Dear creators of FRAME
Thank you very much for providing such a useful and open source tool for stable isotope analyses. I am using your model to detect the source (nitrification/denitrification) of N2O production in a grassland soil.
I used the file “N2Ofrac.csv”
from the real-world
example of Lewicki et al. 2022 for taking into account the isotope fractionation by
reduction of N2O to N2 during denitrification. I suspect that most of my
samples lie in this "denitrification reduction range" according to what I have seen in the dual isotope plots.
So I ran the model and also
got some results, but I am unsure if my input was actually correct. So my
specific questions are:
- In the FRAME sample input data, do
I need to give the d15N-N2O value in relation to the substrate (e.g.
d15-NO3-), so the difference between d15N-N2O and soil d15-NO3- ?
Same question also for d18O-N2O, does it have to be the difference to the
d18O of the soil water (I would have these data available)? In the file for the N2O sources, I put the values from the Supplement of Yu et al. 2020
- I realized that the results are quite sensitive to the column
“stdev” in the data input file. So far I put the long-term stdevs of the
measurement systems (of SP, d15N-N2O and d18O-N2O) in there. Is this correct? Or should I put the stdev of all the
samples at the respective sampling date? Are the results more stable if less but more
representative data points are used (i.e. averages)?
- What exactly does the "r" in the FRAME output for the fractions stand for? Is it the part of the N2O at the site that gets fully reduced to N2? And then the other fractions (bD, nD, fD and Ni), which add up to about 100%, do they give the fraction of the actual processes producing the "remaining" N2O?
It would be really
great if you could give me some advice on the questions above.
Best wishes,
Iris Feigenwinter
Researcher in the Grassland Sciences Group at ETH Zurich, Switzerland