| 03 Jul 2025
Dynamics of inorganic nitrogen cycling, redox conditions, and microbial community composition in a floodplain soil induced by a heavy rainfall after summer drought
Abstract. Climate change increases the frequency of droughts and heavy rainfall events in temperate climates of Central Europe. This study examined the impacts of such extreme weather events on the downward transport of water and inorganic nitrogen species, the main processes governing inorganic nitrogen turnover and the dynamics of the soil microbial community in loamy floodplain soils. To this end, we conducted a field experiment simulating a heavy rainfall event applying isotopically (2H) enriched water on a plot of arable land with fine-textured soil after a prolonged dry period. Subsequently the experimental plots were irrigated with non-labelled water. 50 cm deep soil cores were collected from irrigated and dry control plots at different time points (days 0, 6/8 and 14) and analysed for soil nitrate and ammonium content, soil water content, 2H-enrichment of soil water, microbial community composition based on DNA and RNA, as well as extracellular enzyme activities. Furthermore, redox potentials were recorded in situ along the soil profile on dry and irrigated plots. Whereas the irrigation water of the simulated heavy rainfall event penetrated at least 50 cm deep into the soil along preferential flow paths after only two hours, ammonium and nitrate values remained constant at 50 cm depth despite a significant decrease of nitrate in the topsoil after irrigation. This suggests that some nitrate might have been transported rapidly below 50 cm depth in contrast to ammonium, which is rather immobile in soils due to strong sorption to clay minerals. Ammonium behaved very conservatively and only increased significantly in the topsoil (0–5 cm) in response to the wetting of the dry soil (“Birch effect”). At day 6 and day 14 of the experiment, the non-labelled irrigation water of the daily recurrent moderate irrigation doses resided in the topsoil layer (0–5 cm depth) and only a small fraction penetrated to the subsoil, because the shrinkage cracks present at the beginning of the experiment had closed. In parallel, the in situ redox potential dropped from > +400 mV to below -100 mV relative to the standard hydrogen electrode. Soil nitrate contents decreased when redox potentials were below +100 mV and vice versa. This implies that nitrate formation and transformation after the initial rewetting was governed by soil redox conditions instead of processes induced by the rewetting after drought. Microbial community composition analysis showed no significant differences between the active fraction of the community (RNA) on irrigated and dry plots. Diversity indices on the RNA-level were somewhat higher on dry than on wet plots, although not significantly, which suggests that drying and the following rewetting selected for more adapted taxa. In contrast, activities of the extracellular enzymes β-glucosidase and exochitinase showed significantly higher activities on wet plots than on dry plots, whereas the activity of acid phosphatase did not respond to the irrigation treatment. All enzymes decreased in activity with soil depth. Overall, the heavy rainfall event had only a minor effect on the turnover of inorganic nitrogen species. Nitrogen turnover was mainly governed by redox conditions. The presence of pronounced shrinkage cracks formed after a drought period before the heavy rainfall event, however, allows for rapid nitrate dissipation into the subsoil and eventually to the underlying aquifer. Our study also showed that the soil microbial community in temperate climates reacts in terms of activity and may ultimately be reshaped to represent more resilient taxa regarding drying and rewetting cycles.
General comments:
The authors present results from a field experiment examining water flow, nitrogen cycling, soil redox potential, and microbial community composition responses to initial wet up of dry soil and subsequent smaller irrigation events. The study utilizes advanced techniques such as stable isotope labeling to track the movement of the initial pulse of water added and RNA extractions to characterize the active microbial community. The authors conclude that water movement down preferential flow paths (cracks in dry soil) can potentially lead to nitrate leaching into an underlying aquifer, that following a pulse of nitrogen mineralization after the initial wet up (Birch effect) that soil redox potential drives inorganic nitrogen cycling (“normal redox-driven nitrogen cycling”), and that rewetting shifts the microbial community toward “better adapted taxa”. I found the manuscript to be lacking in compelling motivation (e.g., the knowledge gap identified was diffuse not specific) and that the findings were largely confirmatory of existing knowledge (e.g., water moves down preferential flow paths such as cracks in the soil, the Birch effect is short-lived).
The presentation of results could be improved by framing a narrative around the research questions or study aims. As is, each paragraph of the results section reads as a comprehensive laundry list of results that the reader can get lost in. Also, some paragraphs start with a recap of the methods (e.g. Lines 323-325, 337-340, 376-379, 408-411) or embed results within the context of explaining the methodology (e.g., Lines 307-313, 350). Topic sentences that highlight the overall patterns as relevant to the main takeaways would help orient the reader.
Specific comments:
I do not agree with the authors’ conclusion that their data demonstrate processes switching from a “pulse-driven regime after rewetting to the ‘usual’ redox-driven nitrogen cycling regime governed by oxygen availability” (Lines 561-565). This is an over-simplification of the nitrogen cycling dynamics in the two weeks following wet-up. Interpreting patterns in nitrogen cycling process rates from ammonium and nitrate masses (note that “mass” is reported in Figure 5, not “content” as referred to in the text) is always rife with speculation because there are many different production and consumption processes that can be in play—and in this case, transport of nitrate (and to a lesser extent, ammonium) from overlaying soil can also factor in. While the extracellular enzyme activity response to initial wetting does support the interpretation that a wetting-induced increase in nitrogen mineralization led to the increase in ammonium mass in the top 10 cm of soil, I do not find strong evidence that soil redox subsequently drove the observed ammonium and nitrate masses on T6/T8 and T14. The logic presented on Lines 542-560 to explain the increase in nitrate mass in the wet plots over time is that temporal variation in soil redox reflected changes in oxygen availability associated with daily irrigation events, and that nitrification led to nitrate production whenever soil redox was greater than +100 mV. Without any measurements of nitrification rates, this is all speculation from two sets of soil redox measurements in one wet plot (see comment below). Furthermore, both extracellular enzyme activity (a proxy for nitrogen mineralization or ammonium production rates) and ammonium mass remaining relatively constant from T0 through T14 in the wet plots, suggesting that rates of ammonium consumption (via microbial assimilation or nitrification) also remained relatively constant during this period. This suggests that nitrification rates did not ramp up after T0 to lead to the increased nitrate mass on T6. However, this is all speculation on my part because no process rates were actually measured.
The very curious increase in ammonium and nitrate mass in the dry plots between T0 and T14 highlights how difficult it is to infer controls on nitrogen cycling process rates simply from changes in inorganic nitrogen pool sizes. The authors offer some potential explanations (Lines 569-575), but none seem very convincing.
I understand the technical reasons why the soil redox measurements occurred in very few plots, but this raises some concerns about the validity of drawing major conclusions from the measurements. Soil redox measurements occurred only in three locations (two locations in one wet plot and one location in one dry plot), and the soil redox patterns differed greatly between the two locations in the one wet plot. Therefore, I have concerns about using these measurements to speculate about drivers of nitrogen cycling process rates across all three wet plots. Moreover, the reference electrodes were placed in the glyphosate and irrigation plot, suggesting that they may not have served as a good stable reference for determining the redox potential measured as the voltage difference between them and the redox probes placed in the wet and dry plots.
I found it difficult to interpret the results without seeing data on soil moisture along the depth profiles. While the water mass balance and soil redox potential results were shown in Figures 3 and 4, respectively, and a description of the methods for determining gravimetric soil water content was described in Lines 210-214, I did not see these data reported in the manuscript.
Technical corrections:
Please use paragraph indents, and even line breaks between paragraphs, to make it easier for reviewers to read the manuscript. As is, each page was a solid block of text, and it made it difficult to orient to where a new line of thought was introduced in a new paragraph.