Reply to “Basis for Comparisons of Soil CO 2 Respiration Test Procedures”

This letter is provided in response to a letter (Brinton and Vallotton, 2018) regarding our recent publication (McGowen et al., 2018). We welcome the constructive discussion to advance laboratory methods for assessing soil health indicators. This is why we shared soil samples from our work with Dr. Brinton’s team in hopes of extending efforts to better understand the relationships between CO2 burst methods and the experimental conditions that influence the observed differences. It is expected that different laboratory methods seldom yield the same absolute values but are often correlated with one another. Our goal was to contribute to the science and in-depth understanding for method standardization in soil health assessment. Our conclusion of lower labor requirement for the GC method was based mainly on the time required washing dishes for the methods. For each analysis, the GC method uses a gas vial with a crimp top closure and gray butyl septa. The closure and septa are discarded, leaving only the vial to be washed. The components needed for one analysis using the Solvita (Woods End Laboratories) method include a jar, a rubber gasket, a lid, and a plastic beaker. The washing of these components combined with manual reading of the Solvita paddle contributes to considerable labor requirement. We agree with Dr. Brinton that the Solvita method could run >60 samples per day as permitted by the GC setup, providing that labor and incubation chambers are not limited. However, the XYZ sampler could be programmed to inject water into the vials, allowing the GC to run continuously with precise incubation times after water introduction. Regarding the correlation between the two methods, the slope between the two methods presented in McGowen et al. (2018) was low, but the r2 was 0.90 for the linear range of the Solvita method. Additionally, the r2 for the curvilinear relationship between the Solvita color number and the GC CO2 burst was 0.92 when all data were used. We agree with Dr. Brinton’s team that the low slope is probably the result of differences in headspace volumes and soil sample sizes used in the measurements. We believe that sensitivity analysis is needed to better understand how these two experimental conditions affect emissions per gram of soil. However, in McGowen et al. (2018), we did not attempt to dilute the headspace in the Solvita method as our preliminary experiments using soils within the linear range of the Solvita method did not result in values that were proportional to the Solvita values generated without dilution. In other words, we could not simply use a dilution factor of two when we used half the soil of the original method and arrive at the same value. Therefore, we are currently evaluating the impact of sample size and chamber volume for both the GC method as well as the hydroxide trap method to further understanding. The data provided by Dr. Brinton and his team show clearly that the paddles used for the Solvita method are accurate in providing a measurement of headspace CO2 concentrations based on the relationships between the other CO2 analytical methods presented in their letter. The data also show the discrepancy between the Solvita values as measured in our laboratory and the current method they used. In fact, the data provided in Brinton and Vallotton (2018) comparing the OKSU-SOL (Oklahoma State University data, Solvita) Reply to “Basis for Comparisons of Soil CO2 Respiration Test Procedures”