Water–Salt–Root Interactions in Drip-Irrigated Arid Shelterbelts: Toward Predictive Root-Zone Regulation

Abstract

Arid and semi-arid shelterbelts must provide long-term ecological protection under chronic water scarcity, high evaporative demand, and rising salinization risk, yet management still lacks an integrated framework linking irrigation, root-zone salt dynamics, and woody plant performance. Here, we synthesize evidence on water–salt–root linkages in drip-irrigated shelterbelts and related dryland woody systems from a structured Web of Science Core Collection search (1 January 2000–1 January 2026). The evidence shows that shelterbelt performance is governed not by water or salinity alone, but by a coupled root-zone system: localized irrigation creates moisture–salt heterogeneity, salts accumulate near evaporative fronts and emitter margins, and roots redistribute depth, density, and uptake zones. In hyper-arid saline-drip systems, precipitation may be only ~24.6 to <50 mm yr⁻¹, evaporation > 3000–3639 mm yr⁻¹, groundwater salinity 2.8–29.7 g L⁻¹, active roots 20–80 cm, and salt mainly in the 0–20 cm surface layer. Irrigation thus acts as both the basis of establishment and a source of long-term vulnerability, particularly where saline groundwater or other non-conventional water sources are used. Management options can improve root-zone habitability, but shelterbelt-specific thresholds and integrated indicators remain limited. This review proposes a root-zone-centered framework supporting predictive regulation.

1. Introduction

Drylands are now central to global sustainability rather than peripheral to it. They cover about 46% of the Earth’s land surface and support approximately 3 billion people, yet remain disproportionately exposed to desertification, biodiversity pressure, and climate-driven water instability [1,2,3,4]. Formal protection is also weak: protected-area coverage in drylands is only about 12%, compared with 21% outside drylands, and 95–100% of current natural dryland habitat is projected to experience some degree of land conversion by 2100 [2]. Within these landscapes, shelterbelts and windbreaks are not simply linear plantations or agricultural accessories, but multifunctional ecological infrastructures that reduce wind erosion, stabilize sand surfaces, protect roads and farmland, moderate microclimate, and sustain production and habitat functions under severe environmental stress [5,6]. Their practical value is well established: in Northeast China, shelterbelts contributed 4.68%, 4.28%, and 9.45% to maize-yield increases in high-, middle-, and low-climatic-potential productivity zones, respectively, although the average protected farmland proportion was only 18.28%, far below the approximate 80% optimal level [7,8]. However, the dryland context fundamentally distinguishes arid and semi-arid shelterbelts from more humid woody systems. Dryland vegetation greenness has become increasingly precipitation-sensitive, with a reported sensitivity increase of 0.624% yr⁻¹ in drylands but a decrease of −0.618% yr⁻¹ in wet regions [3]. Unlike those in favorable forest climates, arid shelterbelts are planted at or beyond the hydrological margins of woody persistence, where even modest shifts in water balance can alter long-term viability [9,10].

Dryland tree planting cannot be treated as an undifferentiated restoration good. Recent work has shown that dryland afforestation involves strong trade-offs among carbon goals, biodiversity, hydrology, and ecosystem appropriateness, and that reforestation is often wrongly conflated with restoration in non-forest systems [5,6]. For shelterbelts, the key question is therefore not whether trees are desirable in principle, but when woody belts remain ecologically justified and operationally sustainable. In arid and semi-arid regions, this question is sharpened because shelterbelts are expected to deliver durable protective functions under chronic water deficit, strong evaporative demand, mineralized groundwater, and salinization risk [2,3,4,5]. Even when they improve adjacent cropland microclimate or productivity, the whole-system water balance may remain tight or even unfavorable, depending on configuration, field size, species composition, and irrigation source [7,11]. For example, in Central Asian shelterbelt scenarios assuming 4 ha fields, total crop–shelterbelt water consumption was higher for all investigated crops except pear, indicating that water-resource benefits depend on field geometry, crop type, and tree water use [11]. Thus, shelterbelt research requires a systems perspective in which ecological services, water cost, and long-term carrying capacity are evaluated together [5,11,12].

The central practical constraint is that many arid shelterbelts cannot persist without irrigation, yet the water sources that enable establishment may also intensify long-term vulnerability. Across water-scarce regions, non-conventional water resources, including saline groundwater, recycled wastewater, drainage water, and brackish sources, are becoming increasingly important for agriculture and vegetation maintenance [13,14]. However, non-traditional water sources can cost approximately 1.5–4 times more than traditional supplies, with desalinated seawater reported at the upper end of about USD 3.3 m⁻³ [13]. This transition expands the feasibility of shelterbelt establishment in desert margins and oasis–desert transition zones, but shifts the management challenge from water supply alone to water-quality compatibility, infrastructure reliability, and long-term soil response [15,16]. At the same time, improvements in irrigation technology do not automatically guarantee real water conservation, because gains in application efficiency can be offset by salinity accumulation, rebound use, or poor adoption conditions [17,18]. This is especially important where irrigation water is saline or brackish, because each irrigation event may simultaneously relieve water deficit and add salts to the root-zone system [16,17]. In arid woody systems, these constraints are severe because the goal is persistent ecological protection, not short-term biomass maximization [19]. Irrigation dependence is therefore a structural condition rather than a temporary technical issue, positioning the root zone at the center of both survival and failure.

The root zone becomes the key interface because arid irrigation rarely produces a homogeneous soil environment [20]. Under drip or localized irrigation, water is redistributed into discrete wet bulbs, while salts move toward evaporative fronts, emitter margins, or specific depth intervals; these patterns are modified by soil texture, irrigation interval, groundwater influence, and root uptake [21,22,23]. In drylands, salinization is increasingly recognized as a multi-scale process linking climate, vadose-zone transport, irrigation practice, and groundwater dynamics rather than as a static soil property [24,25,26]. Global studies have shown that plant water stress is governed by measurable hydraulic thresholds: the global mean critical soil-moisture threshold is approximately 0.19 m³ m⁻³, ranging from 0.12 m³ m⁻³ in arid ecosystems to 0.26 m³ m⁻³ in humid ecosystems [27,28,29]. These advances are relevant to shelterbelts because survival is controlled not by bulk water or salt averages, but by whether active roots remain connected to favorable moisture microsites while avoiding recurrent salt accumulation [30,31,32]. A global synthesis of tree water uptake reported seasonal water-uptake-depth shifts in 74 of 120 reviewed studies, while a meta-analysis of root-system architecture screened more than 2000 records and synthesized 70 controlled-condition studies, showing that root adjustment is measurable and highly responsive to water stress [32,33]. Root plasticity, water uptake depth, and above-/below-ground drought adjustment are therefore key biological processes through which irrigation design, salinity, and climate are translated into whole-plant performance [33,34,35].

Despite growing interest in dryland woody systems, the evidence base remains fragmented. (i) Many studies describe irrigation amount, EC_e, groundwater depth, soil moisture, or plant response, but fewer studies translate these variables into comparable soil–root–plant decision criteria. (ii) A substantial part of the mechanistic evidence still comes from orchards, irrigated croplands, and other woody systems rather than from true shelterbelts, creating a transferability problem. (iii) Below-ground processes are less standardized than above-ground observations, although isotopic tracing, HYDRUS-based simulation, ET-based indicators, microtensiometers, and remotely sensed root-zone salinity products are increasingly available. (iv) Most studies remain plot- or site-based, whereas shelterbelt sustainability also depends on stand continuity, protective function, and regional water–salt constraints.

Against this background, this review synthesizes water–salt–root linkages in arid and semi-arid shelterbelts, with particular attention to drip-irrigated and salinity-affected systems. Predictive root-zone regulation is defined here as an adaptive management approach that uses measurable soil, root, plant, and hydrological indicators to anticipate root-zone stress before irreversible shelterbelt decline occurs. Rather than treating water scarcity, salinity, root response, and management practice as separate themes, the review frames them as a coupled root-zone regulation problem. The review addresses three linked questions:

(1)

What environmental constraints define arid shelterbelt survival.

(2)

How water–salt–root mechanisms shape shelterbelt performance.

(3)

How diagnostic indicators, evidence-tiered benchmarks, and management options can support predictive regulation.

Its main contribution is to distinguish direct shelterbelt evidence from transferable evidence and to clarify how quantitative values should be interpreted without overstating their management validity.

2. Methodology

Literature searches were conducted in the Web of Science Core Collection (WoSCC; Clarivate, London, UK), covering the period from 1 January 2000 to 1 January 2026. WoSCC was selected as the primary database because it provides multidisciplinary coverage, relatively consistent bibliographic metadata, and reproducible search fields suitable for transparent literature identification and citation tracking. Searches were conducted using the Topic field, which includes title, abstract, author keywords, and Keywords Plus. No additional language filter was applied during the database search; however, records without sufficient methodological or result information for evidence extraction were excluded during screening. The search strategy combined four concept blocks: shelterbelt or windbreak systems, drip or micro-irrigation, salinity or saline-water constraints, and arid or dryland environmental contexts. The main search string was:

TS = ((“shelterbelt*” OR “windbreak*” OR “protective forest*” OR “tree belt*” OR “woody belt*” OR “buffer strip*” OR “hedgerow*”) AND (“drip irrigation” OR microirrigation OR “micro irrigation” OR “subsurface drip” OR “trickle irrigation” OR fertigation) AND (salin* OR sodic* OR “saline groundwater” OR “brackish water” OR “soil salinity” OR “electrical conductivity” OR EC_e OR EC_w OR SAR OR groundwater OR “water table”) AND (arid OR semi-arid OR hyper-arid OR dryland* OR desert* OR oasis)).

Because shelterbelt-related studies are sometimes indexed under different system labels, two supplementary sensitivity searches were also conducted. The first broadened the system terms to include arid woody plantations, shrubs, and comparable irrigated woody systems; the second targeted sand-control afforestation, desert plantations, shelter forests, and desert-highway vegetation systems. These supplementary searches were used to reduce the risk of missing mechanistically relevant studies on wetted-bulb formation, salt redistribution, root distribution, and salinity-related plant responses. The complete search strategy is provided in Supplementary Material.

All records retrieved from WoSCC were merged in EndNote 20 (Clarivate, London, UK) and deduplicated using title, author, year, and DOI fields, followed by manual verification where necessary. The initial search identified 103 records. After removing 24 duplicates, 79 records were screened by title, abstract, and keywords. Nine records were excluded at this stage because they were clearly outside the review scope, such as studies without an arid or dryland context, without a salinity or groundwater-salinity component, or without a drip/micro-irrigation component. The remaining 70 records were assessed at full-text level. No additional records were excluded after full-text assessment because all 70 records met at least one of the predefined evidence categories described below and contained extractable information relevant to the synthesis. These 70 studies were therefore retained in the final evidence base (Figure 1).

Eligible studies were classified into two evidence categories. The core shelterbelt evidence set included studies that directly addressed arid, semi-arid, or hyper-arid shelterbelts, windbreaks, protective forests, tree belts, shelter forests, or roadside woody shelter systems, and that included a drip or micro-irrigation component together with salinity, sodicity, saline/brackish water, groundwater, or water-table processes relevant to salt accumulation or leaching. To be included in the core set, a study also needed to provide at least one extractable evidence component linking soil moisture–salt heterogeneity, root traits or architecture, and/or tree or stand performance. The transferable background evidence set included studies from arid woody plantations, orchards, shrublands, or comparable irrigated woody systems that did not directly study shelterbelts but provided transferable evidence on wetted-bulb dynamics, salt redistribution, root distribution shifts, salinity thresholds, irrigation responses, modeling approaches, or indicator reporting. The structure of the final evidence base, including its temporal, geographical, and thematic composition, is summarized in Figure 2.

To avoid overextending transferable evidence, the synthesis followed an explicit evidence-use hierarchy. Direct shelterbelt evidence was used as the primary basis for shelterbelt-specific interpretation and management implications. Evidence from related woody systems, including orchards, woody plantations, and riparian shrubs, was used only to support mechanisms, methods, or indicative benchmarks. Evidence from irrigated croplands or general agricultural systems was used mainly as background process evidence or methodological support. When quantitative or management-relevant values were derived from non-shelterbelt systems, they were explicitly treated as transferable evidence and were not presented as direct shelterbelt management thresholds.

For each included study, information was extracted using a structured coding framework. The extracted fields included publication year, country or region, system type, climatic setting, water-source and salinity indicators, groundwater or water-table descriptors, soil moisture and salinity profiles, salt accumulation or leaching patterns, irrigation type and frequency, emitter or wetted-zone descriptors where available, root depth, fine-root distribution, root length density, root biomass allocation, plant growth, survival, photosynthesis, biomass, productivity, water-use indicators, and monitoring or modeling methods. Screening and initial coding were conducted according to the predefined criteria by the first author, and ambiguous cases and evidence classification were checked through discussion among the co-authors. The final evidence base was then synthesized narratively around four themes: environmental constraints, coupled water–salt–root mechanisms, irrigation dependence and management bottlenecks, and options for predictive root-zone regulation.

3. Environmental Context of Arid Shelterbelts

3.1. Aridity Constraints

Arid and semi-arid shelterbelts face one of the most restrictive ecological settings on Earth. Although drylands receive less formal conservation protection than non-drylands, with protected-area coverage of only about 12% compared with 21% outside drylands [2], the key issue here is operational: shelterbelts must stabilize soil, reduce wind erosion, protect transport corridors and farmland, and maintain vegetation cover under chronic atmospheric water deficit. Their establishment therefore begins at the margin of woody plant persistence rather than within a favorable forest environment. This climatic boundary is becoming more unstable rather than less. Dryland vegetation greenness has shown a robust increase in precipitation sensitivity of 0.624% yr⁻¹, whereas wet regions exhibited a decrease of −0.618% yr⁻¹, indicating that dryland vegetation is becoming more tightly coupled to fluctuations in water supply [3]. At the same time, dryland productivity is increasingly shaped by interacting climatic and anthropogenic pressures rather than by rainfall amount alone [4]. Thus, long-term sustainability is constrained by vegetation carrying capacity, water availability, and system multifunctionality, rather than by short-term tree survival alone [5]. For shelterbelts, this means that persistence depends on whether water inputs, stand structure, and soil processes remain within a narrow ecohydrological operating range over time. This range should be interpreted dynamically: warming, precipitation variability, and extreme dry-down events may narrow it by increasing evaporative demand, accelerating salt concentration between irrigation or rainfall events, and reducing the reliability of fixed irrigation intervals [21].

3.2. Salinity and Groundwater

In drylands, water limitation rarely occurs in isolation. It is commonly accompanied by salinity, groundwater mineralization, poor natural leaching, and strong evaporative concentration at or near the soil surface [21,22]. Plant failure in shelterbelts is often determined not only by how much water is present, but also by where salts accumulate relative to the active root zone. Global projections already identify several dryland salinization hotspots, including South America, southern and western Australia, Mexico, the southwestern United States, and South Africa, while also indicating that the magnitude and spatial consistency of future salinity change vary across climate scenarios (Figure 3) [21]. At the same time, recent syntheses emphasize that salinization in drylands is driven by multi-scale interactions among climate, irrigation practice, shallow groundwater, and soil transport processes [23]. Under these conditions, the practical target is not water supply alone, but the coordinated control of soil moisture, salt redistribution, and root-zone habitability [36,37].

Groundwater conditions further tighten this survival window. In arid riparian ecosystems, shallow groundwater depth, lower groundwater salinity, and higher soil moisture are associated with higher vegetation diversity [38]. In another desert riparian system, groundwater depth and soil factors jointly explained 85.8% of vegetation variance, and species richness, above-ground biomass, coverage, community height, foliage projective cover, and leaf area index all declined as groundwater depth increased [39]. Threshold analysis along the Manasi River indicated that riparian vegetation could generally be supported when the water table remained shallower than about 6 m, while diversity peaks occurred at approximately 2–3 m for herbs, 3–5 m for shrubs, and 2–4 m for the overall plant assemblage [40,41]. However, the ~6 m value should be treated as a regional ecological reference rather than a universal groundwater threshold. Its applicability depends on species composition, rooting depth, soil texture, groundwater salinity, capillary-rise capacity, and evaporative demand. In fine-textured or strongly capillary soils, shallow saline groundwater may intensify upward salt flux, whereas coarse aeolian sands may reduce capillary continuity but also provide weak water retention and limited buffering capacity [41]. For arid shelterbelts, groundwater can therefore buffer drought or intensify salinity stress, depending on its depth and mineralization [42,43,44].

3.3. Wind and Soil

Shelterbelts in arid regions are usually built where aeolian stress and episodic sand and dust storm (SDS) hazards are already severe. Their windbreak function is needed precisely because shifting sand, abrasion, burial, and surface erosion are persistent threats to infrastructure and vegetation [45]. At the regional scale, surface-observation analysis across Asia from 2000 to 2022 showed that moderate and severe dust events declined significantly, but this decline was mainly attributed to fewer strong-wind days, with a contribution exceeding 50%; when daily maximum wind speed exceeded 13.0 m s⁻¹, the probability of severe dust events increased [46]. This means that shelterbelt performance must be evaluated under episodic wind extremes, not only under mean wind conditions. In Kansas–Nebraska agricultural windbreaks, protected-field yield increases averaged 16% for soybean and 10% for wheat when significant positive effects occurred [47]. More broadly, shelterbelt height, porosity, orientation, width, continuity, and uniformity jointly determine wind-reduction efficiency; in protected farmland, reported yield benefits commonly range from 10% to 30%, and optimum optical porosity is often around 0.20–0.40 [48]. At the hyper-arid end of this gradient, the Tarim Desert Highway shelterbelt was constructed to protect a corridor crossing the Taklimakan Desert, where severe sand-disaster environments still account for about 37.1% of the highway length [49]. In this system, reported mean annual wind speed is about 2.5 m s⁻¹, maximum instantaneous wind speed reaches 20.0 m s⁻¹, and the number of sand-shifting windy days exceeds 130 days per year [50]. Thus, shelterbelt vegetation must withstand the same erosive forces it is intended to weaken, making the challenge mechanical as well as physiological.

The soil environment compounds this stress. Hyper-arid shelterbelts are often established on coarse aeolian sands with low organic matter, weak aggregation, limited nutrient retention, and poor water-holding capacity [51]. Yet long-term shelterbelt construction can gradually modify these substrates. After 7 years of saline-water irrigation in the Desert Highway shelterbelt, soil aggregate size and stability increased, nutrient accumulation improved, and normal growth of salt-adapted plants was maintained [51]. Aggregate formation has also been shown to strengthen with planting time, irrigation-water salinity, and species effects [52]. Under long-term shelterbelt construction, the 0–5 cm soil layer showed marked improvements in water retention: saturated water content increased by 4.42% and 12.67%, field capacity by 68.9% and 70.41%, and available water content by 87.84% and 70.97% in two shelterbelt microsites relative to shifting sand [53]. Surface salt crusts can further suppress evaporation, with an inhibition efficiency of 33.0% in shelterbelt soil compared with 13.8% in bare sand [54], although these same crusts also reflect active salt accumulation and redistribution around emitters [55]. Thus, shelterbelt soils are gradually transformed by irrigation, deposition, vegetation, and salinity into a more structured but also more management-sensitive root environment.

3.4. Irrigation Dependence

Water dependence in protective forests should be interpreted along a gradient rather than as a single hyper-arid pattern. Many protective forests and shelterbelts worldwide are not designed around saline groundwater drip irrigation, but their long-term stability still depends on whether water availability, stand structure, and management intensity remain compatible. For example, the Great Plains Shelterbelt Project in the United States had established 30,233 shelterbelts by 1942, extending 29,927 km and protecting approximately 1.62 million ha across 30,200 farms; about 80% of the shelterbelts were rated as good or better in 1944, but only 42% remained in good condition by 1954 because of drought, pests, grazing, removal, and insufficient management [48,56]. This example shows that even non-hyper-arid protective forests can decline when water stress and management deficits accumulate. At the opposite end of this gradient, hyper-arid sandy shelterbelts with negligible rainfall and extremely high evaporation may require continuous external water inputs for persistence. One well-documented saline-drip-irrigated system in the Tarim Basin extends roughly 436–437 km in length, is generally 72–78 m wide, and occupies about 3128 ha [50,57]. It is supplied by 108 wells, each irrigating about 3.12 × 10³ m², with groundwater salinity averaging around 5 g L⁻¹ and ranging from about 2.8 to 29.7 g L⁻¹ [58,59]. Mean annual precipitation is only about 24.6 mm or, more broadly, <50 mm, whereas annual evaporation exceeds 3000 mm and reaches 3639 mm in some reports [51,59]. Under such conditions, freshwater-free persistence is unrealistic; external irrigation is not supplementary but foundational [57].

Once shelterbelt persistence depends on external or groundwater-linked water supply, irrigation creates a narrow operational window rather than a simple solution. In the saline-drip system, soil moisture declines approximately linearly within a typical 10-day irrigation cycle, whereas soil electrical conductivity commonly increases from day 2 to day 6 and then decreases from day 6 to day 8, indicating that water depletion and salt concentration occur on overlapping timescales [58]. Soil-water storage can be divided into a 0–20 cm quickly changing layer, a 20–60 cm active layer, a 60–100 cm weakly changing layer, and a deeper regulated layer [60]. Long-term saline drip irrigation mainly intensifies salinity in the 0–20 cm or 0–10 cm surface layer, while deeper soil conditions remain comparatively more buffered [59]. However, this depth pattern should be treated as a site-specific outcome of sandy texture, saline groundwater, emitter placement, and high evaporative demand rather than as a universal shelterbelt rule. A broader water-balance constraint is also evident in semi-arid sand-fixation forests. In southern Horqin Sandy Land, groundwater declined linearly from 1953 to 2009, while land-use change strongly altered regional water consumption; agricultural land and broadleaved forests accounted for more than 40% and 20% of total water consumption, respectively, whereas Pinus sylvestris var. mongolica plantations accounted for only about 12% [61]. Because approximately 95% of P. sylvestris var. mongolica roots were distributed within the upper 100 cm soil layer, declining groundwater and reduced accessible water contributed to plantation decline [61]. These cases show that irrigation dependence is a system-level water-balance problem involving external water inputs, salinity and leaching, groundwater accessibility, vegetation demand, and long-term management.

This leads to the central survival–management paradox of arid shelterbelts: the same irrigation that enables establishment can also become the main pathway for long-term ecological risk. Root distribution in a saline-drip-irrigated sandy shelterbelt is highly sensitive to microsite condition and age; in hardened sand, roots are concentrated mainly in the 0–40 cm layer, whereas deeper rooting occurs in flatter sandy ground and ridge sand [62]. Under different irrigation regimes, saline groundwater can support plant survival, but irrigation amount and interval strongly alter soil-water storage and plant water use [63]. Species also differ in their tolerance and growth under high-salinity drip irrigation, indicating that shelterbelts cannot be managed as uniform tree rows [64]. Overall, sustainable dryland afforestation requires that water demand, salinity control, rooting behavior, and vegetation carrying capacity be addressed together rather than separately [5,36].

4. Coupled Water–Salt–Root Mechanisms in Arid Shelterbelts

4.1. Root-Zone Heterogeneity

As conceptualized in Figure 4, drip irrigation reorganizes the root zone into spatially heterogeneous moisture–salt microsites, linking environmental forcing, wet-bulb formation, root redistribution, shelterbelt persistence, and protective performance. Under drip irrigation, the root zone is not wetted uniformly. Instead, irrigation creates a localized wet bulb, while salts are displaced toward the outer margins and upper evaporative front. In arid environments, this pattern is intensified by strong evaporation, weak natural leaching, and, in many shelterbelt settings, aeolian disturbance. Consequently, water-rich and salt-rich microsites coexist, and plant performance depends on whether roots can exploit the former while avoiding the latter [31,65,66,67]. Modeling and synthesis work further indicate that this heterogeneity is controlled by emitter discharge, irrigation interval, dripline position, soil texture, root distribution, and the degree of compensatory root water uptake, so the same irrigation amount can produce very different root-zone environments across systems [68,69]. For shelterbelts, this heterogeneity matters because root-zone mismatch can reduce canopy continuity and weaken belt-level protective functions.

In saline-drip-irrigated sandy shelterbelts, this heterogeneous pattern is especially distinct because soil water depletion and salt concentration unfold on overlapping timescales. Along the Desert Highway shelterbelt, soil moisture within the upper profile declined sharply during an irrigation cycle, while surface-layer salinity rose as evaporation and plant uptake concentrated salts near the topsoil. For 15-day and 10-day irrigation cycles, topsoil moisture decreased from 27.4% to 2.4% and from 26.4% to 2.7%, respectively, whereas topsoil electrical conductivity increased from 0.64 to 3.32 dS m⁻¹ and from 0.70 to 3.99 dS m⁻¹ [58,59]. Over longer shelterbelt development, irrigation salinity mainly affected the 0–20 cm layer, while soil below 20 cm remained comparatively stable [59]. Around emitters, soil salt crust EC was also spatially anisotropic: high values occurred at 20–30 cm upslope from emitters but shifted to 40–50 cm downslope, confirming that microtopography and emitter position influence the exact location of salt concentration bands [55]. These patterns indicate that drip-irrigated shelterbelts are regulated less by mean soil salinity than by the geometry of moisture–salt segregation in the active rooting zone [62].

Transferable evidence from broader arid drip-irrigated systems supports this mechanism, but should not be treated as direct shelterbelt evidence. Numerical simulations that explicitly incorporated fine-root growth under mulched drip irrigation described a semi-elliptic cylindrical wet bulb and a salt accumulation zone centered around 40–50 cm depth before the next irrigation event [70]. In drip-irrigated orchards, root-weighted soil moisture is commonly higher and root-weighted EC_e lower than arithmetic profile means, indicating that root proliferation preferentially occurs where moisture is higher and salinity lower [71]. In nectarine, root growth decreased above an EC_e threshold of about 4 dS m⁻¹ [71]. This value is an indicative woody-crop benchmark only; for shelterbelts, the transferable point is that the relevant ecological unit is not the bulk soil profile, but the fine-scale overlap among wetted soil, salt exclusion fronts, and root occupancy.

4.2. Root Redistribution and Plasticity

Root plasticity is the biological mechanism that converts water–salt heterogeneity into either persistence or decline. Under drought and salinity, roots do not merely shrink; they reorganize their depth, branching, fine-root morphology, hydraulic function, and uptake distribution [34]. Across woody systems, drought adjustment commonly involves altered fine-root allocation, shifts in rooting depth, and stronger integration between below-ground and above-ground trait responses rather than isolated changes in single traits [34]. Woody plants display distinct fine-root strategies under altered precipitation, and below-ground responses in dry environments are often stronger and more plastic than simple biomass summaries suggest [72,73]. For saline-drip-irrigated shelterbelts, the key question is therefore whether roots can reposition into water-accessible layers before salinity and evaporation degrade them.

For drought- and salt-adapted desert shrubs, such as Haloxylon ammodendron, root redistribution often appears as a vertical decoupling between surface salt accumulation and deeper root activity. In saline-drip-irrigated shelterbelt conditions, H. ammodendron roots are mainly distributed in the 20–80 cm layer, whereas salts are concentrated primarily in the 0–20 cm surface soil [63]. Net photosynthetic rate is most closely related to water status in the main root activity layer rather than to bulk surface conditions [74]. Root system position also varies with substrate condition: in hardened sand, roots are concentrated mainly in the 0–40 cm layer, while in flatter sandy ground and ridge sand they extend more deeply [62]. Thus, shelterbelt plants persist not because salinity is absent, but because active roots remain partly displaced away from the most saline surface band.

Irrigation regime changes this distribution further by altering both carbon allocation and hydraulic behavior. Under irrigation intervals of 1, 2, 4, 8, and 12 weeks, the biomass allocation pattern of H. ammodendron and Calligonum mongolicum changed significantly, and the root-to-stem ratio and feeder-root-to-assimilating-leaf ratio increased under more water-limited desert conditions [75]. For H. ammodendron, leaf-scale photosynthesis remained comparatively stable under 1-, 2-, and 4-week intervals, whereas 12-week intervals significantly inhibited photosynthesis in both species [75]. Apparent hydraulic conductance for H. ammodendron was about 2.7–2.8 mmol m⁻² s⁻¹ MPa⁻¹ under 1–4 week intervals but declined to 2.5 and 2.3 mmol m⁻² s⁻¹ MPa⁻¹ under 8- and 12-week intervals [75]. Irrigation interval therefore reshapes not only water supply, but also the structure and operating efficiency of the plant-hydraulic pathway.

Comparable arid woody-system evidence points in the same direction. In southern Nevada, more than 60% of fine-root length and biomass of Tamarix ramosissima occurred at 20–60 cm depth, where moisture and nutrients were more favorable [76]. Under stronger salinity, fine roots became scarce in the surface layer, and at 200–400 mM NaCl their vertical distribution was clearly constrained by salinity rather than by moisture alone [77]. In the lower Tarim River, Hydrus-1D simulations indicated that 56.8% of T. ramosissima roots were concentrated in the 60–100 cm layer, while the 0–60 cm layer accumulated substantial salt and the zone below 140 cm became aeration-limited due to shallow groundwater [78]. Stable-isotope evidence from the Gurbantünggüt Desert further showed that T. ramosissima shifted from using 75% shallow–middle soil water early in the season to about 80% deep soil water in August and September, and tracer uptake confirmed root access to depths > 500 cm [79]. Together, these findings show that deepening, redistribution, and seasonal switching of uptake sources are central to woody survival in saline arid soils.

4.3. Plant Performance Responses

Because water and salinity are partitioned unevenly in space, whole-plant performance is also spatially mediated. In the Taklimakan Desert, saline groundwater irrigation at 17.5, 25, and 35 mm per event produced species-specific photosynthetic responses, and H. ammodendron was more drought-adapted than C. mongolicum [74]. Net photosynthetic rate was primarily associated with soil moisture status in the main root activity layer, not with mean topsoil conditions. Under higher water availability, C. mongolicum expressed higher photosynthetic potential, whereas H. ammodendron maintained stronger performance under drier conditions [74]. This indicates that root-zone heterogeneity is translated into leaf function through species-specific rooting depth, hydraulic conductance, and osmotic tolerance.

Water flux measurements support the same conclusion. Under identical irrigation treatment, stem sap flow of C. mongolicum on sunny days was 1.5–5.3 times that on dusty days, while H. ammodendron varied by 3.5–5.5 times between the same weather classes [80]. Thus, atmospheric forcing modulates the expression of root-zone access and plant water status. Growth outcomes also remain species-specific. Under saline drip irrigation with groundwater salinity around 4.04 g L⁻¹, three shelterbelt afforestation species all survived, but their growth and ecological suitability differed substantially [64]. Shelterbelt performance therefore emerges from the interaction among root placement, root-zone salt partitioning, meteorological demand, and species-level physiology, rather than from irrigation volume alone.

4.4. Indicators and Diagnostic Metrics

The literature converges on a practical implication: water–salt–root interactions can only be interpreted when soil, root, and plant indicators are measured jointly and at compatible spatial scales. The most informative soil descriptors are EC_w/EC_e, depth-resolved soil moisture, salinity position relative to emitters, groundwater depth, and the timing of salt accumulation between irrigation events [55,58,59]. On the root side, the most useful metrics are rooting depth, fine-root biomass, root length density, and the distribution of active uptake zones [81,82,83,84,85]. Together, these variables form an integrated diagnostic set for evaluating how irrigation, salinity, and root distribution interact in the active root zone (

Table 1. Core diagnostic indicators for water–salt–root assessment.

Indicator Domain	Core Indicator	What It Captures	Recommended Reporting Unit or Form	References
Irrigation water	ECw/irrigation-water salinity	Salt input from irrigation water	dS m−1 or g L−1	[88,89]
Soil water	Depth-resolved soil moisture	Root-zone water availability and dry-down	% or m3 m−3 by depth interval	[58,59]
Soil salinity	ECe or equivalent salinity index	Salt stress intensity in the root zone	dS m−1 by depth interval	[71,90]
Salt position	Salt accumulation position relative to emitter	Moisture–salt segregation geometry	Distance from emitter × depth	[70]
Groundwater	Water-table depth and salinity	Capillary contribution and salinity risk	m; dS m−1 or g L−1	[45]
Root distribution	Rooting depth and depth distribution	Root overlap with moist or saline layers	Depth interval or % by depth	[79]
Fine-root response	Root length density/fine-root biomass	Active uptake zone and root plasticity	cm cm−3, m m−3, or biomass by depth	[91]
Plant water status	Stem or plant water potential	Actual plant water stress	MPa or threshold values	[92,93,94]
Leaf function	Photosynthesis, stomatal conductance, sap flow	Functional response to root-zone conditions	gas exchange or sap-flow metrics	[80]
System outcome	Survival, growth, canopy or protective function	Integrated shelterbelt performance	survival %, growth increment, biomass, service indicators	[95,96]

). Root traits are increasingly recognized as performance indicators rather than merely descriptive morphology, and the methodological literature has emphasized that trait definition, sampling depth, and spatial reference to water and salt fields are essential for comparability [71,86,87]. Isotopic methods and Hydrus-type models are especially valuable because they connect static root distributions to dynamic water uptake under non-uniform stress [78,79,86].

At the same time, diagnostic indicators require quantitative reference ranges if they are to support management interpretation rather than remain descriptive variables.

Table 2. Evidence-tiered representative values and indicative benchmarks for interpreting water–salt–root regulation in arid shelterbelt-related systems.

Parameter	Representative Value/Range	Unit	Evidence Type	Interpretation for Root-Zone Regulation	References
Mean annual precipitation in hyper-arid shelterbelt context	~24.6 to <50	mm yr−1	Direct shelterbelt evidence	Rainfall is insufficient for reliable water supply or salt leaching; irrigation is foundational.	[50,59]
Mean annual evaporation in hyper-arid shelterbelt context	>3000 to 3639	mm yr−1	Direct shelterbelt evidence	High evaporation accelerates post-irrigation drying and upward salt return.	[51]
Salinity of groundwater used for shelterbelt irrigation	2.8–29.7	g L−1	Direct shelterbelt evidence	Irrigation water also acts as a salt source, requiring salt-balance management.	[58]
Irrigation interval used in representative desert shrub experiments	1, 2, 4, 8, 12	weeks	Direct experimental evidence	Irrigation interval affects hydraulic conductance, biomass allocation, and photosynthesis; values are experimental gradients.	[75]
Typical active root layer of Haloxylon ammodendron in shelterbelt system	20–80	cm	Direct shelterbelt evidence; species-specific	Active roots may remain below the most saline surface layer; wetting depth should match uptake depth.	[63,74]
Surface salt-affected layer under saline drip irrigation	0–20	cm	Direct shelterbelt evidence	Surface salinity should be interpreted together with active root depth, not alone.	[55,58,59]
Root-growth salinity threshold reported for nectarine	~4	dS m−1 (ECe)	Transferable woody-crop evidence	Indicative benchmark only; shelterbelt-specific salinity thresholds require validation.	[71]
Groundwater depth generally supportive of riparian vegetation in arid systems	shallower than ~6	m	Transferable arid woody-system evidence	Indicative groundwater-access reference; effects depend on salinity, capillary rise, aeration, and rooting depth.	[40]
Soil-matric-potential trigger suggested for young yellowhorn	−30	kPa	Transferable dryland woody-planting evidence	Example of an establishment-stage irrigation trigger; transfer requires local calibration.	[92]
Soil-matric-potential trigger suggested for bearing yellowhorn	−20	kPa	Transferable dryland woody-planting evidence	Shows that irrigation thresholds may vary by growth stage.	[92]
Water reduction under regulated deficit irrigation with limited yield penalty in woody crops	20–30	% reduction	Transferable woody-crop evidence	Indicative water-saving range; not a direct shelterbelt prescription.	[97]
Yield variation under moderate woody-crop deficit irrigation	usually within ±10	%	Transferable woody-crop evidence	Useful for comparison, but shelterbelts should be evaluated by survival and protective function.	[97]

summarizes representative values reported or inferred from arid shelterbelt-related systems, while distinguishing direct shelterbelt evidence from transferable woody-system or general root-growth benchmarks. For example, mean annual precipitation of only ~24.6 to <50 mm becomes meaningful when considered together with annual evaporation exceeding 3000–3639 mm; this extreme imbalance explains why rainfall cannot meet woody water demand or provide effective salt leaching in hyper-arid saline-drip shelterbelts. Similarly, groundwater salinity of 2.8–29.7 g L⁻¹ shows that irrigation water is also a salt input, while the contrast between a 0–20 cm surface salt-affected layer and a 20–80 cm active root layer of Haloxylon ammodendron illustrates the importance of matching wetting-front depth with active uptake zones. Values transferred from orchards, riparian vegetation, or general root-growth studies are therefore retained only as indicative benchmarks that require shelterbelt-specific validation. Accordingly, Table 2 should be read as an evidence-tiered interpretation aid rather than a management-threshold table: direct shelterbelt values indicate reported system conditions, whereas transferred values indicate hypotheses or reference ranges that require local validation.

5. Operational Bottlenecks in Root-Zone Regulation

5.1. Scheduling Constraints

Once a shelterbelt depends on artificial watering, its long-term viability becomes tightly linked to irrigation amount, interval, and seasonal timing. In some extremely arid shelterbelt systems, substantial water savings are possible without immediate growth penalties. In the Taklimakan hinterland, reducing annual irrigation volume from 420 mm to 201.6 mm had almost no effect on plant growth and could save more than 50% of irrigation water; even a more conservative adjustment could still save more than 30% [98]. This suggests that some shelterbelt systems may operate above minimum biological demand, but it does not justify universal aggressive water reduction. In woody crops more broadly, reducing irrigation by 20–30% often causes only limited yield change—typically within 10%—while increasing water productivity by 10–30% [97]. These values should be treated as transferable, conditional ranges rather than shelterbelt-specific deficit-irrigation thresholds. The operational margin lies between recoverable water reduction, where growth or photosynthetic function can recover after irrigation, and irreversible Under irrigation, where soil moisture, plant water status, or root-zone buffering falls below recovery capacity. Transferable woody-planting evidence, for example, suggests soil-matric-potential triggers of −30 kPa for young yellowhorn and −20 kPa for fruit-bearing yellowhorn, but comparable shelterbelt-specific thresholds remain to be calibrated [99].

Scheduling difficulty becomes even more acute when saline or brackish water is the only available source. In a hyper-arid date palm system in the United Arab Emirates, irrigation depended on saline groundwater and required careful management because groundwater is both increasingly saline and effectively non-renewable at current use rates [88,89]. In brackish-water olive systems in the Negev, long-term use of groundwater with EC ≈ 4.4–4.5 dS m⁻¹ supported cultivation because no economical alternative source existed, yet the long-term suitability of both surface drip irrigation and subsurface drip irrigation remained questionable due to salinity and sodicity risks in the active root zone [100]. These transferable woody-system cases indicate that saline-water scheduling cannot rely on evapotranspiration replacement alone. It should combine evapotranspiration-based demand estimates, soil-moisture or plant-status sensing, EC_w/SAR, leaching fraction or leaching requirement, post-irrigation EC_e, and active root depth under a salt-balance logic: salt introduced through irrigation must be displaced away from fine-root zones before it exceeds species- and soil-specific tolerance [101].

Seasonal timing introduces another constraint. In arid jujube systems, winter irrigation is used not simply to add water, but to regulate salt redistribution before the next growing season. The objective is explicitly to prevent secondary salinization while supporting survival and growth in the following year [102]. This illustrates a wider shelterbelt problem: irrigation events often need to provide water supply, salt flushing, thermal buffering, and next-season conditioning simultaneously. Because these functions do not share the same optimal timing or quota, scheduling should be judged by whether the wetting front, salt peak, and active root layer remain properly aligned across seasons, rather than by irrigation volume alone [103].

5.2. Systems Failure Risks

Drip irrigation is widely favored in drylands because it improves delivery efficiency and localizes water to the root zone, but its performance is vulnerable to system-level failure. Emitter clogging is one of the clearest examples. Under saline-water drip irrigation, calcium–magnesium carbonates were identified as the dominant clogging materials, accounting for 63.3–91.1% of the deposits inside emitters across irrigation-water salinities of 1.0–6.0 dS m⁻¹ [104]. Meta-analysis of reclaimed-water drip systems likewise showed that emitter clogging changes discharge rate and irrigation uniformity and can substantially degrade system performance [105]. In extreme drylands, clogging is also a root-zone regulation risk: reduced discharge or irrigation uniformity alters wet-bulb geometry, leaching uniformity, salt distribution, and plant water supply. Maintenance should therefore focus on discharge uniformity, relative flow reduction, EC_w/SAR, carbonate precipitation risk, filtration, flushing, and chemical cleaning rather than on emitter operation alone [106,107].

The design of the irrigation system is itself another bottleneck. In jujube grown on saline soils under extremely arid conditions, indirect subsurface drip irrigation outperformed conventional surface drip irrigation as a combined water-saving and salt-control strategy. A vertical tube diameter of 75 mm and irrigation volume of 13–14 L plant⁻¹ was identified as the more suitable mode; under this configuration, the average desalination ratio in the 0–50 cm layer at 30 cm from the tube reached 25.2%, average yield was 2579 kg ha⁻¹, and irrigation water-use efficiency reached 3.48 kg m⁻³ [108]. After conversion from long-term flood irrigation to drip irrigation, a jujube orchard under a 660 mm treatment achieved a 7.6% yield increase and 60.3% increase in WUE relative to flood irrigation [109]. These transferable examples show that irrigation modernization can improve root-zone control and water productivity, but system selection should not be based on delivery efficiency alone. It should match soil texture, EC_w/SAR, target rooting depth, required leaching, clogging risk, maintenance capacity, and irrigation legacy. Surface drip is easier to inspect, whereas subsurface or indirect subsurface systems may reduce evaporation and improve salt displacement but increase design and clogging risks.

Long-term water-quality effects further complicate system sustainability. Reclaimed or saline water can keep woody systems productive, but only under cautious management. In citrus, the medium- and long-term feasibility of reclaimed water combined with regulated deficit irrigation was demonstrated, yet appropriate management remained necessary to avoid fruit-quality damage associated with phytotoxic elements [110]. In long-term brackish-water drip-irrigated olive orchards, continued yield decline and even orchard uprooting motivated investigation of root-zone salinity and sodicity patterns, underscoring that system survival can be undermined gradually rather than through abrupt failure [100]. Consequently, sustainability should not be inferred from short-term survival, WUE, or growth alone, but from whether multi-year saline loading can be managed without progressive EC_e/SAR increase, emitter failure, root-zone degradation, or canopy decline.

5.3. Implementation Scale, Policy Support, and Socio-Economic Constraints

Even when water-saving technologies are technically effective, adoption and conservation outcomes are not guaranteed. A recent study identified heterogeneous adoption determinants, including farm size, extension services, capital assets, education, income, access to credit, water cost, groundwater access, drought, and aridity [18]. Thus, irrigation modernization in dryland shelterbelts should be treated not as a simple engineering upgrade, but as a coupled technical, financial, institutional, and maintenance challenge.

Large-scale shelterbelt policy also shows that ecological protection requires sustained public investment. China’s Three-North Shelter Forest Program, for example, is a 73-year project from 1978 to 2050, covers about 42.40% of China’s land area, and has involved a reported total investment of CNY 93.3 billion [111]. Its long-term outcomes include improved wind–sand control, soil erosion reduction, carbon sequestration, grain production, and regional economic output [111]. This example indicates that shelterbelt sustainability depends on decades of coordinated planning, funding continuity, species adjustment, maintenance, and monitoring rather than one-time establishment.

Socio-economic evidence from windbreak and ecological compensation studies further supports this point. In the United States, producer satisfaction with windbreaks was reported as high as 72–99%, but non-adoption was mainly associated with lack of land and upkeep requirements, while removal was driven by poor condition, aging, and conflicts with farming practices [112]. In China’s grassland ecological compensation policy, government investment reached CNY 77.4 billion in 2011–2015 and CNY 93.8 billion in 2016–2020; the policy improved NDVI by about 3.2% and increased herder income, but also exacerbated local income inequality [113]. For arid shelterbelts, management strategies should therefore be evaluated not only by ecological effectiveness, but also by water cost, maintenance burden, and adoptability. Because shelterbelt benefits are often protective and public rather than immediately marketable, long-term regulation will require stable funding, water-allocation coordination, technical extension, monitoring capacity, and mechanisms for valuing ecosystem services. Without these institutional supports, technically sound root-zone regulation may remain difficult to implement at landscape scale.

6. Management Options for Root-Zone Regulation

6.1. Irrigation Strategies

Effective management in arid and semi-arid shelterbelts should be framed as root-zone regulation rather than simple irrigation saving. The central task is to maintain wetting patterns that connect active roots to water while preventing recurrent salt return to the same soil layers [114]. From this perspective, useful options are not isolated technologies, but coordinated controls over irrigation quota, interval, leaching opportunity, and seasonal timing. Reviews of salinity management and partial root-zone drying converge on the same point: localized irrigation can improve water productivity only when the wetting front, salt front, and root activity zone remain dynamically aligned [115]. Where roots already operate under salinity and evaporative stress, water-saving methods that ignore salt displacement may narrow rather than improve the habitable root zone.

Accordingly, irrigation scheduling should be calibrated against soil moisture and salt behavior together, rather than water demand alone. Under mulched drip irrigation in southern Xinjiang, reducing the soil-moisture lower limit expanded both the humid perimeter and the salt-leaching range, but the response was not monotonic. An irrigation amount of about 58 mm lowered root-zone salinity by roughly 58.3% relative to higher-threshold treatments during flowering, yet very low lower limits also reduced irrigation frequency and increased post-harvest salt re-accumulation [116]. Thus, irrigation amount and frequency cannot be optimized separately: larger single events may improve short-term leaching, whereas excessively long intervals can allow salts to rebound during dry-down. This explains why fixed calendar irrigation is rarely optimal under strong evaporative demand. Several core equations and indicator definitions remain useful for interpreting irrigation demand, salinity hazard, leaching requirement, and water-use efficiency (

Table 3. Key equations and indicator definitions relevant to water–salt–root regulation.

Indicator/Equation	Expression	Main Use	Key Caution	References
Crop evapotranspiration	ETc = Kc × ETo	Estimate baseline irrigation demand	Represents standard conditions unless locally corrected	[101]
Soil–water salinity relation under standard leaching assumption	ECe ≈ 1.5 ECw	Screen root-zone salinity from irrigation-water salinity	Assumes standard leaching and root water-use patterns	[117]
Sodium adsorption ratio	SAR = Na+/[((Ca2+ + Mg2+)/2)0.5]	Assess sodicity hazard	Ion concentrations must use consistent chemical units	[118]
Leaching requirement	LR = ECw/(5ECe − ECw)	Estimate water fraction needed for salt leaching	Screening equation; requires field validation in heterogeneous soils	[119]
Water-use efficiency (general)	WUE = Production variable/Water variable	Relate biomass or yield to water consumed or supplied	Numerator and denominator must be explicitly defined	[120]
Intrinsic water-use efficiency	iWUE = A/gs	Assess leaf-scale carbon gain per stomatal conductance	Not interchangeable with field-scale WUE or irrigation efficiency	[120]
Incremental irrigation water-use efficiency	IWUE = (Yi − Yd)/I	Evaluate yield gain attributable to irrigation input	Definitions vary; the formula must be stated explicitly	[121]

), but they should be used as a screening sequence—estimating evapotranspiration demand, checking EC_w/SAR and leaching requirement, and then evaluating WUE or IWUE—rather than as universal prescriptions.

Another scheduling principle is that irrigation thresholds cannot be generalized across soils. Global syntheses now show that critical soil-moisture thresholds are strongly texture-dependent: coarse soils approach water limitation earlier during dry-down, whereas finer soils remain more buffered but become more sensitive to rising atmospheric demand. A new global satellite-derived dataset of soil critical point and permanent wilting point, validated against 1334 stations, further shows that threshold-based irrigation design can now be tied to actual hydraulic context rather than to uniform field-capacity percentages [29,30]. For shelterbelts planted in sandy or aeolian soils, irrigation should usually be triggered earlier than in fine-textured agricultural soils, especially where roots rely on narrow wet bulbs rather than broad moisture reserves [122].

6.2. Root-Zone Reinforcement

Many management options work by reinforcing the root-zone environment, with organic amendments among the most practical examples. Under brackish-water irrigation, compost incorporation increased plant fresh biomass by 64%, dry biomass by 50%, root length by 121.1%, and plant height by 15.8%, while promoting Na⁺ and Cl⁻ leaching and reducing sodicity stress [123]. In mildly saline soils, combined wood vinegar and biochar increased antioxidant enzyme activity by 5.82–37.21% and 11.31–64.95% relative to single amendments, while reducing malondialdehyde by 2.47–51.72%, indicating stronger physiological stress buffering [124].

Amendments are most useful when they simultaneously improve structure, water retention, ion balance, and biological stress resistance. However, they are not universally beneficial. A review on biochar in saline soils emphasized that some materials can inhibit root development, increase alkalinity, or aggravate salt-related problems when feedstock, dose, or soil context are poorly matched [125]. Safe application should therefore be defined not by a universal dose, but by whether the amendment avoids unacceptable increases in pH, EC, SAR, or root-growth inhibition under local soil and irrigation conditions. Management should avoid treating biochar, compost, or microbial inoculation as generic solutions; their value lies in targeted deployment under defined salinity, texture, and irrigation conditions, with follow-up monitoring of EC, SAR, rooting depth, and soil physical recovery [24].

6.3. Species–Water–Salinity Matching

Root-zone regulation also depends on biological matching because species differ in rooting depth, salt tolerance, hydraulic adjustment, and ability to exploit heterogeneous moisture microsites. In semi-arid Mongolian afforestation, an irrigation level of 4 L h⁻¹ was sufficient to produce the highest biomass for most root classes of Populus sibirica and Ulmus pumila, whereas fertilization did not improve soil chemistry and suppressed root growth. In U. pumila, fertilization also reduced rooting depth, branching density, and wind-responsive root plasticity, implying that more nutrient input is not always compatible with mechanically stable shelterbelt establishment [126,127]. Species choice matters as much as input choice. Under drip irrigation with salinity of 8 g L⁻¹, survival reached 86.8% for Tamarix chinensis, 89.6% for Calligonum mongolicum, and 92.4% for Haloxylon ammodendron, all markedly higher than under border irrigation [128]. In H. aphyllum, halotolerant rhizobacteria improved seedling quality under combined salinity and dust stress, suggesting that microbially assisted establishment can complement species selection where abiotic stresses overlap [129].

In practice, the best management option is often not a stronger intervention, but a better fit among irrigation mode, salinity level, soil conditions, and species traits [57]. A practical matching framework should consider at least five linked factors: irrigation-water salinity, expected wetting-front depth, dominant salt accumulation layer, species rooting depth, and required protective function. Shallow-rooted species require more stable near-surface moisture and lower surface salinity, whereas deeper-rooted or halophytic shrubs are better suited where surface salts accumulate but middle or deeper layers remain intermittently wetted.

6.4. Adaptive Monitoring

Management options are becoming increasingly effective when coupled with adaptive monitoring rather than fixed prescriptions. UAV-based monitoring around shelterbelts has already shown that water competition is spatially quantifiable: for cotton adjacent to shelterbelt rows, competition was strongest within 0.1H–1H, and biomass at 0.1H was reduced by more than half, whereas plots at 2H–3H were largely unaffected [96]. Such spatial thresholds can be translated into field actions: zones within 0.1H–1H can be prioritized for supplemental irrigation, spacing adjustment, or crop–shelterbelt configuration redesign, whereas zones beyond 2H–3H may require less direct intervention. This type of information is directly useful for shelterbelt spacing, supplemental irrigation, and mixed-system design because it identifies where the shelterbelt shifts from protection to competition [130,131].

At larger scales, salinity monitoring is moving toward operational decision support. Sentinel-1/2 feature-space models have achieved stable salinity inversion with R² up to 0.568, while UAV–satellite spatiotemporal fusion combined with random forest has achieved R² = 0.94 for daily salinity mapping in arid fields [95,132,133]. Over regional scales, root-zone salinity mapping in the Manas River Basin indicates that management should focus on root-zone salt content rather than surface salinity alone. For shelterbelt application, these remote-sensing outputs should therefore be coupled with field verification of EC_e, soil moisture, active root depth, plant water status, and canopy continuity before being used to guide irrigation or leaching decisions [134,135].

6.5. Practical Application of Predictive Root-Zone Regulation

Figure 5 summarizes predictive root-zone regulation as a practical framework for arid shelterbelt management. To avoid treating irrigation, salinity control, species selection, and monitoring as separate actions, the framework links them through one diagnostic sequence: identify water-source salt input, determine active root depth and soil texture, monitor soil moisture and EC_e by depth, compare the salt accumulation layer with the active root layer, adjust irrigation or root-zone management, and verify outcomes using canopy continuity, survival, growth, and protective performance. This sequence emphasizes that management success depends on maintaining a usable root zone through time, rather than on applying any single technology in isolation.

7. Research Gaps and Future Directions

7.1. Evidence Scope and Transferability

The evidence base remains broader in topic than in mechanism, and broader in transferable systems than in true shelterbelt systems. Recent bibliometric synthesis confirms that shelterbelt and windbreak research is widely distributed across soil conservation, biodiversity, microclimate regulation, and agricultural productivity, but this breadth does not mean that water–salt–root interactions have been studied with equal depth [7,136]. In the literature reviewed here, 41 of the 70 studies (58.6%) were classified as core shelterbelt evidence, whereas 29 studies (41.4%) were classified as transferable background evidence. Many of the most informative process-based studies still come from orchards, cotton systems, and saline irrigated farmlands rather than from mature arid shelterbelts themselves. This creates an unavoidable transferability problem: these systems are useful analogs for wetting-front dynamics, root-zone salinity, and irrigation scheduling, but their objectives differ from those of long-lived woody shelterbelts designed for ecological protection rather than crop production. Accordingly, their reliability for shelterbelt management should be interpreted by evidence tier: direct shelterbelt studies provide the strongest management relevance, closely related woody systems provide mechanistic or threshold-related analogs, and agricultural systems mainly provide background process or methodological support.

Long-term ecological performance is still under-evaluated relative to short-term physiological or soil responses. Recent regional assessments of the China Three-North Shelterbelt Program show that ecosystem functioning has improved overall, but also reveal pronounced spatial heterogeneity, persistent low-service zones in more arid central and western areas, and emerging challenges such as aging stands and functional decline [137]. Likewise, long-term analysis in the Yanqi Basin shows that shifts in irrigation practice can reshape both ecological conditions and irrigation behavior over time, indicating that irrigation change has delayed and system-wide effects rather than only plot-scale short-term outcomes [138]. For shelterbelt studies, this means that short experiments describing one or two irrigation cycles are necessary but insufficient. More chronosequence work, repeated monitoring across stand ages, and explicitly long-term evaluations are needed to determine whether a management option is merely effective in the short term or genuinely sustainable over decades.

7.2. Comparable Thresholds

Dryland systems are controlled less by mean conditions than by threshold behavior during dry-down and salt concentration. Much of the current literature reports irrigation amount, EC_e, soil moisture, or groundwater depth, but relatively few studies define actionable thresholds that connect these variables to root viability, plant function, or management decisions. A new observation-based global map of critical soil moisture estimated an average θcrit of 0.19 m³ m⁻³, varying from 0.12 m³ m⁻³ in arid ecosystems to 0.26 m³ m⁻³ in humid ecosystems [28]. Complementary datasets now provide yearly global estimates of the critical point and permanent wilting point from 2002 to 2023 [29]. These values are useful for framing threshold-based thinking, but they are not shelterbelt-specific irrigation thresholds. Their use in shelterbelt management requires calibration against local soil texture, active root depth, groundwater condition, salinity, and species tolerance.

This gap is also visible in the mismatch between soil indicators and plant indicators. Soil-based control remains dominant, but recent work increasingly shows that plant-based metrics are needed if irrigation is to reflect actual stress rather than assumed stress. In yellowhorn (Xanthoceras sorbifolia), soil-matric-potential-triggered drip irrigation suggested practical SMP values of −30 kPa for young trees and −20 kPa for fruit-bearing trees [92], while almond studies show that stomatal conductance, stem water potential, and photosynthesis can capture irrigation response more directly than soil variables alone [93]. Continuous sensing is also improving: microtensiometers now allow near real-time stem water potential monitoring in woody systems, reducing reliance on intermittent pressure-chamber measurements [94]. These examples should be interpreted as transferable woody-system references rather than direct guidance for arid shelterbelts. The main research need is therefore to develop shelterbelt-specific soil–root–plant thresholds that link EC_e, soil matric potential, active uptake depth, plant water status, canopy continuity, and protective function.

7.3. Below-Ground Process Representation

Although roots are repeatedly invoked as the mediator between water, salt, and whole-plant performance, they remain one of the least standardized components of the evidence base [139]. Perspectives on salt-stress modeling emphasize that realistic prediction requires explicit treatment of root-system architecture, root-mediated salt exclusion, plant-hydraulics, ion compartmentation, and stomatal response, rather than generic salinity coefficients alone [27]. Likewise, syntheses on tree water uptake show that water sourcing is highly dynamic across species, depths, and seasons, while drought-response reviews underline that roots are the first organs to perceive stress and often the first to reorganize under it [32,66]. Yet many studies in saline drylands still infer root response indirectly from soil moisture or canopy performance. Future studies should use more isotope tracing, high-frequency soil sensing, repeated root observations, and hydraulically informed models so that root redistribution and water uptake can be measured rather than assumed. Operationally, this requires reporting rooting depth, root length density, fine-root biomass, active uptake depth, isotope-inferred water-source proportions, and plant hydraulic indicators in forms that can parameterize root water uptake, salinity-stress response, compensatory uptake, and plant-hydraulic functions in HYDRUS-type or coupled soil–plant models.

7.4. Climate-Change Amplification and Compound-Stress Gaps

Current shelterbelt studies still rarely test how climate change may alter water–salt–root coupling rather than how it may simply intensify aridity. Warming, higher evaporative demand, more variable precipitation, and more frequent extremes may change soil dry-down, capillary salt return, wetting-front depth, and salt-peak position [140]. Episodic rainfall or leaching can temporarily displace salts, but subsequent hot dry periods may drive salt rebound toward evaporation fronts, especially where saline groundwater or saline irrigation water is present [141,142]. Thus, fixed irrigation intervals, static EC_e thresholds, and one-time salt-reset practices may become less reliable under future climate variability.

A key gap is the lack of climate-informed experiments and models that evaluate whether wetting-front design, leaching schedules, amendments, and species selection remain effective under compound scenarios of heat, drought, irregular rainfall, shallow saline groundwater, and aeolian disturbance. Future research should link field monitoring with climate scenarios, root-zone water–salt models, and plant-hydraulic indicators to define dynamic rather than fixed management thresholds. Such work is needed to determine when management can adjust to climate variability and when shelterbelt systems may approach operational limits.

7.5. Limitations and Future Research Directions

This review has several limitations that should be considered when interpreting its conclusions. First, although the WoSCC provides relatively consistent metadata and reproducible search fields, relying on a single database may have excluded relevant regional studies, technical reports, non-English publications, and long-term management records. This limitation is important for shelterbelt research because many site-specific experiences are reported in local journals or institutional documents. Future reviews should therefore combine WoSCC with Scopus, Google Scholar, regional databases, and gray literature to improve evidence coverage. Second, although international examples were incorporated where possible, shelterbelt-specific evidence remains geographically uneven, with a strong concentration in Chinese arid and hyper-arid systems. Cases from the Tarim Basin, Taklimakan Desert, and northwestern China provide valuable end-member evidence for saline-drip-irrigated shelterbelts, but their transferability to other drylands depends on climate, soil texture, groundwater depth and salinity, species composition, and irrigation infrastructure. Third, long-term shelterbelt-specific datasets remain limited. Most studies still examine irrigation, salinity, roots, plant performance, or protective function separately, whereas predictive root-zone regulation requires these variables to be monitored jointly. Emerging machine-learning and remote-sensing approaches are already improving multilayer soil-moisture estimation and salinity mapping beyond the surface layer. However, these tools need to be integrated with field measurements of root distribution, plant water status, canopy continuity, and protective performance. Future research should develop multi-scale designs in which field experiments define soil–root–plant response thresholds, remote sensing tracks spatial heterogeneity, and long-term regional analyses evaluate cumulative ecological consequences.

8. Conclusions

This review highlights that the sustainability of arid and semi-arid shelterbelts should be understood as a coupled water–salt–root regulation problem rather than as a matter of water supply or salinity control alone. The main novelty is the proposed predictive root-zone regulation perspective, which links localized irrigation, salt redistribution, active root placement, plant performance, and protective function within a single management framework. The evidence shows that irrigation can enable shelterbelt establishment under severe aridity, but it can also create long-term vulnerability when saline-water input, evaporative salt return, wetting-front geometry, emitter performance, and root distribution are not jointly managed. Therefore, water-saving or salinity-control strategies should be evaluated by whether they preserve a habitable root zone and sustain stand-level protective function.