Rising atmospheric CO2 concentrations, primarily from fossil fuel combustion, are a major driver of climate change. A significant portion of these emissions is absorbed by the terrestrial biosphere, acting as a carbon sink and slowing global warming. However, the strength and distribution of this terrestrial carbon sink are highly uncertain, particularly across different ecosystems and time scales. The seasonal oscillation of atmospheric CO2 in the Northern Hemisphere has increased substantially since the 1960s, linked to intensified photosynthetic activity in northern ecosystems. While initially, warming appeared to enhance carbon uptake, recent evidence suggests a weakening or reversal of this relationship in some regions. This study focuses on the U.S. Corn Belt, a highly productive agricultural region at northern mid-latitudes, to investigate the impact of warming on carbon exchange dynamics, acknowledging the limited research on this crucial area compared to tropical and boreal ecosystems. Previous studies have indicated that agricultural intensification, particularly corn production in the Midwest, has significantly influenced the Northern Hemisphere's CO2 seasonal cycle. However, uncertainties remain regarding the magnitude of this influence and the response of these heterogeneous ecosystems to future warming. The researchers aim to quantify the sensitivity of net CO2 exchange to interannual temperature variations, distinguishing between croplands and natural ecosystems, to project future CO2 sequestration under various warming scenarios.

Existing literature highlights the enhanced seasonal CO2 exchange due to amplified plant productivity in northern ecosystems, supported by observations of increased plant growth in high latitudes. However, studies have noted a recent shift in Eurasian boreal forest greening response possibly associated with warmer and drier summers. While agricultural intensification has been implicated as a major driver of changes in the CO2 seasonal cycle in the Northern Hemisphere, accounting for a substantial portion of the enhanced carbon exchange, uncertainties persist regarding the overall strength of this effect and future climate responses. Studies have explored the impact of warming on temperate forest ecosystems, revealing altered phenology and increased net primary productivity in spring and autumn, suggesting enhanced carbon sequestration. However, the role of northern temperate croplands in the context of the global CO2 seasonal cycle is comparatively less understood.

This study leverages a decadal record (2007-2019) of direct boundary layer CO2 measurements from a tall tower in southern Minnesota (KCMP), representative of the Corn Belt. The data were compared with long-term atmospheric CO2 data from two other Midwest sites (LEF and WBI) from NOAA's Global Greenhouse Gas Reference Network. The researchers used a wavelet transform to extract the long-term CO2 growth rate and detrended CO2 seasonal cycle. A cropland fraction (fcs) was defined as the ratio of corn and soybean land area to the total land area. The hourly CO2 concentrations at KCMP were sampled based on wind direction to distinguish between areas with higher and lower cropland fractions. The long-term growth rate of CO2 was compared across the three tower sites and a continental background site (NWR) to assess regional influences. The CO2 seasonal amplitude and drawdown were analyzed, correlating these metrics with crop production data. Net ecosystem exchange (NEE) data from CarbonTracker 2019 were used to quantify the sensitivity of net CO2 exchange to interannual temperature variations. A multiple linear regression (MLR) was used to determine the sensitivity of ΔCO2 (first-time derivative of the CO2 time series) and NEE anomalies to temperature, water availability (3-month cumulative precipitation), and radiation. A panel data model, integrating climate and NEE anomalies from the three tower sites, was employed to estimate biome-specific (croplands and natural ecosystems) temperature sensitivities. Finally, projected climate data from CMIP5 under RCP4.5 and RCP8.5 scenarios were used to predict future impacts on the CO2 seasonal cycle and net CO2 uptake in the Corn Belt.

The analysis revealed significant spatial gradients in atmospheric CO2 concentrations, strongly influenced by the varying cropland fractions across the study area. While agricultural intensification is considered a crucial driver of increased CO2 seasonal amplitude at decadal to multi-decadal scales, this study demonstrates a decoupling between crop yields and CO2 exchange intensity at the interannual scale. The CO2 seasonal amplitude showed an increasing trend at KCMP and WBI, though not statistically significant. A positive temperature sensitivity of net CO2 exchange (i.e., higher temperatures reduced net CO2 uptake) was observed in July and August for both croplands and natural ecosystems. This positive sensitivity was more pronounced in dry summers, suggesting the importance of water stress alongside temperature. Conversely, a negative temperature sensitivity was observed in June for croplands, possibly reflecting the positive feedback between crop canopy development and photosynthetic capacity during the early vegetative stage. Natural ecosystems showed a weaker response to temperature variations during spring and early summer. Analysis using a panel data model revealed that croplands played a dominant role in the negative NEE temperature sensitivity in June, while both croplands and natural ecosystems showed positive sensitivities in July and August. This analysis revealed that higher summer temperatures decreased corn yields significantly, by approximately 3.0% per degree Celsius increase above a base temperature of 23°C. In contrast, soybean yields did not show a significant relationship with temperature, possibly due to the CO2 fertilization effect. The projected temperature changes under future warming scenarios were used to predict the impacts on the CO2 seasonal cycle and net CO2 uptake. The results suggest a potential reduction in the CO2 seasonal amplitude and net CO2 uptake during the peak growing season, by 10-20% of annual sequestration, though with some offsetting positive effects in spring due to phenological changes.

The findings challenge the paradigm that warming consistently benefits CO2 sequestration in northern mid-latitude ecosystems. The observed decoupling between crop yields and CO2 exchange intensity at the interannual scale highlights the complex interplay of photosynthesis and ecosystem respiration responses to climatic forcings. The dominant negative impact of summer temperatures on CO2 uptake, even in relatively wet summers, points to the importance of temperature stress beyond water limitations. The strong negative relationship between corn yield and July NEE reinforces the crucial role of corn production in mediating the link between temperature and cropland CO2 exchange. The lack of a significant temperature effect on soybean yields underscores the need for more detailed investigations into the interplay of CO2 fertilization and temperature stress in different crop systems. These results highlight the importance of accurately representing the temperature sensitivity of cropland CO2 exchange in carbon cycle models to improve the prediction of future carbon-climate feedbacks.

This research demonstrates that warming temperatures significantly reduce summer carbon sequestration in the U.S. Corn Belt, predominantly due to temperature stress in both croplands and natural ecosystems during the peak growing season. The impact of temperature on corn yields was particularly strong, while soybean yields showed less sensitivity, potentially influenced by the CO2 fertilization effect. Future research should focus on further investigating the complex interactions of climate variables, CO2 fertilization, crop-specific responses, and adaptation strategies in order to refine carbon cycle models and improve our understanding of climate change impacts on agriculture and carbon sequestration.

The study relies on a limited number of tower sites and regional-scale NEE inversions which could limit the generalizability of the findings to the broader U.S. Corn Belt. The analysis does not explicitly account for potential adaptation strategies by farmers in response to temperature changes and their associated impacts on crop phenology and yields. Further, while the study accounts for water availability, it may not fully capture the effects of other factors such as soil nutrient availability or pest pressures on plant productivity and CO2 exchange. The use of projected climate data based on specific climate models introduces uncertainty, and the actual future changes could deviate from the predictions.