the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 18 May 2026
Limited direct impacts of precipitation changes on microbial resource limitation in a moderately saline-alkaline desert steppe
Abstract. While soil enzyme C:N:P ecological stoichiometry is known to be sensitive to precipitation changes in acidic to slightly alkaline grasslands, its response to long-term precipitation changes in moderately to severely saline-alkaline desert steppes remain unclear. In these ecosystems, the triple stress of soil water limitation, nutrient poor, and salinity-alkalinity complicates enzyme response, especially under extreme regimes. Based on a precipitation manipulation experiment (50% reduction, 30% reduction, ambient, 30% increase, 50% increase) initiated in 2014 in a moderately saline-alkaline desert steppe in northwestern China, this study assesses microbial resource limitation and identifies their driving factors by monitoring the monthly dynamics of soil extracellular enzyme C:N:P stoichiometry after 9-year treatment. Enzyme stoichiometry showed limited responses to both reduced and increased precipitation (P > 0.05). However, when responses did occur, they varied depending on the direction and intensity of the precipitation change, as well as the specific index examined. Similarly, enzyme vector length and angle were minimally affected by altered precipitation, with phosphorus being the primary limitation for microbes. The variation in vector length and angle was primarily explained by plant traits and microbial stoichiometry, respectively. Precipitation changes altered vector length and angle by modifying soil properties (moisture, NH4⁺-N concentration, pH), plant traits (diversity, carbon concentration, C:P), and microbial stoichiometry (carbon content, C:N, C:P). Rather than exerting a direct effect on microbial resource limitation, altered precipitation indirectly influenced it through modifying soil resource availability, plant diversity, and the carbon-linked stoichiometry of both plants and microbes.
- Preprint
(1689 KB) - Metadata XML
-
Supplement
(592 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-2023', Anonymous Referee #1, 22 Jul 2026
-
RC2: 'Comment on egusphere-2026-2023', Anonymous Referee #2, 17 Aug 2026
Wang et al. present an interesting study using a long-term precipitation manipulation experiment on inferred microbial resource limitation in a saline-alkaline desert steppe grassland in northwestern China. The study utilizes a field experiment first established in 2014 with five precipitation treatments (+30%, +50%, ambient, -30%, -50%). In 2022, the authors sampled the plots monthly from May to August and measured extracellular enzyme activity along with soil and plant properties. The long duration of the manipulation and the focus on a saline-alkaline desert steppe make the dataset potentially valuable as these systems are less represented in literature.
I have, however, concerns about whether the current statistical analyses are appropriate for the experimental design and whether some of the resulting conclusions are interpreted too strongly. I also have concerns about the comparability of the addition and reduction treatments and the non-independence of some of the derived variables. For these reasons, I would recommend major revisions before the manuscript could be published in SOIL.
General comments:
1. Statistical analysis: The statistical analysis doesn't seem to account for the repeated measurements of the same experimental plots. The experiment consists of 15 plots, three replicates per treatment, with each plot sampled monthly from May to August in 2022. However, the methods state that a two-way ANOVA was used to assess treatment, sampling month, and their interaction, followed by one-way ANOVAs within individual months. There is no mention of repeated measures, a plot-level random effect, or a mixed effects framework. If the four observations from each plot were treated as independent, this would be pseudoreplication and could underestimate uncertainty and affect the reported significance. The authors should clarify the structure used for the statistical tests and, if needed, reanalyze the data using an approach that accounts for repeated measures.
The same issue of non-independence may also apply to the linear regressions, Pearson correlations, variation-partitioning analysis, and PLS path models. The manuscript does not report the sample size used for these analyses or explain whether repeated observations from the same plots were treated as independent. If all plot-by-month observations were included as separate data points, the repeated sampling structure would need to be accounted for here as well. This is particularly important for the reported significance of correlations and path coefficients.
The experiment is also described as a randomized block design, but the block structure is not described in the methods and it is unclear whether block was included in the analysis. Please clarify the design and whether it was incorporated statisically.
2. Comparability of precipitation treatments: The precipitation-addition and precipitation-reduction treatments are not directly symmetrical in their implementation. From 2014 to 2017, precipitation reduction was applied only from May to August, whereas from 2018 onward the reduction treatments were imposed year-round. In contrast, precipitation addition continued to be applied only from May to August. This complicates comparisons between the +30% and −30% treatments and between the +50% and −50% treatments, particularly the interpretation of the asymmetry index, because the treatments differ not only in direction but also in their seasonal timing and treatment history. The authors should clarify the actual precipitation received by each treatment and explain how the differing treatment implementation may affect the interpretation of directional and intensity-dependent responses.
3. Non-independence of derived variables: The relationships between microbial CUE and the vector metrics should be interpreted with caution because these variables are not mathematically independent. Vector length and angle are calculated from the relative activities of the C-, N-, and P-acquiring enzymes, while CUE is also calculated using enzyme C:N and C:P ratios derived from the same underlying enzyme measurements. As a result, the reported regressions between CUE and vector length or angle, as well as the path from vector length to CUE in the PLS model, may partly reflect shared mathematical structure rather than an independent ecological relationship. The authors should acknowledge this dependence and reconsider how strongly these relationships are interpreted. I also suggest referring to CUE consistently as estimated or calculated microbial CUE, since it was not measured directly.
4. Interpretation of treatment and sampling month effects: The manuscript places considerable emphasis on individual treatment differences within particular months, even though the treatment × month interactions reported in Tables 2 and 3 are not significant. I think the interpretation should be led primarily by the overall treatment and interaction tests, with within-month comparisons treated more cautiously where the interaction is not significant. The authors should also clarify what post-hoc procedure and any correction for multiple comparisons were used. More generally, a significant sampling-month effect indicates temporal variation among the four sampling dates, but month itself is only a proxy for multiple changing environmental and biological conditions. The manuscript should therefore avoid interpreting the month effect as direct evidence for specific mechanisms such as hydrothermal conditions, plant growth, or soil-property changes unless those mechanisms are supported by the measured environmental variables.
Specific comments:
1. Line 81: Please check the reported latitude of the study site (37°80′ N). Geographic coordinates expressed in degrees and minutes should have minutes <60, so this appears to be an error.
2. Line 87: The site is described as “moderately saline-alkaline.” Please provide the criteria used for this classification and report the relevant EC range. The manuscript reports pH >8.5, but the basis for the salinity classification is less clear.
3. Line 84/112: The manuscript reports a mean annual precipitation of 323.0 mm for 2014–2022, but the +30% and +50% water additions are calculated relative to a local average annual precipitation of 289.4 mm. Please clarify the origin and time period of the 289.4 mm value and why it was used as the treatment baseline.
4. Line 116-119: Please clarify whether the three 1 × 1 m vegetation subplots were newly selected each month. Because aboveground vegetation was destructively harvested, repeated sampling of the same subplot could affect subsequent measurements.
5. Line 118: The term “Patrick richness index” appears to correspond to species richness. Please define how it was calculated and consider using the more widely recognized term “species richness” throughout.
6. Section 2.4: Please provide the pH of the acetate buffer, substrate concentrations, and details of any fluorescence calibration, quenching, or background corrections used in the enzyme assays. These details are particularly important given the strongly alkaline soils at the study site.
7. Line 307 - The statement that a vector length of 0.61 represents a “theoretical optimum” for microbial C limitation requires clearer explanation and an appropriate reference.
Citation: https://doi.org/10.5194/egusphere-2026-2023-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 161 | 76 | 24 | 261 | 36 | 23 | 22 |
- HTML: 161
- PDF: 76
- XML: 24
- Total: 261
- Supplement: 36
- BibTeX: 23
- EndNote: 22
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The manuscript addresses the effect of precipitation on soil enzymes in acidic to alkaline soils. The experiment was conducted in Northwest China. Rainfall variations altered the length and angle of the vector by modifying soil properties (moisture, NH4⁺-N concentration, pH), plant characteristics (diversity, carbon concentration, C:P), and microbial quantification (carbon content, C:N, C:P). The topic is important, and manuscript fits with the scope of the journal. but it has some weaknesses associated with the presented data and discussion have shortcomings as discussed below. Therefore, the current version of the manuscript needs major revision to be published in (egusphere). There are several issues must be addressed.
Major comments
Overall, the manuscript is excellent, the effort is evident, and the idea is superb, but we would like to enhance the clarity of the research with further elaboration in the results and discussion section.