Search

Soil respiration (Rs), the soil‐to‐atmosphere CO2flux produced by microbes and plant roots, is a critical but uncertain component of the global carbon cycle. Our current understanding of the variability and dynamics is limited by the coarse spatial resolution of existing estimates. We predicted annual Rs and associated uncertainty across the world at 1‐km resolution using a quantile regression forest algorithm trained with observations from the global Soil Respiration Database spanning from 1961 to 2011. This model yielded a global annual Rs estimate of 87.9 Pg C/year with an associated global uncertainty of 18.6 (mean absolute error) and 40.4 (root mean square error) Pg C/year. The estimated annual heterotrophic respiration (Rh), derived from empirical relationships with Rs, was 49.7 Pg C/year over the same period. Predicted Rs rates and associated uncertainty varied widely across vegetation types, with the greatest predicted rates of Rs in evergreen broadleaf forests (accounting for 20.9% of global Rs). The greatest prediction uncertainties were in northern latitudes and arid to semiarid ecosystems, suggesting that these areas should be targeted in future measurement campaigns. This study provides predictions of Rs (and associated prediction uncertainty) at unprecedentedly high spatial resolution across the globe that could help constrain local‐to‐global process‐based models. Furthermore, it provides insights into the large variability of Rs and Rh across vegetation classes and identifies regions and vegetation types with poor model performance that should be prioritized for future data collection. Soils emit large amounts of carbon dioxide to the atmosphere every year via the process of soil respiration, which greatly exceeds emissions from human sources. However, rates of soil respiration are highly variable in space, which limits our ability to balance global carbon budgets and forecast climate change. We used a novel application of a machine learning approach to predict annual rates of soil respiration at high resolution (1 km) across the globe and examined spatial patterns of the associated uncertainty of these predictions. Predictions were made based on how observations of soil respiration were related to climate (annual temperature and annual and seasonal precipitation) and vegetation information. Predicted annual soil respiration and prediction uncertainty varied across ecosystem types and regions. Our predictions suggest that evergreen tropical forests dominate global annual soil respiration emissions. Dryland, wetland, and cold ecosystems had the highest associated prediction uncertainties, suggesting that future soil respiration measurements would be especially useful in these areas. The high spatial resolution of our predictions will help researchers studying the carbon cycle at local to global scales. A high spatial resolution machine learning approach was used for estimating soil respiration across the worldPredictions of soil respiration varied widely across ecosystem classes, and allowed for suitable estimates of heterotrophic respirationAssociated prediction uncertainty was highest in high latitudes and data scarce regions, which should be targets for future measurements

Do the benefits of data sharing outweigh its perceived costs? This is a critical question, and one with the potential to change culture and behavior. Dai et al. (2018, https://doi.org/10.1002/2017JG004277) examined how data sharing is related to scientific impact in the field of eddy covariance (EC) and found that data sharers are disproportionately high‐impact researchers, and vice versa; they also note strong regional differences in EC data‐sharing norms. The current policies and restrictions of EC journals and repositories are highly uneven. Incentivizing data sharing and enhancing computational reproducibility are critical next steps for EC, ecology, and science more broadly. The raw data that scientists generate—typically from experiments and/or observations—have traditionally been kept private, but research funders, journals, and scientists themselves are pushing for change, arguing that taxpayer‐funded data should be “open” (available to all) on both moral and practical grounds. Do influential scientists share more, and does data sharing lead to higher impact? This commentary discusses a recent paper examining these issues in a particular field of ecology and earth system science. Current policies and restrictions are highly uneven, and it is critical to provide proper credit for scientists who freely share their data. Data sharers are disproportionately high‐impact researchers, but the causality is uncertainInstitutional support for data sharing in eddy covariance research is unevenIncentivizing data sharing and enhancing computational reproducibility are critical next steps

Upland forest soils are typically major atmospheric carbon dioxide (CO2) sources and methane (CH4) sinks, but the contributions of root and microbial processes, as well as their separate temporal responses to environmental change, remain poorly understood. This 2‐year study was conducted in a temperate, deciduous forest located on the Chesapeake Bay in Maryland, USA. We used temporal CO2and CH4flux measurements, exclusion‐source partitioning, and an ecosystem‐scale flooding experiment to understand how carbon (C) fluxes, and their root and microbial sources, respond to seasonal and episodic environmental change. We show that the root‐and‐rhizosphere component of soil CO2and CH4flux is significant and that its dependence on soil temperature and volumetric water content (VWC) influences soil C dynamics at seasonal timescales. Experimental flooding shows that CO2and CH4flux responses to episodic moisture change were driven by suppression of soil heterotrophs, while root respiration did not respond to transient hydrologic disturbance. Methane uptake responded strongly to episodic inundation, reinforcing the important role of soil moisture in the short‐term control of the forest soil CH4sink. However, despite the clear seasonality of CH4uptake, as well as its strong response to short‐term experimental inundation, temperature and VWC were weak predictors of CH4uptake at a seasonal timescale. We suggest that CH4consumption in the long‐term may be determined by vegetation, nutrients, microbial communities, or other factors correlated with seasonal changes. Our results indicate that root and microbial sources of both CO2and CH4flux respond differently in timing and magnitude to seasonal and episodic environmental change. The well‐drained soils of temperate forests are major atmospheric carbon dioxide (CO2) sources and methane (CH4) sinks and are in turn an important component of the global forest C sink. However, increasing precipitation intensity and extreme‐event frequency may amplify CO2release and diminish CH4uptake from temperate forest soils, with poorly constrained feedback implications for Earth's forest C cycle. Separating the individual components (roots, fungi, and bacteria) of C flux is crucial to identify the factors controlling forest soil C dynamics under changing environments. In this study, we show that root‐derived CO2and CH4fluxes are sensitive to temperature and moisture content at seasonal timescales but do not respond to episodic environmental changes, such as extreme flooding. In fact, roots acted as a response buffer and enhanced the resilience of temperate forest soil respiration to a short‐term hydrologic disturbance. The muted response of roots may have been because surface roots were not saturated long enough to trigger a physiological response and/or that deeper roots remained unsaturated, mitigating the impacts of flooding on surface roots. Predicting such responses and tipping points requires a greater mechanistic understanding of the cumulative impacts of extreme precipitation and flooding on forest ecology and biogeochemistry. Root dependence on soil temperature and moisture content influences the seasonal dynamics of temperate forest soil CO2and CH4fluxCO2and CH4flux responses to episodic flooding were driven by suppression of soil heterotrophs, not roots and their associated rhizosphereRoots acted as a response buffer and enhanced the resilience of temperate forest soil respiration to a short‐term hydrologic disturbance Root dependence on soil temperature and moisture content influences the seasonal dynamics of temperate forest soil CO2and CH4flux CO2and CH4flux responses to episodic flooding were driven by suppression of soil heterotrophs, not roots and their associated rhizosphere Roots acted as a response buffer and enhanced the resilience of temperate forest soil respiration to a short‐term hydrologic disturbance

Boreal forests store 30% of the world's terrestrial carbon (C). Consequently, climate change mediated alterations in the boreal forest fire regime can have a significant impact on the global C budget. Here we synthesize the effects of forest fires on the stocks and recovery rates of C in boreal forests using 368 plots from 16 long‐term (≥100 year) fire chronosequences distributed throughout the boreal zone. Forest fires led to a decrease in total C stocks (excluding mineral soil) by an average of 60% (range from <10% to >80%), which was primarily a result of C stock declines in the living trees and soil organic layer. Total C stocks increased with time since fire largely following a sigmoidal shape Gompertz function, with an average asymptote of 8.1 kg C m−2. Total C stocks accumulated at a rate of 2–60 g m−2yr−1during the first 100 years. Potential evapotranspiration (PET) was identified as a significant driver of C stocks and their post‐fire recovery, likely because it integrates temperature, radiation, and the length of the growing season. If the fire return interval shortens to ≤100 years in the future, our findings indicate that many boreal forests will be prevented from reaching their full C storage potential. However, our results also suggest that climate warming‐induced increases in PET may speed up the post‐fire recovery of C stocks. Potential evapotranspiration was a significant driver of carbon stocks and their post‐fire recoveryCarbon stocks accumulated at a rate of 2–60 g m−2yr−1during the first 100 years after forest fireIf the fire return interval shortens to ≤100 years, many boreal forests will be prevented from reaching their full carbon storage potential

Initial land cover distribution varies among Earth system models, an uncertainty in initial conditions that can substantially affect carbon and climate projections. We use the integrated Earth System Model to show that a 3.9 M km2difference in 2005 global forest area (9–14% of total forest area) generates uncertainties in initial atmospheric CO2concentration, terrestrial carbon, and local temperature that propagate through a future simulation following the Representative Concentration Pathway 4.5. By 2095, the initial 6 ppmv uncertainty range increases to 9 ppmv and the initial 26 PgC uncertainty range in terrestrial carbon increases to 33 PgC. The initial uncertainty range in annual average local temperature of −0.74 to 0.96 °C persists throughout the future simulation, with a seasonal maximum in Dec‐Jan‐Feb. These results highlight the importance of accurately characterizing historical land use and land cover to reduce overall initial condition uncertainty. International modeling efforts aim to understand global change and its impacts on humans and the environment. Modeling how human activities change vegetation (e.g., forest to cropland), and subsequently the greater environment, is difficult and highly uncertain, yet crucial to understanding impacts of global change. Here we estimate that an uncertainty range in year 2005 global forest cover of 3.9 M km2(9–14% of the total) generates an uncertainty of 6 ppmv in atmospheric carbon dioxide concentration that increases to 9 ppmv by 2095 in a model experiment that aims to restrict global warming forcing to 4.5 W/m2. Furthermore, local surface temperature uncertainties range from −0.57 to 0.72 °C and persist throughout the 21stcentury. Future studies of global change and its impacts on humans and the environment need to quantify and reduce such land cover uncertainties. Historical land conversion uncertainty can dramatically influence results of subsequent future projections via initial condition uncertaintyA 3.9x106km2difference in 2005 forest area (9–14% of total forest) increases initial CO2uncertainty by 50% from 2005 to 2095The associated initial uncertainty in local annual average temperature persists with a 2005–2094 average range of −0.57 to +0.72 °C