The role of education in biodiversity conservation: Can knowledge and understanding alter locals’ views and attitudes towards ecosystem services?

Abstract

Climate change, human population growth and land use change are among the most important threats to nature and ecosystem services. Local appreciation of ecosystem services and knowledge of how the foundation of these services is affected by local livelihoods are important for the sustainability of natural resources and thus may fundamentally affect human well-being. In this study, we tested how student knowledge, attitudes and views towards ecosystem services and biodiversity changed due to an education programme. We assessed the before and after knowledge, attitudes, and views of 180 level-two secondary school students in the Ngorongoro District, Tanzania. Findings suggested an increased proportion of students could identify ecosystem services after the education programme, with increased knowledge of ecosystem services and their importance being attributed to developing understandings of negative outcomes caused by climate change, human population growth and land use change. The increased awareness of ecosystem services gained via the education programme suggests that education programmes can be an important tool in conservation biology. Therefore, we recommend that education programming dealing with threats to ecosystem services and biodiversity should be included in school curricula, especially for students close to protected areas.

Keywords:

• Attitudes
• biodiversity conservation
• education
• ecosystem services
• secondary schools

Introduction

The most critical challenges facing the world today are climate change, pollution, human population growth, overexploitation of natural resources and habitat loss (IPBES 2019). Together, these challenges are threatening the existence and sustainable use of natural resources for present and future generations and thus jeopardizing the persistence of ecosystem services (IPBES 2019).

Ecosystem services comprise contributions from natural resources to human well-being, and are typically categorized into four categories: (1) Provisioning (such as food, livestock, and water). (2) Regulating (such as carbon storage, climate regulation, pollination, protection from flood and erosion). (3) Cultural (such as recreation, education, religious identity, and heritage). (4) Supporting (including nutrient recycling and allowing other services to be possible) (Arowolo et al. 2018; Crossman et al. 2013; Gashaw et al. 2018; Kindu et al. 2016; Shiferaw et al. 2019; Silvestri et al. 2013).

As the human population increases, more resources are needed from ecosystems to sustain human activities. More extreme and unpredictable weather conditions in Africa, for example, may be attributed to global climate change related to human activities, which consequently affect ecosystems and their services (IPCC 2014). Equally, in various parts of Africa, some people are more dependent on ecosystem services from their near surroundings than elsewhere, and these services have shown to be more sensitive to climate variability. In Serengeti National Park (SNP), for example, climate change is more pronounced due to its dependence on seasonal patterns of rainfall (Ritchie 2008) than surrounding areas. The wet season has become drier, while the dry season is wetter (Ogutu et al. 2010). Due to seasonal rainfall, the water quality and quantity vary (Gereta 2010a), which impacts the whole ecosystem.

Cognition about and overall assessment of nature will vary with individual differences in attitudes (Ajzen 2001). Attitude has been defined as a way of thinking or feeling about something acquired through experience while, knowledge can be described as ‘facts, information and skills acquired through experience or education’ (McAleese 2000).

Education in biodiversity conservation is an important factor in shaping peoples’ knowledge of and attitudes towards biodiversity conservation, and hence should be a part of the curriculum in schools (Kideghesho, Røskaft, and Kaltenborn 2007), at both primary and secondary levels. Generally, performing education in biodiversity conservation programmes for children can be important in acquiring personal knowledge as well as changing their attitudes and behaviors (USAID 2000; WWF 2010) towards biodiversity and ecosystem services. Research has shown that in similar regions and settings, younger generations benefit from learning about sustainable use of natural resources in the future, such as by focusing on biodiversity conservation and responsibilities for preserving nature (Mutisya, Kipgetich, and Rono 2015). Moreover Mutisya, Kipgetich, and Rono (2015) explain how Kenyan schools that establish curricula that foster positive values and attitudes towards wildlife contribute to conserving biological heritage in the future. Additionally, where nature-based education programmes have been implemented in other parts of Africa, children have used this environmental knowledge positively within their communities (USAID 2000; WWF 2010).

We see the most important aim of nature-based education programmes as positively changing children’s behavior towards the environment (Pooley & O'Connor, 2000), by improving understanding and attitudes about protecting local environments (Kellert and Berry 1987). Our assumption is that improved understanding of the importance of ecosystem services leads to an increased awareness of sustainable use of such services (Dalum 2013; Kideghesho, Røskaft, and Kaltenborn 2007; Shingote 2011; St John, Edwards-Jones, and Jones 2010). Education in biodiversity conservation programmes can improve knowledge and skills (Bogner 2002; Christensen, Rowe, and Needham 2007; Iozzi 1989; Jaus 1984; Mangas, Martinez, and Pedauyé 1997; Smith-Sebasto and Cavern 2006), and lead to critical thinking, although knowledge per se does not necessarily lead to changes to more environmentally conscious behavior (Jensen 2002; Urban and Martin 2015). However, studies have indicated that attitudes may change with knowledge and behavior (Karris et al. 2020; Martinis et al. 2018; Røskaft et al. 2007; Thompson and Mintzes 2002).

In contrast to many children in Western schools, some African children do not learn about wildlife ecology and the types and importance of natural resources found locally (Erhabor and Don 2016; EU. 2022), despite the fact that their community or family livelihoods may depend more heavily on the local environment (Kioko, Kiringe, and Wahungu 2010). A lack of formal education regarding wildlife ecology and natural resources found in their surroundings has been identified as an issue in this regard (Erhabor and Don 2016). However, it is critical that such knowledge is also accessible to local people (Hariohay et al. 2018), including in schools and other contexts, such as tourism (Gereta 2010b). Educational programmes with an emphasis on biodiversity and ecosystem services have proven to be effective in improving children understanding regarding their surroundings (Aguirre-Bielschowsky, Freeman, and Vass 2012). In Tanzania, the context for this study, 17% (1,920,000) of pupils are enrolled in secondary schools (EPDC_NEP 2018). The aim of this study was 1) to study the current level of knowledge about biodiversity and ecosystem services among local secondary school students surrounding Serengeti National Park, and 2) to offer a short introduction course about biodiversity and ecosystem services to positively affect knowledge, views, and attitudes towards the sustainable use of natural resources.

Research design

Study area

The study was carried out in Ngorongoro Districts within the Loliondo game-controlled area, northeast of Serengeti National Park located in northern Tanzania, comprising an area of 14 763 km² (Lyamuya et al. 2016). The education programme was conducted in six villages named as follows: Wasso (Loliondo Secondary School), Ololosokwan (Emanyata Secondary School), Soitsambu (Soitsambu Secondary School), Samunge (Samunge Secondary School), Mholo (Digodigo Secondary School), and Monic (Lake Natron Secondary School) (hereafter referred to as ‘study schools’; Figure 1). The study schools were public, except from one private, Emanyata Secondary School. The study took place for six weeks during the period from August to October 2016. The study sites were chosen because of their proximity to large, protected areas, while the surrounding area can be categorized as multiple land use allowing coexistence between wildlife and livestock (e.g. Ngorongoro Conservation Area Authority and Loliondo Game Controlled Area). The Serengeti ecosystem is approximately 30 000 km^2, and includes game reserves, a conservation area and Serengeti National Park, where the altitude above sea level ranges from 910 to 1820 m (Zimmermann 2012).

Figure 1. Serengeti ecosystem showing the six study schools indicated with black dots: Loliondo, Emanyata, Soitsambu, Samunge, Digodigo and Lake Natron, located in northern Tanzania (main map). A map of Tanzania (inserted to the lower left) shows the location of the Serengeti ecosystem in northern Tanzania.

The educational programme facilitated by AfricanBioServicess at the Tanzania Wildlife Research Institute is described in detail in an associated Master’s thesis, which forms the basis for this article (Børresen 2016). In brief, approval for the environmental education project was agreed beforehand, and involved using questionnaires, lectures, and posters to help students understand the value of natural resources, and develop knowledge and understanding to inform decision making processes about responsible environmental behaviors. Copies of the lesson plans, slide decks, visuals (i.e. posters) and research instruments are included in Appendices to the thesis, while a short summary is presented on a project blog - https://africanbioservices.eu/my-teaching-adventure-in-tanzania/.

Sampling and data collection

Data to evaluate the impact of the educational programme were collected using a questionnaire survey given to 30 students at second level from each of the six boarding schools (aged 14–20 years).

To examine changes in knowledge, views, and attitudes before and after the education programme, the terms ‘pre-test’ and ‘post-test’ were used. During the first day, 30 students completed a questionnaire that aimed to assess their existing knowledge, views, and attitudes prior to the study programme. On the second and third day, students were taught about ecosystem services and biodiversity. On the fourth day of the education programme, the same 30 students completed the same questionnaire as they were presented the first day to examine whether we could detect a significant increase in knowledge on students’ views and attitudes. The education programme was performed in the same manner in all the study schools, using the identical questionnaires and the same lectures. However, the lectures differed slightly at each study school, adapted to their local conditions. The study programme was presented by a researcher in English and then translated into Kiswahili to minimize the effects of varying language skills. The interviewed students often used the Kiswahili language, as it is their first language.

The complete questionnaire consisted of 11 questions, 12 statement questions, and 7 statements about threats to ecosystem services. It was divided into five main categories (a-e):

1. General information about the respondent and interview (date of interview, school (Loliondo, Emanyata, Soitsambu, Samunge, Digodigo or Lake Natron), gender (male or female) and tribe (Maasai, Sonjo, Chaga, Iraq, or other tribes).

2. Knowledge about ecosystem services and biodiversity.

3. Knowledge about the consequences of threats to ecosystem services (pictures illustrating climate change, land use change and human population growth were used).

4. Statements where the students had to tick ‘yes’ if they agreed to the statement and ‘no’ if they disagreed.

5. Threats to ecosystem services. The respondents ticked ‘yes’ if they believed the alternative was a threat and ‘no’ if they disagreed.

Chi-square tests were used to determine disparity in how the respondents answered the questions that involved categorical data. Binary logistic regression analyses were used to test the relative importance of different independent variables explaining knowledge, attitudes, and views. All statistical analyses were conducted using the Statistical Package for Social Science (SPSS) version 23.

Results

Demographic information

In total, 355 questionnaires were completed during the education programme, 180 students answered the pre-test questionnaire, and among those 175 of the students answered the post-test questionnaire. The gender distribution was relatively even (51.7% males versus 48.3% females). The average age was 15.9 years (SD = 1.3, n = 346), and the most represented tribes were: (1) Maasai (24.0%), (2) Sonjo (23.4%), (3) Chaga (9.9%), and (4) Iraq (9.3%). The 27 other tribes were merged into one category referred to as ‘Other tribes’ (33.3%).

Knowledge about ecosystem services and biodiversity

Significantly more students answered correctly to the statement ‘ecosystem services are human-made services from nature that will always exist’ post-test than pre-test (Table 1). A similar pattern was found for both alternatives: ‘ecosystem services are benefits that nature provides to humans for free’ and ‘ecosystem services are human-initiated services’ (Table 1).

Table 1 . Percentage of the students who answered the question ‘What are ecosystem services?’ correct pre-test versus posttest¹. We report parameter estimates from chi-square tests.

What are ecosystem services?	Pre-test % correct	Post-test % correct	% change	χ2	P
Human made services from nature that will always exist	35.0	80.6	45.6	75.4	<0.001
Benefits that nature provides to humans for free	61.7	90.3	28.6	39.6	<0.001
Human-initiated services	75.6	92.0	16.4	17.6	<0.001
N	180	175

In binary logistic regression models, we used different definitions of ecosystem services as dependent variables (see Table 1) whereas, test (pre- versus post-test), tribe (five tribes as described in the methods), gender (male versus female) and school identity (categorical with six levels), were all used as independent variables. There were consistently more correct answers in the post-test versus pre-test (P < 0.001). However, in one question (‘benefits that nature provides to human for free’), there were also significant differences between gender of respondents (P = 0.003). Males had significantly more knowledge about this pre-test (73.4%) than females (49.4%; χ²= 10.9, df = 1, P = 0.001), while females showed the same level of knowledge as males post-test (90.9% males versus 89.4% females; χ²= 0.11, df = 1, P = 0.741).

The highest increases from pre- to post-test in the statements regarding ecosystem services (Table 2) were found for the following alternatives: ‘food’, ‘timber’ and ‘land for settlement’ (Table 2). Furthermore, a significant increase in knowledge for all the other ecosystem services, ‘water’, ‘firewood’, ‘medicines’ and ‘grazing’, was also found during the post-test questions (Table 2). A significant increase in knowledge also occurred for the non-ecosystem services ‘air balloon’, ‘tourism’ and ‘pollution’ (Table 2), evaluated by pre-test questions, while no significant increase in knowledge was found for others.

Table 2. Percentage of the students who identified ecosystem services and non-ecosystem services correctly pre-test and post-test. We report the percentage change in correct answers from pre- to post-test and parameter estimates from chi-tests assessing whether this change was larger than expected.

Which service nature provides?	Pre-test % correct	Post-test % correct	% change	χ2	P
Food	70.6	90.1*	19.5	20.9	<0.001
Water	73.9	86.3	12.4	8.52	0.004
Timber	69.4	85.1**	15.7	12.2	<0.001
Firewood	69.4	80.6	11.2	5.85	0.016
Medicines	73.9	84.6	10.7	6.14	0.013
Grazing	58.9	72.0	13.1	6.74	0.009
Land for settlement	62.8	84.6	21.8	21.6	<0.001
Air balloon	82.2	92.6	10.4	8.59	0.003
Tourism	24.4	72.0	47.6	80.4	<0.001
Pollution	85.0	94.9	9.9	9.46	0.002
N	180	175

* sample size 171 and ^**sample size 174 due to missing values

Binary logistic regression analyses were conducted with the alternatives of highest significance (‘food’, ‘timber’, ‘land for settlement’ and ‘tourism’ as expressed in Table 2) as dependent variables. The independent variables were ‘pre-test/post-test’, ‘tribe (recoded)’, ‘gender’ and ‘secondary school’. ‘Pre-test/post-test’ was the only independent variable that was significant for all the alternatives: ‘food’, ‘land for settlement’, ‘tourism’ (all P < 0.001) and ‘timber’ (P = 0.001). Females had significantly more knowledge about the ‘tourism’ pre-test than males (18.1% males versus 31.8% females: χ²= 4.51, df = 1, P = 0.034). Both males and females had knowledge at a higher-level post-test (68.2% males versus 76.5% females, post-test values); however, at this stage, the difference between males and females was not significant (χ²= 1.48, df = 1, P = 0.224).

A significant increase in the number of students who identified biodiversity as important because ‘it provides everything we need to survive’ was shown post-test (from 75.0% before to 92.0% after; χ² = 18.5, df = 1, P < 0.001). Thereafter, binary logistic regression analyses were conducted using as dependent variable: ‘it provides everything we need to survive’. The independent variables included were: ‘pre-test/post-test’, ‘tribe (recoded)’, ‘gender’ and ‘secondary school’. ‘Pre-test/post-test’ was the only variable that was statistically significant (P < 0.001). Thus, none of the other variables affected the responses.

Knowledge of consequences from climate change

Significantly more students identified ‘changes in temperature’ because of climate change pre-test than post-test (from 76.7% before to 85.1% correct after; χ² = 4.12, df = 1, P = 0.042). There was no significant change in the number of students who identified ‘reduced human livelihood’ (from 48.3% before to 52.6% correct after; χ² = 0.64, df = 1, P = 0.425) and ‘unpredictable weather’ (from 72.8% before to 64.0% correct after; χ² = 3.17, df = 1, P = 0.075) as consequences of climate change from pre- to post-tests. A binary logistic regression analysis conducted with ‘changes in temperature’ as a dependent variable showed that among the independent variables, ‘pre-test/post-test’, ‘tribe (recoded)’, ‘gender’ and ‘secondary school’. ‘Pre-test/post-test’ was the only variable that was statistically significant (P = 0.038).

Knowledge of consequences from human population growth

During the post-test, a significant increase in knowledge was found in the number of students who identified ‘land use change’ and ‘more cattle’ as consequences of human population growth (Table 3) than in the pre-test. A high proportion of students identified ‘overexploitation of resources’ because of human population growth during the pre-test, but there was no significant change in the post-test (from 71.7% before to 68.0% correct after; χ² = 0.57, df = 1, P = 0.452).

Table 3 . Percentage of the students who identified consequences of human population growth correctly pre-test and post-test.

Consequences of human population growth	Pre-test % correct	Post-test % correct	% change	χ2	P
Land use change	58.3	69.3	11.0	4.99	0.026
More cattle	32.8	54.3	21.5	16.7	<0.001
N	180	175

Binary logistic regression analyses, where the dependent variables were the consequences in Table 3 and the independent variables were ‘pre-test/post-test’, ‘tribe (recoded)’, ‘gender’ and ‘secondary school’, were conducted. ‘Gender’ was a statistically significant contributor explaining the variation for the alternative ‘land use change’ (P = 0.030). Males had more knowledge than females in both pre-test (63.8% males versus 51.8% females; χ²= 2.67, df = 1, P = 0.102) and post-test (73.9% males versus 65.9% females; χ²= 1.31, df = 1, P = 0.252); however, neither was statistically significant. Thus, in total (both pre- and post-test), this difference between the two sexes was close to significant (χ²= 3.70, df = 1, P = 0.054).

Knowledge of threats to biodiversity and ecosystem services

Significantly fewer students answered ‘yes’ to ‘climate change, drought’ as a threat to biodiversity and ecosystem services post-test compared with pre-test (Table 4). In contrast, there was a significant increase in the number of students who answered ‘yes’ to ‘human population growth, infrastructure development’, which was a threat to biodiversity and ecosystem services post-test (Table 4). No difference between the two tests was found regarding ‘tourism’ as a threat to biodiversity and ecosystem services (from 19.4% before to 21.7% correct after; χ² = 0.28, df = 1, P = 0.597). A significant increase in the number of students who answered ‘no’ to ‘wildlife’ and ‘carnivores’ was a threat to biodiversity and ecosystem services between the two tests (Table 4).

Table 4. Percentage of respondents who answered correctly to the following alternatives from the question: ‘What are threats to biodiversity and ecosystem services?’ correctly pre-test and post-test.

Answer	Pre-test % correct	Post-test % correct	% change	χ2	P
Climate change, drought	87.2	77.7	−9.5	5.57	0.018
Human population growth, infrastructure development	53.9	82.9	29	34.3	<0.001
Wildlife	75.0	87.4	12.4	8.95	0.003
Carnivores	68.9	92.0	23.1	29.9	<0.001
N	180	175

Binary logistic regression analyses, where the alternatives in Table 4 were the dependent variables and ‘pre-test/post-test’, ‘tribe (recoded)’, ‘gender’ and ‘secondary school’ were the independent variables, were conducted. The independent variable ‘pre-test/post-test’ was statistically significant in all cases (‘climate change, drought’ (P = 0.027), ‘human population growth, infrastructure development’ (P < 0.001), ‘wildlife’ (P = 0.003) and ‘carnivores’ (P < 0.001)). However, in one case (‘human population growth, infrastructure development’), ‘secondary school’ was also a statistically significant contributor in explaining the variation (P = 0.021). There was a significant difference in knowledge post-test compared with pre-test. Thus, Soitsambu (96.7%) and Digodigo (96.7%) had the best level of knowledge, unlike Loliondo (69.0%), Emanyata (75.9%), Samunge (83.8%), and Lake Natron (74.1%) (χ²= 14.5, df = 5, P = 0.013).

A high percentage of students knew that ‘new infrastructure and pollution’ (from 79.9% before to 75.4% correct after; χ² = 1.02, df = 1, P = 0.314) and ‘harvesting of wildlife’ (from 71.7% before to 70.7% correct after; χ² = 0.04, df = 1, P = 0.839) were threats to ecosystem services pre-test, and there was no significant change in post-test. There was a significant decrease in the number of students who identified ‘unpredictable weather’ as a threat between the two tests (Table 5). In contrast, a significant increase was found in the number of students who identified ‘overexploitation of resources due to human population growth’ and ‘human diseases’ between the two tests (Table 5). For the alternatives ‘schools and education’ and ‘protected areas’, the results showed a significant increase in the number of students who answered that, these were threats post-test compared with pre-test (Table 5).

Table 5. Percentage of respondents who answered correctly to the alternatives from the question ‘What are threats to ecosystem services?’ pre-test and post-test.

Question	Pre-test % correct	Post-test % correct	% change	χ2	P
Unpredictable weather	91.1	80.5	−10.6	8.28	0.004
Overexploitation of resources due to human population growth	83.7*	93.1*	9.4	7.65	0.006
Human diseases	38.8*	51.7	12.9	5.97	0.015
Schools and education	92.8	79.4*	−13.4	13.3	<0.001
Protected areas	88.3**	69.0	−19.3	19.6	<0.001
N	180	174

* sample size 175, 178 and ^**sample size 179 due to missing values

Binary logistic regression analyses were conducted, where the dependent variables were the cases in Table 5 and the independent variables were ‘pre-test/post-test’, ‘tribe (recoded)’, ‘gender’ and ‘secondary school’. ‘Pre-test/post-test’ was the only variable that was statistically significant in all cases (‘unpredictable weather’ (P = 0.005), ‘overexploitation of resources due to human population growth’ (P = 0.010), ‘human diseases’ (P = 0.022), ‘schools and education’ (P = 0.001) and ‘protected areas’ (P < 0.001)). However, ‘secondary school’ was significant for both ‘human diseases’ (P = 0.013) and ‘schools and education’ (P = 0.007).

For ‘human diseases’, Lake Natron had a lower knowledge level pre-test than the others, and the difference between the study schools was statistically significant (Loliondo (56.7%), Emanyata (41.4%), Soitsambu (50.0%), Samunge (31.0%), Digodigo (36.7%) and Lake Natron (16.7%); χ²= 12.7, df = 5, P = 0.027). There was a higher level of knowledge for all the study schools, except for Loliondo. Posttest, the difference between the study schools was also statistically significant (Loliondo (34.5%), Emanyata (55%), Soitsambu (66.7%), Samunge (62.1%), Digodigo (60.0%) and Lake Natron Secondary School (29.6%); χ²= 13.6, df = 5, P = 0.018). For ‘schools and education’, all the study schools showed a high level of knowledge pre-test, and there was no significant difference between the secondary schools (Loliondo (90%), Emanyata (83.3%), Soitsambu (96.7%), Samunge (93.3%), Digodigo (93.3%) and Lake Natron (100.0%); χ²= 7.38, df = 5, P = 0.194). However, post-test, the difference between the study schools was statistically significant, and Loliondo and Samunge Secondary Schools showed lower knowledge levels than the other study schools (Loliondo (65.5%), Emanyata (75.9%), Soitsambu (90.0%), Samunge (66.7%), Digodigo (93.3%) and Lake Natron (85.2%); χ²= 12.8, df = 5, P = 0.025).

In one case (‘protected areas’), ‘tribe (recoded)’ was also a statistically significant contributor in explaining this variation (p = 0.001). The difference between the tribes was significant pretest; thus, the Maasai and Sonjo tribes showed lower knowledge levels than the others (Maasai (80.5%), Sonjo (79.1%), Chaga (94.1%), Iraq (100.0%) and other tribes (95.1%); χ²= 11.3, df = 4, P = 0.024). All the tribes showed lower knowledge levels post-test; thus, the difference among the tribes was not significant (Maasai (60.5%), Sonjo (62.5%), Chaga (77.8%), Iraq (58.8%) and other tribes (80.4%); χ²= 7.10, df = 4, P = 0.131).

Statements

Almost all the respondents answered ‘yes’ to ‘I depend on ecosystem services’ (from 97.8% before to 97.1% yes after; χ² = 0.15, df = 1, P = 0.697) and ‘I think it is important to protect ecosystem services’ (from 100.0% before to 99.4% yes after; χ² = 1.03, df = 1, P = 0.310) pre-test; thus, there was no significant change in knowledge from pre- to post-test. There were no significant changes in the students’ answers to the statements ‘I want ecosystem services to be present in the future’ (from 98.3% before to 98.3% yes after; χ² = 0.001, df = 1, P = 0.972) and ‘threats to the ecosystem affect me’ (from 76.7% before to 70.9% yes after; χ² = 1.50, df = 1, P = 0.221) post-test compared with pre-test; however, a high percentage answered ‘yes’ to both of these statements. However, there was a significant increase in the number of students who answered ‘yes’ to ‘ecosystem services are mostly here for human use’ post-test (Table 6). Few students agreed to the statement ‘ecosystem services are not important to me’, and there was no significant change post-test (from 2.2% before to 6.3% yes after; χ² = 3.62, df = 1, P = 0.057). Few students also agreed to the statement ‘it is not important for me to protect ecosystem services’, but there was a significant increase in the number of students who agreed to this post-test (Table 6). Significantly fewer students answered ‘yes’ to ‘I think it is important to share knowledge about ecosystem services’ post-test than pre-test (Table 6).

Table 6. Percentage of respondents who answered ‘yes’ to statements about nature and ecosystem services pre-test and post-test.

Statement	Pre-test % yes	Post-test % yes	% change	χ2	P
Ecosystem services are mostly here for human use	78.3	88.0	9.7	5.90	0.015
It is not important for me to protect ecosystem services	3.4*	8.6	5.2	4.32	0.038
I think it is important to share knowledge about ecosystem services	100.0	97.1	−2.9	5.22	0.022
N	180	175

Binary logistic regression analyses were conducted, where the statements in Table 6 were the dependent variables. The independent variables were ‘pre-test/post-test’, ‘tribe (recoded)’, ‘gender’ and ‘secondary school’. ‘Pre-test/post-test’ only gave statistical significance for the following statements: ‘ecosystem services are mostly here for human use’ (P = 0.016) and ‘it is not important for me to protect ecosystem services’ (P = 0.049). For the statement ‘it is not important for me to protect ecosystem services’, ‘tribe (recoded)’ was also a statistically significant contributor in explaining the variation (P = 0.006). The difference between the tribes was not significant pre-test (Maasai (9.5%), Sonjo (0.0%), Chaga (0.0%), Iraq (0.0%) and other tribes (3.3%); χ²= 7.50, df = 4, P = 0.112), but the difference between the tribes was statistically significant post-test. Post-test, the Maasai, Iraq and Chaga showed fewer positive attitudes compared with the others (Maasai (18.6%), Sonjo (5.0%), Chaga (11.1%), Iraq (17.6%) and Other tribes (0.0%); χ²= 13.5, df = 4, P = 0.009).

Discussion

Our findings suggest to which extent an education programme may contribute to gaining knowledge about ecosystem services and biodiversity.

Knowledge about ecosystem services and biodiversity

There was a significant increase in the number of students who acquired knowledge about ecosystem services during the education programme. This was revealed by comparing the post-test and pre-test as seen in results chapter. This finding corroborates with former studies showing that adding wildlife conservation programmes into primary and secondary school curricula can result in increased conservation knowledge (Bitanyi et al. 2012; Kollmuss and Agyeman 2002). There was also a significant increase in the ability to differentiate between ecosystem services and non-ecosystem services (Larson, Green, and Castleberry 2009; Lyamuya et al. 2016). However, during pre-test questions many students did not understand that tourism activities were an ecosystem service. This might be due to the common belief that tourism is something for ‘rich white people’ (Norton, 1996; Trapp-Fallon, 2003).

Knowledge of consequences from climate change

Many students gained knowledge that ‘changes in temperature’ is a consequence of climate change. Thus, our findings suggest that education led to an increase in knowledge (Andersen 2018; Larson, Green, and Castleberry 2009; Lyamuya et al. 2016). However, there were no significant changes in the students’ answers for ‘reduced human livelihood’, ‘unpredictable weather’ or the alternatives that were not consequences. This finding may indicate that the respondents are not directly affected by climate change in their daily life. Some Americans do not believe in climate change mostly being based on personal experience Cox, Navarro-Rivera, and Jones (2014). Scepticism was among others based on the fact that ‘weather has not changed/still cold’ and that there is natural variation in temperature (Cox, Navarro-Rivera, and Jones 2014). Additionally, a study by Ojala (2012) showed that to cope with climate change, young people use a variety of strategies, ranging from contributing to solving the problem with daily life activities to placing less emphasis on how serious climate change occurs. In addition, the results could be supported by Erdogan (2011) or that indigenous people have a broad knowledge base that has been transferred between generations (Gadgil, Berkes, and Folke 1993) and used to cope with emerging challenges.

Knowledge of consequences from human population growth

There was an increase in respondents answered that ‘land use change’ and ‘more cattle’ are consequences of human population growth (post-test). Here we have shown that education has led to an increase in knowledge, as reported by different authors (Kollmuss and Agyeman 2002; Larson, Green, and Castleberry 2009; Lyamuya et al. 2016). Moreover, the response for ‘overexploitation of resources’, because of human population growth, being identified as a threat during pre-test, was an indication of the value of pre-existing knowledge in the community. This has to be acknowledged for knowledge may have been transferred and retained through generations (Gadgil, Berkes, and Folke 1993). This was observed in this study, and help in making an emphasis on the importance of education programmes to be a continuous activity to nurture enough knowledge to young generations.

Knowledge of threats to biodiversity and ecosystem services

Many students identified ‘climate change, drought’ as a threat to biodiversity and ecosystem services during the pre-test. There were also significantly fewer respondents who answered ‘yes’ to whether ‘unpredictable weather’ was a threat to ecosystem services. The reduction in the respondents’ answers could be explained by the respondents using a coping strategy that put less emphasis on how serious climate change might be (Ojala 2012). Additionally, many students answered that ‘human population growth, infrastructure development’ and ‘overexploitation of resources due to human population growth’ are threats to ecosystem services (Andersen 2018; Larson, Green, and Castleberry 2009; Lyamuya et al. 2016). Furthermore, a significant number of respondents recognized that both ‘wildlife’ and ‘carnivores’ were not threats to biodiversity and ecosystem services during the post-test. Similarly, Røskaft et al. (2007) found that education might improve local attitudes towards carnivores. Bitanyi et al. (2012) stated that to ensure long-term wildlife survival, continual education programmes that improve locals’ knowledge are important (Kideghesho 2006).

Statements

Most of the respondents had positive attitudes and views towards ecosystem services, but it was hard to determine exactly whether knowledge affected the students’ attitudes and views during the pre-test.

There was no significant change during the post-test in the students’ answer to the statement ‘plant and animals are mostly here for human use’. Røskaft et al. (2007), who studied attitudes towards carnivores in Norway, concluded that fear towards large carnivores and negative attitudes are not equivalent. Fear can be reduced by knowledge, but this may not influence other conditions of negative attitudes (Røskaft et al. 2003; Røskaft et al. 2007). This finding can indicate that the students do not have a more positive view regarding plants and animals, even though they gained more knowledge about wildlife. Domestic animals can also be harmed by diseases and parasite transmission from wildlife, and because of this, local people can retain negative attitudes towards wildlife, even though communities can benefit from wild animals (Gereta and Røskaft 2010). This finding might help explain why the students did not have more positive attitudes towards plants and animals. More students think that ecosystem services are mostly here for human use (Bogner 2002). This result also indicates that the students would have benefited from a longer education programme and more lessons (Lindemann-Matthies 2002). However, Kideghesho, Røskaft, and Kaltenborn (2007) argued that education might not be positive for conservation initiatives, unlike Røskaft et al. (2003), who found that educational programmes are good management strategies.

The two statements ‘I think it is important to protect ecosystem services’ and ‘it is not important for me to protect ecosystem services’ gave contrary results. This finding might be based on the fact that it is harder to understand negative statements (‘it is not important for me to protect ecosystem services’) than positive ones. Patten (2016) found that it is easy to become confused when responding to negatively worded statements, and negatives are easily overlooked. Additionally, Johnson, Bristow, and Schneider (2011) findings suggest that ‘it is untrue that negatively worded questions do not adversely affect the pattern of responses to a survey question’. Nevertheless, the results indicate that most of the respondents believed that it is important for themselves to protect ecosystem services. Many of the students answered that they think it is important to protect ecosystem services. Few students agreed to the statement ‘ecosystem services are not important to me’, even though there was no significant change during the post-test. This result might indicate that the students think it is important to protect ecosystem services and that ecosystem services are important to themselves (Christensen, Rowe, and Needham 2007). In addition, Erdogan (2011) found that an ‘ecology-based nature and education programme contributed significantly to children’s responsible environmental behaviour’.

Factors explaining knowledge, views, and attitudes

The analyses showed that ‘pre-test/post-test’ was the variable that had the most impact on the results. Previous studies support that education programmes can contribute to more knowledge (Andersen 2018; Bitanyi et al. 2012; Hariohay et al. 2018; Kollmuss and Agyeman 2002; Lyamuya et al. 2016). In addition, Holmern, Nyahongo, and Røskaft (2007) concluded that in villages outside Serengeti National Park, wildlife education in biodiversity conservation programmes, along with the development of better primary and secondary education, can contribute to fewer misconceptions and more tolerance of depredation. In some cases, ‘gender’, ‘tribe’ and ‘secondary school’ were also significant contributors in explaining some of the variations. Males and females had in a few cases unequal knowledge of different nature conservations. Mmassy and Røskaft (2013) found that in recognizing bird species, gender was an important criterion; males had more knowledge about such species than women, which can explain why ‘gender’ was significant. Most of the findings in this study shows that gender did not have any effect on attitudes towards protected areas. Mmassy and Røskaft (2013) experiencing wild environments are an important factor for being able to recognize selected bird species. Among the tribes they tested, the Maasai tribe had more knowledge about bird species. A positive learning environment has a positive impact on performance development in secondary school (Wiborg et al. 2011). According to this, dissimilar teaching environments might have had different effects on students’ performance. Furthermore, both home and school environments can influence students’ attitudes towards science. This finding could be an explanation for why ‘tribe’ and ‘secondary school’ in some cases were statistically significant. However, the results for Loliondo secondary school might also be explained by the random sampling method, which is different from the other study schools.

Conclusion and implications for future research and teaching practice

Our results indicated that the present education programme led to an increase in knowledge about ecosystem services and biodiversity and their drivers of change among secondary school students. Furthermore, our results showed that the students improved their views and attitudes towards nature and natural resources. Such positive effects of educational programmes in schools around protected areas would imply substantial importance for the appreciation and conservation of natural resources in and around these vulnerable areas. Although the students’ knowledge improved with education and the students showed positive attitudes, there is no guarantee that the knowledge and attitudes will persist over time. Possibly, one can only hypothesize that this education programme could contribute to knowledge, attitudes and views that are more stable and that will persist. It would be important to test the effect of a similar education programme that last for a longer teaching period and/or have more lectures. Based on the experience from the current study, we recommend similar education programmes to other schools close to important biodiversity hotspots. Moreover, knowledge about wildlife conservation programmes should be included in curriculum for both primary and secondary schools to improve education (Lyamuya et al. 2016; Nyahongo 2010) for both present and future generation.

Acknowledgements

We thank the Commission of Science and Technology in Tanzania (COSTECH) and Tanzania Wildlife Research Institute (TAWIRI) for approving our proposal and granting permission to undertake this study. We acknowledge at least two anonymous reviewers as well as Dr. Franco Peniel Mbise for comments that highly improved this article. We extend our thanks to Dr. Emmanuel Masenga, Loliondo Education Officer, and headmasters in six secondary schools for assistance and organizing this study. As all participants in this study are anonymous, no ethical clearance from authorities in Tanzania or Norway was necessary. However, permissions to do the study in different schools were given by the school principal.

Disclosure statement

The authors declare no conflict of interest in publishing this article.

Additional information

Funding

This study received funding from the AfricanBioServices Project through the European Union’s Horizon 2020 research and innovation programme under Grant Agreement No. 641918 (AfricanBioServices).

Notes on contributors

Solveig T. Børresen

Solveig Trøen Børresen completed her Master’s education in Education Science with fieldwork related to this article. She is currently a high school teacher.

Rehema Ulimboka

Rehema Ulimboka is a PhD student in education biology at NTNU. She has been a lecturer at University of Dodoma and is interested in how local people deal with conservation issues.

Julius Nyahongo

Julius William Nyahongo is a Professor in Socio-ecology and Principal of College of Education at the University of Dodoma. His main interest is human-wildlife interactions, particularly the offtake of bushmeat.

Peter S. Ranke

Peter Sjolte Ranke is currently a chief engineer at NTNU. He has a leading role in a large project collecting data on metapopulation dynamics of house sparrows in northern Norway. In Tanzania he is involved in projects related to bird communities, wildlife population trends and human wildlife interactions.

Gine Roll Skjaervø

Gine Roll Skjærvø is a chief engineer at NTNU dealing with teaching and natural resources management. She did a PhD on human demography in Norway and is currently involved in projects dealing with human wildlife interactions in Tanzania.

Eivin Røskaft

Eivin Røskaft is a professor in evolutionary ecology at NTNU. His PhD was on behavioral ecology of crows, but he has worked in Africa for more than 30 years in projects related to human wildlife interactions. He has trained more than 40 African students to PhD or MSc levels.