Soil respiration is an important ecosystem process and fluctuates daily and seasonally with abiotic and biotic factors, inevitably leading to uncertainty in the estimation of daily, monthly or seasonal means from short-term measurements. The measurement strategies are poorly understood and are commonly analyzed using noncontinuous measurements. The objectives of this study were to evaluate the optimum measurement time and frequency of soil-respired CO2 (Rs) and its δ13C (δs) value, as well as soil temperature (ST) and soil water content (SWC) during the growing season. To provide more refined recommendations, we performed long-term continuous measurements of Rs and δs value with a multichannel and dual-cycle observation system under non-steady-state conditions in an irrigated oasis agroecosystem in Northwest China. The daily optimum measurement time of Rs was 9:00–10:00, which accounted for 89.0% of the total measurement days and was within the previously recommended time scale of 9:00–12:00. δs showed little diurnal variation, and each hourly mean approximately represented the daily mean. The Rs and δs measured at 9:00–11:00 did not significantly differ from the daily mean, yet the soil temperature (ST) and soil water content (SWC) measured at 9:00–11:00 were lower than the daily mean values. The number of days needed to obtain values ±10% for Rs, δs, ST and SWC was 101, 23, 13, and 25, respectively (95% confidence interval (CI)), which sharply decreased to 11, 3, 1, and 3 days, respectively (95% CI), when the desired accuracy expanded to ±30% for Rs, δs, ST and SWC in the growing season. That means the measurement interval of Rs and δs were about is 14 days and 53 days during the growing season in agroecosystems (95% CI, ±30% mean). Overall, this study analyzed the optimum measurement times for Rs and δs at daily, monthly and seasonal time scales and provided guidelines for improving measurement frequencies under noncontinuous measurements.
Read full abstract