the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Ideas and perspectives: Errors associated with the gross nitrification rates in forested catchments calculated from the triple oxygen isotopic composition (Δ17O) of stream nitrate
Weitian Ding
Urumu Tsunogai
Fumiko Nakagawa
Abstract. A novel method to quantify the gross nitrification rate (GNR) in each forested catchment using the triple oxygen isotopic composition (Δ17O) of stream nitrate eluted from the catchment has been proposed and used in recent studies. However, the equations used in the calculations assumed homogeneous Δ17O values of nitrate being metabolized through either assimilation or denitrification within the forested soil layers without particular discussions. The GNR estimated from the Δ17O of stream nitrate using the equations was more than six times the actual GNR in our simulated calculation for a forested catchment where the Δ17O values of nitrate being metabolized in the soil were heterogeneous and showed a decreasing trend with increasing depths. Therefore, we should verify that the Δ17O values of nitrate being metabolized are homogeneous in forested soils or estimate the possible range of errors using Δ17O of stream nitrate to estimate the GNR.
- Preprint
(624 KB) - Metadata XML
-
Supplement
(65 KB) - BibTeX
- EndNote
Weitian Ding et al.
Status: closed
-
RC1: 'Comment on bg-2022-236', Anonymous Referee #1, 09 Feb 2023
Nitrate is one of important N forms available for microbe and plant. However, it almost impossible to quantify nitrate production rate (nitrifcation) at an ecosystem level, especially for forest ecosystems, where there are very large spatial and temporal variations. A novel method to quantify gross nitrification at an ecosystem level has been originally proposed by Tsunogai et al. (2011, Biogeosciences) by using nitrate triple oxygen isotopes. Herein, we define it as the nitrate oxygen isotope method. Tsunogai et al. had successfully applied this method to quantify gross nitrcation rates for lake systems in Japan. Later, this method was adopted by Fang et al. (2015, PANS), Huang et al. (2019, EST) and Hattori et al. (2019, Sci. Total Environment) to quantify annual gross nitrication rates for forests. The manuscript submitted by Weitian Ding et al. questioned its appliction of this method to forest soils. The authors of this manuscript suggested that the nitrate oxygen method is only suitable for the environment, like lake system, where the distribution of 17O of nitrate is homogeneous. To demonstrate it, the authors calculated gross nitrification rates to be 83.6 kg N/ha.yr when assuming that the distribution of 17O of nitrate is heterogeneous in forest soils from the top to subsoil, while to be 13.0 kg N/ha.yr when assuming that it is homogeneous, respectively, according to results from a Japanese forest reported by Hattori et al. The authos concluded that it should verify that 17O values of nitrate being metabolized are homogeneous in forest soils. The manuscript was clearly written and the topic is within the scopes of Biogeosciences. However, the authors may have misundstood the assumptions for the nitrate oxygen isotope method when applying to forest soils. Thus, the conclusions made by the authors cannnot be supported.
First, it is not necessary to assume that the distribution of 17O of nitrate along the soil profile is homogeneous or heterogeneous when apply the nitrate oxygen isotope method to forest soils. In fact, the assumption of the method is that the plants or microbes access the same nitrate source in forest soil with denitrifiers (Fang et al., 2015, PNAS). This assumpition is not identical to the assumption that the spatial distribution of 17O of nitrate along the soil profile is homogeneous, as demonstrated by the authors (Fig. 2).
Second, it is not correct to assume that the distribution of 17O of nitrate along the soil profile is homogeneous (Fig. 2). 17O has been rarely measured along the soil profile. The one and only study shows a sharp decrease in 17O of nitrate in the top soil and remains relatively constant in the soil from 25 to 95 cm (Hattori et al., 2019, Sci. Total Environment). Thus, the assumption made by the author was not supported the field observation.
Third, I agree that it is the distribution of 17O of nitrate along the soil profile is highly hetrogeneous, as nitrification is dominant in surface soils, and deposited nitrate may enter soil from the forest floor. However, it is not correct to assume that 17O of nitrate decreased linearly with soil depth (Fig. 1). The field observation by Hattori et al did not support this assumption. This may be main reason for unrealstically low gross nitrification rate (13 kg N/ha.yr) as calculated by the authors, in the study forest with modate to high N deposition (16 kg N.ha.yr). Nitrification must be strongly active in this forest, which was supported by high soil nitrate concentrations and a large seasonal varation in 15N and 17O of nitrate (Fig. 3 of Hattori et al.).
In fact, the nitrate oxgen isotope method admit high heterogeneity of soil nitrfication in both spatially and seasonally. And it is difficult and almost impossible to capture these heterogeneities. However, these heterogeneities can be integrated to streamwater. The nitrate oxygen isotope method takes this advantage of it.
Citation: https://doi.org/10.5194/bg-2022-236-RC1 - AC1: 'Reply on RC1', Weitian Ding, 12 Mar 2023
-
RC2: 'Comment on bg-2022-236', Joel Bostic, 10 Feb 2023
General Comments:
Ding et al. present a simulation to show overestimation of gross nitrification rates (GNR) as calculated using triple oxygen isotope ratios of nitrate in precipitation and streamwater. This simulation demonstrates that the assumption of uniform D17O values in soils can result in erroneously high gross nitrification rates. The uniform D17O in soils assumption has been applied in previous studies in which triple oxygen isotopes of nitrate were used to quantify GNR, which highlights the importance of Ding et al.’s findings. I have questions regarding the assumptions of Ding et al.’s simulation, however, and am unsure if their simulation is equivalent to the watershed-scale GNR that is presented in other studies.
I suggest that Ding et al. elaborate on the assumptions of their simulation. For example, is it realistic to assume that D17O decreases linearly with depth? In this scenario, the nitrate (NO3) flux down the soil profile would consist of mostly unprocessed atmospheric NO3 (at least until halfway down the depth profile). This does not appear to align with the empirical data used in the simulation from Hattori et al. How would the simulation results change if D17O dramatically decreased from deposition to the first soil depth used in the simulation? Existing data show that D17O of NO3 can be low in the uppermost soil horizons (<5 per mille in the upper 7 cm, Yu and Elliott 2018; 2 per mille +/- 1.1 per mille from 0-30 cm, Costa et al., 2011), which suggests that the assumption of a linear decrease of D17O from deposition to “water” (i.e., the lowest soil depth in the simulation; Figure 1) is not necessarily valid. I also request the authors elaborate on the assumption that there is no biological processing in the water layer. Is this water layer the stream? Or is it intended to be the saturated zone of the soil? If it is the latter, is it reasonable to assume there is no processing along flowpaths between the saturated zone and the stream? How would the simulation results change if this assumption was violated?
While Ding et al. present a simulation that suggests inaccurate GNR estimates produced by the triple oxygen isotope approach at the watershed scale, I suggest the authors expand the scope of their premise to make it more broadly relevant. GNR, as calculated by soil scientists in soil cores (which has been done with D17O, Yu and Elliott 2018, along with many others using d15N), are not equal to gross nitrification rates as calculated at the watershed scale using streamwater NO3. I think this difference needs to be made clear and this manuscript represents a potential outlet to do so. For example, if D17O is a conservative tracer, a decrease in D17O of NO3 between deposition and streamwater requires addition of or dilution by terrestrial NO3, which has D17O = 0 (or approximately equal to 0). By this logic, Hattori’s D17O data indicate that the streamwater NO3 measured must have had terrestrial NO3 added (i.e., nitrification) or have been diluted by terrestrial NO3 along its flowpath from deposition to stream. I think this example illustrates the difference between GNR as measured in a soil core and GNR as measured at the watershed scale. Perhaps it would be more appropriate to give a different name to the watershed scale metric that can be calculated using eq. 6.
Specific comments:
-Lines 16-17: needs reference. Is NO3 often limiting? Or NH4? Or just N more generally?
-Line 24: I suggest a different word than “determined”. Perhaps “quantified” or “estimated”
-Line 24: What is a “water environment”? A stream? A pond? A lake? Soil water?
-Line 30-31: I suggest providing a range of D17O values of atmospheric NO3. 26 per mille is a good average, but it can be much lower.
-Line 32: What does “almost stable” mean?
-Eq. 1: I suggest a different subscript for “water”. Readers might see “water” and think you are measuring the triple oxygen isotopes of O in H2O.
-Line 46: Riha et al. 2014 was not a forested catchment. It was an urban watershed study.
-Line 54: reference for NO3 being homogenous?
-Lines 95-99: Where were the soil data collected in relation to the stream for Hattori et al.? Is it reasonable to assume there was no processing along the flow paths from soil to stream?
-Line 136: Are you missing a word at the end of this sentence? Perhaps add “sources” to the end.
References:
Costa AW, Michalski G, Schauer AJ, Alexander B, Steig EJ, Shepson PB. 2011. Analysis of atmospheric inputs of nitrate to a temperate forest ecosystem from δ17o isotope ratio measurements. Geophysical Research Letters 38.
Yu and Elliott, 2018. Probing soil nitrification and nitrate consumption using Δ17O of soil nitrate. Soil Biology and Biochemistry. https://doi.org/10.1016/j.soilbio.2018.09.029Citation: https://doi.org/10.5194/bg-2022-236-RC2 - AC2: 'Reply on RC2', Weitian Ding, 12 Mar 2023
Status: closed
-
RC1: 'Comment on bg-2022-236', Anonymous Referee #1, 09 Feb 2023
Nitrate is one of important N forms available for microbe and plant. However, it almost impossible to quantify nitrate production rate (nitrifcation) at an ecosystem level, especially for forest ecosystems, where there are very large spatial and temporal variations. A novel method to quantify gross nitrification at an ecosystem level has been originally proposed by Tsunogai et al. (2011, Biogeosciences) by using nitrate triple oxygen isotopes. Herein, we define it as the nitrate oxygen isotope method. Tsunogai et al. had successfully applied this method to quantify gross nitrcation rates for lake systems in Japan. Later, this method was adopted by Fang et al. (2015, PANS), Huang et al. (2019, EST) and Hattori et al. (2019, Sci. Total Environment) to quantify annual gross nitrication rates for forests. The manuscript submitted by Weitian Ding et al. questioned its appliction of this method to forest soils. The authors of this manuscript suggested that the nitrate oxygen method is only suitable for the environment, like lake system, where the distribution of 17O of nitrate is homogeneous. To demonstrate it, the authors calculated gross nitrification rates to be 83.6 kg N/ha.yr when assuming that the distribution of 17O of nitrate is heterogeneous in forest soils from the top to subsoil, while to be 13.0 kg N/ha.yr when assuming that it is homogeneous, respectively, according to results from a Japanese forest reported by Hattori et al. The authos concluded that it should verify that 17O values of nitrate being metabolized are homogeneous in forest soils. The manuscript was clearly written and the topic is within the scopes of Biogeosciences. However, the authors may have misundstood the assumptions for the nitrate oxygen isotope method when applying to forest soils. Thus, the conclusions made by the authors cannnot be supported.
First, it is not necessary to assume that the distribution of 17O of nitrate along the soil profile is homogeneous or heterogeneous when apply the nitrate oxygen isotope method to forest soils. In fact, the assumption of the method is that the plants or microbes access the same nitrate source in forest soil with denitrifiers (Fang et al., 2015, PNAS). This assumpition is not identical to the assumption that the spatial distribution of 17O of nitrate along the soil profile is homogeneous, as demonstrated by the authors (Fig. 2).
Second, it is not correct to assume that the distribution of 17O of nitrate along the soil profile is homogeneous (Fig. 2). 17O has been rarely measured along the soil profile. The one and only study shows a sharp decrease in 17O of nitrate in the top soil and remains relatively constant in the soil from 25 to 95 cm (Hattori et al., 2019, Sci. Total Environment). Thus, the assumption made by the author was not supported the field observation.
Third, I agree that it is the distribution of 17O of nitrate along the soil profile is highly hetrogeneous, as nitrification is dominant in surface soils, and deposited nitrate may enter soil from the forest floor. However, it is not correct to assume that 17O of nitrate decreased linearly with soil depth (Fig. 1). The field observation by Hattori et al did not support this assumption. This may be main reason for unrealstically low gross nitrification rate (13 kg N/ha.yr) as calculated by the authors, in the study forest with modate to high N deposition (16 kg N.ha.yr). Nitrification must be strongly active in this forest, which was supported by high soil nitrate concentrations and a large seasonal varation in 15N and 17O of nitrate (Fig. 3 of Hattori et al.).
In fact, the nitrate oxgen isotope method admit high heterogeneity of soil nitrfication in both spatially and seasonally. And it is difficult and almost impossible to capture these heterogeneities. However, these heterogeneities can be integrated to streamwater. The nitrate oxygen isotope method takes this advantage of it.
Citation: https://doi.org/10.5194/bg-2022-236-RC1 - AC1: 'Reply on RC1', Weitian Ding, 12 Mar 2023
-
RC2: 'Comment on bg-2022-236', Joel Bostic, 10 Feb 2023
General Comments:
Ding et al. present a simulation to show overestimation of gross nitrification rates (GNR) as calculated using triple oxygen isotope ratios of nitrate in precipitation and streamwater. This simulation demonstrates that the assumption of uniform D17O values in soils can result in erroneously high gross nitrification rates. The uniform D17O in soils assumption has been applied in previous studies in which triple oxygen isotopes of nitrate were used to quantify GNR, which highlights the importance of Ding et al.’s findings. I have questions regarding the assumptions of Ding et al.’s simulation, however, and am unsure if their simulation is equivalent to the watershed-scale GNR that is presented in other studies.
I suggest that Ding et al. elaborate on the assumptions of their simulation. For example, is it realistic to assume that D17O decreases linearly with depth? In this scenario, the nitrate (NO3) flux down the soil profile would consist of mostly unprocessed atmospheric NO3 (at least until halfway down the depth profile). This does not appear to align with the empirical data used in the simulation from Hattori et al. How would the simulation results change if D17O dramatically decreased from deposition to the first soil depth used in the simulation? Existing data show that D17O of NO3 can be low in the uppermost soil horizons (<5 per mille in the upper 7 cm, Yu and Elliott 2018; 2 per mille +/- 1.1 per mille from 0-30 cm, Costa et al., 2011), which suggests that the assumption of a linear decrease of D17O from deposition to “water” (i.e., the lowest soil depth in the simulation; Figure 1) is not necessarily valid. I also request the authors elaborate on the assumption that there is no biological processing in the water layer. Is this water layer the stream? Or is it intended to be the saturated zone of the soil? If it is the latter, is it reasonable to assume there is no processing along flowpaths between the saturated zone and the stream? How would the simulation results change if this assumption was violated?
While Ding et al. present a simulation that suggests inaccurate GNR estimates produced by the triple oxygen isotope approach at the watershed scale, I suggest the authors expand the scope of their premise to make it more broadly relevant. GNR, as calculated by soil scientists in soil cores (which has been done with D17O, Yu and Elliott 2018, along with many others using d15N), are not equal to gross nitrification rates as calculated at the watershed scale using streamwater NO3. I think this difference needs to be made clear and this manuscript represents a potential outlet to do so. For example, if D17O is a conservative tracer, a decrease in D17O of NO3 between deposition and streamwater requires addition of or dilution by terrestrial NO3, which has D17O = 0 (or approximately equal to 0). By this logic, Hattori’s D17O data indicate that the streamwater NO3 measured must have had terrestrial NO3 added (i.e., nitrification) or have been diluted by terrestrial NO3 along its flowpath from deposition to stream. I think this example illustrates the difference between GNR as measured in a soil core and GNR as measured at the watershed scale. Perhaps it would be more appropriate to give a different name to the watershed scale metric that can be calculated using eq. 6.
Specific comments:
-Lines 16-17: needs reference. Is NO3 often limiting? Or NH4? Or just N more generally?
-Line 24: I suggest a different word than “determined”. Perhaps “quantified” or “estimated”
-Line 24: What is a “water environment”? A stream? A pond? A lake? Soil water?
-Line 30-31: I suggest providing a range of D17O values of atmospheric NO3. 26 per mille is a good average, but it can be much lower.
-Line 32: What does “almost stable” mean?
-Eq. 1: I suggest a different subscript for “water”. Readers might see “water” and think you are measuring the triple oxygen isotopes of O in H2O.
-Line 46: Riha et al. 2014 was not a forested catchment. It was an urban watershed study.
-Line 54: reference for NO3 being homogenous?
-Lines 95-99: Where were the soil data collected in relation to the stream for Hattori et al.? Is it reasonable to assume there was no processing along the flow paths from soil to stream?
-Line 136: Are you missing a word at the end of this sentence? Perhaps add “sources” to the end.
References:
Costa AW, Michalski G, Schauer AJ, Alexander B, Steig EJ, Shepson PB. 2011. Analysis of atmospheric inputs of nitrate to a temperate forest ecosystem from δ17o isotope ratio measurements. Geophysical Research Letters 38.
Yu and Elliott, 2018. Probing soil nitrification and nitrate consumption using Δ17O of soil nitrate. Soil Biology and Biochemistry. https://doi.org/10.1016/j.soilbio.2018.09.029Citation: https://doi.org/10.5194/bg-2022-236-RC2 - AC2: 'Reply on RC2', Weitian Ding, 12 Mar 2023
Weitian Ding et al.
Weitian Ding et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 553 | 147 | 17 | 717 | 56 | 12 | 13 |
- HTML: 553
- PDF: 147
- XML: 17
- Total: 717
- Supplement: 56
- BibTeX: 12
- EndNote: 13
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1