The team’s fieldwork covered 16 plots across four land-use systems, each plot a 500-square-meter sampling unit designed to capture the internal variability of these ecosystems. In the coffee systems, located between 1,060 and 1,080 meters above sea level, the researchers measured every tree within their plots, recording trunk diameter at breast height, total height, and wood density. These measurements fed into generalized allometric regression equations that estimate aboveground biomass from trunk dimensions and wood basal density, with species-level wood density values drawn from the Global Wood Density Database. Where species data were unavailable, the team used genus or family averages. Belowground biomass was estimated using root-to-shoot ratios developed for tropical highland forests, and a standard carbon fraction of 50 percent was applied to convert dry biomass into carbon, a figure consistent with IPCC default values of 0.47 to 0.50 for tropical woods.

The results are unambiguous. Primary forests stored approximately 133.74 tonnes of aboveground carbon per hectare and 40.12 tonnes of belowground carbon per hectare, equivalent to roughly 638.05 tonnes of carbon dioxide per hectare when the carbon is converted to its CO2 equivalent. The coffee systems, by comparison, stored dramatically less. Among them, the agroforestry coffee system with mixed shade trees performed best, followed by coffee grown with Inga, while the monoculture system consistently registered the lowest values across every carbon metric. The statistical models, fitted with Gamma distributions and generalized linear mixed models in R, achieved adjusted R-squared values exceeding 99 percent, confirming that the differences in carbon storage were overwhelmingly associated with the type of ecosystem rather than random variation.

Biodiversity followed the same gradient. Using the Shannon index, Simpson index, and Pielou evenness index, the researchers found that primary forests hosted the highest species richness and the most balanced distribution of individuals across all metrics, with statistically significant differences at p < 0.01. The agroforestry coffee system maintained a species richness of 6.25, though its lower evenness value of 0.42 suggested that a few dominant species shaped the community. Coffee grown with Inga showed lower richness at 2.0 but moderate evenness at 0.58, while the monoculture system presented minimal diversity, reflecting a simplified community dominated by very few species. The primary forest’s superior diversity may be linked to its higher elevation of 1,625 meters, where cooler temperatures and greater humidity favor the coexistence of more plant species.

When the researchers translated carbon storage into economic terms using Peru’s social cost of carbon of 7.1 US dollars per tonne, the primary forest’s value reached approximately 949.52 dollars per hectare in aboveground carbon alone, with additional valuations of 284.86 dollars per hectare in belowground carbon and 4,530.15 dollars per hectare when total CO2 storage was considered. The coffee systems, by contrast, registered values far below these figures, with the agroforestry system performing best among them and the monoculture worst. The economic gap between primary forest and any coffee system is so wide that the researchers describe forest conversion as generating significant losses in carbon-related benefits, losses that compound with every additional hectare cleared.

A sensitivity analysis tested how these valuations would shift under alternative carbon pricing scenarios. At low prices of 5 and 20 dollars per tonne of CO2, the economic differences between coffee systems narrowed considerably. But from 50 dollars per tonne upward, the gap widened, with the agroforestry coffee system standing out as the highest-value option among the coffee treatments, followed by coffee with Inga and then the monoculture. At 100 dollars per tonne, the greater structural complexity of the agroforestry system translated into a clearly visible relative economic advantage, suggesting that carbon markets could increasingly reward farmers who maintain shade trees and diverse canopy structures within their coffee plantations.

A principal component analysis integrating the ecological and economic data revealed that the first two axes explained 94.07 percent of total variation, with the first component capturing a robust gradient from simplified monoculture systems to complex, carbon-rich primary forests. The biplot positioned primary forests at the high end of total CO2, evenness, Shannon diversity, and Simpson index values, while the agroforestry coffee system clustered near species richness, and the monoculture sat at the opposite end of the gradient. Coffee with Inga occupied an intermediate position, suggesting a transitional state between low and high ecological complexity. This multivariate structure reinforces the biodiversity-functioning theory: ecosystems with greater species diversity tend to store more carbon because diverse communities use resources more complementarily, build more complex forest structures, and are more resilient to environmental disturbances.

The findings carry direct implications for climate policy in Peru and beyond. The researchers note that forests are expected to contribute roughly 10 percent of the mitigation required to limit global warming to 1.5 degrees Celsius, which would require carbon prices to reach approximately 281 dollars per tonne of CO2 by 2055. By comparison, the average carbon sequestration cost in northwest China has been estimated at 4.77 dollars per tonne, well below the 7.1 dollars per tonne applied in this study. Protecting forests in the Peruvian Amazon may require investments ranging from 7.97 to 10.45 dollars per tonne of CO2, figures that support the implementation of conservation strategies and increase the economic value of standing forests. The study’s authors argue that these valuations provide a technical basis for incorporating coffee agroforestry systems into Payments for Ecosystem Services schemes and for aligning coffee production models with Peru’s Nationally Determined Contributions under international climate agreements.

The research also speaks to the principles of REDD+ initiatives, which aim to conserve forests and increase carbon stocks through economic incentives. The observed transition from monoculture coffee to agroforestry systems represents a trajectory compatible with REDD+ objectives by simultaneously increasing stored carbon and biological diversity. The economic potential associated with conserving tree cover, rising from 174.06 dollars per hectare in monoculture to 4,530.15 dollars per hectare in primary forest, illustrates how market mechanisms such as PES and REDD+ can generate effective incentives when supported by strong institutions, transparent monitoring systems, and equitable distribution of benefits among local actors. In Latin America, which constitutes one of the main global suppliers of forest carbon credits, agroforestry systems hold particular promise for generating credits due to their CO2 capture capacity, though certification continues to face challenges related to additionality, permanence, and transparency of projects.

The authors acknowledge that their observational comparison carries limitations. The primary forest site sat more than 500 meters higher than the coffee systems, an altitudinal difference that could influence carbon dynamics, though previous research has shown that the loss of aboveground biomass from forest conversion dominates the carbon balance at all elevational levels. The models achieved high statistical fit, but uncertainty remains associated with ecosystem variability, biomass measurement, and the selection of allometric equations, meaning small changes in carbon estimates could modify the economic valuation of the ecosystem service. Nevertheless, the study’s central message is clear: conserving native forests remains the most effective strategy for maintaining carbon storage, biodiversity, and ecosystem services in the Peruvian Amazon, while promoting sustainable agroforestry practices offers a complementary pathway that rewards farmers for keeping trees on the land. As Peru grapples with continued forest loss, these findings provide the quantitative evidence needed to make standing forests economically competitive with the coffee that threatens to replace them.

Subject of Research: Carbon sequestration economics of forest-to-coffee conversion in the Peruvian Amazon

Article Title: Converting forests to coffee cultivation (Coffea arabica) significantly reduces organic carbon sequestration payments in the Peruvian Amazon

Article References: Converting forests to coffee cultivation (Coffea arabica) significantly reduces organic carbon sequestration payments in the Peruvian Amazon. (n.d.). https://doi.org/10.3389/frevc.2026.1905945

Image Credits: AI Generated

DOI: 10.3389/frevc.2026.1905945

Keywords: carbon sequestration, Peruvian Amazon, coffee agroforestry, REDD+, payments for ecosystem services, biodiversity, Converting, forests, coffee, cultivation, Coffea, arabica

The research, led by Firew Bekele Abebe of Addis Ababa University and Haramaya University together with colleagues at Stockholm University and the Swedish University of Agricultural Sciences, focused on the Gomma and Gera districts in the Jimma zone, a landscape of forest fragments, crop fields, grazing lands and coffee farms. Traditionally, Arabica coffee in its native range has been cultivated only between roughly 1,200 and 2,000 meters above sea level, a narrow band dictated by the plant’s acute sensitivity to temperature. Above that line, the highlands were considered unsuitable. Yet when the researchers surveyed sites between 2,007 and 2,358 meters, they found thriving young coffee agroforests established by smallholder farmers who had made a calculated bet on a warming world.

To understand why farmers were moving uphill, the team conducted semi-structured interviews with thirty farmers who had recently established coffee agroforestry in this novel elevational range, paired with thirty neighboring farmers who had not yet made the switch. The results were striking in their consistency. Every single interviewed farmer cited climate change as a major reason for establishing coffee at these previously inhospitable heights, and 93 percent specifically pointed to rising temperatures. Two farmers added that rainfall, once too abundant for coffee, had declined enough to make cultivation viable. Crucially, this was not a story of government mandates or external pressure; it was farmers responding accurately to meteorologically real changes they had observed on their own land.

Climate was not the only driver. Just over half of the farmers said they wanted a higher and more stable income, since coffee is a perennial cash crop whose global prices are currently high. Nearly a third reported another, more surprising motivation: crop raiding. Baboons, monkeys and wild pigs were devastating their maize, teff and sorghum fields, and converting annual cropland into shaded coffee agroforests offered both a new revenue stream and a partial escape from wildlife losses. The idea itself spread through social networks, with 40 percent of farmers crediting neighbors as the source of the idea and nearly half naming neighbors as their primary source of knowledge about how to grow coffee, a pattern that hints at powerful social contagion in land-use change.

The consequences of this upward shift for biodiversity were a central question of the study, because agroforestry can cut two very different ways for conservation. When coffee is established inside existing forests, management typically thins the canopy and reduces the structural complexity that forest-associated species depend on. When it is established on open land, however, it functions as a form of ecological restoration, adding tree cover, shade, carbon storage, soil fertility and erosion control to landscapes that previously had none. The fate of the novel coffee zone therefore hinged on a deceptively simple question: were the new agroforests carved out of remaining forest, or planted on open land?

The answer was overwhelmingly the latter. Seventy-three percent of the surveyed farmers had established their coffee agroforests on open land such as former cropland or grazing areas, a pattern the researchers attribute to the scarcity of forest in the novel range and to state ownership of most remaining forest, which limits farmers’ ability to convert it. Even more encouraging were the vegetation surveys. In 30-by-30-meter plots at each site, the team counted, identified and measured every tree and shrub with a trunk diameter of at least five centimeters, along with saplings. Forest-derived and open-land-derived agroforests showed statistically indistinguishable species richness, Shannon diversity and abundance for both adult woody plants and saplings, a finding that contradicts the common assumption that agroforests established on open land are biodiversity-poor imitations of the real thing.

There were differences, to be sure. Trees in forest-derived agroforests were thicker at breast height, reflecting the presence of old, mature individuals, and community composition differed subtly between the two types, with species such as Croton macrostachyus, Ficus sur and Millettia ferruginea more abundant in forest-derived plots and Vernonia amygdalina more common in open-land plots. But the most notable feature was the extensive overlap in composition. Unlike open-land agroforests elsewhere in the tropics, which are often dominated by a handful of non-native shade trees, the Ethiopian agroforests in the novel range were built from a diverse set of native species, either planted from seeds and saplings gathered in nearby forest or regenerated naturally from seed banks and adjacent woodland. Management choices, whether trees were planted, allowed to regenerate naturally, or both, left little imprint on the resulting communities, suggesting that native tree recruitment in this landscape is remarkably resilient.

The forward-looking part of the study may prove the most consequential. When the researchers asked the thirty neighboring farmers without coffee whether they planned to establish it, 87 percent said yes, and all but one of those planned to plant on open land rather than in forest. Every one of these prospective adopters expected a better income, and half anticipated no challenges at all beyond the well-known burden of high initial investment costs, which was also the only disadvantage cited by established farmers. The researchers identify two positive feedback mechanisms that should accelerate adoption. First, farmers imitate successful neighbors, so each new agroforest becomes a visible advertisement for the next. Second, a crop-raiding spiral may take hold: as more of the landscape converts to woody agroforestry, wild animals gain more shelter and raiding pressure on remaining annual crop fields intensifies, pushing still more farmers toward coffee.

The implications ripple outward from the Jimma highlands. Modeling studies have warned that 39 to 59 percent of the land currently suitable for Arabica coffee in Ethiopia could become unsuitable by the end of the century, threatening the livelihoods of millions of smallholders and the genetic reservoir of the crop itself. An upward migration of coffee farming has long been proposed as an adaptation pathway, and this study provides the first detailed empirical evidence that the migration is already underway, driven not by top-down policy but by farmers’ own accurate perceptions of a changing climate. Because the new agroforests are overwhelmingly established on open land and are rich in native woody species, the shift is expected to increase tree cover, biodiversity and ecosystem services, including carbon sequestration, shade, soil water regulation and erosion control, at higher elevations of the Afromontane landscape, an area that has historically suffered some of the region’s highest deforestation rates precisely because coffee agroforestry was absent there.

The authors are careful to note the caveats. The full landscape-scale consequences for forestation will require quantifying natural forest cover, mapping tenure rights and assessing how regulations are observed in practice, and complementary studies at the lower edge of the coffee range are needed to understand the complete regional picture. Hidden social costs, including increased household labor burdens and effects on children’s schooling documented in earlier work, also deserve attention. Yet the core finding stands as a rare piece of good news at the intersection of climate change, agriculture and conservation: in the birthplace of Arabica coffee, a warming world is nudging an entire farming system upslope, and the farms it leaves behind on the open hillsides are filling with native trees. What looks like retreat may, in this corner of Ethiopia, be a quiet expansion of the forest itself.

Subject of Research: Climate-driven upward shift of Arabica coffee agroforestry and its reforestation consequences in southwestern Ethiopia

Article Title: Climate-mediated upward shift of coffee agroforestry and associated reforestation in Arabica coffee’s native range

Article References: Abebe, F. B., Hylander, K., Nemomissa, S., Bekele, T., Zewdie, B., & Tack, A. J. M. (2026). Climate-mediated upward shift of coffee agroforestry and associated reforestation in Arabica coffee’s native range. Ambio. https://doi.org/10.1007/s13280-026-02470-3

Image Credits: AI Generated

DOI: 10.1007/s13280-026-02470-3

Keywords: climate change, coffee agroforestry, Arabica coffee, Ethiopia, Afromontane forests, reforestation, biodiversity, ecosystem services, smallholder farmers, elevational shift, crop raiding, tree cover