Kilimanjaro SES comprises the southern slopes of the biophysical unit of Mount Kilimanjaro, located in Tanzania, and its associated actors, including smallholder farmers, business owners, and nature conservationists. It covers approximately 4,000 km² from the lowlands at 700 meters above sea level (masl) to the summit at 5,895 masl (Fig. 1). The mountain's elevational gradient supports a range of ecosystems, making it a biodiversity hotspot (A. Hemp & Hemp, 2018). This includes a vast supply of regulating, material, and non-material Nature’s Contributions to People (NCP), which meet the needs and interests of various actors. Kilimanjaro SES supports the livelihoods of over one million people (National Bureau of Statistics, 2023), with the Chagga people being the dominant ethnic group (Mtallo & Rubagumya, 2015).

Kilimanjaro’s conservation history dates back to the early 20th century, during colonial rule, when the foothills were officially designated as a game reserve and a forest reserve (Durrant, 2004 and references therein; Kivumbi & Newmark, 1991). In 1941, the social buffer forest, also known as the half-mile forest strip (HMFS) of approximately 88 km², was established to enable local communities to manage and benefit from the forest reserve (Kivumbi & Newmark, 1991). When authority over the HFMS changed repeatedly, local communities’ access to forest-supplied NCP was restricted. In 1973, Kilimanjaro National Park was established above the forest reserve (2,700 m asl), with the dual objectives of nature conservation and tourism. The park management follows a fortress conservation model, with strict protection and non-consumptive resource use (Durrant & Durrant, 2008; Holyroyd, 2016; Newmark, 1991). In 2005, overexploitation of forest-supplied NCP led to the incorporation of the remaining forest reserve and HFMS into the National Park, restricting local access (Frömming, 2009; Holyroyd, 2016; Sébastien, 2010). The ban was partially lifted between 2014 and 2020, permitting only women to collect firewood and feed NCP, but it was fully reinstated in 2021 (personal communication with Kilimanjaro National Park staff).

Important economic activities in the Kilimanjaro SES include agriculture (Soini, 2005b). In the lowlands (700–1,000 masl), small-scale maize and beans farms owned by Chagga people dominate (Misana et al., 2003; Soini, 2005a), while the highlands (1,200–1,800 masl) feature both commercially grown coffee plantations leased to investors and a traditional Chagga agroforestry system (Kihamba in Chagga) in small-sized homegarden (Fernandes et al., 1985; A. Hemp, 2006). The homegardens reflect the bio-cultural significance of smallholding in Kilimanjaro SES (C. Hemp et al., 2018; Sébastien, 2010) where coffee, banana, and a diverse mix of trees, shrubs, and crops, such as yams, are combined in a multi-layered, highly biodiverse structure (A. Hemp, 2006; Mdoe & Wiggins, 1997). Since the 1980s, homegardens have gradually shrunk in size and shifted in crop composition: Maize has gained prominence due to declining ownership of lowland farms and changing dietary preferences (Ichinose et al., 2020), while coffee cultivation has decreased due to falling market prices (Soini, 2005a, 2005b), and overall vegetation cover has declined due to land fragmentation and increased settlement (Misana et al., 2012; Said et al., 2021).

Besides farming, nature-based tourism in Kilimanjaro SES is crucial for the local and national economy. The mountain is among Tanzania’s top destinations, with climbing as the main attraction (Kilungu et al., 2019; Tanzania National Parks, 2023). It generates substantial revenue, reported up to US$51 million in park fees in 2013 (Christie et al., 2014; Musa, G et al., 2015). Locally, tour guides and operators get their salaries and tips (Christie et al., 2014; Peaty, 2012) by providing guided hikes inside KINAPA and other nature-based activities, such as visiting waterfalls or engaging in bio-agricultural experiences with the Chagga outside the park (Kilungu et al., 2019). Moreover, conservation efforts extend beyond park boundaries, with government, local and international community-based and non-governmental conservation organizations planting trees and protecting forests (Gross et al., 2025; C. Hemp et al., 2018).

Despite these conservation efforts, anthropogenic drivers threaten both the ecological integrity and people’s quality of life (Peters et al., 2019; Said et al., 2019). These drivers include land cover changes such as the expansion of settlements and agricultural land, as well as shrinking per capita land due to population growth and inheritance-driven land division (Grossmann, 2008; Peters et al., 2019; Said et al., 2021; Soini, 2005b). Together with climatic changes such as decreased precipitation and rising temperatures, these drivers fundamentally contribute to hydrological shifts in the region (A. Hemp, 2009), exacerbating the risk of forest fires (A. Hemp, 2009), and reducing water availability essential for agriculture and human well-being in the region (Agrawala et al., 2003; Said et al., 2019).

We obtained research permits from the Tanzania Commission for Science and Technology (COSTECH < COSTECH number removed for review process>) and the Tanzania Wildlife Research Institute (TAWIRI). The research also followed the ethical guidelines of the Norwegian National Committee for Research Ethics in the Social Sciences and the Humanities and the standards recommended by the German Data Forum (2017). All research participants provided written consent, acknowledging prior information about the study’s objectives, consent to audio recordings, and permission to appear in photographs.

We thematically analyzed the qualitative data collected from the workshops and interviews using MAXQDA 2022 (VERBI, 2018). First audio recordings from small group discussions, plenary sessions during the workshops, and interviews were transcribed and translated from Swahili into English. Guided by our research questions, we then inductively coded the data through an iterative process of code generation and theme identification (Braun & Clarke, 2006). For example, from the scenario-developing exercise in workshop 2, we coded desirable components perceived by the participants in three themes: environmental protection, land use and agriculture management, and sociocultural preservation (Table S6). Similarly, we reviewed all proposed management actions, grouped similar concepts, and selected a category name that best represented the specific actions (Table S5). For example, we named ‘sustainable agriculture and diversification’ as the category for all actions related to agricultural diversification, sustainable farming practices, off-farm livelihood alternatives, and youth involvement in agriculture. Participants’ ranking points for the effectiveness and achievability of the management actions were combined and averaged across actors and presented as the mean effectiveness score in Table 1. We also extracted exemplary quotes supporting participants' perspectives, which we anonymised using actor group codes and randomly assigned two-digit numbers, e.g., SF15 = smallholder farmers, BS17 = business owners, NC08 = nature conservationists (Table S5).

For qualitative ranking of perceived changes on key SES elements, NCP supply and quality of life dimensions for each scenario conducted in workshop 2, we summarized the narratives in Table S7 and numbered the rankings: strongly increase (+ 2), increase (+ 1), no change (0), decrease (-1), and strongly decrease (-2) to generate a heatmap in Fig. 5.

The demographic information of the research participants showed that 65% of them were aged between 25 and 44 years, representing youth and middle-aged participants, while the rest were older individuals, up to 75 years old. Majority had vocational or undergraduate education, and fewer held postgraduate degrees. Ninety-five per cent of the participants were men, reflecting the gender composition of the positions they represented in institutions or companies. Among invited farmers, men attended as they are often considered the heads of households. Additionally, 65% of participants were native to the region, and 80% had lived there for over 15 years (Table S3).

Participants identified five key elements of Kilimanjaro SES: water resources, biodiversity, land use and agriculture, tourism, and traditions and culture (Table S4). They reported significant changes in all elements over the past 40 years (Table S5). For example, participants recalled a past where wildlife was more present and natural systems more intact: “As a child, I remember seeing porcupines in our garden. We also used to hunt pigeons, but now you rarely see such wildlife” (BS17). Participants also remembered higher vegetation cover and abundant water resources: “When I grew up in a shanty town in the late 1980s, we even had a little channel (Mfongo in Chagga) running through our backyard, but now there is no water to flow in these channels” (NC08). Further, they mentioned low-volume tourism and that the Chagga traditions and culture fostered a close relationship between people and nature: “You had an attachment, like ‘oh my Kihamba’, you look after it, you have a sense of ownership and custodianship, especially on the mountain, these are resources supplied by God” (BS17).

Participants perceived deforestation and biodiversity loss as resulting from population growth, which led to increased demand for land for agricultural expansion —a primary form of land use change. Water resources such as springs and rivers were perceived to have either disappeared or become polluted. They also noted that climatic change contributed to the drying of springs and altered river flow patterns, resulting in seasonal flow variations. Additionally, participants noted that the erosion of traditions and culture weakened societal connections to nature (Table S4). On the other hand, participants noted an increase in tourism, possibly due to improved facilities and infrastructure, as well as a diversification of tourism products. Similarly, while water resources were noted to decline, improvements in water supply infrastructure increased household access to water.

The outcomes of these changes varied among different actors. For example, business owners in agriculture and tourism, as well as tourists, benefited from developments such as improved water supply and tourism infrastructure, as well as diversified tourism products. In contrast, other actors, such as smallholder farmers, resorted to unsustainable farming practices, e.g., farming near riverbanks due to decreased per capita land availability. These changes also altered NCP provision. Moreover, they led to an increased loss of cultural identity and a decline in connectedness with nature, particularly among the younger generation, due to westernization and lifestyle changes (Table S4). Figure 3 illustrates the current conditions of Kilimanjaro SES as perceived by participants in Workshop 1.

The following sections provide brief descriptions of each scenario, focusing on key SES elements, NCP supply, quality of life dimensions, and the actors perceived as winners and losers. Table S7 provides a summary of these characteristics for each scenario.

The Crowded Kilimanjaro scenario was characterized by a high population, ineffective legal and policy frameworks, and shrinking per capita land, leading to resource overexploitation and the potential displacement of people (Fig. 4, top left). In this scenario, food and other goods were imported, while protected areas faced increasing pressure from human encroachment. TEK related to agriculture was retained primarily in fragmented homegardens. This scenario indicated a decline of key SES elements, most notably regulating NCP, and a deterioration in the quality of life dimensions for most actors (Fig. 5). While short-term economic benefits supported industries and businesses through activities like logging, most actors became losers in the long term (Table S7).

The Urbanized Kilimanjaro scenario envisioned a future of significant urban expansion alongside rapid development in infrastructure and technology (Fig. 4, top right illustration). The scenario was characterized by improved social infrastructure and the adoption of advanced agricultural technologies. A technology-driven approach to conservation was implemented, including CCTV surveillance and fencing around Kilimanjaro National Park, except in designated wildlife corridors. Tourism flourished with the introduction of cable cars, bringing the mountain experience to more visitors. The scenario showed a declining trend in vegetation cover in areas outside the Park, along with other key SES elements, such as water resources, biodiversity, traditions, and culture. Additionally, material, regulating, and non-material NCP also declined (Fig. 5). By contrast, the amount of tourism and markets seemed increased. Yet, some of the quality of life dimensions for actors associated with markets and tourism seemed to deteriorate since the employment opportunities for those in the mountain business, and porters decreased due to a cable car (Table S7).

The Chagga Traditional Life scenario envisioned a positive future rooted in environmental stewardship (Fig. 4, bottom left illustration). The scenario is characterized by a low population and a strong reliance on TEK, which fostered a culture of environmental stewardship. Sustainable land-use planning guided the development of agriculture through homegardens. Ecotourism and cultural tourism associated with the Chagga community thrived, although the social and communication infrastructure remained underdeveloped. Traditional customs and practices were upheld, reinforcing moral values within the close-knit community. The scenario demonstrated the preservation of key SES elements and a high supply of NCP, although access to roads and the educational dimensions of quality of life were perceived as decreasing (Fig. 5 and Table S7).

The Kilimanjaro Techno Life scenario envisioned a future of advanced technological development and new infrastructure (Fig. 4, bottom right illustration). The scenario was characterized by a low population density, with prioritized agricultural production and industries focused on processing produce, such as coffee roasting and packaging. Conservation focused only inside the National Park, while technological development of the water system, such as water harvesting and innovative irrigation systems, helped mitigate the negative impacts of climatic change. Overall, the scenario had a positive outlook for most SES elements, including material and regulating NCP and the different dimensions of quality of life (Fig. 5), yet participants anticipated a decline in traditions and culture (Fig. 5, Table S7).

In workshop 3, participants identified three aspects that are desirable for the future of Kilimanjaro SES: (1) environmental protection, (2) land use and agriculture management, and (3) cultural preservation. For environmental protection, participants identified the conservation of biodiversity and ecosystems as a desirable element of the future, which required other elements, such as the reforestation of degraded lands and the efficient management of water resources. Concerning land use and agricultural management, participants reported a need for sustainable agricultural practices, effective land use planning, and socioeconomic development. Finally, regarding preserving Chagga culture, participants reported on desirable elements such as social harmony and conserving traditional practices. Moreover, participants identified undesirable elements that need mitigation actions, including limited financial resources, inadequate social infrastructure, displacement, and unemployment. Table S6 in the Supplementary Material provides an overview of all desirable and undesirable elements, along with exemplary quotes from the discussion.

Participants identified seven broad categories of management actions to achieve the desirable futures. These ranged from legal and policy-related actions at the national level to environmental stewardship requiring local collective action. They included political support and conservation enforcement, sustainable agriculture and diversification, environmental education and awareness, community-led environmental stewardship, reforestation programs, traditions and culture preservation, and land use planning (Table 1). No important differences emerged across actor groups or between actions in terms of perceived effectiveness and feasibility of the management actions. However, all participants ranked political support and conservation enforcement highest, followed by sustainable agriculture and diversification. While not exhaustive, participants also identified actors responsible for implementing the management actions from the national level to village leaders and local communities. For example, actions such as reforestation and community-led stewardship were perceived to require collective actions at a local level, while others, like land use planning, required national-level financial investment and institutional support. In addition, participants identified a lack of financial resources and limited awareness as the most common barriers to implementing the management actions, as detailed in Table 1.

Participants perceived that for actions related to political support and conservation enforcement, there is a lack of coordinated, participatory, and community engagement in their implementation despite having supportive laws, policies, regulations, and guidelines for environmental protection and conservation. Participants raised concerns that local communities are excluded from decision-making and are at the receiving end.

Regarding sustainable agriculture and diversification, participants mentioned initiatives like the Kweme (Oyster nuts) project, which aims to avoid deforestation while earning income, as oyster nuts require trees for support. In line with this, participants emphasized diversification through beekeeping and off-farm livelihood alternatives to reduce pressure on land. Participants also placed a special focus on engaging youth in agriculture by increasing their involvement and fostering pride in agricultural heritage.

Participants also noted that continuous mass media campaigns were needed to raise awareness about environmental issues and conservation. For example, they suggested that social media campaigns with celebrity engagement to advocate for conservation could be effective in gaining the attention of young people. In addition, participants identified potentially productive actions, including hands-on environmental education. “We could engage our children in conservation through activities such as planting trees at home and making sure to report progress to the teacher. Children have faith in their teachers and can abide by their words rather than their parents” (NC05).

Furthermore, participants highlighted the importance of preserving traditions and culture through community-led environmental stewardship. Participants proposed the idea of a living village exhibit, where most villagers promote sustainable agriculture and conservation. This includes maintaining local varieties of bananas in homegardens, many of which have medicinal significance, and planting native tree species. “Having a museum displaying Chagga traditional way of life and artifacts would not be enough to counteract the rate at which we are losing our culture... It will be important to make conscious decisions as a community to showcase our culture and preserve our traditions in our everyday life” (SF15).

Participants acknowledged ongoing reforestation programs by local communities, public and private institutions, and conservation and development NGOs in Kilimanjaro SES. They mentioned tree planting programs near water sources and homesteads, and the establishment of tree plantations to satisfy material NCP demand and reverse the environmental degradation (Table S5). “We need a balanced ecosystem where human needs from nature can be met while we conserve, but accessing the half-mile strip forest in Kilimanjaro National Park is banned, people should focus on planting trees to meet their demand for fodder, poles and timber in their lands” (NC09).

Participants also noted that land use planning was crucial, as many villages lacked proper plans, leading to environmental degradation. Participants acknowledged government efforts to conserve water sources through laws prohibiting agricultural activities within 60 meters of a riverbank or water source. Table S5 in Supplementary Material lists management actions under each category and exemplary quotes from the participants.

“The conservation of a substantial proportion of the world’s biodiversity and NCP largely depends on the customary institutions and management systems of Indigenous Peoples and Local Communities (IPLCs)” (IPBES et al., 2019, p. 914). This highlights the importance of local institutions and knowledge systems, also reflected in our PSP in Kilimanjaro SES, where transformation towards desirable futures emerged as a social process rooted in the lived experiences of the actors. PSP in Kilimanjaro SES unraveled who holds “the socioculturally mediated capacity to act” (Ahearn, 2001, p.112) or agency, where it is expressed, and how it shapes the desirable futures. This was evident in the four scenarios developed in workshop 2, which all had three commonalities. First, the presence of the national park boundary shaped perceptions of actors regarding where their agency was limited and who held that agency. The PSP process revealed that participants largely imagined futures within the bounds of the existing institutional arrangement of the Kilimanjaro National Park (Fig. 4). Across all four scenarios, the Park’s continued existence was assumed, and its boundary was treated as fixed. No alternative governance models or uses for the National Park were envisioned. This situation may reflect, on the one hand, a general acceptance of the Park's continued existence, given its role in providing key NCP, such as climate regulation and water (Durrant & Durrant, 2008), and the fact that our PSP began at present, when the Park and its borders are well established. On the other hand, it may also indicate a perceived lack of opportunity to imagine alternative futures with other conservation models, such as community-based management or a Biosphere Reserve. As a result, the agency was perceived as being limited to areas outside the Park. This finding aligns with literature on the fortress conservation model (Goldman, 2003; West et al., 2006), which shows that protected areas often constrain local agency, as actors are not meaningfully included in decision-making (Andrade & Rhodes, 2012). This can lead to local communities focusing on spaces where they can exert greater autonomy (Morrison et al., 2019), such as outside the National Park. Nevertheless, current conservation policies in Tanzania direct park authorities to develop specific programs, such as support for community-led conservation initiatives and livelihood options, or conservation education, to ensure the engagement of local communities in conservation outside the protected area boundaries (Durrant & Durrant, 2008). Moreover, these policies include a benefits-sharing program through which local communities engage in Community Conservation Services, also known locally as “Ujirani mwema” in Swahili (Durrant & Durrant, 2008).

Second, the presence of homegardens in all scenarios suggests a strong sense of agency in areas below the park boundary. This reflects how embedded practices and local knowledge play a central role in shaping desirable futures (Fig. 4). In the Chagga Traditional Life scenario, the dominant presence of the homegardens was accompanied by the preservation of TEK, and non-material NCP such as cultural heritage and identity or connectedness with nature (Fig. 5), while the Crowded Kilimanjaro only included some remnants of these. These homegardens, deeply rooted in Chagga tradition, are the bright spots for sustainable and self-sufficient futures (Bennett et al., 2016; A. Hemp, 2006). Their multi-layered structure of trees and bushes integrated with crops such as bananas, coffee, and yams, among others, represents a long-standing form of land use that is ecologically resilient and culturally meaningful (Fernandes & Nair, 1986; A. Hemp, 1999, 2006) and they are highly valued by smallholder farmers (Sanya et al., 2025). As Kothari (2008, p. 23) concluded, “…the future of conservation lies, at least partly, in its past,” the management of homegardens for desirable futures in Kilimanjaro SES similarly reflects this conclusion. The Chagga homegardens are living archives of culture, TEK, and intergenerational care. These agroforestry systems, passed down through generations, reflect a deep-rooted commitment to tradition, where maintaining a homegarden is not only a livelihood activity but also a marker of identity. Within these spaces, families gather, rituals are held, and stories are shared, reinforcing ties between people and nature. The homegardens preserve a wealth of TEK, including medicinal plants used in animal husbandry and human health, local banana varieties with healing properties, and management techniques that have long supported sustainable land use (A. Hemp, 1999, 2006; Mollel et al., 2017). The revival of crops like Kweme (Oyster nuts), which require trees for support, is helping reconnect people with forgotten uses while halting deforestation and biodiversity loss. The homegardens provide new meanings to old practices, and knowledge is not only conserved but also informs desirable futures.

Third, across all scenarios, technology was perceived as a tool for exercising agency by either enhancing traditional knowledge or enabling more sustainable practices, thereby demonstrating its crucial role in shaping desirable futures. Across scenarios, participants recognized the need for technological innovations, such as water harvesting systems, greenhouses, improved road access, and processing facilities for their farming products (Fig. 4). Interestingly some of the technological developments such as water harvesting systems were envisioned for the management of homegardens. This points to the importance of braiding TEK with modern technological knowledge for fostering sustainable futures (Bennett et al., 2025; Ray, 2023; Tengö et al., 2017).

“Any governance system designed to regulate complex biological systems must have as much variety in the actions that it can take as there exists in the systems that are being regulated” (Ostrom, 1998). This quote speaks directly to the challenge of governing complex SES, like Kilimanjaro where desirable futures reflect a high degree of complexity shaped by multiple interdependent elements such as the impact of diverse drivers of change on land use, biodiversity, water resources, traditions and culture, NCP supply or actors’ quality of life (Table S4; Fig. 5). To achieve the consented desirable futures, multiple actors operating at different governance levels are required (Table 1) since no single actor possesses all the knowledge, tools, and capacity to govern SES complexity effectively. This result aligns with the need for polycentric governance, where multiple, overlapping decision-making centers operate with some degree of autonomy but are linked through cooperation, coordination, and mutual accountability (Ostrom, 2006). Importantly, polycentricity extends governance beyond formal institutions to encompass informal interactions and relationships crucial to achieving desirable futures, including information sharing, coordination, problem-solving, and even internal conflict resolution (Biggs et al., 2012, 2015).

Similarly, the PSP and backcasting applied in this study showed visions for desirable futures centered around environmental protection, land use and agriculture management, and cultural preservation. These results align with a recent study that showed the need to integrate multiple components across both ecological and social dimensions when designing desirable futures (Juri et al., 2025). Moreover, the management actions discussed in workshop 3 reinforced the complexity of Kilimanjaro SES, as actions ranged from the need for political support and conservation enforcement to sustainable agriculture and diversification, as well as reforestation programs (Table 1). The management actions were mapped onto different levels of governance, ranging from local to national levels (Table 1). For example, regarding environmental protection, management actions included reforestation and community-led stewardship, which were perceived as requiring the engagement of local communities and TEK at a local level. In contrast, land use planning necessitated financial backing and a higher-level policy commitment. This example illustrates the importance of involving various governance levels that also illustrates the need for formal (laws, policies) and informal institutions (customs, collective action, TEK) to represent the actors’ diverse interests and needs (Chaffin et al., 2014; Martín-López & Montes, 2015; Morrison et al., 2019; Ostrom, 2006). These results support the claims for a governance system that embraces institutional diversity and polycentricity to foster desirable futures, which also aligns with the IPBES Transformative Change Assessment’s emphasis on inclusive, accountable, and adaptive governance as essential for addressing the root causes of biodiversity loss and enabling context-specific solutions across scales (O’Brien et al., 2025).