Figure: Western Rwanda Landscape. Photo by Ping Sun.

The following is a repost from the blog, ‘Ecologically Speaking’. It has been lightly modified to fit our own formatting requirements. The original post was written by Jacqueline Poertner, William Apollinaire, and Vicky Temperton and can be found here. 

_________  

“If you want to go fast – go alone. If you want to go far – go together.” 

– origin unknown, often attributed to an African proverb 

There is something very exciting going on in the realms of ecosystem restoration and you (yes, you!) should know about it! Last year, the research team of A living lab for social-ecological restoration in western Rwanda, part of a larger project dedicated to social-ecological systems informing ecosystem restoration in rural Africa, launched its living lab in Rutsiro. 

There’s even an article on this blog about the launch! You’ll find it here along with more info on the background and the previous stages of the project (it might be helpful to read that article first to get the full picture). But now, let’s catch you up on recent happenings and give you a glimpse of what’s in store for the project’s future. So, lean back and enjoy the journey through the process of the living lab in Rutsiro! 

A quick recap anyway

If you’ve read the launch article, you’ll know that the living lab is all about bridging science and practice in Rwanda’s efforts for ecosystem restoration and reconnecting local communities with native tree species within an Agroforestry setting. For that, the team is following a process of co-creation to implement social-ecological experiments on farmers’ land in Rutsiro, that essentially involves planting trees as part of an Agroforestry approach to restoration. 

Who is behind the team? There are around 50 transdisciplinary stakeholders from across sectors working together in two different roundtables, each of which is based on two governance models, each representing a distinct restoration community of practice. One group is NGO-driven, uniting representative restoration actors from government, NGOs, academia and the private sector. The other one is almost entirely farmer-driven: local farmers, who are also traditional healers, beekeepers, teachers or carpenters, alongside community leaders and local extension officers. 

The communities have been involved in regular workshop sessions in which members first started by establishing codes of conduct, strategies on how to communicate, expectations and potential contributions. What’sespecially interesting for us now is what happened in the last workshops in February this year. 

Setting the scene: The grand selection of tree species

In February, each group had a workshop in which they co-designed the restoration experiments by ranking and prioritizing the native and exotic trees species suitable for the area for the restoration experiments. As a result, both governance models now have a list of prioritized tree species. These species are around 80 % native and 20 % exotic, which is quite radically different compared to the fact, that so far, most trees planted in western Rwanda are not native, which leads to loss of biodiversity and ecosystem services and depauperate landscapes (Nyiramvuyekure et al. 2026). 

Interestingly, the two groups did the same workshops in parallel, but came up with a slightly different list of tree species they want to plant. Each governance model is assigned to a specific region with living lab sites having been selected for doing the restoration experiments – including the tree planting. 

Let us walk you through the workshop day by day

William Apollinaire, Vicky Temperton and (for one day) Stefan Sieber attended the workshops, as scientists and principle investigators of the Rwanda research project. 

The first day of the workshop – held on separate dates for each governance model – took place in a conference room. This was where the magic of selecting and prioritizing the tree species happened based on the values stakeholders assigned to each species. The values ranged from ecological to socioeconomic criteria, survival rates, availability of planting material and compatibility with the most commonly grown crops in western Rwanda. At the end, the ranked list of tree species the groups want to plant saw the light of day. 

Enjoy the following impressions of how the first day looked like for both governance models: 

Government, NGOs, Academia, Research institutions and the Private Sector are discussing the value of each of the agroforestry trees.

The next day, the groups got into action out in the field. They went out to their living lab sites and met with the local landowners of the land where the trees are going to be planted, including meeting up with scientists involved in the research project, some of whom were visiting the living lab to connect to this aspect of the overall research project. 

Group photo of Living Lab Roundtable community members with landowners of the Living Lab site and the scientists of the Rwanda Restore Project.

These field meetings were crucial, first for the communities to connect to the scientists but also to agree with landowners on the selected species as well as the proportion of native and exotic species to be used in the living lab experiments. After some comments, suggestions and adjustments, a final agreement was reached. And there it was – the blueprint of the living lab experiments, co-created between local stakeholders and scientists! 

An impression of the second day out in the fields:

Dr Apollinaire William is facilitating a discussion between Actor-based Living Lab Roundtable members (NGOs, Government, Academia, Private Sector) with landowners at the Kamaranzara site, one of the sites in Teba Cell, in Rutsiro District, where some of the Living Lab experiments will be conducted in September-October 2026.

What made these workshops special

Not only was this trip key in determining the ongoing process of the restoration experiments, but it was also, and maybe even first and foremost, very valuable in terms of interpersonal exchange. When you hear Dr. William Apollinaire, the Postdoc researcher and coordinator of the living lab research in western Rwanda, speaking about his highlights of the workshops in February, it soon becomes clear that the workshops as well as the whole process mean a lot to everyone who is involved in the living lab. The true jewel was seeing how much the stakeholders are engaged. The participatory process, the involvement and empowerment of the people. Moments like spontaneously singing and dancing together, enjoying and exchanging cultural aspects. All of this leads to a successful co-creation. According to Dr. William Apollinaire, the stakeholders are fully engaged and eager to see the outcome of the living lab. “It was interesting to see how, if people are empowered enough, they can engage actively in the collaborative or participatory process.”

But how do you get to this point of engagement?

It wasn’t difficult to get people on board and sustain their engagement long-term, as the initiators of the living lab selected people for the roundtable who have already been involved in restoration activities. The key, however, is to value everyone who is part of the process. To value everyone’s voice. This way, an empowering environment was created in which every stakeholder could feel respected. Part of this is also organizing the logistics to make sure that everyone is involved, for instance, by facilitating communication. Understanding everyone’s expectations and goals. Every stakeholder being part of the design as well as the planning and implementation process: That is how co-creation works. That is how this momentum of engagement came to life.

What’s in store for the future?

Now you know what’s currently going on in the living lab, but we certainly don’t want to withhold the exciting next steps that are coming up our way. Right at this moment, the trees, which the two governance models have decided on, are developing in a nursery of local restoration stakeholders. They are growing and waiting to be planted in October this year, but before the planting happens, the farmers will learn how to prepare the land for the planting in September. The kick-off of the planting campaign in October will be a big event where the trees will go into the ground and we’re going to hear different speeches from people from the government and local communities. Excitingly, the national media is going to be invited to cover the event and who knows, maybe you might even read an article about it on this blog, so stay tuned!

While the trees are growing

In the meantime, the stakeholders will, together with the scientists, define socio-ecological indicators of success to be measured during the experiments and how those are going to be monitored. Examples of these indicators of success might be improving the nutrition of local communities as well as reducing medical costs. Ecosystem restoration is not just about the trees but also about shrubs and other plants that are incorporated alongside them, with nutritional and medicinal value. Species that have these values and are going to be planted include, for instance, Albizia gummifera, Carapa grandiflora, Vernonia amygdalina, Milletia dura, Ricinus communis, passion fruit, local papaya, avocado, lemon, myrianthus, and chayote.

Attention! The initiators of the living lab are also writing a book about agroforestry trees in Western Rwanda and their many benefits and uses. It will highlight especially native species and their potential but also include non-native species which are still useful for ecosystem restoration. Watch this space in the future for more information on the book. The aim in publishing this book is to contribute to connecting local people and NGOs to the wide array of, especially native trees they have in at their disposal.

What happens after the planting?

Don’t forget that the planting of around 80 % native species in the experiments is very different to the usual non-native tree planting in Western Rwanda. Therefore, it will be very interesting to see the outcome of the experiments. After some time, the two governance models will be compared, and the stakeholders will continue to be engaged by learning how to monitor the sites. The overarching goal is for the communities to have full ownership of the restoration in the end and that they keep monitoring it themselves in the future.

One last thing we want to give you to take with you on your way

Let’s go back to this sentence you already read at the beginning:

“If you want to go fast – go alone. If you want to go far – go together.”

As you probably noticed from this article, the co-creation of ecosystem restoration is, most importantly, a joint learning process. This might take a long time. But it will be most definitely worth it! Only by including everyone involved, designing and walking the path together, will the outcome match what everyone envisioned. Or, to say it in the words of Dr. William Apollinare: “That is how we achieve success in scaling up and upgrading the restoration experiments from research to practice!”.

Thank you for reading and hopefully, you’ll be back for the next update of the living lab in Rutsiro!

Citation: 

Nyiramvuyekure, V., Fischer, J., Kaplin, B. A., Mukuralinda, A., & Temperton, V. M. (2026). Woody vegetation diversity remains low after extensive forest landscape restoration efforts in a western Rwandan landscape. Biological Conservation, 317, 111812. https://doi.org/10.1016/j.biocon.2026.111812

Figure: Western Rwanda Landscape. Photo by Ping Sun.

By Molly Parker, on Sun et al. (2026) Home‑garden connectivity rather than tree‑cover connectivity facilitates biodiversity in fragmented tropical forest landscapes, https://doi.org/10.1007/s10980-026-02352-9 (full citation at the bottom of this post). 
 

In tropical biodiverse landscapes, forest loss, fragmentation, and degradation are major drivers of biodiversity loss. Forest landscape restoration interventions, a popular strategy to address these issues, most often focus on tree planting and overlook other tree-rich systems such as home gardens (HG). This recent study in western Rwanda’s Afromontane rainforest landscape by Sun et al. (2026) aims to assess these two land use types and their respective roles in supporting landscape connectivity and biodiversity. 

The study area (Fig. 1), the Afromontane rainforest landscape of western Rwanda, has dominant land use types of tree plantations and HGs. This recent study by Sun et al. (2026) aimed to assess their (tree plantations’ and HGs’) respective roles in supporting landscape connectivity and biodiversity. Researchers collected data from 91 field sites to examine how connectivity of tree cover and HG connectivity related to the richness and diversity of woody plants and birds. 

Fig. 1. Study area in the Western Province of Rwanda, covering seven administrative districts from north to south: Rubavu, Nyabihu, Rutsiro, Ngororero, Karongi, Nyamasheke and Rusizi. The blue dotted line marks the Congo-Nile watershed divide (Congo-Nile Crest), while protected national park areas are highlighted in green color.

The results (Fig. 2) showed that tree cover connectivity was negatively correlated with the richness and diversity of woody plants and birds, suggesting that increased connectivity through mostly exotic trees didn’ttranslate to habitat connectivity for biodiversity. However, HG connectivity was positively correlated with the richness and diversity of woody plants and birds. 

Fig. 2. Scatterplots showing the relationships between tree-cover connectivity and home-garden connectivity, with (A) woody plant richness; (B) Shannon diversity index of woody plants; (C) bird richness; (D) Shannon diversity index of birds represented by a color gradient ranging from blue (low richness) to red (high richness). Points represent different land matrix types: circles for home gardens and triangles for tree plantations.

 

This is an important finding as the more common restoration practices are on increasing tree cover. This research challenges the usual approach, suggesting that maybe restoration should begin focusing on connectivity of HGs since their connectivity seems to function as a network of ecological steppingstones supporting both plant and bird communities across the fragmented landscape.  

The authors recommend that conservation efforts continue and that restoration be scaled up to address degraded forests. Specifically, they recommend protecting remaining natural forests, promoting the regeneration of secondary forests dominated by indigenous tree taxa, and actively expanding biodiverse agroforestry systems like HGs. The findings make a compelling case that incorporating HGs into restoration planning could be essential for maximizing forest landscape restoration effectiveness. 

Citation: 

Sun, P., Baumann, M., Kaplin, B. A., Kuemmerle, T., & Fischer, J. (2026). Home-garden connectivity rather than tree-cover connectivity facilitates biodiversity in fragmented tropical forest landscapes. Landscape Ecology, 41(5), 78. https://doi.org/10.1007/s10980-026-02352-9 

Figure: Exemplary landscape at the sampling sites. Photo by Verene Nyiramvuyekure.

By Molly Parker, on Nyiramvuyekure et al., 2026, Woody vegetation diversity remains low after extensive forest landscape restoration efforts in a western Rwandan landscape,https://doi.org/10.1016/j.biocon.2026.111812 (full citation at the bottom of this post). 

Rwanda, like many biodiversity-rich tropical countries, is facing land degradation, biodiversity loss, and climate change, all of which are exacerbated by anthropogenic disturbances and further driven by historical aspects which have placed enormous strain on the country’s landscapes and natural resources. With the aim of providing ecological and social benefits and mitigating climate change impacts, forest landscape restoration has been implemented country-wide. Despite these efforts, the extent to which these efforts are actually restoring the diversity, structure, and function of ecosystems has remainedlargely unknown. Nyiramvuyekure et al. (2026) examined this by looking at woody vegetation species structure and diversity across different land use types in western Rwanda, using Gishwati-Mukura National Park as a reference site. 

This study took place across four districts in western Rwanda (Rubavu, Nyabihu, Rutsiro, and Ngororero) which were further classified into five social-ecological clusters (Fig. 1).  

Fig. 1. Study area and identified clusters; Cluster 1: Mountain pastures; Cluster 2: Lowland roads; Cluster 3: Houses and trees; Cluster 4: High mountains; Cluster 5: Remote regions.

Using a quasi-experimental design, the researchers surveyed 159 sites across four land use types within the study: agriculture mosaics, tree patches, pastures, and homegardens. Species composition was documented by classifying each species as native or exotic and assigning it to one of three functional groups: pioneer, forest specialist, or generalist. An initial hypothesis by the authors, based on known patterns of management intensity and planting history in the region, was that different land use types would restore woody vegetation diversity and structure to varying extents.  

The findings showed significant variation in woody species richness and compositions across land use types. The reference sites within the national park had 100% native woody species, which was significantly more native species than the other land use types (Fig. 2). 

Fig. 2. Woody species classification based on the origin (native or exotic species).

 

Of the non-reference sites, homegardens had the highest species richness and diversity, while agricultural mosaics and pastures showed intermediate richness, and tree patches had the lowest (Fig. 3).  

Fig. 3. (Figure 5 in article) Boxplot of species richness within land-use type: A: Agriculture Mosaic, H: Homegardens, P: Pastures, and T: Tree patches.

 

Regardless of extensive restoration efforts, the landscape in western Rwanda continues to be dominated by exotic species such as Eucalyptus and Pinus. Although these species provide numerous crucial socio-economic benefits (timber, fuel, erosion control, etc.), these often come at the cost of biodiversity and ecological resilience. Homegardens stand out in this study as they are traditionally overlooked in restoration initiatives, but here they showed the highest native woody species richness of the non-reference land use types. This suggests there may be a greater role for homegardens to play in conservation strategies.  

The study highlights the complexities of forest landscape restoration in western Rwanda. It shows that woody species richness remains low in restored sites compared to reference sites even after significant restoration. Based on their findings, the authors recommend using more native species in restoration efforts, including homegardens in restoration strategies, and conducting additional research on the drivers of species selection in restoration. It is not news that balancing biodiversity conservation efforts with socio-economic needs of local communities is a major challenge. It is, however, becoming clear that more inclusive and adaptive restoration approaches are needed not only in Rwanda, but in other regions facing similar pressures and problems.  

Citation: 

Nyiramvuyekure, V., Fischer, J., Kaplin, B. A., Mukuralinda, A., & Temperton, V. M. (2026). Woody vegetation diversity remains low after extensive forest landscape restoration efforts in a western Rwandan landscape. Biological Conservation, 317, 111812. https://doi.org/10.1016/j.biocon.2026.111812