More than 17,000 tree species are at risk from rapid global change

Trees are fundamental to global biodiversity and ecosystem services. However, the accelerating pace of human activities during the Anthropocene is severely threatening tree diversity through habitat loss, fragmentation, degradation, and over-exploitation. This poses a significant risk of extinction or decline for numerous tree species, triggering cascading effects across trophic levels. Effective tree conservation requires an intensive, targeted approach, starting with accurate assessments of current threats. While the Global Tree Assessment (GTA) has provided valuable conservation statuses for most tree species, significant limitations remain, including a large number of Data Deficient species and the lack of consideration for recent temporal changes in threat intensity and exposure. This research aims to address these limitations by employing a data-driven approach to quantify recent changes in six major threats to trees: crop agriculture expansion, tree cover decline (overexploitation), urban built-up area expansion, deforestation, changes in burned area, and climate change (measured via temperature, vapor pressure deficit (VPD), and precipitation). The study uses species' extent of occurrence (EOO) to calculate exposure to threats, independent of species' sensitivity or adaptability. By analyzing the congruence between recent changes in threat exposure and IUCN conservation status categories, the study aims to prioritize species for expert-based IUCN Red List re-evaluation, ultimately strengthening conservation efforts.

Existing automated conservation status assessments typically do not incorporate recent temporal changes in threat exposure. Studies considering these temporal dynamics, such as projections of increased extinction risk for Chinese woody species, highlight the importance of changing threat landscapes. The Global Tree Assessment (GTA) identified nine threats to tree species, but 7,700 species remain Data Deficient. Furthermore, the impact of some threats, especially climate change, is difficult to diagnose, leading to potential underestimation of risk. The need for periodic expert-based IUCN Red List re-evaluations, coupled with challenges in assessing threats to rare, remote species, underscores the need for a data-driven approach to complement and improve these assessments. Previous research has highlighted the high exposure of global tree diversity to human pressures and identified conservation priorities for endangered trees facing multiple threats, emphasizing the urgency for improved threat assessment and prioritization.

The study used occurrence records from five open-access databases (GBIF, BIEN, DryFlor, RAINBIO, ALA) for 41,835 tree species. After data cleaning and standardization, rates of recent threat exposure were calculated for 32,090 species. Species' extents of occurrence (EOOs) were estimated using minimum convex polygons, refined by removing climatically unsuitable areas and water bodies. Six threats were analyzed using various datasets from Google Earth Engine (GEE), including cropland expansion (2003-2019), tree cover decline (2000-2015), built-up area expansion (2000-2020), deforestation (2000-2020), burned area change (2001-2020), and climate change variables (2000-2019) from CHELSA-BIOCLIM+. Rates of recent change were calculated differently for unidirectional (land-use changes) and bidirectional (fire and climate change) threats. Unidirectional changes were calculated as the percentage of EOO changing per year. For bidirectional changes, median-based linear models were used to determine rates of change. The congruence between these rates of change and IUCN Red List conservation status categories was then analyzed. Species exceeding the 95th percentile of a threat's rate of change were prioritized for IUCN Red List re-evaluation. Hotspots of highly exposed species were identified through mapping.

The analysis revealed a substantial underestimation of threats to tree species globally. Of the 32,090 species analyzed, 54.2% (17,393 species) showed major and increasing exposure to at least one threat. However, only 8.7% of these were listed as threatened on the IUCN Red List. The study highlights a significant discrepancy, especially for Data Deficient (57.8% of those in the study) and Not Evaluated (58.4%) species. The largest changes in threats were due to deforestation and tree cover decline. Maximum annual rates of change were identified for specific species, demonstrating the severity of threats. Lower rates of change were observed for cropland agriculture, urban development, and burned area. Climate change significantly impacted species' extents, with warming temperatures and increased drought being prevalent. Correlations between land-use changes were identified, with overlaps between tree cover decline and other threats such as climate change and deforestation. All IUCN conservation status groups were present across the entire exposure gradient for most threats, including extreme ends, highlighting the widespread impact of these threats and the inadequacy of current assessments. Species with small extents were more likely to be threatened or not evaluated, but threatened and vulnerable species of all extent sizes were found at extremes of exposure. Climate change rates intensified in 2010–2020 compared to the previous decade. The study identified 17,393 unique tree species (54.2% of those assessed) requiring prioritization for expert-based assessment due to significant threat exposure. Hotspots of highly exposed species were mainly located south of the Tropic of Cancer, in South America, Africa, and parts of Asia. The distribution of relative numbers of species exposed to high rates of recent global change showed different patterns than absolute numbers, with prominent hotspots in subtropical regions of China, the Canadian Arctic Archipelago and Northern Russia. Hotspots varied for each threat, but were mostly located in equatorial regions. Synergies between threats were also observed, with some being causal (e.g., agricultural expansion leading to decreased fire activity). A considerable number of species (22.3%) were highly exposed to multiple threats, with some species facing up to five out of six analyzed threats. Spatial overlays revealed areas where species are heavily exposed to multiple threats, primarily in Southeast Asia, West Africa, and parts of South America.

The findings demonstrate a significant global underestimation of threats to tree species. The mismatch between threat exposure and current IUCN Red List statuses suggests that many species classified as Not Threatened are at risk and need re-evaluation. The study highlights the underestimated impact of climate change on trees, especially considering the increasing evidence of forest dieback and reduced resilience. The unequal distribution of threats creates hotspots of highly exposed species, emphasizing the need for targeted conservation efforts. Synergies between threats further exacerbate the risks, highlighting the complexity of assessing and managing these threats. The study's methodology offers a valuable tool for prioritizing species for IUCN Red List re-evaluations, improving the efficiency and effectiveness of conservation actions and policy changes.

This study provides a comprehensive data-driven assessment of recent changes in anthropogenic threats to tree species, revealing a significant underestimation of risk in current IUCN Red List assessments. The identification of ~17,000 tree species experiencing increasing threat exposure underscores the urgent need for improved conservation strategies. The methodology employed offers a flexible, transparent approach to prioritize species for expert-based re-evaluation, enhancing the efficiency of conservation efforts. Future research should investigate the discrepancies between threat exposure and IUCN statuses, further refining the integration of data-driven approaches and expert knowledge in conservation assessments. The methodology is readily adaptable to other taxa, offering potential as an early warning system for biodiversity conservation under global change.

The study acknowledges limitations in data availability, particularly for some GTA-identified threats (livestock farming, energy production, wood plantations, invasive species), which could not be included due to insufficient suitable data. While the study uses a refined EOO to improve threat assessment, the method might still underestimate the true extent of some species, especially those with limited occurrence records. The method is limited in its ability to capture the impact of indirect or complex threats. The selected thresholds for prioritization were not literature-defined, and this may affect species ranking in future assessment. Additionally, the study's focus on recent change might not fully capture the cumulative impact of long-term, slow-onset threats.