コーヒー農林業システムにおける地上部バイオマスと炭素蓄積量のLiDARベース評価：異なる標高における気候緩和と小規模農家の生計への示唆

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:LiDARと混合効果モデルを組み合わせた農林業炭素計測の手法論と、農園スケールの較正必要性に関する実証的知見を提供する。

🏢実務担当者:コーヒー等農産物のサプライチェーン炭素会計や自然資本評価において、LiDARを活用した低コストMRVの可能性を検討できる。

🏛政策担当者:小規模農家の炭素市場・REDD+参加を促進するMRV政策設計において、LiDARベース手法の費用対効果と較正要件を考慮すべき。

Coffee agroforestry systems (CAS) mitigate climate change, but their structural complexity hinders accurate carbon quantification. We combined aerial LiDAR and field data to assess vegetation height, aboveground biomass (AGB), and carbon stocks across three altitudinal strata (1,277–1,710 m a.s.l.) in six CAS plots (comprising 1,559,858 LiDAR-derived observation cells) in northern Peru. Mean canopy height (2.47–8.79 m), AGB (1.55–40.59 Mg ha⁻¹), and carbon (0.73–19.08 Mg C ha⁻¹) varied widely; coefficients of variation often exceeded 100%, reflecting high heterogeneity. A mixed-effects logarithmic model (LiDAR-derived height as a fixed effect; farm as a random effect) revealed a significant positive height-biomass relationship (β₁ = 2.400; p < 0.001). The model's estimated parameters provide quantitative evidence of the importance of farm-level random effects. While the global fixed-effect slope for the height-biomass relationship was 2.400, the mixed-effects framework allowed this scaling rule to deviate by farm. For instance, the model quantified a distinct random slope of 3.065 for one specific farm. This represents a 27.7% increase in the rate of biomass accumulation per unit of canopy height compared to the global average. This quantitative divergence in allometric scaling across farms provides direct evidence that local conditions significantly alter biomass dynamics, justifying a random-effects structure over traditional fixed-effect models. Random effects captured local variations driven by management and species composition, with one farm showing distinct accumulation dynamics (β₁ = 3.065). Because of partial dependence between LiDAR metrics and field-calibrated biomass estimates, model performance reflects internal structural consistency rather than independent predictive validation. Consequently, LiDAR height should be interpreted as a robust structural indicator rather than an entirely independent indicator. Nevertheless, LiDAR-derived height reliably indicates AGB in CAS, though model transfer requires farm-scale calibration. This non-destructive, spatially explicit approach strengthens carbon monitoring in tropical agroforestry. By reducing the costs and uncertainties of Measurement, Reporting, and Verification (MRV), this workflow lowers the barriers for smallholder integration into carbon markets and REDD+ programs, directly linking accurate carbon accounting to improved livelihood incentives. .

• openalex https://doi.org/10.1016/j.tfp.2026.101489first seen 2026-10-09 04:39:02