Coffee is a global commodity crop which shapes tropical landscapes through its particular agricultural system (Harvey et al., 2021). Coffee landscapes are highly valuable for global climate and biodiversity (Jha et al., 2014; Pendrill et al., 2019). Coffee farming and processing is a major source of income for local populations (Harvey et al., 2021). Depending on shade levels, its cultivation may preserve local ecosystems, but it can also degrade them (Jha et al., 2014; Martin et al., 2020). Climate change threatens to alter local land use and production systems. Adaptation to and mitigation of the effects of climate change will help secure local livelihoods, and have been shown to have global implications through teleconnections (Rahn et al., 2014; Pendrill et al., 2019; Verburg et al., 2019).

Climate-smart landscapes (Scherr et al., 2012) have been established as a general guiding framework for effective climate action, and include multifunctional landscapes. Multifunctional landscapes deliver multiple ecosystem services throughout each landscape, maximizing contributions to human and other species' well-being and achieving system sustainability (Fischer et al., 2006; Wu, 2013; Hölting et al., 2019). However, despite their relevance for climate action, a more finely-grained framework and review has yet to be developed. Such an approach should structure applicable practices and connect local action to benefits at local, landscape and global scales (Harvey et al., 2021).

We use coffee landscapes as an example of a tropical commodity crop adapting to climate change, as it is of major importance for local livelihoods and biodiversity impacts, especially in Latin America. This lays foundations for establishing more effective links with local policy and communities, and with international donors, when aiming to sustainably adapt commodity landscapes to climate change.

The socio-ecological systems framework describes human-nature interaction and sees human actions as strongly embedded in and interacting with local and global dynamics. For land systems, this framework integrates land use as a connection between social and ecological systems, whilst these elements are embedded in and interact with the global earth systems (GLP, 2005). Coffee farming systems can be seen as a complex socio-ecological systems (Schröter et al., 2018; Malik et al., 2019). Interactions occur between (i) the social system, embedding socio-cultural values and economics, (ii) the ecological system of tropical landscapes, and (iii) the local land-use system of coffee farming, forming and being formed by landscape and culture. Meanwhile, these arising coffee landscapes are coupled to (iv) the global earth system, especially through the impacts of climate change and the landscapes' importance for sustaining global biodiversity, whilst also being influenced through (v) international market dynamics.

The social system of coffee farming embeds the local economy, oftentimes strongly depending on this commodity crop, and the related local cultural values. Globally, for about 12.5 million farmers, coffee farming is the primary source of income (Browning, 2019), of whom many are smallholders (Pham et al., 2019). Coffee farming provides jobs and contributes to poverty alleviation and to the economies of the countries where it is produced (Pham et al., 2019). Including the farmers' families, an estimated 60–120 million people's livelihoods depends on coffee farming (Sachs et al., 2019). However, in many countries coffee farming is also deeply rooted in local culture and heritage: The coffee cultural landscape of Colombia, for example, was listed in the UNESCO World Heritage List² in 2011 (Silva, 2017), showing its value as local and global cultural heritage. Likewise, coffee as a locally-produced crop can be deeply integrated in cultural rituals, as, for example, the Ethiopian coffee ceremony shows (Hirons et al., 2016).

The ecological system is partly characterized by the local abiotic and biotic conditions that interact with local coffee farming. Abiotic conditions such as temperature and precipitation influence coffee quality and yields (Pham et al., 2019). Moreover, altitude and climate influence which coffee species can be farmed: Robusta better withstands higher temperatures but needs sufficient precipitation or additional irrigation (Descroix and Snoeck, 2009). Meanwhile, Arabica coffee, which is more highly valued due to its milder taste, is generally farmed at higher altitudes (Descroix and Snoeck, 2009).

Land use connects the social system with the ecological tropical environment in which coffee farming is embedded, and also shapes the natural environment. Pollinators and pest control agents are important for coffee yield and quality, and at the same time their abundance and diversity is influenced by the way in which local farming systems are managed (Tscharntke et al., 2011). The characteristics of the land-use system always include the coffee plant as a woody perennial but vary with production intensity (Moguel and Toledo, 1999): In extensive systems, coffee is farmed as an understory crop in multispecies farming systems. However, in more intensive systems diversity of companion crops is gradually reduced and overstory trees (especially shade trees) are increasingly mono-species (Jha et al., 2014; Martin et al., 2020). In sun-grown coffee systems they are entirely absent. Systems with higher degrees of shade can store more carbon, are more climate-resilient, can deliver more diverse ecosystem services, and contribute more strongly to nature and biodiversity conservation (Tscharntke et al., 2011; van Rikxoort et al., 2014; Pham et al., 2019). However, lower levels of shade offer higher yields, at least if abundance of water is sufficient (Schnabel et al., 2018). Recognition of this tight socio-ecological interaction and its importance for biodiversity can be seen in various coffee production countries: Apart from the coffee cultural landscapes in Colombia, also Man and the Biosphere Reserves in Mexico³, El Salvador⁴, and Ethiopia (UNESCO, 2011) explicitly acknowledge the key influence of coffee farming on the landscape and its biodiversity.

The earth system is mainly influenced by the impact of coffee farming on the global tropics, whilst impacting coffee landscapes through market dynamics and recently also climate change: The tropics harbor the highest biodiversity worldwide, where climate changes poses special risks for such biodiversity (Pimm et al., 2014; IPBES, 2019). More than 19 of the currently recognized 36 global biodiversity hotspots are situated in the tropics, and many of these in major coffee production countries such as Brazil, Vietnam, Colombia, and also the Mesoamerican region. At the same time, tropical rainforests are probably the most emblematic carbon sink globally. However, as coffee farming areas expand or shift to new areas, these forests and the biodiversity they preserve may be compromised through deforestation (Läderach et al., 2017).

International market demands and dynamics may induce such expansion of coffee farming areas (Pendrill et al., 2019; Harvey et al., 2021), and thereby illustrate a further important interaction within the socio-ecological system. The coffee market is buyer-driven, with power located in consuming countries, mainly located in the global north (Grabs and Ponte, 2019). Prices are volatile and pressure is passed on along the value chain to upstream actors, compromising profit margins and economic viability of coffee traders and growers (Grabs and Ponte, 2019). Examples for interactions of global trends in coffee production with local landscapes can be seen in the 1990's boom and bust cycles or also recent altitudinal migration due to climate change and consumers' preferences. The boom and bust cycles were caused by global market deregulation, combined with booms after local frosts in Brazil that caused global coffee prizes to surge, and induced overproduction on new cultivation areas, leading up to busts in later years (Eakin et al., 2009). More recently, the example of high-grown coffee illustrates the direct influence global markets can have on local ecosystems: This coffee is increasingly demanded and more highly valued due to a milder, more aromatic taste, incentivizing shifts of coffee farms within landscapes toward areas in higher altitudes, especially as, with climate change, lower areas become unsuitable for coffee farming (CIAT, 2017). However, such altitudinal migration thereby compromises local natural habitats and rainforests that remain in these higher altitudes (Läderach et al., 2017). Thus, we begin to see how climate change acts as a new and additional driver, causing great distortion in the socio-ecological system of coffee farming.

Our research will show how the adoption of CS-practices in coffee farming systems contributes to local, landscape and global benefits. However, considering the multitude of available practices, instead of trying to link each individual practice to different benefits, we classified practices in seven functional groups (for example “soil characteristics” or “water management”, see section Review findings), according to the functional traits of the farming system to which these practices contribute. Then, based on these functional groups, we established links to the broader benefits provided by the adoption of those different functional groups (for example “habitat connectivity” or “water quality”).

Our research design is based on an iterative, exploratory literature review (Arksey and O'Malley, 2005; Baumüller, 2018) and consists of three successive major steps: We (i) identified the different CS-practices applicable to coffee farming systems, (ii) grouped them into different functional groups that describe the main functional trait that these practices target within the farming system, to then (iii) perform an additional review, that identifies benefits that are provided at local, landscape and global scales when adopting CS-coffee practices.

In detail, in a first step, our iterative reviews started from the basic element of climate-smart landscapes: the adoption of CS-practices. We reviewed literature from the year range of 2010–2020, drawing on Web of Science, Scopus and Google Scholar as search engines and including scientific and gray literature. Our understanding of CS-practices is based on the official definition by the Food and Agricultural Organization by the United Nations (FAO, 2013), therefore including both, climate adaptation and climate mitigation practices, as long as positive livelihood effects for the farmer can be assumed.

A specific search for CS-practices in coffee yielded very few results: results were limited to individual case studies and gray literature, without taking further into account interactions between practices or landscape effects. Therefore, we defined the search scope somewhat more broadly: on the one hand, we broadened the agricultural field, including also CS-practices (search terms: “climate-smart practices”; “climate-smart agriculture” AND practices) in agriculture in general, and then subsequently refined listings to only include practices applicable to coffee farming systems. We checked for duplicates and then reviewed title, and if relevant, abstract of the papers, before reading the papers more in detail. For Google Scholar, we only reviewed the first 300 results for each query. To refine CS-practices to be applicable to coffee farming systems, the specific characteristics of coffee-farming systems were taken into account, including its characteristic as a woody perennial that can be farmed as an understory crop. Furthermore, practices had to be applicable in the tropics, were coffee is farmed.

On the other hand, we broadened the type of practice, including also climate mitigation (“climate change” AND mitigation AND coffee) or climate adaptation practices (“climate change” AND adaptation AND coffee) but here only including sources specifically aiming at coffee farming systems. We identified seven papers as especially relevant to list CS-practices or practices that offer combined adaptation and mitigation potential in coffee farming (find listing of practices in the Supplementary Material S1) and from which then our subsequent identification of functional groups started from.

In the second step, we distilled major functional groups from the listings of CS-practices for coffee-farming systems. These groups correspond to the main functional trait that the practices address within the coffee farming system. Functional groups (such as “Natural habitats”) then allowed us to link functional traits that are improved through the adoption of CS-practices (such as “Structural elements and natural habitats”) to broader benefits of coffee-farming systems that have been reported throughout scientific literature on coffee-farming systems.

Therefore, in a third step, and based on these functional groups, we then performed another review, to identify contributions to local, landscape and global benefits. For each functional group we checked for corresponding papers that report on elements in coffee-farming systems or similar farming systems (for example cacao) presenting connections between the identified functional groups and benefits at local, landscape or global scale. Here, we based our work in the principles of landscape ecology (especially Tscharntke et al., 2005, 2012; Fischer et al., 2006). Also, we took into account the conceptual advances on landscape benefits and landscape multifunctionality (Scherr et al., 2012; and for landscape services Termorshuizen and Opdam, 2009; Bastian et al., 2014). These benefits were then reported in a structured form, applying the Millennium Ecosystem Assessment (MEA) framework for ecosystem service assessment, distinguishing between regulating, provisioning and cultural ecosystem services (MEA, 2003).

Adopting CS-practices in coffee farming systems provides benefits that reach beyond the farm scale. Instead, the benefits help progress toward multifunctional landscapes and landscape and global sustainability, especially when coordination across the landscape takes place. In the tropics, and especially Latin America, coffee-farming landscapes face vast distortions due to climate change. By more widely introducing CS-practices in coffee farming, local actors can capitalize on the benefits provided to (i) promote especially those CS-practices that offer local benefits most needed in the local context, (ii) connect to international donors and value-chain actors, that have an interest in contributing to preserve the globally-valued landscape services these coffee systems provide (or which could be improved through the adoption of CS-practices), and (iii) coordinate and support coffee farmers to maximize effectiveness of their individual climate-adaptation processes. Within this context, our collection and classification of existing climate-smart practices in coffee farming can help to select most appropriate practices for local adaptation needs. Then, our review and the linking of functional groups to benefits further allows for efficient communications with policy makers and “best-fit” selection of CS-practices to be introduced in the coffee-farming systems.

However, there are three additional considerations: firstly, regarding possible conflicts that may arise; secondly, the need to coordinate action at the landscape scale; and thirdly, financing needs and lead time for introducing the appropriate practices.

Tensions can exist within coffee landscapes as it is traded globally but impacts landscapes locally. Coffee is a global commodity crop that shapes local culture and landscapes, trading local ecosystem services for financial returns, as Schröter et al. (2018) have shown. With coffee being traded on a buyer-driven market (Grabs and Ponte, 2019), external dynamics shape coffee landscapes, causing potential tensions within the landscape regarding land-use scarcity and pathways of future farm management. The benefits found at landscape scale, facilitated by the integration of CS-coffee practices on farm, show that opportunities exist to reconcile export-orientated farming with local well-being. Thereby potential land-use tensions within the landscape can be alleviated.

However, some CS-coffee practices themselves may lead to tensions within the region, especially irrigation: although beneficial for the individual farmer, irrigation may compromise water availability for other users within the landscape. When precipitation patterns change, this will also negatively affect groundwater recharge, thus if farmers purely rely on introducing irrigation systems to increase their climate-smartness, groundwater levels will sink. Therefore, coordination at landscape scale will be needed, bearing in mind that landscape services depend on landscape wide implementation and spatial patterns.

Thus, landscape-wide coordination is not only needed “to enhance field-level benefits of [sic] climate-smart practices, […] secure ecosystem functions, and […] enhance the effectiveness of climate mitigation efforts” (Scherr et al., 2012; p. 5) but is also of great importance to mitigate potential conflicts between different land users and with civil society throughout the landscape.

Climate-smart coffee landscapes can act as a possible entry point for sustainable landscape governance and climate adaptation. To avoid the risks that commodity-centric landscape governance may pose (c.f. Bastos Lima and Persson, 2020) landscape managers should combine perspectives from inside and outside the coffee commodity crop sector, carefully balancing power relations. The multi-stakeholder platform proposed by Scherr et al. (2012) might be an adequate agora to facilitate this coordination. In addition, livelihood opportunities outside the coffee sector will have to be identified, as some coffee farmers will face transformative adaptation where local capacity or climate change intensity make incremental or systemic adaptation insufficient. Here, too, a wider approach to developing multi-stakeholder platforms that integrates coffee and non-coffee actors will be needed for a smoother transition to alternative sectors within the landscape, connecting social systems and offering alternative livelihood opportunities.

Lastly, practices such as shade-tree planting will require a certain lead time or need financial capital to be implemented. Payment measures like carbon credits (e.g., Gold Standard⁵), recognizing the benefits of carbon storage and carbon mitigation may finance some of the actions, but additionally financial instruments will need to be established. Similarly, payments for ecosystem services may finance transition as improved ecosystem deliverance has been highlighted here. Already existing practices like in Costa Rica for landscape coordination and transition payments (introduction of NAMA coffee⁶) may serve as examples here.

Our research has shown the complexity of interactions in the socio-ecological system of coffee farming. It becomes clear that though climate change threatens to distort the socio-ecological system of coffee farming, the introduction of CS-practices and climate-smart landscapes can safeguard local livelihoods.

We propose a classification of the seven previously-listed functional groups for CS-practices in coffee farming, highlighting functional traits these practices address within the system of coffee farming. These groups help to understand underlying mechanisms and allow targeting specific functional traits at the farm and landscape level to advance toward developing and harnessing landscape multifunctionality. Additionally, if implemented in a coordinated way throughout the landscape, the adoption offers landscape and global benefits that include (non-exhaustively) improved water quality, biodiversity conservation and habitat connectivity, as well as stabilized regional climate patterns.

Our considerations regarding transition toward climate-smart landscapes can help to mitigate potential conflicts and trade-offs within landscapes. Having outlined how local action and adoption of CS-practices help to sustain landscape services, different land-users can better understand and negotiate which landscape services they value particularly or want to prioritize and how to allocate burdens and rewards for conservation or sustainable actions. This can include for example transfer payments such as payments for ecosystem services directed to benefit local coffee farmers.

Trade-offs from especially threatening possible adaptation scenario can be avoided if opportunities for in-field climate adaptation are realized: Farmers performing altitudinal migration in cooler climates threaten native forests that remain there. Likewise, if they abandon coffee farming the existing agroforestry systems of coffee farms will be degraded as they are generally replaced by annual farming systems. Instead, financial and knowledge support from external actors can enable farmers to also implement those CS-practices that may be more costly or require specialized knowledge, but at the same time preserve the valuable landscape services needed within the landscape.

Policy makers in coffee farming regions now need to facilitate participative processes on prioritization of vital landscape services (c.f. Brandt et al., 2017 as an example). Based on these, locally-relevant functional groups of CS-practices in coffee farming can be identified. For the adoption of the corresponding functional groups, interactions are needed between researchers, proposing innovative financial instruments, policy-makers and cooperatives, disseminating knowledge and offering support, and practitioners, actively being involved in a process of co-creation.

Further research needs to design and support processes for actual realization of such landscapes, where the implementation of appropriate CS-coffee practices is linked to landscape-wide processes of integrated landscape management, seizing on the particular opportunities of the coffee farming sector. In this respect, financing options may especially be studied further. As a first step, one may think of a climate-smart coffee label, which may be oriented in the example of NAMA café. Whilst long debated, landscape certification is still scarcely found in practice (Flinzberger et al., 2020) but further on could provide a useful mechanism to finance shaping climate-smart landscapes (c.f. Ghazoul et al., 2009; Williams-Guillén and Otterstrom, 2014; Tscharntke et al., 2015; Deans et al., 2018; Harvey et al., 2021). Social science is needed to facilitate developing inclusive multi-stakeholder platforms that enable climate services (Findlater et al., 2021) to become on-ground climate action, integrating the concept of climate-smart (coffee) landscapes,. Also, using insights from social sciences and studying cultural and social landscape services can reveal interconnectedness and further beneficial effects of climate-smart coffee landscapes on psychological well-being and sense of belonging, amongst others. Moreover, although applying the concept of climate-smart landscapes may increases sustainability within farming landscapes, the reduction of trade-offs and climate impacts along the value chain shall not be forgotten, as here great carbon-reduction potentials persist (Kilian et al., 2013). In coffee, the focus will then need to be laid especially on implementing more sustainable ways of processing (Wainaina et al., 2020), which will also need to be considered in climate-smart landscapes for landscape-related commodity crops other than coffee.

Our review demonstrated how adopting climate-smart practices in coffee-farming systems provides local, landscape and global benefits. Benefits include improved water quality, biodiversity conservation and habitat connectivity, as well as stabilized regional climate patterns. These benefits show that opportunities exist to reconcile export-orientated farming with local ecosystem services. When landscape wide coordination is assured for, the multiple interests in coffee landscapes can be met and conflicts avoided. Local farmers, policy-makers and global donors must unite to improve uptake of climate-smart coffee practices in a coordinated way, to thereby augment and safeguard coffee-farming's socio-ecological system along with associated local, landscape and global benefits.

Initial idea was developed by CB. Material preparation, data collection, and analysis were performed by PS. The first draft of the manuscript was written by PS, and CB commented on previous versions of the manuscript. All authors contributed to the study conception and design, read, and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.