities since hunting will be less viable in CAMPFIRE areas. However, this can be improved by promoting the establishment of community game ranches to ensure active management and presence of resident animals in CAMPFIRE areas. Moreover, the CAMPFIRE programme mostly focusses on hunting, which also limits the amount of revenue the communities get. Product diversification and value addition, for example, ecotourism and curio shops offer an opportunity to enhance community benefits. Additionally, governments need to engage stakeholders who include local communities, and make joint decisions about how PAs should be gazetted and managed. This will help governments take proper account of local community needs when setting up PAs so as to ensure positive PA staff–community relationships and produce long-lasting results for both conservation and local communities.

While the community perceived benefit-sharing to have a significant influence on their relationship with PA staff, benefit-sharing was unable to explain PA staff–community relationships from the PA staff perspectives, most likely due to the fact that communities adjacent to PAs assume they should have some rights to wildlife resources and therefore should benefit from them. Moreover, PA staff is mostly concerned with conservation, and some of them are not local residents, hence, the issue of community benefits might not be of interest to them. Our findings from community sample concur with other authors who reported that benefit-sharing has significant influence on PA staff–community relationships (Adams & Hulme 2001; Hutton et al. 2005; Kideghesho et al. 2007; Tessema et al. 2010). Benefits to the communities can further be improved through developing transparent systems for benefit sharing, good governance systems, and improved community involvement. Most of the study communities in the present study are currently not directly benefiting from tourism in the PAs. This situation could be improved by allocating a certain percentage of revenue from tourism to the communities and/or allocating lease sites for photographic tourism within the PAs for community enterprises under public–private community partnership arrangements, hence, resulting in enhanced collective benefits to the community and infrastructural improvements within communities. Capacity building of local communities is another action that could be taken to enhance skills of local people with future potential benefits such as improved employability in higher paying jobs, empowering local people to start small tourism ventures, and also enhanced skills to effectively manage natural resources within communities.

Communication between PA staff and communities, and community involvement in conservation were significant in influencing PA staff–community relationship from both the communities and PA staff perspectives. Currently, only Gonarezhou National Park has a community liaison officer. Increasing the number of community liaison officers can enhance communication and ultimately, the relationship between PA staff and communities. Where ineffective communication exists, trust between communities and PA staff is low. If communication between PA staff and communities is not improved, the relationship between the two can also be difficult to mend. Our results on communication between PA staff and communities corroborate those of Ormsby and Kaplin (2005) who reported that difficulty of communication between Masoala National Park authorities and adjacent communities in Madagascar could have led to conflicts and negative relationships. Communication between PA staff and communities could be improved through engagement of community liaison officers by the PAs who act as mediums between PA staff and communities. Moreover, enhanced community involvement in conservation could improve PA staff–community relationships (Liu et al. 2010; Ebua et al. 2011).

Community perceptions on tourism, conservation, and PA staff were all significant in influencing PA staff–community relationship. The negative perceptions signify communities’ low levels of trust and satisfaction levels with PA staff, which indicates negative relationships between PA staff and local communities. Mutanga et al. (2015b) recorded negative perceptions of PAs by the communities attributed to limited financial benefits from tourism in Umfurudzi Park and Gonarezhou National Park. Allendorf (2010) suggested that community’s perceptions are a major component of the PA staff–community relationship. PA management, therefore, need to address the negative perceptions in order to improve the relationship between PA staff and communities through extending more benefits to the communities, for example, employing local people and enhancing access to natural resources such as thatching grass.

Similarly, our results showed that PA staff perceptions of communities had significant influence on their relationship with the communities. Where illegal activities, such as illegal hunting, livestock grazing in PA, uncontrolled fires, and encroachment into PAs are concerned, and where communities are always at loggerheads with PA staff, PA staff often have a negative perception of local communities, hence the negative PA staff–community relationships (Milgroom & Spierenburg 2008; Holmes 2013). As a result, PA staff tend to use force over local communities (Laudati 2010) thereby fuelling negative relationships between PA staff and the local communities. Thus, enhancing local community participation in conservation, and increased interaction of PA staff and local community through conservation initiatives would help improve PA staff–community relationships.

Neither problems caused by PA existence to communities which include crop raiding and livestock depredation by wild animals, safety to human lives, and confrontations with PA staff, nor problems caused by communities such as poaching, habitat destruction, and encroachment had an influence on PA staff–community relationships. Cawston Ranch and some boundary section of the northern Gonarezhou National Park are fenced hence human–wildlife conflicts are minimal in these areas and/or sections. Similarly, as reported by Gandiwa et al. (2012), in some PAs in Zimbabwe, fences had to be erected between wildlife areas and villages as a way of minimising human–wildlife conflicts. Our findings are contrary to earlier studies that have recorded cases of human–wildlife conflict (Muboko et al. 2014; Matema & Andersson 2015) and other conservation-induced costs that leads to communities’ low satisfaction with PA staff (Shibia 2010; Snyman 2012). However, it may be that respondents, especially local community members, were not open in divulging sensitive information on illegal activities affecting PAs, hence, the less representation of the influence of these factors on PA staff–community relationships.

Our findings suggest that all the factors, but one (problems caused by PA existence to communities which include crop raiding and livestock depredation by wild animals, as well as problems caused by communities such as poaching and habitat destruction), are important for conservation relationships. However, improving factors that have the potential of enhancing PA staff–community relationships which are communication, community perceptions of tourism, community perceptions of conservation, community perceptions of PA staff, benefit-sharing, community involvement in conservation, and PA staff perceptions of communities requires both financial and non-monetary resources such as time and skills. Determination from the involved parties is also essential for long-term natural resources conservation. The same goes for decreasing the effects of the history of PA creation, which, if allowed to increase can worsen PA staff–community relationships. By aiding the appreciation and understanding of some of the underlying factors that can contribute to either negative or positive relationships, these findings can have an important bearing on PA budgets and the general allocation of resources. Issues to consider in resource allocation planning include investing in the establishment of effective mechanisms for the transparent exchange of information and ironing out of grievances between PAs and local communities, for example, through employing community liaison officers. Also, of importance is capacity building mainly focussed on provision of support to communities especially with training in entrepreneurship and livelihood activities like poultry projects to reduce dependency on wildlife resources. Investing in ongoing training for PA staff so that they understand how best to deal or interact with local communities is also important. These can help provide both economic and non-monetary benefits to communities and most importantly can be instrumental in forming positive perceptions of tourism, conservation, and PA staff by the communities. Training for PA staff can help improve PA staff perceptions of communities.

Our findings are also important in the broad PA management as they help in addressing important issues on the complexity of interactions between nature and society. This underlines the importance of striking a balance between respecting community needs, expectations, and decision-making on one hand, and the commonly used methods of imposing terms and processes on communities to attain conservation goals, on the other hand (Krause et al. 2013). The discussed factors thus can help in informing parameters within which the roles of PAs and communities in conservation relationships can be defined. Both conservation agencies and communities need to pay more attention to the highlighted factors to nurture positive relationships. Conservation agencies can do this through compensating for community displacement during PA creation, providing opportunities for community involvement in tourism, and improving communication channels between PA staff and communities. Communities, on the other hand, can make efforts for community members to actively get involved in conservation so as to reduce illegal activities that negatively impact PAs such as veld fires and illegal hunting to improve PA staff perceptions of them. They can also find alternative sources of income, for example, through diversifying the CAMPFIRE programme and revenue options by offering tour guiding services and selling curios to tourists so as to reduce direct dependability on PA resources. This is necessary to gain and maintain both parties’ support for positive PA staff–community relationships and ensure long-term sustainability of wildlife conservation.

5. Conclusion

We conclude that communities generally perceived the relationship they had with the PAs to be negative while PA staff generally perceived a positive relationship with the communities. From the community perspectives, it is evident that seven of the eight tested factors had an influence on PA staff–community relationships. Problems caused by PA existence to adjacent communities had no significant influence on PA staff–community relationships. From the PA staff perspective, four of the six tested factors had significant influence on PA staff–community relationships while benefit-sharing and problems caused by communities to PAs had no significant influence on PA staff–community relationships. Communication, perceptions (both communities’ and PA staff’s), community involvement in conservation, history of PA creation, and problems caused by both PA existence to the communities or by communities to PAs had the same effect on PA staff–community relationship from both samples. Only benefit-sharing had a different effect on PA staff–community relationship from the two perspectives.

We recognise that besides the factors that were addressed in this present study, there could be other factors influencing PA staff–community relationship. Thus, there is need for multidisciplinary research on other factors that potentially influence PA staff–community relationships, since gaining an in-depth understanding of factors influencing these relationships is increasingly becoming important in promoting harmonious PA staff–community relationship and biodiversity conservation in general.

Acknowledgements

We thank Chinhoyi University of Technology for financial support, Zimbabwe Parks and Wildlife Management Authority, Cawston Ranch management, all the PA staff involved, the Ministry of Local Government, Urban and Rural Development, the Districts’ Authorities, and the respective Chiefs for support and permission to carry out this study in their respective areas of jurisdictions, and Felicity Mutanga for helping with data collection and capturing. We are grateful to two anonymous reviewers for the valuable comments and suggestions.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Chinhoyi University of Technology.

Appendix 1. Demographic profiles of respondents. Values are numbers of respondents, and percentages in parenthesis

Demographic profiles of community respondents, n = 938
Gender		Household size
Male	541(57.7)	<5	404(43.1)
Female	397(42.3)	6–10	381(40.6)
Age(years)		>10	153(16.3)
18–25	177(18.9)	Total number of livestock
26–35	237(25.3)	<5	485(51.7)
36–45	201(21.4)	6–10	238(25.4)
46–55	107(11.4)	11–15	84(9.0)
56–65	131(14.0)	16–20	50(5.3)
66–75	70(7.5)	21–25	26(2.8)
76+	15(1.6)	26–30	20(2.1)
Highest level of education		>30	33(3.5)
No formal education	159(170)	Level of income per year
Primary education	442(47.1)	<US$1000	824(87.8)
Secondary education	308(32.8)	US$1000–$2000	68(7.2)
Adult education	13(1.4)	US$2001–$3000	18(1.9)
College diploma	11(1.2)	US$3001–$4000	7(0.7)
University graduate	5(0.5)	US$4001–$5000	6(0.6)
Number of years stayed in the village		US$5001–$6000	5(0.5)
<2	39(4.2)	>$6000	10(1.1)
3–10	134(14.3)	Neighbouring protected area
11–20	330(35.2)	Umfurudzi	74(7.9)
21–30	153(16.3)	Gonarezhou	278(29.6)
31–40	137(14.6)	Matusadona	281(30.0)
41–50	52(5.5)	Cawston Ranch	305(32.5)
>50	92(9.8)
Demographic profiles of PA staff respondents, n = 133
Gender		Job position
Male	113(85.0)	Park Manager	1(0.8)
Female	20(15.0)	Assistant Park Manager	2(1.5)
Age(years)		Senior Game Ranger	7(5.3)
18–25	22(16.5)	Game Ranger	110(82.7)
26–35	64(48.1)	Community Liaison Officer	1(0.8)
36–45	22(16.5)	General Hand	7(5.3)
46–55	14(10.5)	Skinner	2(1.5)
56–65	8(6.0)	Driver	1(0.8)
66–75	3(2.3)	Chef	1(0.8)
76+	0	Mechanic	1(0.8)
Highest level of education		Protected area
No formal education	8(6.0)	Umfurudzi	22(16.5)
Primary education	29(21.8)	Gonarezhou	37(27.8)
Secondary education	74(55.6)	Matusadona	28(34.6)
Adult education	0	Cawston Ranch	46(21.1)
College diploma	14(10.5)
University graduate	8(6.0)
Number of years working in the Protected area
<5	49(36.8)
6–10	52(39.1)
>10	32(24.1)

Appendix 2. Descriptive statistics for dependent and independent variables for the community data sample

Item	Code	Mean	Standard deviation	Skewness	Kurtosis
Dependent variable
Protected area (PA) staff–community relationship
Trust
I believe that the Park Management/staff have good intention	TR 01	2.68	2.37	0.99	−0.75
I have faith/trust in the promises or statements of Park staff	TR 02	2.43	2.12	1.26	0.05
I believe the Park has a lot of experience and knows best with regards to wildlife conservation	TR 03	4.41	3.51	9.38	195.13
Fair principles seem to guide the Park’s behaviour	TR 04	3.62	2.55	0.24	−1.66
Category mean	TR	3.29	1.88	1.80	16.02
Commitment
I am proud to tell others that I support wildlife conservation	COMT 01	4.44	2.74	−0.32	−1.77
I really care about the fate of this Park	COMT 02	3.97	2.69	0.01	−1.83
I am willing to put in a great deal of effort in order to help wildlife conservation be successful	COMT 03	4.27	2.69	−0.21	−1.79
I feel that the Park is trying to maintain a long-term commitment to our community	COMT 04	3.23	2.53	0.98	−0.74
Category mean	COMT	3.98	2.11	−0.10	−1.35
Control mutuality
The Park really listens to what our community has to say	C/MUT 01	2.66	2.31	0.98	−0.74
The Park believes the opinions of our community are important	C/MUT 02	2.63	2.28	1.02	−0.63
The Park gives our community enough say in its decision-making process	C/MUT 03	2.28	2.05	1.42	0.49
Generally speaking, both the Park and ourselves are satisfied with the decision-making process	C/MUT 04	2.46	2.18	1.20	−0.17
Category mean	C/MUT	2.51	1.74	0.95	−0.24
Satisfaction
I feel our community is important to the Park	SAT 01	3.40	2.59	0.42	−1.61
Both the Park and ourselves benefit from our relationship	SAT 02	2.50	2.20	1.16	−0.26
Our relationship with the Park is good	SAT 03	2.93	2.45	0.76	−1.18
In general, I am satisfied with our relationship with the Park	SAT 04	2.78	2.38	0.89	−0.92
Category mean	SAT	2.90	1.85	0.71	−0.65
Overall relationship	REL	3.17	1.57	0.57	−0.47
Independent variables
History of PA creation
Forced relocation	HIS 01	5.26	2.49	−0.93	0.94
Fences and fines approach	HIS 02	5.36	2.42	−1.00	−0.79
Restrictions on the use of natural resources	HIS 03	5.58	2.25	−1.24	−0.16
Wildlife attacks on crops, livestock, and humans	HIS 04	5.93	2.10	−1.70	1.14
Decline in traditional practices as a result of restricted access	HIS 05	5.55	2.31	−1.20	−0.30
Category mean	HIS	5.54	1.64	−1.00	0.11
Communication
Park owners/management communicate well with us	COMN 01	3.05	2.48	0.67	−1.30
Game ranchers communicate well with us	COMN 02	2.87	2.39	0.80	−1.06
There is a regular communication network	COMN 03	2.91	2.40	0.77	−1.10
The Park normally holds meetings with us	COMN 04	2.22	2.08	1.48	0.59
We are consulted in decision-making	COMN 05	2.12	1.96	1.66	1.32
Category mean	COMN	2.63	1.62	0.81	−0.27
Community perceptions of tourism
I would be happy to see more tourists here	T/PERC 01	4.25	2.76	−0.18	−1.85
I would be happy if my children worked in the tourism industry	T/PERC 02	5.31	2.44	−0.98	−0.83
Tourism benefits the whole community	T/PERC 03	3.30	2.62	0.49	−.1.59
My family has more money because of tourism	T/PERC 04	2.07	1.94	1.68	1.36
Because visitors want to experience our culture, tourism strengthens our cultural tradition	T/PERC 05	3.07	2.53	0.65	−1.35
Tourists respect our culture and traditions	T/PERC 06	3.04	2.45	0.68	−1.25
Tourism offers financial opportunities for me that have adequately offset my losses from conservation	T/PERC 07	2.13	2.02	0.65	1.16
Category mean	T/PERC	3.31	1.50	0.27	−0.77
Community perceptions of conservation
It is important to protect plants and trees in the Park	C/PERC 01	5.36	2.49	−1.05	−0.78
It is important to protect wild animal species in the Park	C/PERC 02	5.49	2.38	−1.18	−0.43
People who poach should be punished	C/PERC 03	4.91	2.61	−0.66	−1.41
It is good this land is protected	C/PERC 04	4.35	2.70	−0.27	−1.77
I think the Park was created for the betterment of the community	C/PERC 05	3.83	2.71	−0.11	−1.83
I am happy that my village boarders the Park	C/PERC 06	3.67	2.70	0.22	−1.80
Category mean	C/PERC	4.60	1.70	−0.38	−0.82
Community perceptions of PA staff
The Park staff are generally helpful and understand our problems, needs and expectations	PS/PERC 01	2.27	2.13	1.43	0.39
The Park staff are friendly to us	PS/PERC 02	2.35	2.15	1.35	0.28
The Park staff are open to our suggestions and concerns regarding development and conservation programmes	PS/PERC 03	2.31	2.10	1.38	0.34
The Park staff encourage us to participate in conservation programmes	PS/PERC 04	3.01	2.47	0.69	−1.27
The Park staff respect our input or appreciate our efforts	PS/PERC 05	2.34	2.08	1.33	0.26
Category mean	PS/PERC	2.46	1.65	1.09	0.30
Problems caused by PA existence to the community
Damage caused by wild animals	PROB 01	5.30	2.49	−0.96	−0.94
Confrontations with conservation authorities	PROB 02	4.86	2.52	−0.58	−1.43
Threats to human safety	PROB 03	5.18	2.42	−0.84	−1.06
Restrictions on livestock grazing areas	PROB 04	4.99	2.50	−0.70	−1.28
Restrictions on firewood collection	PROB 05	4.66	2.59	−0.44	−1.61
Restrictions on animal hunting	PROB 06	4.69	2.62	−0.47	−1.59
Restrictions on fishing	PROB 07	4.80	2.57	−0.55	−1.50
Category mean	PROB	4.93	1.72	−0.47	−0.83
PA benefits to the community
Provides building material	BEN 01	2.12	2.08	1.65	1.05
Conservation awareness	BEN 02	3.26	2.56	0.51	−1.52
Provides job opportunities	BEN 03	2.44	2.21	1.22	−0.17
Skills development programmes	BEN 04	2.09	1.95	1.69	1.37
Hunt animals or plants for food and medicine	BEN 05	1.77	1.67	2.29	3.96
Provides game meat from culled animals	BEN 06	2.11	2.03	1.66	1.14
I earn money from tourism	BEN 07	1.91	1.83	2.00	2.54
Category mean	BEN	2.25	1.28	1.27	1.57
Community involvement in conservation
The Park values our input a lot	C/INV 01	2.29	2.16	1.39	0.27
Our local/traditional knowledge is important to the Park	C/INV 02	2.76	2.37	0.92	−0.87
Park management trusts us to make meaningful contributions to the management processes	C/INV 03	2.26	2.05	1.44	0.54
We participate and receive benefits from tourism	C/INV 04	2.07	1.96	1.75	1.52
Increased number of goods and services obtained from local businesses	C/INV 05	2.70	2.41	0.99	−0.80
Number of displays of traditional culture or customs within the PA by local communities	C/INV 06	1.84	1.75	2.15	3.31
Category mean	C/INV	2.32	1.38	1.11	0.71

The dependent variable PA–community relationship was coded REL and was assessed using trust (TR), commitment (COMMT), control mutuality (C/MUT), and satisfaction (SAT). The independent variables were coded HIS for history of creation of PAs, COMN for communication, T/PERC for community perceptions of tourism, C/PERC for community perceptions of conservation, PS/PERC for community perceptions of PA staff, PROB for problems caused by PA existence, BEN for benefit-sharing, and C/INV for community involvement in PA management.

Appendix 3. Descriptive statistics of dependent and independent variables for the PA staff data sample

Item	Code	Mean	Standard deviation	Skewness	Kurtosis
Dependent variable
Protected Area (PA) staff–community relationship
Trust
I believe that the community has good intention	TR 01	4.21	2.16	−0.14	−1.32
I feel a strong faith in the integrity of the community	TR 02	4.29	1.76	−0.19	−.074
I believe the community has invaluable knowledge with regards to wildlife conservation	TR 03	4.61	2.08	−0.37	−1.26
Ethical principles seem to guide the community’s behaviour	TR 04	4.58	1.93	−0.43	−0.78
Category mean	TR	4.44	1.38	0.17	−0.77
Commitment
The PA values its relationship with the community	COMT 01	4.91	1.94	−0.62	−0.72
I really care about the fate of this community	COMT 02	4.92	1.87	−0.84	−0.36
There is much to be gained by supporting this community	COMT 03	4.71	1.81	−0.31	−0.95
I feel that the community is trying to maintain a long-term commitment to our PA	COMT 04	3.92	1.94	0.02	−1.02
Category mean	COMT	4.61	1.24	−1.03	0.97
Control mutuality
The community really listens to what the PA has to say	C/MUT 01	4.31	1.90	−0.16	−0.85
The community believes the opinions of the PA are important	C/MUT 02	4.15	1.91	−0.16	−0.97
The community gives the PA enough say in its decision-making process	C/MUT 03	4.66	2.0	−0.42	−0.93
Generally speaking, both the community and the PA are satisfied with the decision-making process	C/MUT 04	4.27	1.86	−0.2	−1.0
Category mean	C/MUT	4.35	1.48	−0.22	−0.31
Satisfaction
I feel the PA is important to the community	SAT 01	5.02	2.20	−0.79	−0.86
Both the community and the PA benefit from our relationship	SAT 02	4.86	1.9	−0.75	−0.41
Our relationship with the community is good	SAT 03	4.27	1.98	−0.25	−1.06
In general, I am satisfied with our relationship with the community	SAT 04	4.4	1.92	−0.26	−1.05
Category mean	SAT	4.6	1.23	−0.33	−0.34
Overall relationship	REL	4.5	1.0	−0.27	−0.01
Independent variables
History of PA creation
Forced relocation	HIS 01	4.30	2.19	−0.25	−1.42
Fences and fines approach	HIS 02	4.60	1.86	−0.52	−0.70
Restrictions on the use of natural resources	HIS 03	4.83	1.83	−0.46	−0.90
Wildlife attacks on crops, livestock, and humans	HIS 04	5.01	1.74	−0.62	−0.39
Decline in traditional practices as a result of restricted access	HIS 05	4.46	2.0	−0.25	−1.32
Category mean	HIS	4.64	1.30	−0.2	−0.91
Communication
The PA communicates well with the community	COMN 01	5.44	1.76	−1.29	−0.97
The community communicates well with us	COMN 02	5.31	1.60	−0.52	−0.88
There is a regular communication network	COMN 03	4.20	1.92	−0.15	−0.98
The Park normally holds meetings with the community	COMN 04	4.57	1.98	−0.41	−1.05
The community is consulted in decision-making	COMN 05	3.81	2.32	0.1	−1.56
Category mean	COMN	4.66	1.28	−0.27	−0.27
PA staff perceptions of the community
The community is generally keen to learn about wildlife conservation	COM/PERC 01	4.67	2.01	−0.64	−0.85
Whenever we hold training programmes for the community, they are usually supportive and they come in their numbers	COM/PERC 02	4.10	1.79	−0.14	−0.84
The community understands the importance of wildlife conservation	COM/PERC 03	4.32	1.65	−0.24	−0.72
The community wants to participate in any tourism activities	COM/PERC 04	4.08	1.9	−0.18	−1.09
The community is innovative and always come up with exciting programmes for the tourists	COM/PERC 05	3.14	1.78	0.62	−0.54
Category mean	COM/PERC	4.06	1.15	−0.24	−0.68
Conservation problems caused by the community
Loss of wild animal species due to poaching	PROB 01	5.44	1.78	−0.84	−0.6
Reduction in wild animal habitat due to encroachment	PROB 02	4.96	1.85	−0.95	−0.07
Community collaboration with external commercial poachers	PROB 03	4.88	1.9	−0.57	−0.7
Over-reliance on natural resources	PROB 04	4.36	1.72	−0.41	−0.6
Resistance to taking up any income generating tourism activities	PROB 05	3.75	1.75	0.28	−0.49
Resistance to attending any training programmes in wildlife conservation	PROB 06	3.77	1.87	0.18	−0.85
Reluctant to upgrade their educational qualifications so that they can be employed in higher level positions within the park	PROB 07	4.78	1.91	0.4	−0.9
Category mean	PROB	4.56	1.01	−0.59	0.57
PA benefits to the community
Building material	BEN 01	3.42	2.10	0.23	−1.34
Conservation awareness	BEN 02	4.66	1.83	−0.5	−0.59
Job opportunities	BEN 03	4.77	1.93	−0.43	−0.98
Skills development programmes	BEN 04	3.98	1.92	0.06	−1.12
Free hunting of animals or plants for food and medicine	BEN 05	3.24	2.21	0.48	−1.24
Game meat from culled animals	BEN 06	3.62	2.15	0.15	−1.41
Earning money from tourism	BEN 07	3.29	2.04	0.37	−1.18
Category mean	BEN	3.85	1.17	−0.05	−0.47
Community involvement in conservation
The PA values community input a lot	C/INV 01	4.41	2.08	−0.33	−1.14
The community’s local/traditional knowledge is important to the Park	C/INV 02	4.72	1.98	−0.52	−0.83
PA management trusts the community to make meaningful contributions to the management processes	C/INV 03	4.23	1.91	−0.03	−0.98
The community participates and receive benefits from tourism	C/INV 04	3.55	2.07	0.26	−1.26
Increased number of goods and services obtained from local businesses	C/INV 05	3.88	1.94	0.05	−1.01
Number of displays of traditional culture or customs within the PA by local communities	C/INV 06	3.11	1.91	0.52	−0.74
Category mean	C/INV	3.98	1.20	−0.35	0.07

The dependent variable PA–community relationship was coded REL and was assessed using trust (TR), commitment (COMMT), control mutuality (C/MUT), and satisfaction (SAT). The independent variables were coded HIS for history of creation of PAs, COMN for communication, COM/PERC for PA staff perceptions of the community, PROB for problems caused by the community, BEN for benefit-sharing, and C/INV for community involvement in PA management.