Tracing sources and turnover of soil organic matter in a long-term irrigated dry forest - a non-exchangeable hydrogen isotope approach

Soil organic matter (SOM) originates from various sources such as foliar litter, roots and microbial (e.g. fungal) components. The relative sources contribution represents one of the key unknowns in SOM dynamics. Our study aimed to explore whether stable isotope ratios of non-exchangeable hydrogen (H_n) bound to organic matter can be used to differentiate SOM sources, since natural ²H_n abundance can strongly differ between root and foliar tissues. We also investigated if long-term irrigation with ²H-depleted water in a pine forest can be used to track H_n incorporation into organic matter inputs and eventually in the soil pools.

In a 17-year-long irrigation experiment in a dry pine forest, we assessed variations in natural abundance of ²H_n, ¹³C, and ¹⁵N in SOM sources (foliar litter, fine roots, fungal mycelia), decomposing litter, soil (organic layers and uppermost 5 cm-mineral soil) and particle-size fractions. We then applied a Bayesian mixing model (including δ²H_n,δ¹³C, and δ¹⁵N) to estimate the relative sources contribution to SOM.

Natural ²H_n abundance was significantly higher in roots vs. foliar litter (up to +39‰), and in fungal mycelia vs. roots (up to +41‰). Results from Bayesian mixing model suggest that foliar litter contributed to approximately 68 ± 10% of SOM in organic layers and in coarse particulate organic matter (POM). Foliar litter and roots contributed similarly to upper 2 cm of mineral soil (46 ± 11%), while 2-5 cm of mineral soil were largely derived from roots (61 ± 13%). Fungal mycelia contributed to 18 ± 8% of mineral-associated organic matter (MOM), while only to 1-2% of coarse and fine POM. Bayesian mixing models provided only a general indication of the sources contribution to SOM, also considering that isotopic signatures shifted during decomposition. Measurements of isotope signatures in microbial necromass might allow a more accurate assessment of the different SOM sources contribution.

The δ²H_n depletion of soil water under irrigation was paralleled by a comparable decrease in δ²H_n of roots (~12‰). In comparison, the natural ²H_n abundance in fresh needles and foliar litter decreased less strongly (~ 7‰ and 4‰, respectively), likely due to photosynthetic adjustments that may have counterbalanced the irrigation water ²H-depletion. Similar to soil water ²H-depletion, δ²H_n values in coarse POM were 11‰ lower in irrigated vs. dry plots, suggesting that nearly all organic H_n turned over or exchanged with soil water in less than two decades. In contrast, δ²H_n values in fine POM and MOM decreased only by 3‰ under irrigation, which indicate that these fractions comprise slower cycling H_n pools.

Our study showed that the natural ²H_n abundance represents a promising tool to differentiate among SOM sources. While ¹³C and ¹⁵N did not clearly separate between roots and foliar litter, H_n isotopic signatures allowed a good discrimination between SOM sources. In addition, long-term irrigation can provide a potential in situ ²H-labelling of SOM, which may help to examine organic H_n turnover rates across SOM pools.