the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 21 Aug 2026
Irrigation Water Redistribution and Competition Mechanisms Between Cropland and Shelterbelts in Arid Regions: Insights from Stable Water Isotopes data
Abstract. Protective forests play an important role in maintaining the stability of arid oasis agroecosystems; however, water exchange and competition between shelterbelts and cropland under limited water availability remain poorly quantified. This study used stable hydrogen and oxygen isotopes to investigate irrigation-water redistribution and water-use interactions in a maize (Zea mays L.)–poplar (Populus alba L.) shelterbelt system in the Minqin Oasis, Hexi Corridor. The results showed that irrigation water was transferred from cropland to shelterbelts through both soil lateral movement and root-mediated uptake. Across four irrigation events, modeled transpiration-related water exchange between farmland and shelterbelts reached 84.49 mm, accounting for 19.56 % of applied irrigation water. Irrigation-induced lateral transfer into shelterbelt soil averaged 13.26 % of irrigation depth, with the largest contribution occurring in the 20–60 cm soil layer (5.49 %). Maize and poplar exhibited substantial overlap in water-source use, with a mean proportional similarity index of 73.35 %, indicating strong potential competition for soil water resources. These findings reveal that irrigation water redistribution links cropland and shelterbelts through coupled hydrological processes and highlight the need to jointly consider crop water demand and shelterbelt water consumption in arid oasis management.
- Preprint
(1536 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-4760', Anonymous Referee #1, 20 Sep 2026
-
AC1: 'Reply on RC1', Guofeng Zhu, 28 Sep 2026
Public Author Reply to Referee #1
Manuscript: egusphere-2026-4760
Dear Referee #1,
Thank you for the careful and constructive review. We have considered all of your comments and have prepared the corresponding revisions. If invited to submit a revised manuscript after the interactive discussion, these changes will be incorporated in full. Below we provide a concise point-by-point response, focusing on how each issue will be addressed and how the interpretation of the main results will be adjusted.
Major comments
Comment 1: The study relies on isotopic compositions obtained using the LI-2100 system, which can be regarded as an automated equivalent of cryogenic vacuum extraction. Recent studies have demonstrated that bulk stem water extracted by cryogenic methods may differ isotopically from mobile xylem sap due to within-stem isotopic heterogeneity. Because the primary conclusions regarding maize-poplar water-source partitioning and competition depend on stem-water isotopic composition, the authors should discuss the potential implications of this methodological limitation and evaluate how it might affect estimates of deep-soil-water uptake and water-source overlap.
Response: We agree. We will explicitly identify the analyzed stem water as bulk extracted water rather than selectively collected mobile xylem sap. The revised Discussion will explain that a bulk-water–sap-water isotopic offset could shift the plant-water endmember used in IsoSource and alter the inferred shallow-, middle-, and deep-soil contributions and the proportional similarity (PS) index. Because mobile xylem sap was not independently sampled, we will not apply an unsupported correction; instead, deep-soil-water contributions and maize–poplar water-source overlap will be interpreted as conditional on the isotope composition of vacuum-extracted bulk stem water. Recent references addressing stem-water heterogeneity and extraction uncertainty will also be added. The Abstract will describe the PS result as substantial water-source overlap rather than as direct evidence of competition.
Comment 2: The paper provides numerical estimates but lacks a rigorous uncertainty assessment and independent validation. The authors should provide a sensitivity or uncertainty analysis to evaluate the influence of isotope measurement error on the reported transfer estimates. What confidence intervals exist around the 13.26% transfer estimate? Furthermore, the authors assume that rainfall and irrigation water enter the soil without isotopic fractionation. Given the climatic conditions of the study area, evaporation after irrigation can be substantial and near-surface isotopic enrichment is common. Discuss the uncertainties arising from this assumption.
Response: We will strengthen the uncertainty assessment in three ways. First, the mean summed lateral-transfer ratio of 13.26% will be reported with an approximate t-based 95% confidence interval of 12.02–14.50% across the four irrigation events (n = 4). Second, the analytical precision of the isotope measurements (±0.6‰ for δ²H and ±0.2‰ for δ¹⁸O) will be distinguished explicitly from event-to-event uncertainty. Third, the Methods will no longer state that rainfall and irrigation water undergo no fractionation; instead, these inputs will be treated as conservative tracers over the first-day post-irrigation interval, while evaporative isotope enrichment will be identified as an unresolved structural uncertainty, particularly for the 0–20 cm layer. Accordingly, the lateral-transfer values will be interpreted as model-based estimates conditional on the stated assumptions.
Comment 3: Another assumption is that location-related differences in modelled evapotranspiration primarily reflect access to farmland water rather than microclimatic, structural, or physiological differences between tree groups. However, higher evapotranspiration could also arise from differences in canopy structure, wind exposure, radiation, and other site-specific factors. The article acknowledges that these estimates depend on several assumptions. A more detailed discussion of these uncertainties would strengthen the manuscript.
Response: We agree and will further narrow this interpretation. The 84.49 mm difference will be retained only as a modeled edge–far evapotranspiration contrast and will not be treated as a quantitative estimate of root-mediated uptake of farmland irrigation water. The revised manuscript will explicitly discuss possible effects of canopy structure, wind exposure, incident radiation, local microclimate, and physiological status, as well as the absence of sap-flow validation. Future work will be framed around paired sap-flow measurements together with position-specific radiation, wind, and canopy observations to separate farmland-water effects from local environmental and structural controls.
Line-specific comments
Comment 1: Description of the study site;The first explicit mention that the study site is located in China (excluding the authors' affiliations) does not appear until page 5. Moreover, this information is conveyed only through the map and its caption rather than in the main text. Please consider adding this information to Section 2.1, "Site Description".
Response: We will revise Section 2.1 so that the main text explicitly identifies Minqin Oasis as being in the Hexi Corridor, Gansu Province, northwestern China, rather than relying on the map alone.
Comment 2: Line 87: Perhaps the Köppen climate classification could be added here.
Response: We will add the regional climate classification as cold desert (BWk) and cite Beck et al. (2018).
Comment 3: Line 102 Please improve the figure caption by explicitly referring to panels (a), (b), (c), etc. In addition, panel (d) should include a clearer description of the data series. The data are from 2021 and 2023; however, the figure gives the impression that it represents two consecutive years. Please also add the positions of the Trime-TDR access tubes to the experimental layout shown in panel (c).
Response: We will revise the caption to describe panels (a)–(d) explicitly, distinguish the May–September observation periods in 2021 and 2023 so that they are not interpreted as consecutive years, and add the schematic positions of the 15 Trime-TDR access tubes to panel (c).
Comment 4: Lines 116 and 121 The same information is reported twice. Please retain one occurrence and remove the other.
Response: We will remove the duplicate description from Section 2.2 and retain the detailed 10-cm interval, 0–100 cm sampling description in Section 2.2.1.
Comment 5: Line 165 A citation to Craig (1961) would, in my opinion, be appropriate here.
Response: We will add Craig (1961) to the stable-isotope delta-notation statement in Section 2.3.1.
Comment 6: Line 165 A citation to Craig (1961) would, in my opinion, be appropriate here.
Response: We will reorganize Section 3.2 by moving the detailed event-specific infiltration and lateral-transfer estimates into a new Table 2. The main text will retain only the key layer means and principal patterns, while Fig. 3 will continue to show their variation.
Comment 7: Line 730 I could not find this reference cited in the main text.
Response: We will cite Yang et al. (2015) explicitly in Section 2.3.3 where the layerwise mass-balance formulation and its extension to lateral transfer are described, and retain the reference in the bibliography.
We thank the reviewer again for these comments. Collectively, the revisions will make the assumptions, uncertainty, and evidential limits of the isotope-based and evapotranspiration analyses more explicit while preserving the site-specific hydrological value of the dataset.
Citation: https://doi.org/10.5194/egusphere-2026-4760-AC1
-
AC1: 'Reply on RC1', Guofeng Zhu, 28 Sep 2026
-
RC2: 'Comment on egusphere-2026-4760', Ying Zhao, 22 Sep 2026
This manuscript addresses a relevant question for water-limited oasis agriculture and brings together soil-water observations, stable isotopes, root distributions, and a simple water-balance approach in a maize–poplar shelterbelt system. The field setting is valuable and the data could support a useful site-specific assessment of irrigation-associated water redistribution and seasonal water-source overlap. However, several central interpretations currently extend beyond what the measurements can demonstrate. I recommend major revision.
1. The 84.49 mm edge–far evapotranspiration difference should not be interpreted as quantified root-mediated uptake of farmland irrigation water.
The manuscript treats the difference in modelled evapotranspiration between shelterbelt-edge trees and trees farther from the field as transpiration-related water exchange. However, the FAO-56 Penman–Monteith framework provides reference evapotranspiration; it is not clear how this was converted into location-specific tree transpiration. Please provide the full parameterization and all inputs that differed between the two tree groups, including any canopy, aerodynamic, vegetation, or water-stress adjustments. It is also necessary to explain how comparability between the two groups was established.
Without independent validation, such as sap-flow measurements, tracers, or a direct demonstration of lateral root access to the cropland, the 84.49 mm value cannot be treated as the amount of irrigation water taken up from the field by shelterbelt roots. At most, it is a modelled edge–far evapotranspiration contrast that may be consistent with greater water access at the field edge. The title, abstract, Fig. 6, discussion, and conclusions should all be revised accordingly if direct supporting evidence is unavailable.
2. The interpretation of the 13.26 % lateral-transfer estimate requires a clearer control volume, area basis, and isotope-identifiability analysis.
The reported 13.26 % value is the sum of layer-specific, irrigation-depth-normalized estimates. It is not yet an area-integrated fraction of the total irrigation volume, because the farmland and shelterbelt areas have not been converted to a common reference area. In addition, the layerwise calculation attributes post-irrigation changes in soil-water storage and isotope composition to lateral transfer while other processes may also affect the signal.
Please define the control volume, reference area, time window, measured quantities, and assumed fluxes for each calculation. The result should be described consistently as a model estimate of layer-specific lateral redistribution, rather than as a directly quantified fraction of the applied irrigation water.
The isotope-based source partitioning also needs stronger diagnostics. For each relevant sampling period and location, please show whether the three aggregated soil-depth sources are distinguishable in dual-isotope space relative to their variability and analytical precision. The IsoSource settings, accepted solution ranges, and model uncertainty should be reported. Where source endmembers substantially overlap, conclusions about depth-specific uptake or lateral transfer should be reduced accordingly.
3. Water-source overlap should be kept distinct from demonstrated crop–tree competition.
A proportional similarity index of 73.35 % shows overlap in estimated source use, but it does not by itself demonstrate a competitive effect on maize. Demonstrating competition would require evidence of resource depletion or a crop response, such as differences in soil-water potential, plant water status, transpiration, growth, or yield relative to an appropriate control.
The manuscript should therefore frame its result as substantial water-source overlap and a potential risk of crop–tree competition, rather than as confirmation of competition mechanisms. The management implications should similarly be presented as hypotheses to be tested in controlled irrigation or shelterbelt-design experiments. This revision would preserve the practical relevance of the study while aligning the claims with the available evidence.
Overall, the manuscript could make a useful contribution after its quantitative claims and management implications are brought into line with the evidence actually provided.
Citation: https://doi.org/10.5194/egusphere-2026-4760-RC2 -
AC2: 'Reply on RC2', Guofeng Zhu, 28 Sep 2026
Public Author Reply to Referee #2
Manuscript: egusphere-2026-4760
Dear Referee #2,
Thank you for the detailed and constructive assessment. We agree that several interpretations in the original manuscript should be brought into closer alignment with what the measurements can directly support. We have prepared corresponding revisions and, if invited to submit a revised manuscript after the interactive discussion, will incorporate them in full. The principal changes are to treat the 84.49 mm result as a modeled edge–far evapotranspiration contrast, to define the 13.26% result explicitly as an irrigation-depth-normalized model estimate, and to distinguish water-source overlap from demonstrated crop–tree competition.
Comment 1: The 84.49 mm edge–far evapotranspiration difference should not be interpreted as quantified root-mediated uptake of farmland irrigation water.
Response: We agree. Section 2.3.6 will clarify that the FAO-56 Penman–Monteith equation provides reference evapotranspiration (ET₀), representing atmospheric evaporative demand over a standardized reference surface rather than direct tree transpiration. The meteorological inputs used in the calculation will be stated explicitly, and we will clarify that tree-group-specific canopy, aerodynamic, vegetation, and water-stress parameters were not independently constrained. We will also make clear that comparability between shelterbelt-edge trees and trees farther from the farmland was not independently demonstrated, because canopy structure, wind exposure, incident radiation, local microclimate, and physiological status may also contribute to the contrast. In the absence of sap-flow measurements, tracer validation, or direct measurements of lateral root access, the 84.49 mm value will therefore be interpreted only as a modeled edge–far evapotranspiration contrast, not as quantified root-mediated irrigation-water uptake. The title, Abstract, Section 2.3.6, Section 3.1, Fig. 2, Fig. 6, Discussion, and Conclusion will be revised consistently. The 19.56% value will be retained only as the numerical equivalence of the contrast relative to applied irrigation depth, not as an irrigation-water uptake fraction.
Comment 2: The interpretation of the 13.26 % lateral-transfer estimate requires a clearer control volume, area basis, and isotope-identifiability analysis.
Response: We agree that the 13.26% value should not be interpreted as an area-integrated fraction of total irrigation volume. The revised Methods will define the control volume as the monitored 0–100 cm soil profile represented by the 0–20, 20–60, and 60–100 cm layers, with the calculation based on pre- and post-irrigation water storage and isotope composition over the first-day post-irrigation interval. For each event, layer-specific lateral-transfer depths will be summed and normalized by the corresponding irrigation depth. Because no common farmland–shelterbelt reference-area conversion is applied, the result will be described consistently as an irrigation-depth-normalized model estimate of layer-specific lateral redistribution. The 13.26% mean will also be reported with an approximate t-based 95% confidence interval of 12.02–14.50% across the four events. The isotope mass-balance assumptions will be clarified by treating rainfall and irrigation inputs as conservative tracers over the first-day interval and by identifying post-irrigation evaporative enrichment as an unresolved structural uncertainty.
For the plant water-source partitioning, Section 2.3.2 will report the IsoSource settings explicitly: a 1% source increment and a mass-balance tolerance of 0.1‰, following Phillips and Gregg (2003). All source-proportion combinations satisfying the mass-balance criterion within this tolerance will be treated as feasible solutions. The revised Discussion will state that aggregation into three depth intervals can mask within-layer isotope heterogeneity and that overlap among source signatures can broaden the feasible-solution space and reduce source identifiability. Accordingly, the reported source proportions will be interpreted as conditional estimates within the three-source framework rather than as uniquely resolved depth-specific uptake fractions, and depth-specific conclusions will be reduced where endmember overlap is substantial.
Comment 3: Water-source overlap should be kept distinct from demonstrated crop–tree competition.
Response: We agree and will distinguish these concepts consistently throughout the manuscript. Section 2.3.4 will state explicitly that the proportional similarity index quantifies overlap in relative resource use and does not itself measure competitive effects. The Results and Abstract will therefore describe the mean PS value of 73.35% as substantial water-source overlap rather than as evidence of demonstrated competition. The manuscript title will be revised from “Competition Mechanisms” to “Water-Use Interactions.” Management implications concerning irrigation scheduling and shelterbelt design will be presented as testable hypotheses requiring controlled irrigation and shelterbelt-design experiments, rather than as demonstrated measures for reducing crop–tree competition. The Conclusion will be revised to preserve the same distinction between observed source overlap and potential competitive interactions.
We thank the reviewer again for these comments. The planned revisions will retain the practical relevance of the field observations while ensuring that the quantitative claims, uncertainty statements, and management implications remain within the evidential limits of the measurements.
Citation: https://doi.org/10.5194/egusphere-2026-4760-AC2
-
AC2: 'Reply on RC2', Guofeng Zhu, 28 Sep 2026
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 170 | 73 | 87 | 330 | 77 | 71 |
- HTML: 170
- PDF: 73
- XML: 87
- Total: 330
- BibTeX: 77
- EndNote: 71
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The paper's main value lies in combining isotope evidence, soil-water observations, and a simple water-balance framework to quantify crop-shelterbelt interactions in an important oasis agricultural system. The practical relevance is high because the Minqin Oasis is one of the most water-limited agricultural environments in China. The management implications are therefore important, even if the methods themselves are not fundamentally new.
However, I see several weaknesses here, particularly in the discussion of the results themselves and in the determination of uncertainties for the modelled values.
The study relies on isotopic compositions obtained using the LI-2100 system, which can be regarded as an automated equivalent of cryogenic vacuum extraction. Recent studies have demonstrated that bulk stem water extracted by cryogenic methods may differ isotopically from mobile xylem sap due to within-stem isotopic heterogeneity. Because the primary conclusions regarding maize-poplar water-source partitioning and competition depend on stem-water isotopic composition, the authors should discuss the potential implications of this methodological limitation and evaluate how it might affect estimates of deep-soil-water uptake and water-source overlap.
The paper provides numerical estimates but lacks a rigorous uncertainty assessment and independent validation. The authors should provide a sensitivity or uncertainty analysis to evaluate the influence of isotope measurement error on the reported transfer estimates. What confidence intervals exist around the 13.26% transfer estimate? Furthermore, the authors assume that rainfall and irrigation water enter the soil without isotopic fractionation. Given the climatic conditions of the study area, evaporation after irrigation can be substantial and near-surface isotopic enrichment is common. Discuss the uncertainties arising from this assumption.
Another assumption is that location-related differences in modelled evapotranspiration primarily reflect access to farmland water rather than microclimatic, structural, or physiological differences between tree groups. However, higher evapotranspiration could also arise from differences in canopy structure, wind exposure, radiation, and other site-specific factors.
The article acknowledges that these estimates depend on several assumptions. A more detailed discussion of these uncertainties would strengthen the manuscript.
Line-specific comments
Description of the study site
The first explicit mention that the study site is located in China (excluding the authors' affiliations) does not appear until page 5. Moreover, this information is conveyed only through the map and its caption rather than in the main text. Please consider adding this information to Section 2.1, "Site Description".
Line 87
Perhaps the Köppen climate classification could be added here.
Line 102
Please improve the figure caption by explicitly referring to panels (a), (b), (c), etc. In addition, panel (d) should include a clearer description of the data series. The data are from 2021 and 2023; however, the figure gives the impression that it represents two consecutive years. Please also add the positions of the Trime-TDR access tubes to the experimental layout shown in panel (c).
Lines 116 and 121
The same information is reported twice. Please retain one occurrence and remove the other.
Line 165
A citation to Craig (1961) would, in my opinion, be appropriate here.
Line 308
There are perhaps too many numerical values presented directly in the text, which makes the section difficult to follow. Consider moving some of this information to a table.
Line 730
I could not find this reference cited in the main text.