In the 45 years after legislation of the Clean Air Act, there has been tremendous progress in reducing acidic air pollutants in the eastern United States, yet limited evidence exists that cleaner air has improved forest health. Here, we investigate the influence of recent environmental changes on the growth and physiology of red spruce (Picea rubens Sarg.) trees, a key indicator species of forest health, spanning three locations along a 100 km transect in the Central Appalachian Mountains. We incorporated a multiproxy approach using 75-year tree ring chronologies of basal tree growth, carbon isotope discrimination (∆ C, a proxy for leaf gas exchange), and δ N (a proxy for ecosystem N status) to examine tree and ecosystem level responses to environmental change. Results reveal the two most important factors driving increased tree growth since ca. 1989 are reductions in acidic sulfur pollution and increases in atmospheric CO , while reductions in pollutant emissions of NO and warmer springs played smaller, but significant roles. Tree ring ∆ C signatures increased significantly since 1989, concurrently with significant declines in tree ring δ N signatures. These isotope chronologies provide strong evidence that simultaneous changes in C and N cycling, including greater photosynthesis and stomatal conductance of trees and increases in ecosystem N retention, were related to recent increases in red spruce tree growth and are consequential to ecosystem recovery from acidic pollution. Intrinsic water use efficiency (iWUE) of the red spruce trees increased by ~51% across the 75-year chronology, and was driven by changes in atmospheric CO and acid pollution, but iWUE was not linked to recent increases in tree growth. This study documents the complex environmental interactions that have contributed to the recovery of red spruce forest ecosystems from pervasive acidic air pollution beginning in 1989, about 15 years after acidic pollutants started to decline in the United States.

We conducted a meta-analysis of carbon and oxygen isotopes from tree ring chronologies representing 34 species across 10 biomes to better understand the environmental drivers and physiological mechanisms leading to historical changes in tree intrinsic water use efficiency (iWUE), or the ratio of net photosynthesis (Anet) to stomatal conductance (gs), over the last century. We show a ∼40% increase in tree iWUE globally since 1901, coinciding with a ∼34% increase in atmospheric CO2 (Ca), although mean iWUE, and the rates of increase, varied across biomes and leaf and wood functional types. While Ca was a dominant environmental driver of iWUE, the effects of increasing Ca were modulated either positively or negatively by climate, including vapor pressure deficit (VPD), temperature, and precipitation, and by leaf and wood functional types. A dual carbon–oxygen isotope approach revealed that increases in Anet dominated the observed increased iWUE in ∼83% of examined cases, supporting recent reports of global increases in Anet, whereas reductions in gs occurred in the remaining ∼17%. This meta-analysis provides a strong process-based framework for predicting changes in tree carbon gain and water loss across biomes and across wood and leaf functional types, and the interactions between Ca and other environmental factors have important implications for the coupled carbon–hydrologic cycles under future climate. Our results furthermore challenge the idea of widespread reductions in gs as the major driver of increasing tree iWUE and will better inform Earth system models regarding the role of trees in the global carbon and water cycles.

Carbon use efficiency (CUE) represents how efficient a plant is at translating carbon gains through gross primary productivity (GPP) into net primary productivity (NPP) after respiratory costs (R a ). CUE varies across space with climate and species composition, but how CUE will respond to climate change is largely unknown due to uncertainty in R a at novel high temperatures. We use a plant physiological model validated against global CUE observations and LIDAR vegetation canopy height data and find that model-predicted decreases in CUE are diagnostic of transitions from forests to shrubland at dry range edges. Under future climate scenarios, we show mean growing season CUE increases in core forested areas, but forest extent decreases at dry range edges, with substantial uncertainty in absolute CUE due to uncertainty in R a . Our results highlight that future forest resilience is nuanced and controlled by multiple competing mechanisms.

No abstract

Plants transfer, on average, 0.33 Pg carbon from the atmosphere to the biosphere each day, removing nearly one-third of anthropogenic CO 2 emissions annually (Friedlingstein et al., 2020). This enormous transfer of CO 2 is accompanied by similarly large, unavoidable losses of H 2 O vapor (Schlesinger & Jasechko, 2014), which plants have evolved to minimize through a combination of morphological (e.g. waxy cuticle, sunken stomates) (Wang et al., 2022) and physiological strategies (e.g. stomatal aperture control) (Ainsworth & Rogers, 2007; Cowan & Farquhar, 1977). While plant morphological traits are relatively static and change over longer time scales (Franks & Beerling, 2009, but see Pérez-Ramos et al., 2019), physiological processes controlling leaf gas exchange (e.g. net photosynthesis, A net ; stomatal conductance to H 2 O, g s ) are highly dynamic and rapidly respond (e.g. in seconds to minutes) to changes in environmental stimuli such as light, temperature, aridity, and atmospheric