Long-term satellite monitoring reveals contrasting trajectories of Macaronesian laurel forests across oceanic islands
Year: 2026
Type: article
Abstract: Macaronesian laurel forests are relict, cloud-dependent ecosystems of high conservation value, currently restricted to fragmented island patches and threatened by climate change, disturbance and land-use legacies. However, spatially explicit assessments of recent forest-condition trajectories remain scarce at multi-archipelago scales. Here, we assessed long-term spatial and temporal trajectories of laurel forests across Madeira and the Canary Islands from 2000 to 2025 and evaluated their geographical, disturbance and climatic context. Annual MaxEnt models generated a habitat suitability index (HSI) from MODIS time-series predictors, interpreted here as a remotely sensed indicator of habitat condition. Temporal trends within current forest stands were quantified using Mann–Kendall tests and Sen’s slope, locally benchmarked against temperature and precipitation series and related to elevation, forest loss and island using Random Forest regression. Results revealed marked spatial heterogeneity and no uniform regional trajectory. EVI (Enhanced Vegetation Index) was the dominant predictor throughout most of the time series, while negative trends were particularly evident in parts of Madeira, La Gomera and Tenerife. Local climate series in selected declining areas showed significant warming of 0.019–0.036°C yr⁻¹, whereas precipitation generally declined. The Random Forest yielded an OOB-R² of 0.454 and an OOB-RMSE of 0.00054, interpreted as measures of internal model fit. Island identity showed the highest variable importance, indicating substantial between-island heterogeneity, followed by forest loss and elevation, which showed predominantly negative and non-linear partial responses. Overall, recent laurel-forest trajectories were strongly island-specific, highlighting the value of long-term satellite monitoring for identifying areas requiring finer-resolution assessment, field validation and targeted conservation action.
Source: Trees Forests and People
Institutions: Universidad de La Laguna, Instituto Pirenaico de Ecología
Language: English
Subfield: Ecology
Field: Environmental Science
Domain: Physical Sciences
SDGs: Life on land, Climate action
Open Access status: gold
Estimated APC paid: $1,680
OpenAlex ID: W7223897232