Henning Teickner , Klaus-Holger Knorr

Hodgkins et al. (2018) used mid infrared spectra (MIRS) to make inferences about the stability of peat against decomposition along a latitudinal gradient from the tropics to northern latitudes. Recently, we have shown that these spectral prediction models are biased and their predictions more uncertain than considered in Hodgkins et al. (2018) (Teickner and Knorr (2022)). Here, we show what consequences this bias and uncertainty and additional neglected uncertainty sources have for the main conclusions in Hodgkins et al. (2018).

In particular, we find that:

1. Larger aromatic contents may be necessary to stabilize tropical near-surface peat against decomposition and aromatics may accumulate at a higher rate with depth in tropical peatlands than estimated by Hodgkins et al. (2018).

2. Larger uncertainties indicate that also larger differences in aromatic contents between (sub)tropical and high latitude peat than estimated by Hodgkins et al. (2018) are possible, also between deeper peat.

3. More research should focus on how estimates of carbohydrate and aromatic contents from MIRS may be confounded by other organic matter fractions, in particular proteins. As a first step, this requires accurate concepts to name organic matter fractions and variables used in the interpretation of MIRS.