The characterisation of functional-trait dynamics of vegetation from remotely sensed data complements the structural characterisation of ecosystems. In this study we characterised for the first time the spatial heterogeneity of the intra-annual dynamics of the fraction of absorbed photosynthetically active radiation (FAPAR) as a functional trait of the vegetation in Prades Mediterranean forest in Catalonia, Spain. FAPAR was derived from the Multispectral Instrument (MSI) on the Sentinel-2 satellite and validated by comparison with the ground measurements acquired in June 2017 at the annual peak of vegetation activity. The validation results showed that most of points were distributed along the 1:1 line, with no bias nor scattering: R² = 0.93, p < 0.05; with a root mean square error of 0.03 FAPAR (4.3%). We classified the study area into nine vegetation groups with different dynamics of FAPAR using a methodology that is objective and repeatable over time. This functional classification based on the annual magnitude (FAPAR-M) and the seasonality (FAPAR-CV) from the data on one year (2016–2017) complements structural classifications. The internal heterogeneity of the FAPAR dynamics in each land-cover type is attributed to the environmental and to the specific species composition variability. A spatial autoregressive (SAR) model for the main type of land cover, evergreen holm oak forest (Quercus ilex), indicated that topographic aspect, slope, height, and the topographic aspect x slope interaction accounted for most of the spatial heterogeneity of the functional trait FAPAR-M, thus improving our understanding of the explanatory factors of the annual absorption of photosynthetically active radiation by the vegetation canopy for this ecosystem.

Protected areas (PAs) in South America play a crucial role in safeguarding biodiversity and ecosystem services. However, the widespread presence of roads within PAs poses a major yet largely overlooked threat to their ecological integrity. In this study, we used the most recent and consistent global roads dataset to evaluate the extent and spatial distribution of road networks within South American PAs. Our analysis revealed approximately 150,000 km of roads—equivalent to 3.5 times the Earth’s circumference. This extensive network exacerbates habitat fragmentation, increasing the number of PA patches from 1,230 to over 43,000, with 75% of these fragments smaller than 1 km². Consequently, seven out of nine terrestrial biomes on the continent have their protected interiors, on average, less than 4 km from the nearest road or unprotected surroundings. Roads infiltrate PAs through multiple entry points (13 per PA, on average), with 36% classified as highways or primary roads facilitating rapid transit and 20% paved, further amplifying their vulnerability to potential anthropogenic stressors. Only remote regions, such as flooded grasslands and savannas, Sub-Antarctic forests, and large areas of the Amazon basin, exhibit low road presence within PAs. Immediate conservation action is essential, with a focus on transparent road monitoring and management to maintain the ecological functions of these vital landscapes and ensure the long-term conservation of nature.

Understanding vegetation responses to climate variability is essential for assessing long-term ecosystem dynamics. Leaf Area Index (LAI) is a widely used variable to characterise vegetation state and productivity. However, attributing observed global LAI trends to specific climatic drivers remains challenging due to non-linear interactions, strong spatial heterogeneity, and scale-dependent processes.This study is conducted within the framework of the PROFECIA project, which aims to improve the monitoring and interpretation of vegetation responses to climate change by combining remote sensing observations and artificial intelligence techniques. We analyse global LAI trends over the period 1982–2022 using the GEOV2-AVHRR long-term satellite record and examine their relationship with trends in key climatic variables obtained from the ERA5 reanalysis, including temperature, precipitation, radiation, and several indicators of water availability and drought conditions. All trends are computed consistently over the 1982-2022 temporal record to ensure a homogeneous assessment of long-term vegetation–climate relationships at the global scale.The vegetation–climate relationships are modelled using a suite of machine learning approaches, including tree-based methods and neural networks, designed to capture non-linear responses across diverse climatic and ecological conditions. Particular emphasis is placed on the role of the training strategy: different spatio-temporal sampling schemes are evaluated to assess their impact on model performance, robustness, and generalisation capability when analysing long-term trends at the global scale.To move beyond purely predictive modelling, the study systematically applies explainable artificial intelligence (XAI) techniques to interpret the trained models. Methods such as SHAP-based attribution and partial dependence analyses are used to quantify the relative contribution of individual climatic drivers to observed LAI trends and to examine how these contributions vary across regions and time periods.Overall, this work highlights the importance of combining robust machine learning training strategies with interpretability tools to improve the attribution of long-term vegetation trends to climatic drivers, providing new insights into global vegetation–climate interactions over the last four decades.

Forests are under increasing pressure from climate change, land-use transformation, and human activities, threatening their capacity to regulate climate, store carbon, and sustain biodiversity. Detecting long-term changes in forest canopy structure and productivity is therefore essential. Leaf Area Index (LAI) is a widely used proxy of vegetation structure and function and serves as a key indicator of forest productivity. It captures greening associated with canopy densification and potential carbon uptake, and browning linked to forest degradation, disturbance, or loss. However, global assessments of forest greening and browning and their spatial determinants remain limited.Here, we present a global analysis of forest LAI trends over the last ~25 years using the High-Quality Reprocessed MODIS LAI dataset (HiQ-LAI; 8-day, 500 m; 2000–2024). The analysis is restricted to pure forest pixels derived from Hansen’s 30 m global forest cover map. Trends were quantified using Sen’s slope estimator and their significance assessed with the Mann–Kendall test. Globally, 21 % of forest areas exhibited significant greening, while 8 % showed browning, revealing strong regional contrasts. Browning patterns in tropical and subtropical forests are predominantly associated with land-use change, whereas in boreal regions they are largely driven by fire disturbances. In contrast, greening hotspots extend beyond climatic and CO₂ fertilization effects and strongly overlap with intensively managed forests, including plantation-dominated regions in China and Europe.Our findings demonstrate that land-use change, forest management, and disturbance regimes are key spatial determinants of observed forest greening and browning patterns. However, these processes remain underrepresented in many Earth system models. By providing a spatially explicit global baseline, this study supports improved representation of human and disturbance processes in climate–vegetation modelling and informs conservation, climate mitigation, and sustainable forest management strategies under accelerating global change.