Their message is clear: ecosystem restoration will play only a limited role in mitigating climate change. But that doesn’t make restoration any less important, however, for protecting biodiversity, strengthening ecosystem resilience, and locally adapting to climate change.

A holistic approach to global restoration

Previous studies on carbon sequestration potential from ecosystem restoration focused on forests and total carbon stocks, suggesting restoration could offset up to two-thirds of carbon emissions. But these estimates were built on imprecise, uncertain  and unrealistic assumptions, for example about land availability for restoration or policy feasibility.

Tölgyesi, Temperton and colleagues took a broader view. They:

• modelled restoration potential across four major ecosystems: forests, shrublands, grasslands, and wetlands; by using a broad database compilation with high-resolution satellite data.
• applied machine learning to predict the potential cover percentages of native ecosystem types to terrestrial locations using climatic, soil and topographic predictors
• estimated carbon sequestration using annual rates, not total stocks, over the timeframe from 2030–2100 for in total 12 biome-ecosystem combinations (e.g. temperate forests, tropical grassland).
• filtered the land available for restoration by excluding areas that are naturally intact, built-up, intensively farmed, or low in productivity (e.g., polar or arid regions).
• factored in future climate scenarios and ecosystem state transitions, which may cause losses in existing carbon stocks.

Here’s what the study uncovered

The study estimates that restoring the maximum available area under current climate conditions could sequester 96.9 gigatons of carbon (Gt C) by 2100. Seems like a lot, doesn’t it? The reality check shows: That’s just 17.6% of total anthropogenic emissions to date, or between 3.7% and 12.0% of projected future emissions (so, depending on the four used global emissions scenarios, so-called Shared Socioeconomic Pathways).

But, and this is the kicker, when restoration is matched to future climate conditions and takes into account the expected state transitions of ecosystems (e.g. forest converting to savannah) the carbon benefit drops to nearly zero. That seems pretty sobering at first. But this realistic assessment is extremely important and holds opportunities for climate and nature protection. Why? Read on.

An important comparison: forests vs. open ecosystems

A major strength of this paper is that it goes beyond trees. Grasslands, shrublands, and wetlands (open ecosystems) are often overlooked, yet they store substantial amounts of carbon, particularly underground, have a higher albedo, and are more resilient to fire and drought.

In the most feasible and realistic restoration scenario of this study, about 58% of carbon gains come from forests, while 42% come from open ecosystems. This balanced view helps avoid the mistake of planting trees where they don’t belong – thoughtless actions that happen currently and can harm biodiversity and local nutrient and water cycles.

Policy implications: less carbon, more resilience

So, what should we take from this?

1. Very important point: Ecosystem restoration is still crucial, just not as a silver bullet for climate change.
2. Restoration should be pursued for biodiversity, ecosystem resilience, and local climate change adaptation.
3. Site prioritization matters: the researchers identified specific 100×100 km priority zones where restoration could yield the highest carbon benefit, including temperate areas, such as American prairies and central Asian steppes and not only formerly prioritized tropical rainforest regions.

What we need is a shift in mindset

The authors conclude that restoration should be repositioned: from a tool to offset emissions, to a strategy for climate adaptation, biodiversity protection, and ecosystem service support. This is also integrated in important policies and agendas like the EU Nature Restoration Law from 2024 and the UN Decade on Ecosystem Restoration. But it requires clearer communication about what restoration can and cannot deliver.

Rather than chasing carbon credits, we should restore ecosystems to help humans and nature adapt together – to an uncertain climate future.

This study marks an essential milestone, not only for research at Leuphana but also for the global restoration science community. It sets a new benchmark for how restoration potential should be assessed: with ecological nuance, spatial realism, and climate foresight.

You can read and share the paper here: https://www.nature.com/articles/s41561-025-01742-z

The article from Leuphana’s school of sustainability about the study can be found here: https://www.leuphana.de/en/institutions/faculty/sustainability/news/single-view/2025/08/06/new-scientific-evidence-on-ineffective-and-unjust-climate-policies.html

Info: Functional richness & functional divergence
Functional Richness (FRich) describes the range of different functional traits found in a community. A high FRich means many ecological strategies are present, suggesting greater ecosystem adaptability. Functional Divergence (FDiv) measures how species’ traits are distributed within that range. High FDiv indicates that species are functionally distinct, occupying different niches, which is often linked to resource specialization or strong competition. Together, these metrics help assess biodiversity beyond species counts, focusing on how species function.

What are funtional traits – and why do they matter?

Functional traits are measurable properties of plants, such as leaf thickness, nutrient content, or wood density, that influence how plants interact with their environment. These traits determine important processes like photosynthesis, water use, nutrient cycling, and carbon storage. Understanding how these traits vary helps scientists assess how forests function, how resilient they are to change, and how they may respond to climate stressors like droughts or rising temperatures.

Most Earth System Models hdo neglect the diverse and heterogeneous tropical forest biome by representing it as a largely uniform ecosystem. By this oversimplification, the accuracy of predictions about ecosystem functioning, climate feedbacks, and biodiversity resilience is limited.

Info: Earth System Models
Earth System Models (ESMs) are complex computer simulations that integrate physical, chemical, and biological processes across the atmosphere, biosphere, hydrosphere, and geosphere to understand and predict how the Earth system responds to natural and human-induced changes. They track the flow of energy, water, carbon, and nutrients to predict changes in climate, vegetation, and biogeochemical cycles.

A unique effort to map global canopy traits

In response, the researchers of the study undertook a comprehensive analysis of tropical forest canopy traits. They combined  field-collected data from more than 1,800 vegetation plots and tree traits with Sentinel-2 satellite remote-sensing, terrain, climate and soil data, to predict variation across 13 morphological, chemical and structural traits of trees and to map the functional diversity of forests across the tropical Americas, Africa, and Asia. The forest sites under study span a total of almost 800 hectares, covering diverse climates and landscapes.

Infobox: Sentinel-2 satellite imagery
Sentinel-2 is a pair of Earth observation satellites from the European Space Agency (ESA). They provide high-resolution optical imagery every 5 days, capturing data in 13 spectral bands. This allows scientists to monitor vegetation, land use, water bodies, and more. In ecology, Sentinel-2 is crucial for detecting plant traits, forest structure, and environmental change at fine spatial scales.

Distinct forest identities: Americas, Africa, Asia

The analysis reveals strong biogeographical differences in canopy functional traits:

American tropical forests exhibit the highest functional richness, meaning they span a broader range of trait combinations. This reflects the high species diversity and environmental heterogeneity, found in the tropical Americas.
African forests, by contrast, show the highest functional divergence, indicating a more specialized pattern of resource use. This might be driven by long-term environmental pressures such as historical drought.
Asian tropical forests (including parts of Australia) display high average values for traits like leaf size, water content, and nutrient concentrations, likely linked to the dominance of the Dipterocarpaceae family in Southeast Asia.
These trait distributions suggest that each region has evolved distinct canopy strategies shaped by evolutionary history, soil fertility, rainfall seasonality, and past climate conditions.

Wet vs. dry

Dry forests (e.g., in Brazil’s cerrado or African savannas) have species with high SLA and nutrient-rich, fast-turnover leaves. This indicates acquisitive strategies optimized for rapid growth during short wet periods.
Wet forests (like Amazonia or Borneo) show traits associated with conservative strategies with thicker, denser leaves and higher carbon investment.

Implications for science, conservation, and climate models

This study improves the realism of Earth System Models by providing detailed trait maps. It identifies regions with high uncertainty or data gaps, guiding future field work (especially in parts of Africa and Asia). Furthermore, the study showcases the rising importance of AI and remote sensing technologies in mapping plant traits and biodiversity on a large scale. But while these tools are powerful, they are meant to complement – not replace – classic ecological methods like field sampling and species identification. To really understand how functional diversity changes over time, we still need to keep investing in good old-fashioned fieldwork, which then feeds into more advanced models and predictions. The insights of this fascinating study can serve as a foundation for forecasting changes in functional forest composition under shifting climates and contribute to building a more process-based and accurate ecological modelling framework across different spatial and temporal scales.

You want to learn more about the study and its findings? Then read the full article here: https://www.nature.com/articles/s41586-025-08663-2#citeas

An experiment bringing the mountain to the valley

To simulate the effects of rising temperatures, the researchers transplanted highland plant communities from the German Alps to warmer, lower-elevation regions. Over four years, the researchers analysed how the transplanted communities changed in terms of species composition, functional identity (characteristics like leaf traits and resource-use strategies) and diversity (a measure of the variety of functional traits within a community). Therefore, they asked questions such as: How do communities change under warmer conditions? Which species are “winners” and which are “losers”? And how does the immigration of warm-adapted species affect the functional identity and diversity of communities as well as ecosystem functions and processes?

What makes this approach unique is that entire plant communities were integrated into natural environments. This allowed the researchers to account for real interactions with local species and environmental influences, providing more realistic results than e.g. artificial warming chambers. Factors like competition with native plant species and the effects of natural soils were thus included.

What are the findings, and what do they mean?

The study measured functional morphological and biochemical leaf traits, enabling the calculation of community-weighted trait means, functional richness, and functional divergence. The focus on functional traits provided insights into how species interactions and resource-use strategies could shift in response to climate warming.

The findings highlight the dynamic yet vulnerable nature of alpine ecosystems. Within just four years, the transplanted communities underwent significant changes. These communities first gained species richness, attributed to the immigration of warm-adapted lowland species and the “lag phase” of highland specialists, meaning their delayed disappearance. Highland specialists initially resisted extinction through strategies like vegetative reproduction (e.g., via rhizomes).

Warm-adapted species, originating from lower regions, exhibited different “strategies”: they grow faster, absorb more nutrients, and compete stronger for light. Over time, this could threaten the survival of cold-adapted specialists, which rely on slow growth and efficient resource use. For instance, Poa alpina (alpine meadow-grass), present in nearly all transplanted plots in the first year, had nearly vanished by the fourth year.

Another striking result is the increasing similarity between the transplanted and lowland communities. The mountain thus loses part of its unique identity, becoming similar to the lowlands with regard to species composition.

A call to action for protecting montane and alpine ecosystems

Overall, the results confirm the stated hypotheses that climate warming leads to significant changes in species richness, composition, and functional traits in mountain meadow communities. Highland plants adapted to cold temperatures and nutrient-poor soils are at risk of disappearing due to climate change. These species are not only ecologically but also culturally important. The study emphasizes that alpine ecosystems worldwide face similar threats, emphazising the strong connection between climate change and biodiversity loss.

To preserve these valuable mountain ecosystems, targeted conservation measures are essential. These include reducing greenhouse gas emissions to mitigate the effects of climate change and developing strategies to protect and support the survival of sensitive highland specialists.

If you are interested in the study in more detail, you can find it here: https://onlinelibrary.wiley.com/doi/full/10.1111/jvs.13280

Ecosystem restoration has become a global priority due to accelerating land degradation, biodiversity loss, and climate change. However, the ecological, social, and interconnected social-ecological impacts of restoration efforts remain insufficiently understood. The DFG research project A social-ecological systems approach to inform ecosystem restoration in rural Africa (2023-2028) aims to develop a comprehensive framework, approaching ecosystem restoration from a social-ecological systems perspective for understanding the mechanisms involved in generating different restoration outcomes.

Focusing on western Rwanda, a global restoration leader, it seeks to generate both place-based insights and transferable knowledge for restoration efforts worldwide. The research team consists of an interdisciplinary team and includes scientists from Leuphana (Institute for Social-Ecological Systems (SESI) and Institute of Ecology (IE)), the Universities of Göttingen and Kassel as well as the FU Berlin. The work is divided into eight interconnected sub-projects, organized into four clusters (see below). The research unit combines post-hoc assessments, participatory experiments, and future scenario planning to provide a holistic understanding of restoration dynamics. The findings shall advance restoration science and social-ecological research, directly benefitting restoration efforts in Rwanda, and offering global insights for improving restoration practices worldwide.

One of the sub-projects (SP7), headed by Vicky Temperton from the Leuphana Institute of Ecology and Stefan Sieber (Leibniz-Zentrum für Agrarlandschaftsforschung (ZALF) e.V.), focuses on establishing a living lab to bridge science and practice in Rwanda’s restoration efforts. While past restoration has largely relied on Eucalyptus monocultures, new initiatives promote native tree species to enhance biodiversity and ecosystem services. Using a transdisciplinary approach, SP7 collaborates with stakeholders to co-design, co-produce, and co-evaluate restoration solutions. Scientific experiments will be conducted in two governance models, allowing comparison of their impacts on biodiversity, resilience, and livelihoods.

The following article was originally written and published by Dr. William Apollinaire on the ecosystem restoration blog.

After the Rwanda Restore project kick-off in Kigali in January 2024 and a restoration stakeholder conference held there from February 19–21, 2025, the SP7 research unit officially launched the Living Lab in Rutsiro District, western Rwanda, on February 25, 2025.

42 participants attended the workshops from academia, various governmental and non-governmental institutions as well as local communities. The workshops aimed to establish a Living Lab Roundtable and define two governance models for the Living Lab, along with its code of conduct and communication strategy. Further, the workshop discussed opportunities and challenges for the current, mid-term and long-term future of restoration in Rutsiro, using a Three Horizons Approach, for restoration strategic planning in Rutsiro.

Organized into small group discussions, participants defined the code of conduct, communication principles and strategy and the three horizons for the restoration in the area. At the end of the workshop, participants visited Living Lab sites in Gihango Sector. During the visit, the research team hiked through the hilly landscape to observe various land uses and assess the potential for restoration interventions. They were accompanied and guided by local field assistants and farmer group members, who provided valuable insights.

A community workshop was organized in two cells of the Gihango District: Teba and Shyembe. The field trip brought together representatives of farmers groups and cooperatives, carpenters, traditional healers, beekeepers, and farm owners of plots where the Living Lab sites sit.

Individual visits were also arranged to meet model farmers who have enhanced their nutrition by integrating food plants such as Chayote, Passion fruit, Avocado, Pineapple and Cucumber into agroforestry systems, particularly within their home gardens, for their consumption or the market. In some cases, farmers combine a variety of fruit plants and legumes within the same plot.

The study selection and the social network analysis have been completed in 2024. The roundtable has now been set up, the ongoing process before the end of this year will cover the co-design of field trials, definition of impact areas, delineation and registration of demonstration sites, including the signing of sustainability agreement with farm owners within the Living Lab sites. Further, the roundtable of stakeholders will soon develop the indicators of success.

You can find out more about the research project and the sub-projects here: https://ecosystemrestoration.net/subprojects/