Drought‐induced tree mortality is expected to increase globally due to climate change, with profound implications for forest composition, function and global climate feedbacks. How drought is experienced by different species is thought to depend fundamentally on where they access water vertically below‐ground, but this remains untracked so far due to the difficulty of measuring water availability at depths at which plants access water (few to several tens of metres), the broad temporal scales at which droughts at those depths unfold (seasonal to decadal), and the difficulty in linking these patterns to forest‐wide species‐specific demographic responses. We address this problem through a new eco‐hydrological framework: we used a hydrological model to estimate below‐ground water availability by depth over a period of two decades that included a multi‐year drought. Given this water availability scenario and 20 year long‐records of species‐specific growth patterns, we inversely estimated the relative depths at which 12 common species in the forest accessed water via a model of water stress. Finally, we tested whether our estimates of species relative uptake depths predicted mortality in the multi‐year drought. The hydrological model revealed clear below‐ground niches as precipitation was decoupled from water availability by depth at multi‐annual scale. Species partitioned the hydrological niche by diverging in their uptake depths and so in the same forest stand, different species experienced very different drought patterns, resulting in clear differences in species‐specific growth. Finally, species relative water uptake depths predicted species mortality patterns after the multi‐year drought. Species that our method ranked as relying on deeper water were the ones that had suffered from greater mortality, as the zone from which they access water took longer to recharge after depletion. Synthesis. This research changes our understanding of how hydrological niches operate for trees, with a trade‐off between realized growth potential and survival under drought with decadal scale return time. The eco‐hydrological framework highlights the importance of species‐specific below‐ground strategies in predicting forest response to drought. Applying this framework more broadly may help us better understand species coexistence in diverse forest communities and improve mechanistic predictions of forests productivity and compositional change under future climate.

Fires determine vegetation patterns, impact human societies, and are a part of complex feedbacks into the global climate system. Empirical and process-based models differ in their scale and mechanistic assumptions, giving divergent predictions of fire drivers and extent. Although humans have historically used and managed fires, the current role of anthropogenic drivers of fires remains less quantified. Whereas patterns in fire–climate interactions are consistent across the globe, fire–human–vegetation relationships vary strongly by region. Taking a data-driven approach, we use an artificial neural network to learn region-specific relationships between fire and its socio-environmental drivers across the globe. As a result, our models achieve higher predictability as compared to many state-of-the-art fire models, with global spatial correlation of 0.92, monthly temporal correlation of 0.76, interannual correlation of 0.69, and grid-cell level correlation of 0.60, between predicted and observed burned area. Given the current socio-anthropogenic conditions, Equatorial Asia, southern Africa, and Australia show a strong sensitivity of burned area to temperature whereas northern Africa shows a strong negative sensitivity. Overall, forests and shrublands show a stronger sensitivity of burned area to temperature compared to savannas, potentially weakening their status as carbon sinks under future climate-change scenarios.

Overlapping habitats and sharing of resources between elephants and people has led to intense elephant‐human conflicts, especially crop depredation by elephants, across elephant‐range countries. While raiding agricultural crops, elephants face numerous threats from people through chase, injury and the risk of death which could enhance the associated energetic costs, ultimately elevating their stress levels. We hypothesized that crop‐raiders (in the human‐production habitat) would show higher faecal glucocorticoid metabolite ( fGCM ) levels, a proxy of stress‐response, as compared to nonraiders (in protected forests). To study this, 208 faecal samples were collected from crop‐raiding elephants in a human‐production habitat and 394 samples from nonraiding elephants in protected forests during 2013 and 2015. Contrary to our expectation, fGCM levels were significantly higher in nonraiding than in crop‐raiding elephants of both sexes. As one of the possible factors for lower fGCM in elephants inhabiting the human‐production habitat, the influence of benefits obtained from foraging here was assessed. For this, the difference in vegetation greenness (standing biomass) between the human‐production habitat and the protected forests was analysed from remotely‐sensed Normalized Differential Vegetation Index ( NDVI ), and further confirmed by measuring dietary quality (faecal nitrogen (N) content and C:N ratio as proxies for crude protein). Interestingly, higher NDVI values (greater biomass availability), higher N content and lower faecal C:N ratio (indicating higher protein content in the diet) of elephants in the human‐production habitat suggested enhanced nutritional levels here as compared to protected forests. Further, there were significant correlations between faecal C:N ratio (positive) or N content (negative) and fGCM levels. These findings suggest that crop‐raiding comes with the benefits of a superior quality diet which may help in reducing human‐induced stress‐response in elephants inhabiting or foraging within human‐production habitats.