|This is the second time that I review this manuscript. The authors have greatly improved the manuscript by addressing my previous comments on the FG data, the experimental design, and the data interpretation. My conclusion after the 2nd round of review is that this study is among the first attempts to directly capture photo- and thermal degradation-induced CO2 and CO fluxes in the field. It presents convincing data to support that thermal degradation can be a significant source of CO emission in dryland ecosystems. Thanks to the revision, I also gained a much better understanding of their experimental design. But the design is intrinsically limited: there was only one set of transparent vs opaque measurements; in other words, there was no replication in the field to detect photodegradation-induced fluxes. |
I am struggling to make a decision/recommendation on this manuscript. On the one hand, the authors’ data and arguments on CO flux and thermal degradation are sound. This result will be a valuable contribution to the community, and the manuscript will be well cited because of it. On the other hand, the conclusion on photodegradation is potentially misleading, if the readers do not realize the limitation of the design and do not fully digest the results. Eventually, the former outweighs the latter.
In my opinion, this manuscript still needs further revision. A few important details about the experimental design should be double-checked. Data presentation and discussion on the photodegradation-induced fluxes can also be improved. I do realize it is the second round of revision. Thus, my comments are kept at a minimum:
- The reported UV intensities in the laboratory were unrealistic. Fluorescence lamps, like the Philips (only one ‘l’ in Philips) lamps used in this study, cannot generate UV radiation as strong as 45 W m-2 nm-1. It is usually on the order of 0.1-1 W m-2 nm-1. In a laboratory setting, however, the actual radiation intensity would be even lower, because the radiation intensity drops exponentially with distance. For example, I am using the UVB-313EL (~40W) lamps from the Q-Lab, which is comparable to the Philips TL 40W model. Based on a broad-band radiometer (UVP, LLC), UVB intensity is ~30 W m-2 when the radiometer is just underneath the lamp. But it drops to 15 and 2.5 W m-2, when the distance to the lamp increases to 10 and 45 cm, respectively. Without knowing the exact distance between the lamps and the grass sample (the bottom, NOT the top of the 25-cm cylinder used in this study), it is impossible to estimate the actual UV intensity received by the sample. Their field UV intensity seems reasonable to me. But it is clear that the laboratory UV intensity is only comparable with, if not lower than, the field intensity.
- Specify the instrument that was used to measure UV intensity. Irradiance in the unit of W m-2 nm-1 appeared to be collected by a monochromator. Many previous studies report UV intensity using broadband radiometers (Brandt et al. 2010, Kirschbaum et al. 2011, Lee et al. 2012). Irradiance here can be integrated over a range of wavelength (e.g. 290-320 nm for UVB) so that it can be compared to previous work.
- This issue with UV intensity may seem trivial. However, authors stated that the laboratory experiment had 20 times more UV radiation than natural conditions. Because they did not find differences in CO2 emission between UV exposure and dark in the laboratory, they further concluded that photodegradation did not play an important role in emitting CO2 in the field. This line of reasoning has flaws because of the over-estimation of UV intensity in the laboratory.
- I want to point out that the manuscript did not present data on the effects of UV radiation on CO2 and CO fluxes in the laboratory. Similar comment has been made by both reviewers previously. The idea behind our comments is that we cannot be certain about the author’s conclusion (lines 11, 354-356), unless we see the data. I would like to further argue that the data have to be presented in a meaningful way. The authors included figures (pages 10 and 11) on this topic in their responses to reviewers’ comments. These two figures raise several additional questions: 1) the big error bars suggest that the authors lumped a lot of data together to draw the bar charts. Given the design was not overly complicated, and the number of replicates was extremely low (2), it makes sense to present all the raw data (e.g. point chart with different treatments on the x axis); 2) the highly varied data in the CO2 figure are not consistent with the data in figure 6 that report the temperature effects on fluxes. Specifically, CO2 flux in the dark treatment of the photodegradation experiment (21 degree C) varies between -6 to 4 nmol m-2 min-1; while, CO2 flux in figure 6 (dark, 20 degree C) are all positive and between 2 to 6 nmol m-2 min-1. With similar environmental conditions, why are the data so different between the two experiments? It raises the question whether the design is appropriate for studying photodegradation-induced fluxes. At a bare minimum, I encourage the authors to plot new figures and include them in the supplementary material.
- With the above comments, I wonder whether the authors are still convinced that photodegradation did not significantly contribute to C fluxes in the field. The authors referred to the result of Rutledge et al. 2010 (1 umol m-2 s-1) multiple times in their discussion. Please be aware that this number is an estimation, not a measured flux rate. Personally, I consider it as the upper limit of the photodegradation-induced flux. In fact, results from previous laboratory studies were one order of magnitude lower than the Rutledge estimation (see King et al. 2012 for a review on this).
Below are my interpretations of some of the data. I want to state clearly that my recommendation on this manuscript has nothing to do with these interpretations, because they can be wrong. Please do NOT feel obligated to address them in the revision. I just hope these interpretation and comments can be useful for future studies on photodegradation.
- The CO2 flux data from August 6th to 8th in the authors’ responses (page 18) suggest a detectable impact of chamber design on CO2 flux. We notice a bell-shaped curve of CO2 flux. When CO2 was in the process of reaching its daily maximum, it was always higher in the transparent chamber than in the covered chamber. Similarly, when CO2 dropped in the afternoon, it was also higher in the transparent chamber. When CO2 was around its maximum, however, the relationship reversed. Sometimes, transparent chamber had over 0.5 umol m-1 s-1 higher CO2 flux than the covered chamber. This difference is within the order of magnitude of the flux reported by Rutledge et al. 2010. Unfortunately, without replication, it is impossible to know whether the differences between two types of chambers are meaningful.
- Photo- and thermal degradation can explain the pattern above, assuming it is meaningful. Field CO2 flux comprises biotic respiration and abiotic flux (photo- and thermal). From morning to noon, abiotic flux increases with radiation, contributing to higher flux in the transparent chamber. Around noon, radiation increases temperature to a level that limits respiration. The temperature-induced decline in respiration is bigger in magnitude than the abiotic flux. Thus, transparent chamber has lower flux than covered chamber. Once litter and soil cool in the afternoon, the contribution of abiotic flux becomes detectable.
- This study suggests that it is critical to monitor the temperature in air, chamber, and soil during similar photodegradation studies. The study also suggests another method of detecting photodegradation fluxes by combining flux gradient and flux chamber methods, especially when the footprint of the FG method is homogeneous. The lack of replication, however, limits the ability of this study to detect the radiation-induced flux in the field. This study would have been powerful, if transparent and covered chambers could be compared using multiple chamber locations over a long period of time. It highlights the difficulty in measuring radiation-induced fluxes.