Integrating Microtopographic Engineering with Native Plant Functional Diversity to Support Restoration of Degraded Arid Ecosystems

Abstract

Active restoration structures such as microtopographic water-harvesting designs are widely implemented in dryland ecosystems to improve soil moisture, reduce erosion, and promote vegetation recovery. We assessed the combined effects of planted species identity, planting diversity (mono-, bi- and multi-species mixtures), and micro-catchment (half-moon) structures on seedling performance and spontaneous natural regeneration in a hyper-arid restoration pilot site in Sharaan National Park, northwest Saudi Arabia. Thirteen native plant species, of which four—Ochradenus baccatus, Haloxylon persicum, Haloxylon salicornicum, and Acacia gerrardii—formed the dominant planted treatments, were established in 18 half-moons and monitored for survival, growth, and natural recruitment. Seedling survival after 20 months differed significantly among planting treatments, increasing from 58% in mono-plantings to 69% in bi-plantings and 82% in multi-plantings (binomial GLMM, p < 0.001), indicating a positive effect of planting diversity on establishment. Growth traits (height, collar diameter, and crown dimensions) were synthesized into an Overall Growth Index (OGI) and an entropy-weighted OGI (EW-OGI). Mixed-effects models revealed strong species effects on both indices (F_12,369 ≈ 7.2, p < 0.001), with O. baccatus and H. persicum outperforming other taxa and cluster analysis separating “fast expanders”, “moderate growers”, and “decliners”. Trait-based modeling showed that lateral crown expansion was the main driver of overall performance, whereas stem thickening and fruit production contributed little. Between 2022 and 2024, half-moon soils exhibited reduced electrical conductivity and exchangeable Na, higher organic carbon, and doubled available P, consistent with emerging positive soil–plant feedbacks. Spontaneous recruits were dominated by perennials (≈67% of richness), with perennial dominance increasing from mono- to multi-plantings, although Shannon diversity differences among treatments were small and non-significant. The correlation between OGI and spontaneous richness was positive but weak (r = 0.29, p = 0.25), yet plots dominated by O. baccatus hosted nearly two additional spontaneous species relative to other plantings, highlighting its strong facilitative role. Overall, our results show that half-moon micro-catchments, especially when combined with functionally diverse native plantings, can simultaneously improve soil properties and promote biotic facilitation, fostering a transition from active intervention to passive, self-sustaining restoration in hyper-arid environments.

1. Introduction

Restoring vegetation in hyper-arid environments presents a major ecological challenge due to multiple interacting abiotic constraints. These include extreme water scarcity, nutrient-poor soils, coarse soil textures with limited water-holding capacity, and large diurnal temperature fluctuations, all of which severely restrict natural regeneration [1,2,3]. Sandy and gravelly substrates, which dominate many hyper-arid landscapes, retain little moisture, while surface crusting further limits infiltration and seedling establishment [4]. Soil instability from compaction, erosion, and crusting leads to moisture and nutrient loss, creating inhospitable conditions for germination [4,5]. As a result, natural recovery is extremely slow, rendering passive restoration largely ineffective.

Dryland ecosystems are highly sensitive to changes in soil physical conditions, and the degradation of soil structure is one of the primary drivers prompting the need for restoration as the set of strategies used to improve degraded conditions [6]. Processes such as soil compaction, loss of aggregation, and reduced porosity severely alter the soil’s capacity to support vegetation, thereby impairing root elongation, water infiltration, and nutrient acquisition [7,8]. In degraded soils, where compaction and structural collapse prevail, restricted root growth limits plants’ ability to access moisture and nutrients from deeper layers [9,10]. Consequently, restoration interventions frequently target the reversal of these degradation effects, with improving soil structure emerging as a key strategy to enhance seedling establishment, particularly during the critical early stages of restoration.

Microtopographic water harvesting has emerged as an effective, scalable solution to these limitations. Simple structures such as micro-catchments (half-moon shape or semicircle), embankments, contour ridges, and infiltration pits intercept runoff, enhance infiltration, and trap sediments and organic matter, thereby increasing localized soil moisture and fertility [11,12,13,14]. Over time, these structures can also improve soil physical properties by reducing crust formation and promoting biological activity [15]. Their effectiveness, however, depends strongly on underlying soil characteristics: coarse soils drain rapidly, while compacted fine soils can impede infiltration [16,17]. Optimizing such interventions requires integrating hydrological and biological considerations.

Plant community composition is another key determinant of restoration outcomes [18,19]. Ecological theory and empirical evidence show that species diversity can enhance ecosystem functioning through niche complementarity, efficient resource use, and increased resilience to environmental stress [20,21,22]. In drylands, mixed-species plantings that exploit different rooting depths and stress-tolerance strategies often outperform monocultures in survival and biomass production (e.g., Pywell et al. [18]). Plants contribute by increasing litter deposition, shading the soil surface, reducing evaporation, and enhancing microbial activity, all of which promote nutrient cycling and improve soil properties [23,24,25]. This balance between competition and facilitation is further influenced by soil structure, which shapes resource availability [26,27].

Achieving persistent restoration therefore requires integrating hydrological engineering with functional trait-based species selection. Structural traits such as lateral crown expansion, canopy density, and rooting depth are critical for buffering microclimatic extremes, stabilizing soils, and facilitating the establishment of other species [28,29,30,31]. Species with extensive branching structures or broad crowns and deep root systems, such as Ochradenus baccatus and Haloxylon persicum, can function as nurse shrubs plants by creating favorable microhabitats that retain moisture, reduce soil exposure, thereby supporting spontaneous recruitment [32,33]. In contrast, shallow-rooted or slow-growing species like Ephedra alata are less competitive under severe water limitation [34,35]. The stress-gradient hypothesis (SGH) provides a useful framework for understanding these interactions, predicting that facilitative effects dominate under high abiotic stress [36,37]. In hyper-arid restoration contexts, facilitation by dominant nurse species can accelerate succession, enhance perennial recruitment, and shift vegetation from ephemeral-dominated communities to more persistent perennial assemblages [38,39].

In this context, the present study focuses on evaluating the performance of native seedlings planted under varying species configurations within half-moon soil structures in a hyper-arid landscape. Beyond assessing seedling survival, growth, and early establishment success, the research also examines passive restoration dynamics through the monitoring of spontaneous colonization and natural recruitment processes. The approach emphasizes how soil structural improvements mediated by water harvesting and organic matter accumulation interact with species diversity to shape restoration trajectories. A series of microtopographic modifications was applied to alter surface hydrology and create localized resource-rich microsites, enabling the development of a model for suitable land transformation strategies. By integrating active planting interventions with passive recovery mechanisms, this study aims to provide an evidence-based framework for enhancing ecological resilience and restoration efficiency in degraded drylands. Ultimately, micro-catchment structures create favorable microhabitats that facilitate plant establishment, while native species diversity enhances stability and promotes self-sustaining regeneration.

2. Materials and Methods

2.1. Study Area

This study was conducted within a 100-hectare protected pilot site inside Sharaan National Park (NP), a key research area for testing ecological restoration strategies in hyper-arid environments. The Sharaan NP covers approximately 1542 km² in the AlUla County, northwestern Saudi Arabia (Figure 1). The pilot site is one of the first controlled restoration zones in the region. It offers a secure, disturbance-free environment, excluding grazing and other anthropogenic impacts, thereby enabling reliable long-term monitoring of vegetation responses under extreme climatic conditions. Restoration experiments were conducted on abandoned farmlands (AF), which represent severely disturbed habitats with loamy-clayey soils [40]. This site was deliberately selected because it represents one of the most common forms of land degradation in arid regions. The land was cultivated for several decades before abandonment. Although the area is currently protected, it retains the ecological legacy of historical cultivation and subsequent abandonment, including severe topsoil loss.

The climate is hyper-arid, characterized by extremely low and high annual rainfall, averaging 15.9 mm in 2021, 52.5 mm in 2022, 73.7 mm in 2023, and 40.9 mm in 2024 (data from the on-site Vantage Pro2 weather station). Rainfall occurs mainly between November and February, often as irregular storm events. Temperature ranges from 4 °C in winter to 38.9 °C in summer, with a mean annual temperature of 28.4 °C. The average wind speed recorded during this period was 10.4 km·h⁻¹. Soils are predominantly sandy to loamy, reaching depths of up to 1.5 m, and are developed over Cambrian sandstone substrates.

2.2. Experimental Design and Half-Moons Construction

Half-moons were established on representative degraded land within the pilot site. Each structure measured approximately 9 m in length, 4.5 m in width, and 1 m in depth. Plots were positioned along topographic contours to maximize interception and retention of runoff during rainfall events. This design aimed to enhance in situ water availability by passively harvesting rainwater, an essential strategy for seedling establishment and survival in arid environments with highly erratic precipitation. Because the pilot site is a strictly protected experimental zone within Sharaan NP, the total area that could be manipulated was limited to six replicates in this first phase of the project.

Prior to planting, a soil amendment mixture composed of one part peat moss to two parts sand (v/v) was incorporated into each planting pit. Sand was used to improve soil structure, enhance infiltration, reduce compaction, and increase permeability while maintaining compatibility with native soil characteristics. This mixture aimed to retain moisture while providing a more suitable rooting environment for native seedlings. Species selection prioritized ecological function, drought tolerance, and potential role as pillar or nurse species in the restoration process.

2.3. Vegetation Treatments and Planting

A total of thirteen native plants representing shrubs, grasses, forbs, and small trees were selected for the restoration experiment (

Table 1. Native species selected for restoration trials in hyper-arid environments, their functional traits, ecological roles, and key references.

Scientific Name	Common Description/Functional Traits	Ecological Role/Relevance	Key References
Ochradenus baccatus	Evergreen shrub tolerant to arid and saline conditions; common colonizer of degraded lands	Acts as a pioneer species; improves soil moisture, micro-climate, and surface stability	[30,41,42,43]
Haloxylon persicum	Deep-rooted perennial tree with high sand-binding and windbreak capacity	Stabilizes dunes; enhances water infiltration and soil structure in sandy desert habitats	[34,41,42,43]
Acacia gerrardii	Drought-tolerant, nitrogen-fixing tree with deep root system	Enhances soil fertility; provides shade, litter, and structural diversity	[34,35,41,42]
Senna italica	Perennial leguminous subshrub/herb; nitrogen-fixing and tolerant to dry, nutrient-poor soils	Improves soil fertility and ground cover; useful in understory restoration and erosion control	[44,45,46]
Ziziphus nummularia	Deep-rooted, thorny shrub tolerant to drought and grazing; produces edible fruits	Functions as a nurse plant; provides fodder, fruits, and habitat for fauna	[42,44,47]
Periploca aphylla	Leafless climbing shrub with photosynthetic stems; highly drought-tolerant	Adds structural complexity; supports associated fauna and contributes to rocky slope cover	[41,42]
Haloxylon salicornicum	Perennial dwarf shrub adapted to extreme aridity and nutrient-poor soils	Survives in harsh conditions; contributes to soil stabilization and ecosystem resilience	[42,43]
Lycium shawii	Spiny, drought-tolerant shrub with deep roots and photosynthetic stems	Provides browse and shelter for wildlife and livestock; contributes to soil stabilization	[41,42,47]
Ephedra alata	Low, broom-like, drought-tolerant shrub with jointed photosynthetic stems and shallow fibrous roots	Occupies rocky and sandy microsites; contributes to surface stabilization and provides shelter and nectar/pollen for fauna	[41,42,47]
Leptadenia pyrotechnica	Leafless, deep-rooted shrub with extreme drought and heat tolerance	Acts as an important dune stabilizer; provides fodder and microhabitat in sandy deserts	[42,44,45]
Retama raetam	Woody legume with strong facilitative nurse-plant traits	Facilitates establishment of other species; improves micro-habitat and soil conditions	[33,36,38,48]
Salvadora persica	Evergreen, deep-rooted halophytic shrub/small tree; highly tolerant to salinity, drought, calcareous soils	Stabilizes soils in saline depressions and wadis; provides browse, fruits, and resources for wildlife and livestock	[41,42,45,47]
Calligonum comosum	Shrub with extensive root system and flexible branches; highly adapted to mobile sands	Key dune- and sand-binder; improves microhabitat conditions and reduces wind erosion	[42,44]

) based on their regional ecological relevance and their use in local restoration. Experimental plots were established in December 2023. The experiment comprised three planting diversity treatments: mono-planting (one species), bi-planting (two species), and multi-planting (three or more species). Each diversity treatment was replicated in six half-moon micro-catchments, giving a total of 18 plots. The multi-planting treatments were designed to explore how higher diversity levels influence plant establishment, complementarity, and microhabitat modification.

Each half-moon micro-catchment was planted with a fixed density of 20 individuals. In monospecific treatments, all individuals belonged to the same species. In bispecific treatments, the individuals were divided equally between the two species. In multi-planting treatments, the individuals were distributed evenly across the species included in the mixture, with the exact arrangement randomized to avoid positional bias. The diversity treatment was designed to evaluate how minimal yet realistic levels of species diversity influence plant establishment. Mono-planting allows us to quantify the performance of each species individually under identical microhabitat conditions, whereas mixed plantings made possible to evaluate complementarity or competition within the same half-moon.

Planting treatments were allocated to half-moon structures using a completely randomized design. For each replicate block, we generated a random assignment list that allocated each of the 18 treatments to one micro-catchment. To avoid a plantation-like geometric pattern and mimic natural field heterogeneity, micro-catchments were spaced irregularly around a nominal 6 m center-to-center distance (±0.5 m variation), and the positions of individual plants within each half-moon were randomized. This approach ensured that the spatial configuration reflected realistic ecological conditions rather than a controlled silvicultural layout.

The species used were carefully selected based on their native status, ecological relevance, and proven adaptability to the extreme environmental conditions of the AlUla region. Species were randomly assigned to plots within each treatment to minimize spatial bias and ensure experimental robustness. This randomized design allows reliable assessment of diversity effects on seedling survival, growth performance, and spontaneous recruitment.

2.4. Seedling and Spontaneous Vegetation Monitoring

Planting was conducted in December 2023. Seedlings were monitored twice after planting: the first measurement was taken two months after planting in February 2024, allowing sufficient time for good plant establishment, and the second in August 2025. Key performance indicators included survival rate, vertical growth (height), radial thickening (collar diameter) and lateral expansion (crown axes). Crown size was quantified by measuring two perpendicular crown diameters on the ground: the maximum crown diameter and a second diameter perpendicular to this axis. Both measurements were taken from drip line to drip line of the live canopy using a measuring tape and are expressed in (cm). In addition to planted species, spontaneous regeneration of native plants was recorded within and around the half-moon plots. These naturally recruiting species, emerging without direct planting, serve as indicators of improved microsite conditions and soil health.

2.5. Soil Sampling and Analyses

To characterize baseline soil conditions and detect temporal changes associated with micro-catchment restoration, soil samples were collected from the half-Moon restoration area in 2022 (pre-planting) and 2024 (post-establishment). In each sampling year, composite soil samples were taken from the upper 0–30 cm layer, representing the biologically active zone most responsive to micro-catchment hydrological effects.

Five subsamples were collected from randomly selected half-moon structures and homogenized to form a representative sample for laboratory analysis. Soil pH was measured in both distilled water (1:2.5 soil-to-water ratio) and KCl solution using a calibrated pH meter. Electrical conductivity (EC) was determined using a conductivity meter, while soil resistivity was measured using a standard multi-parameter probe. Soil organic carbon (SOC) was quantified using the Walkley–Black dichromate oxidation method, and total nitrogen was determined via the Kjeldahl digestion technique. Available phosphorus (P₂O₅) and exchangeable cations (Ca, Mg, K, Na) were measured by ICP-OES following ammonium acetate extraction. Calcium carbonate (CaCO₃) content was assessed using the Bernard calcimeter method, and cation exchange capacity (CEC) was estimated from ammonium displacement. Soil texture was determined using the hydrometer method. These analyses provided a baseline for evaluating how micro-catchments influence soil fertility, salinity, and physical conditions over time.

2.6. Statistical Analyses

For analyses of planting diversity effects on survival, planting diversity (mono, bi, multi) was treated as a fixed factor and plot (18 plots; 6 replicates per diversity level) as a random intercept. For each planted individual, vegetative growth was quantified between the initial census (T₀) and final census (T₁). Four structural traits were measured at both dates: plant height (H, cm), collar diameter (CD, cm), short crown axis (CrS, cm) and long crown axis (CrL, cm). For each individual i, growth in each trait was calculated as the difference between the two censuses (e.g., Δheight = height_T₁ − height_T₀). An Overall Growth Index (OGI) was then defined as the sum of growth across all four traits:

OGI_i = ΔH_i + ΔCD_i + ΔCrS_i + ΔCrL_i,

OGI is expressed in centimeters (cm) and integrates vertical growth (height), radial thickening (collar diameter) and lateral expansion (crown axes). Positive OGI values indicate net structural gain, whereas negative values indicate net dieback. To reduce subjectivity in trait weighting and to emphasize traits with higher discriminatory power among individuals, we also derived an entropy-weighted index (EW-OGI). All four growth variables were standardized using min–max normalization before entropy analysis. Let x_ij represent the normalized value of species i for trait j. A probability matrix ${\mathit{p}}_{\mathit{i}\mathit{j}}=$ x_ij/∑_i x_ij was then constructed for each trait, where x_ij is the growth value of trait _j in individual _i (j ∈ {ΔH, ΔCD, ΔCrS, ΔCrL}). Trait-specific entropy values were calculated as

$${\mathit{p}}_{\mathit{i}\mathit{j}}={\displaystyle \frac{{\mathit{x}}_{\mathit{i}\mathit{j}}}{{\mathit{\Sigma }}_{\mathit{i}}{\mathit{x}}_{\mathit{i}\mathit{j}}}}$$

The entropy of each trait was computed as

$${\mathit{E}}_{\mathit{j}}=-\mathrm{k}{\sum }_{\mathit{i}=1}^{\mathit{n}}{\mathit{p}}_{\mathit{i}\mathit{j}}\ln \left({\mathit{p}}_{\mathit{i}\mathit{j}}\right), \mathit{k}={\displaystyle \frac{1}{\mathit{ln}\left(\mathit{n}\right)}}$$

where n is the number of individuals. Divergence (dj) and weights (wj) were calculated. The differentiation coefficient of each trait was computed as:

$$d_{\mathit{j}}=1-{\mathit{E}}_{\mathit{j}}$$

$${\mathit{w}}_{\mathit{j}}={\displaystyle \frac{{\mathit{d}}_{\mathit{j}}}{{\mathit{\Sigma }}_{\mathit{j}}{\mathit{d}}_{\mathit{j}}}}$$

Entropy analysis revealed differing contributions among traits, with collar growth receiving the highest weight, followed by crown-long and crown-short growth, while height growth contributed least. The final EW-OGI for each individual was computed as:

$$\mathrm{EW}-\mathrm{OGI}=\mathit{w}\mathit{H}\cdot \left(\Delta \mathit{H}\right)+\mathit{w}\mathit{C}\mathit{L}\cdot \left(\Delta \mathit{C}\mathit{L}\right)+\mathit{w}\mathit{C}\mathit{S}\cdot \left(\Delta \mathit{C}\mathit{S}\right)+\mathit{w}\mathit{C}\mathit{D}\cdot \left(\Delta \mathit{C}\mathit{D}\right)$$

EW-OGI is dimensionless and ranges approximately from 0 to 1, with larger values indicating better overall growth performance. Species effects on growth performance were evaluated using generalized linear mixed-effects models (GLMMs). All statistical analyses were performed with R software (R version 4.3.1).

Spontaneous regeneration species were identified following the taxonomic keys of Collenette [49] and Chaudhary [42]. Nomenclature verification and updates were conducted using the Plants of the World Online database [50]. Life span categories and taxa classifications followed White and Leonard [51] and Boulos and Al-Dosari [52]. Regeneration was evaluated using species richness, Shannon diversity, and life-form composition.

3. Results

3.1. Seedling Survival Across Planting Treatments

Seedling survival differed markedly among planting diversity treatments after 16 months of monitoring (Figure 2). Multi-planting plots exhibited the highest mean survival rate of 82% ± 5.4 SE, indicating strong establishment and resilience under hyper-arid conditions.

Bi-planting plots showed intermediate survival at 69% ± 6.1 SE, whereas mono-planting plots had the lowest survival at 58% ± 7.3 SE (Figure 3).

The GLMM revealed a significant effect of planting treatment on seedling survival (p < 0.001). Post hoc pairwise comparisons using Tukey-adjusted contrasts on the logit scale confirmed that multi-planting plots had significantly higher survival than both mono-planting (p < 0.01) and bi-planting (p < 0.05), whereas the difference between mono- and bi-planting treatments was not significant (p ≈ 0.14). Thus, higher planting diversity clearly enhanced seedling survival under the half-moon restoration design.

3.2. Growth Performance and Spatial Heterogeneity

The entropy-weighted index (EW-OGI) (dimensionless, ≈0–1) had an overall mean of 0.51 ± 0.09 (mean ± SD; range 0.004–0.814). A mixed model with species as fixed and plots as random again estimated negligible plot-level variance, and the species effect was highly significant (F_12,369 = 7.21, p < 0.001, η² = 0.19), mirroring the pattern observed for OGI. By species level the mean EW-OGI values (

Table 2. Species-level growth performance of planted shrubs under half-moon restoration conditions. Shown are the number of sampled individuals (n) and species means for height growth (mean_dH, cm), collar diameter growth (mean_dCollar, cm), short crown radius growth (mean_dCrownS, cm), long crown radius growth (mean_dCrownL, cm), simple overall growth index (mean_OGI, cm; sum of all four growth components), and entropy-weighted overall growth index (mean_EW_OGI, dimensionless).

Species	Mean_dH (cm)	Mean_dCollar (cm)	Mean_dCrownS (cm)	Mean_dCrownL (cm)	Mean_OGI (cm)	Mean_EW_OGI
O. baccatus	41.62	11.71	66.23	74.31	193.86	0.61
H. persicum	27.74	12.06	42.26	37.00	119.06	0.56
S. italica	7.00	9.80	44.50	42.50	103.80	0.54
A. gerrardii	24.93	5.86	29.83	33.21	93.83	0.52
H. salicornicum	21.00	8.05	17.23	19.43	65.72	0.51
L. pyrotechnica	8.11	6.40	21.35	20.06	55.91	0.50
L. shawii	15.05	5.87	19.41	18.34	58.68	0.50
P. aphylla	18.17	3.76	30.00	24.33	76.26	0.50
Z. nummilaria	16.00	2.99	33.33	26.80	79.12	0.50
R. raetam	12.25	5.43	15.47	13.53	46.68	0.49
E. alata	16.79	2.73	18.36	17.07	54.95	0.48
S. persica	−4.00	4.68	11.50	2.50	14.68	0.46
C. comosum	−35.75	−5.48	−30.75	−58.50	−130.48	0.32

) ranged from 0.61 ± 0.08 (mean ± SD) in O. baccatus and 0.56 ± 0.14 in H. persicum to 0.32 ± 0.15 in C. comosum, with most other species exhibiting intermediate means clustered around 0.48–0.52 (e.g., A. gerrardii 0.52 ± 0.09, R. raetam 0.49 ± 0.08, L. shawii 0.50 ± 0.05).

Post hoc Tukey comparisons showed that O. baccatus and H. persicum generally outperformed most co-occurring species, whereas C. comosum had significantly lower EW-OGI than nearly all other taxa. These patterns underscore pronounced interspecific variation in growth responses, with a small subset of species emerging as particularly well suited to half-moon restoration structures.

The heatmap (Figure 4) visualizes spatial heterogeneity in species performance across the 18 micro-catchment plots, expressed through the OGI. Distinct patterns emerge that reinforce the trait-based and cluster analyses. Ochradenus baccatus markedly outperformed all other species, achieving an index of nearly 50, driven by rapid crown expansion and strong height increment. Haloxylon persicum and S. italica followed, exhibiting intermediate performance values around 26–30, reflecting balanced structural development and good adaptation to moisture-harvesting plots.

Species such as A. gerrardii, Z. nummularia, and P. aphylla demonstrated moderate growth, while H. salicornicum, L. shawii, L. pyrotechnica, and E. alata exhibited relatively low indices (<17), indicating slower establishment rates. The overall pattern underscores species-specific growth strategies, where shrubs with greater lateral crown development and moderate height growth (notably O. baccatus) show the highest adaptation potential under arid restoration conditions.

3.3. Trait-Based Drivers

Trait–performance modeling identified crown expansion as the principal predictor of overall growth success (Figure 5). The model explained about 10.5% of the variation in OGI (R² = 0.105).

Crown long axis (36.0%, p = 0.018) and crown short axis (28.4%, p = 0.065) at planting contributed most strongly to variation in OGI (

Table 3. Trait coefficients indicating the relative impact of morphological and reproductive attributes on the Overall Growth Index in the half-moon restoration experiment.

Trait	Coefficient (Impact on Growth Index)	Interpretation
Crown long growth	36.0	The strongest driver species with large, long-axis crown expansion show the highest overall growth success.
Crown short growth	28.4	The short-axis crown expansion also significantly boosts success; canopy development in both directions matters.
Height growth	13.0	Vertical growth contributes substantially but less than crown spread.
Collar growth	0.3	Stem thickening has a smaller, secondary role.
Flowers and fruits	14.9 and 7.4	Reproductive output is statistically independent of structural success in this experiment.

), followed by flower production (14.9%, p = 0.018), while collar diameter and fruit production had relatively minor and statistically non-significant effects.

3.4. Functional Clustering

Cluster analysis identified three distinct groups of planted species based on their growth-performance profiles (

Table 4. Growth-performance clusters of restored vegetation species based on OGI, EW-OGI, and structural profiles.

Cluster	Description	Mean OGI	Mean EW-OGI	Structural Profile
Cluster 2—“Fast Expanders”	O. baccatus, H. persicum, S. italica	138.91	0.57	Very strong growth in both crown axes and height; highly vigorous, dominant species. Rapidly occupying space and contributing most to early structural development of the restored vegetation
Cluster 0—“Moderate Growers”	A. gerrardii, Z. nummilaria, P. aphylla, L. shawii	76.97	0.50	Steady but conservative growth; balanced height and crown expansion, suitable for stability.
Cluster 1—“Decliners/Poor Performers”	C. comosum, E. alata	–37.76	0.40	Negative or stagnant growth; limited suitability for rapid structural recovery; possibly stress-sensitive or poorly adapted to local conditions.

). These clusters differed in their mean OGI values and associated structural characteristics.

Cluster 2 (“Fast Expanders; green markers) included O. baccatus, H. persicum, and S. italica, with an average performance of 138.91 cm for mean OGI and 0.57 for mean EW-OGI (Figure 6). Cluster 0 (“Moderate Growers”; orange markers) comprised A. gerrardii, Z. nummularia, P. aphylla, and L. shawii, with intermediate average performance of 77.0 cm for mean OGI and 0.50 for mean EW-OGI. Cluster 1 (“Decliners” or “Poor Performers”; blue marker) included C. comosum and E. alata, showing the lowest average performance, with −37.8 cm for mean OGI and 0.40 for mean EW-OGI.

3.5. Spontaneous Species Recruitment

Spontaneous regeneration within the restoration plots was dominated by perennial species, which accounted for approximately 67% of total species richness, whereas ephemeral species represented around 33% (Figure 7 and Figure 8).

Spontaneous regeneration within the half-moon plots is largely dominated by perennial species (Figure 9), with roughly 3–8 perennial species recorded per micro-catchment (orange bars). Ephemeral species (blue bars) are present in fewer plots and in lower richness (0–6 species), but their contribution varies strongly among micro-catchments. Some plots (e.g. H1, H3, H10, H18) are composed almost entirely of perennials, while others (H7, H11, H15, H16) support both the highest total richness and a larger ephemeral component, indicating particularly favourable microsite conditions.

Perennial dominance increased with planting diversity, from 60% in mono-planting, to 63% in bi-planting, and 80% in multi-planting plots (

Table 5. Composition of ephemeral and perennial species across planting treatments and associated successional interpretation.

Planting Type	Median% Ephemeral	Median% Perennial	Pattern/Interpretation
Mono-planting	≈40%	≈60%	Early-successional mix. Ephemerals are still substantial, indicating active regeneration and incomplete canopy closure.
Bi-planting	≈37%	≈63%	Transitional stage. Slight decline in ephemeral dominance; perennials increasingly stabilize microhabitats.
Multi-planting	≈20%	≈80%	Mature restoration state. Strong perennial dominance and structural vegetation maturity; low ephemeral share shows stable, self-sustaining cover.

; Figure 10A). Although these differences were not statistically significant (p = 0.11), the observed trend suggests improved microsite conditions with increased planting diversity. Ephemeral dominance was highest in mono-planting plots (median ~40%), slightly lower in bi-planting (~37%), and lowest in multi-planting (~20%) (Table 5; Figure 10B).

The Shannon diversity index (H′) varied slightly across planting treatments (Figure 11), with bi-planting exhibiting the highest mean diversity, followed by mono-planting, and multi-planting showing slightly lower overall values. These differences were not statistically significant (p = 0.205). Multi-planting plots supported higher perennial dominance but lower Shannon diversity, while mono- and bi-planting plots showed slightly higher evenness.

3.6. Soil Improvement

Soil analyses revealed notable improvements in the physicochemical environment of half-moon micro-catchments between 2022 and 2024 (

Table 6. Comparative soil analysis showing temporal changes in salinity, nutrient availability, organic matter, and physicochemical properties in the half-moon micro-catchments (2022–2024).

Parameter	2022	2024
pH (H2O)	8.64	9.47
pH (KCl)	7.75	8.59
CaCO3 (%)	8.1	7.6
CEC (Cation Exchange Capacity)	13.8	14.0
Ca (exchangeable mg/kg)	12.31	11.56
P (available P2O5 mg/kg)	12.0	24.0
K (exchangeable mg/kg)	1065	1050
Mg (exchangeable mg/kg)	1138	980
Na (exchangeable mg/kg)	1008	380
Conductivity (µS/cm)	90	16
Soil Organic Carbon (%)	2.4	3.5
Total N (mg/kg)	370	280
Resistivity (Ω·m)	1111	6250
Texture (from separate row)	13% clay, 65% silt, 13% sand	Same

). Electrical conductivity and exchangeable sodium declined sharply, indicating reduced salinity and sodicity and suggesting enhanced leaching and improved soil structure within half-moons. Soil organic carbon increased from 2.4% to 3.5%, reflecting organic matter accumulation from litter input and root turnover as vegetation developed. Available phosphorus also doubled, likely through dust deposition and increased biological activity, whereas total nitrogen declined, consistent with rapid plant uptake during early establishment. Although soil pH increased slightly toward the alkaline range, this trend is common in arid soils and did not appear to restrict shrub performance.

4. Discussion

4.1. Integrating Hydrological Design and Ecological Function

This study suggests that combining microtopographic engineering with native plant functional diversity can enhance both planted seedling performance and spontaneous regeneration in hyper-arid landscapes. Such integrative approaches are crucial for ecosystem restoration under extreme climatic limitations, where abiotic stress severely constrains natural regeneration [1,2,3,4,5,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,48,53,54,55,56,57].

The half-moon plot structures appeared to modify surface hydrology by capturing runoff and promoting localized infiltration, which improved water availability and reduced erosion. These outcomes align with numerous field studies across African and Asian drylands showing that micro-catchments enhance soil water storage, soil fertility, and plant biomass [13,58,59]. By reducing evaporative losses and increasing infiltration, create microhabitats with higher resource concentration, where plant establishment and early growth can proceed even under low rainfall conditions. In our experiment, this hydrological improvement translated into higher survival and growth rates, particularly within diverse plantings.

The interplay between hydrological engineering and plant functional traits underpins a synergistic mechanism of restoration: the physical design initiates favorable abiotic conditions, while biological diversity amplifies feedback that stabilizes and sustains the restored microsite [5,15]. Within the constraints of our sample and study duration, our results support the principle that engineering structures should be coupled with ecological processes, rather than applied as stand-alone interventions. Without biological integration, micro-catchments may fail to sustain vegetation long-term once sedimentation fills the basins or water pulses decline [17]. The introduction of functionally diverse native species has the potential to transform restoration structures into self-regulating patches capable of supporting successional development but longer-term data are required to confirm this trajectory.

4.2. The Role of Species Identity and Functional Traits

Among the planted species, O. baccatus and H. persicum emerged as the apparent superior performers, exhibiting high growth and survival rates within the timeframe of the study. These patterns underscore the importance of species-specific functional traits in determining restoration success under moisture-limited conditions. Ochradenus baccatus displayed a wide lateral canopy and dense evergreen foliage that efficiently shaded soil, reduced thermal stress, and minimized evaporative losses—traits that are advantageous for enhancing soil microclimate and moisture retention [30,31]. Its high OGI and apparent facilitative effect on spontaneous recruitment suggest that it functions as a nurse shrub, improving establishment conditions for other species, including both planted and naturally colonizing taxa. Nevertheless, these conclusions are based on a restricted set of plots and a relatively short monitoring period, and should be validated in larger and longer-term experiments. Conversely, species such as E. alata and L. pyrotechnica exhibited low OGI values and limited vegetative expansion. These differences likely reflect contrasting rooting strategies and water-use efficiencies [34,35]. Shallow-rooted species may be disadvantaged in micro-catchments where water pulses infiltrate deeply, whereas deep-rooted or fast-growing shrubs can exploit transient moisture. However, we did not directly measure root depth or water uptake, so these mechanistic interpretations remain hypothetical.

The entropy-weighted OGI strengthens the assessment of species performance by incorporating the informational value of each growth parameter. Crown-long diameter received a higher entropy weight (0.5904) than height growth (0.4096), indicating that lateral canopy expansion provided more discriminative power for evaluating shrub growth in hyper-arid environments. Ecologically, this reflects the importance of horizontal crown development for light interception, hydraulic redistribution, and microhabitat amelioration in desert shrubs [60,61]. Species such as O. baccatus and H. persicum benefited strongly from this weighting, highlighting their apparent suitability for restoration initiatives.

Conversely, species with minimal canopy growth contributed less to landscape stabilization, explaining the lower EF-OGI values for R. raetam and the negative response of C. comosum. These results suggest that half-moon restoration may favor species with rapid crown expansion, supporting their role in early-stage facilitation processes in degraded desert ecosystems [62]. The dominance of lateral crown expansion as a predictor of performance highlights the adaptive value of structural and morphological traits over purely physiological ones during early establishment. Broad canopies enhance resource capture and shading, fostering soil organic matter accumulation and reducing erosion [28,36]. Such architectural traits therefore act as ecosystem engineering attributes [29], initiating a positive feedback loop between vegetation structure and soil function.

4.3. Facilitation Under Extreme Stress: Testing the Stress-Gradient

The results provide strong empirical support for the stress-gradient hypothesis (SGH), which predicts that positive interactions dominate under high environmental stress [28,53]. Under hyper-arid conditions, planted shrubs with greater biomass and canopy development create localized refuges that buffer soil temperature, reduce wind desiccation, and trap organic matter. These microhabitats enhance the germination and survival of spontaneous species, particularly perennials [1,32], as observed in this study. However, we did not measure microclimate or resource availability directly, so our interpretation of facilitation mechanisms remains inferential.

Higher spontaneous species richness and perennial dominance within plots exhibiting high OGI values indicate that facilitative effects outweigh competitive interactions at this early stage. Similar facilitation-driven recruitment patterns have been documented in Mediterranean and Sahelian drylands, where nurse shrubs such as R. raetam and A. tortilis support the establishment of subordinate species [37,38]. However, facilitation intensity can vary with life stage and resource availability [28]. In the present study, facilitation appears strongest in the establishment phase, where physical amelioration by canopy covers likely dominated over competition for water or nutrients. As vegetation matures, competition may increase, leading to dynamic shifts along the facilitation–competition continuum [63]. Thus, long-term monitoring is necessary to determine whether facilitation persists or transitions toward more neutral or competitive interactions as soil conditions improve. Given the limited number of plots and short temporal window, the strength and generality of the observed facilitation patterns should be interpreted cautiously. Future work with larger sample sizes, replicated across sites and years, would allow more robust testing of SGH predictions in hyper-arid contexts.

4.4. Temporal Improvement in Soil Properties

The substantial improvements in soil properties observed between 2022 and 2024 provide a plausible mechanistic explanation for the strong growth performance detected through both the OGI and the entropy-weighted (EW-OGI). The pronounced reductions in electrical conductivity and exchangeable sodium indicate enhanced leaching and lower salinity and sodicity, resulting in improved aggregate stability and deeper infiltration of scarce rainfall pulses [64]. These physical changes are consistent with the higher EW-OGI values in species capable of exploiting deeper and more stable moisture regimes, such as O. baccatus and H. persicum, which emerged as the strongest performers in both growth indices. However, soil data were only available for two sampling years, and we lacked untreated control areas, so we cannot fully disentangle treatment effects from inter-annual climatic variability.

The increase in soil organic carbon and available phosphorus suggests the early development of biogeochemical feedbacks associated with litter input, root turnover, and microbial activation [65,66], which likely enhanced nutrient availability and water-holding capacity. These processes reinforce the growth advantages observed in species with rapid crown expansion, as reflected by the high entropy weight assigned to crown-long diameter growth in the EW-OGI. In contrast, species with shallow root systems or limited canopy development (E. alata, L. pyrotechnica, C. comosum) exhibited lower OGI and EW-OGI values, consistent with their reduced ability to capitalize on improved moisture conditions within half-moons.

The decline in total nitrogen between sampling years further supports the interpretation that fast-growing shrubs rapidly assimilated available N as part of early successional dynamics [5]. Although soil pH increased toward the alkaline range, no negative effects were observed in the EW-OGI responses, indicating strong tolerance of calcareous substrates among the planted species [67]. Nevertheless, all these soil-related interpretations derive from a limited temporal snapshot and modest sample size; additional measurements over longer periods and across more plots would be needed to confirm the persistence and generality of these trends.

Together, these patterns suggest that the half-moon structures functioned as effective ecosystem-engineering interventions, simultaneously improving soil hydrological and nutrient conditions and amplifying the performance of species whose functional traits align with the enhanced micro-environment [29,68]. The convergence of soil improvements and trait-driven growth quantified objectively through EW-OGI highlights the importance of selecting species with rapid lateral crown expansion, efficient water use, and deeper rooting capacity to accelerate early restoration success in hyper-arid ecosystems, while emphasizing the need for caution in extrapolating beyond the specific context and timescale of this study.

4.5. Successional Pathways and Transition to Passive Restoration

The observed shift from ephemeral to perennial dominance across planting treatments marks a fundamental step in ecological succession. Perennial species stabilize soil, accumulate litter, and create semi-permanent microhabitats, thereby enhancing soil fertility and water infiltration [39]. In this study, multi-species plots exhibited up to 80% perennial dominance (Table 6), indicating that higher planting diversity may accelerate successional progression. This pattern aligns with findings from other arid systems showing that structural and functional diversity promote resilience and stability [69,70]. However, the relatively short monitoring period and limited spatial replication mean that these successional signals should be considered preliminary. The reduction in ephemeral abundance suggests a transition from early, disturbance-adapted assemblages toward persistent, self-reinforcing communities. This dynamic indicates that the micro-catchments are beginning to support passive restoration processes, where natural regeneration sustains vegetation recovery without further human intervention [71]. Yet, given the strong inter-annual variability in rainfall and recruitment in hyper-arid environments, longer-term data are required to confirm whether this trajectory is maintained through dry years.

The positive but non-significant correlation between OGI and spontaneous richness (r = 0.29) still implies a biologically meaningful, though statistically weak, relationship between planted performance and natural colonization potential. As the lack of statistical significance, limited sample size, and potential site heterogeneity constrain the strength of inference. Increasing the number of half-moons or expanding sampling across additional restoration sites would strengthen the robustness of future correlation analyses. In hyper-arid environments, where seed banks are sparse and dispersal limited, even small improvements in microclimate or soil structure can have outsized effects on regeneration [5,72,73]. Hence, facilitating species like O. baccatus may function as biotic catalysts, transforming abiotic micro-catchment improvements into self-sustaining vegetation dynamics. Over time, these pioneer–nurse interactions could drive the emergence of complex shrubland communities analogous to pre-disturbance reference ecosystems, but this remains to be demonstrated empirically.

5. Conclusions

Although this study captures early-stage restoration dynamics, long-term monitoring is needed to assess how micro-catchment–plant feedbacks evolve over decades and under varying climatic conditions. Our conclusions are based on a limited number of plots, a restricted set of species, and a relatively short time frame; they should therefore be viewed as preliminary. Specifically, quantifying belowground processes such as root architecture, soil microbial activity, and organic carbon accumulation would provide deeper insight into the mechanisms driving sustained ecosystem recovery. Additionally, the second phase of the restoration study will integrate baseline soil nutrient parameters as covariates to better account for residual heterogeneity and strengthen the mechanistic interpretation of plant–soil relationships. While qualitative indicators such as species richness and life-form composition were suitable for detecting early spontaneous recruitment, future work will incorporate quantitative measures, including plant density and percent cover, to more accurately characterize regeneration intensity and vegetation structural development. Remote sensing and drone-based photogrammetry could offer valuable tools for scaling up this analysis to regional restoration programs.

In conclusion, our results indicate that micro-catchments can create microhabitats conducive to plant establishment, while native species diversity appears to enhance resilience and facilitate natural regeneration. The apparent dominance of O. baccatus as a facilitative species underscores the potential importance of trait-based species selection in dryland restoration. The study provides an initial, scalable model for transitioning degraded deserts from engineered stabilization toward more self-sustaining recovery, thereby contributing to the growing body of evidence guiding restoration science in extreme environments.