Extreme drought events can be considered as representations of the upper tails of climate variability distributions, whose probability is expected to increase due to trends in global climate change. These drought events can transform forest ecosystem structure and function if they exceed tree survival thresholds. As such, regional-scale tree mortality has the potential to impact feedbacks between the biosphere and earth system processes by affecting carbon dynamics, landscape surface properties, and watershed hydrology. For example, a massive pine die-off across western Canada caused the forest to shift from a C sink to a C source. Recently, regional-scale, heat- and drought-induced tree mortality has been documented around the world. Anticipating ecosystem impacts of global change requires investigation of the temperature sensitivity, patterns, and physiological mechanism of tree drought mortality. Quantifying potential impacts of tree mortality on earth system processes requires research into mortality effects on carbon, energy, and water budgets at regional levels. Here we highlight recent research, and identify key uncertainties and research needs to predict the occurrence and anticipate the impacts of global-change-mediated, drought-induced tree mortality. We focus on pinyon pine mortality in the southwest USA as a detailed case study.  
Results/Conclusions

Combined results from observations and recent experiments reveal that drought-induced tree mortality is highly temperature sensitive. The 2002/2003 pinyon pine die-off in the southwest US occurred after an unusually warm drought that was less dry than previous cooler droughts that did not cause widespread mortality. Experiments with transplanted pinyon trees subjected to simulated drought demonstrated a causal link, that elevated temperatures (+4.3°C) reduced survival time by 28%. Comparing co-occurring pinyon and ponderosa pine species mortality responses to this recent warm drought, reveals potential alternative outcomes of global change, which include ecotone shifts and range-wide population crashes. Experimental investigation of the mechanism of mortality in pinyon pine suggests carbon starvation (whereby stomatal closure reduced water loss and prevented photosynthesis while respiration demand continued) and possibly impeded carbon transport via lack of phloem function, but not hydraulic failure of water transport. Despite this work, the uncertainty around mortality responses still limits our ability to predict the likelihood and anticipate the impacts of tree die-off. Studies are needed that explore tree death physiology for a wide variety of functional types, connect patterns of mortality with climate events, and quantify the impacts on carbon, energy, and water flux.