Agriculture produces large emissions of carbon dioxide (CO₂), nitrous oxide (N₂O), and nitric oxide (NO), especially in high-temperature agroecosystems, where management approaches for reducing these emissions are needed. A promising management solution to increase water infiltration and reduce trace gas emissions is subsurface drip irrigation, a method which increases rhizosphere access to water and nitrogenous fertilizers. In a multi-year field study, we compared per-yield irrigation and soil emissions for flood- and drip-irrigated field plots in southern California during two seasons and between two forage crops differing in fertilizer requirements: alfalfa (Medicago sativa L.) and sudangrass (Sorghum bicolor ssp. Sudanese). We monitored soil climate and emission responses to irrigation using a custom array of automated chambers connected to trace gas analyzers that measured gas fluxes continuously every 30 minutes. We found that, compared to flood-irrigated fields, drip irrigation in alfalfa increased yield by 7%, decreased irrigation demand by 11%, and decreased CO₂ emissions by 59%, N₂O by 14%, and NO by 27%. Drip irrigation in sudangrass increased yield by 6%, decreased irrigation by 68%, increased CO₂ emissions by 3%, and decreased both N₂O and NO emissions by 62%. In both crops, differences between irrigation types were strongest in the summer when flooded soil produced the largest pulses of N₂O and NO relative to small drip-irrigated pulses. As agriculture continues to intensify in warmer climates, implementation of subsurface drip irrigation can help reduce agroecosystem contributions to climate change and air pollution while increasing crop yields.

Harvested biomass and irrigation data were collected at field scale; for these variables, we calculated water use efficiency (WUE) of productivity by dividing harvested dry biomass by total irrigation applied per field per harvest. We batch-processed instantaneous soil flux, temperature,  and moisture data for each emissions collection campaign using methods adapted from Andrews and Jenerette (2020) Plant and Soil. Instantaneous fluxes of CO₂, N₂O, and NO were calculated as the regression coefficient of linear increase in gas concentration during the 2.5-minute active measurement period, corrected for soil collar dimensions and atmospheric parameters following the Ideal Gas Law (Davidson et al.,  2000). Soil temperature and moisture were averaged over this same period as well. Instantaneous fluxes of each gas were compiled and integrated with instantaneous soil temperature and moisture measurements at 30-minute resolution using a publicly-accessible R script (https://github.com/handr003/TraceGasArray). Logistical constraints for trace gas measurements forced a decision to not replicate treatments but allowed us to measure high-resolution emission trajectories with replication across different field rows. With time series of these high-resolution measurements, we could interpret interactive responses to rapid changes in soil conditions, such as during re-wetting, that could be missed or misinterpreted at coarser temporal scales. We extracted the magnitude and timing of each peak flux and climate parameter, calculated as the maximum instantaneous measurement recorded over an irrigation event. We also extracted per-irrigation mean temperature, moisture, and daily emission values, the latter of which were calculated as the integrated area under each time series curve using linear trapezoidal method divided by the duration of time following the irrigation event and prior to the next one. 

To link soil emissions to field-scale irrigation and yield measurements, we extracted per-yield emissions values for each trace gas and field. We multiplied each chamber’s mean per-irrigation daily emissions by the length of each harvest period to calculate each chamber’s per-harvest emissions. We then divided this per-harvest emission estimate by per-area yield to calculate the amount of each trace gas emitted per amount yield as estimated by each chamber. These calculations are similar to those used to estimate greenhouse gas intensity (GHGI), a measure which has been used to evaluate the global warming impact of irrigation in agriculture (McGill et al., 2018). Representing emissions in this way allowed us to compare the effectiveness of drip irrigation across time and between crops by taking into account differences in harvest cycles and plant physiological differences.