Tropical forests are vital for global ecosystem services (ES), acting as carbon sinks and crucial biodiversity habitats. However, the economic consequences of climate change impacts on these services remain largely unexplored. This study focuses on Central America, a region with unique old-growth forests experiencing significant climate change pressures and land-use changes, including deforestation and illegal logging. The region's complex topography presents a challenge for climate impact modeling, requiring high-resolution analysis to accurately capture the nuanced effects of climate change on vegetation growth and distribution. Existing studies often lack the fine-scale resolution needed for accurate modeling and often segregate ecological and economic modeling of climate change impacts, thus hindering effective policy decisions. This study aims to bridge these research gaps by integrating high-resolution vegetation modeling with economic valuation of ecosystem services, thus providing critical insights for policymakers. Specifically, the study addresses four key research questions: 1) How will climate change affect climate regulation and habitat services? 2) What is the approximate economic value of these services, and what economic costs could arise from climate change? 3) How sensitive are these values to national price levels and temporal discounting? 4) Where are the hotspots of change that require immediate action?

The global importance of tropical forests for carbon sequestration and biodiversity is well-established. Afforestation efforts aim to offset carbon emissions, but deforestation continues, threatening biodiversity. Climate change affects vegetation, triggering biome shifts and impacting ES. The economic impacts on both commercial forestry and non-use values of natural forests are potentially severe. Central America, with its unique old-growth forests, experiences climate change impacts and could become a future hotspot. Growing land-use pressures exacerbate the issue. Previous studies highlight topography's influence on tropical forest communities and productivity, but large-scale studies often use coarse spatial resolutions due to limitations in high-resolution future climate scenarios and computational resources. Existing analyses on the economic impacts of climate change on Central American forest ES are limited, conducted at either local or continental scales, and often separating ecological and economic modeling. The need for translating ecological outcomes into monetary values is crucial for providing accessible results to decision-makers, raising awareness of ES losses, and relating them to economic indicators like GDP. This study seeks to address these limitations.

This study employs a two-pronged approach: 1) high-resolution vegetation modeling using the dynamic global vegetation model (DGVM) LPJ-GUESS at a 1 arc-second resolution, and 2) economic valuation of climate regulation and habitat services. The model simulates vegetation growth and dynamics from 1985–2100 under two global climate scenarios (GFDL-ESM4 and IPSL-CM6A-LR) and two socio-economic pathways (SSP1-2.6 and SSP3-7.0) to encompass both low and high climate forcing. Net ecosystem exchange (NEE) in t CO2 is used as an indicator for climate regulation, and biome stability serves as an indicator for habitat services. Economic ES values are calculated using social costs of carbon and benefit transfer from existing studies. Habitat is considered a supporting service, valued through its contributions to other ES. The sensitivity of economic estimates to regional price levels and discounting (2% or no discounting) is explored. Results are presented as maps showing stable regions and hotspots of change, economic estimates per country in relation to national GDPs, and hotspots of economic costs. The landforms approach within LPJ-GUESS improves the representation of topographic effects on climate (temperature, solar radiation, soil depth), but some artefacts from the original 0.5-degree resolution of the input precipitation data persist. The analysis primarily focuses on climate-driven changes; anthropogenic impacts such as land-use change are beyond the study's scope. Economic valuation uses cost-based valuation for climate regulation (social costs of carbon) and benefit transfer for habitat services. A global variant with uniform prices and a nationalized variant adjusted based on per capita GDP are considered. Biome stability is calculated by comparing the projected biomes for each year (2071-2100) against the reference period (1985-2014). Habitat service valuation utilizes a benefit transfer approach from a broad database of ecosystem service values, considering a variety of tropical ecosystems to reduce bias. Values are adjusted for inflation to 2015 and discounted (2% or not) for future periods.

Simulations across four climate scenarios (two GCMs and two SSPs) revealed regional trends in CO2 sequestration and biome stability. High biome stability was projected in areas consistent with current tropical rainforest distribution. Decreasing biome stability was observed in dry forest biomes and parts of the Yucatan peninsula. Hotspots where both CO2 sequestration and biome stability decreased covered 4.0–8.7% of the study area. SSP126 scenarios showed greater CO2 sequestration decreases, while SSP370 scenarios exhibited larger biome stability losses. Economic costs of reduced ES provision ranged from $29–313 billion/year for habitat and $1–65 billion/year for climate regulation. Discounting significantly affected habitat cost estimations, increasing them sixfold compared to scenarios without discounting. National price adjustments also influenced cost estimates; using uniform global prices yielded estimates approximately 1.6 times higher than those using nationally adjusted prices. The strongest climate forcing (SSP370) resulted in higher habitat costs, except for IPSL scenarios. Country-level analysis revealed that Belize, Nicaragua, and Honduras experienced the most substantial economic losses, with some scenarios predicting losses exceeding their national GDPs. Habitat costs generally exhibited right-skewed distributions, except for El Salvador and Belize. The consideration of discounting shifted habitat cost distributions toward normal distributions. Economic hotspots, areas with losses in both services, varied based on climate forcing and pricing approach. The most significant cost increases were observed in northwestern regions, the Yucatan peninsula, and montane regions under the global price/no discount scenario. The application of national prices changed the distribution of hotspots, primarily affecting higher-income economies.

The study successfully adapted LPJ-GUESS for Central American conditions, though some limitations remain due to precipitation data resolution. The valuation techniques used are common but have inherent uncertainties, addressed by considering a range of assumptions (discounting, global/national prices). The projected biome shifts are more conservative than some previous regional simulations, suggesting that a significant portion of current rainforests is at lower risk, although areas with lower stability, particularly montane forests, exhibit greater risk. Increased CO2 uptake under SSP370 compared to SSP126 is attributed to CO2 fertilization, although its magnitude remains uncertain. Decreasing biome stability poses significant threats to biodiversity. Projected ES changes represent substantial economic values, even the lowest estimate of $51 billion/year. Discounting, price level, and climate scenario magnitude all influence economic estimates. Habitat costs generally exceeded climate regulation costs. Discounting had opposite effects on carbon costs and habitat values. The study reveals unequal cost distribution among countries, potentially exacerbating existing inequalities. Global uniform pricing better reflects ecological reality, while national pricing reflects actual payment likelihoods.

This study provides crucial insights into climate change effects on Central American forest ecosystems and their economic implications. The integration of high-resolution vegetation modeling and economic valuation offers valuable information for conservation planning. While uncertainties remain, the findings highlight substantial potential economic losses and underscore the need for integrative approaches in payment for ecosystem services (PES) schemes that consider both carbon sequestration and biodiversity conservation. Future research should refine model inputs and explore alternative valuation techniques to enhance the accuracy of these projections.

The study primarily focused on climate-driven changes, omitting anthropogenic influences like land-use change. The valuation techniques (cost-based valuation for climate regulation, benefit transfer for habitat services) have inherent uncertainties. The resolution of the precipitation data used in the model limited the accuracy of the representation of spatial heterogeneity. There is a reliance on benefit transfer for habitat valuation, which may contain biases.