*Article* **Effect of Rice Straw and Stubble Burning on Soil Physicochemical Properties and Bacterial Communities in Central Thailand**

**Noppol Arunrat 1,\*, Sukanya Sereenonchai 1, Chakriya Sansupa 2, Praeploy Kongsurakan <sup>3</sup> and Ryusuke Hatano <sup>4</sup>**


**Simple Summary:** Fire is traditionally used by farmers for clearing fields when the fallow period is short. Soil chemical properties changed significantly after burning and returned to prefire levels after 1 year. *Bacillus*, HSB OF53-F07, *Conexibacter*, and *Acidothermus* abundances increased immediately after burning and then significantly declined, with lower levels 1 year after burning. *Anaeromyxobacter* and Candidatus *Udaeobacter* dominated at 1 year after burning. Burning under high soil moisture conditions and within a very short time caused no effect to the bacterial soil communities.

**Abstract:** Rice straw and stubble burning is widely practiced to clear fields for new crops. However, questions remain about the effects of fire on soil bacterial communities and soil properties in paddy fields. Here, five adjacent farmed fields were investigated in central Thailand to assess changes in soil bacterial communities and soil properties after burning. Samples of soil prior to burning, immediately after burning, and 1 year after burning were obtained from depths of 0 to 5 cm. The results showed that the pH, electrical conductivity, NH4-N, total nitrogen, and soil nutrients (available P, K, Ca, and Mg) significantly increased immediately after burning due to an increased ash content in the soil, whereas NO3-N decreased significantly. However, these values returned to the initial values. Chloroflexi were the dominant bacteria, followed by Actinobacteria and Proteobacteria. At 1 year after burning, Chloroflexi abundance decreased remarkably, whereas Actinobacteria, Proteobacteria, Verrucomicrobia, and Gemmatimonadetes abundances significantly increased. *Bacillus*, HSB OF53- F07, *Conexibacter*, and *Acidothermus* abundances increased immediately after burning, but were lower 1 year after burning. These bacteria may be highly resistant to heat, but grow slowly. *Anaeromyxobacter* and Candidatus *Udaeobacter* dominated 1 year after burning, most likely because of their rapid growth and the fact that they occupy areas with increased soil nutrient levels after fires. Amidase, cellulase, and chitinase levels increased with increased organic matter levels, whereas β-glucosidase, chitinase, and urease levels positively correlated with the soil total nitrogen level. Although clay and soil moisture strongly correlated with the soil bacterial community's composition, negative correlations were found for β-glucosidase, chitinase, and urease. In this study, rice straw and standing stubble were burnt under high soil moisture and within a very short time, suggesting that the fire was not severe enough to raise the soil temperature and change the soil microbial community immediately after burning. However, changes in soil properties due to ash significantly increased the diversity indices, which was noticeable 1 year after burning.

**Keywords:** paddy field; soil organic carbon; soil total nitrogen; microbial diversity; fire

#### **1. Introduction**

Rice straw comprises a major agricultural residue in Thailand. However, large amounts of rice straw are left in fields after harvesting, which are then often burned

**Citation:** Arunrat, N.; Sereenonchai, S.; Sansupa, C.; Kongsurakan, P.; Hatano, R. Effect of Rice Straw and Stubble Burning on Soil Physicochemical Properties and Bacterial Communities in Central Thailand. *Biology* **2023**, *12*, 501. https://doi.org/10.3390/ biology12040501

Academic Editors: Daniel Puppe, Panayiotis Dimitrakopoulos and Baorong Lu

Received: 27 February 2023 Revised: 22 March 2023 Accepted: 22 March 2023 Published: 26 March 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

to prepare the land for the following crops [1,2]. The production of rice straw in Thailand has been estimated at over 20 million tons per year [3]. Among the rice cultivation areas in Thailand, the central region has the highest potential, with two harvests a year. According to the agricultural statistics of Thailand, the total area of the first rice harvest in central Thailand was 1.31 million ha in the crop year of 2020/2021 (3.74 tons ha−<sup>1</sup> in average yield), whereas the second rice harvest area covered 0.54 million ha (4.33 tons ha−<sup>1</sup> in average yield) [4]. This could be explained by its location in the floodplains of the Chao Phraya river basin, which facilitates water availability, thus, encouraging farmers to prepare their lands for the following crops as soon as possible. In this sense, burning is the method of choice for rapidly eliminating rice straw and stubble. If left in the field, rice straw and stubble can represent obstacles for land preparation, and using fire is also the most efficient method for controlling weeds and pests [5,6].

Direct burning in fields significantly changes the soil temperature, moisture, and organic matter content, especially in the topsoil layer [7,8], resulting in an abrupt decline in microbial biomass and diversity [9]. Mickovski [10] reported that burning rice straw and stubble resulted in an increase in soil temperatures by up to 50–70 ◦C in the uppermost 0 to 3 cm of soil, causing a 77% decrease in heterotrophic microorganisms. Biederbeck et al. [11] also found that bacterial populations in the topsoil layer (0–2.5 cm) were reduced by >50% in rice straw burning areas compared to areas where the rice straw remained in the fields. Furthermore, Kumar et al. [12] reported that paddy straw burning reduced the populations of bacteria, fungi, actinomycetes, phosphate-solubilizing microorganisms, potassium-solubilizing microorganisms, cellulose, and microbial enzymes.

Changes to the physical and chemical properties of soil after burning also affect the structure of soil bacterial communities due to the properties of rice straw ash [13]. Duan et al. [14] stated that rice straw ash is alkaline and mainly contains potassium oxide and silicon dioxide; silicon plays a crucial role as a biological stimulant for plant growth [15]. The ash causes changes in the soil pH, affecting the abundance of soil microorganisms. Guo et al. [16] and Zhao et al. [17] reported that changing the soil pH resulted in changes in the abundance of Acidobacteria in rice–wheat cropping systems, whereas the abundances of Proteobacteria, Gemmatimonadetes, and Nitrospirae in soil were modified due to changes in nitrogen, phosphorus, and potassium levels. Avoiding the burning of rice straw by incorporating it into the soil can improve soil nutrient contents by increasing carbon and nitrogen concentrations, thus, affecting the structure of soil bacterial communities. In a study by Zhao et al. [17], the population of Proteobacteria increased following the addition of rice straw due to an increase in soil nitrogen levels. Currently, the incorporation of rice straw into soil is not always cost-effective for farmers in central Thailand, and, thus, burning is the most effortless and cheapest practice for clearing fields when the fallow period is short. However, the effects of burning rice straw and stubble on soil properties and soil bacterial communities remain poorly understood. We hypothesized that soil nutrients would increase, but that soil bacterial communities would decrease after a fire. The objectives of the present study were as follows: (1) to determine the effects of fire on soil organic carbon (SOC), soil total nitrogen (STN), and soil nutrients in paddy fields; (2) to examine changes in the composition and diversity of soil bacterial communities as a result of rice straw and stubble burning; and (3) to analyze the relationship between soil nutrients and soil bacterial communities. The results of this study provide the scientific knowledge for the minimization of postfire risks in paddy fields, which could contribute to implementing suitable management practices for maintaining biodiversity and ecosystem functioning.

#### **2. Materials and Methods**

#### *2.1. Study Site*

The study was conducted in the Taluk subdistrict, Sapphaya district, Chainat province, central Thailand. The area is located in the floodplain of the Chao Phraya river basin and has a tropical savanna (Köppen 'Aw') climate, with an average annual temperature of 27.8 ◦C. April and May are the hottest months, with maximum temperatures in the range

of 38–40 ◦C, whereas the lowest temperatures occur during December–February (18–23 ◦C). The average annual precipitation ranges from 1000 to 1200 mm. The highest precipitation is usually recorded in September, with a total monthly precipitation of 200–300 mm and 10–18 rainy days. The soil belongs to the Chainat series (Cn), which consists of fine, mixed, active, nonacidic, and isohyperthermic Aeric (Vertic) Endoaquepts, and is predominantly dark grayish brown and dark gray. The pH ranges from 5.5 to 6.5, and the soil texture is characterized by silty clay or clay loam. The slopes range from 0 to 2%.

#### *2.2. Field Management Practices and Fire Measurements*

To avoid the effects of variation in environmental conditions, five adjacent farmed fields were investigated in the Taluk subdistrict. The "RD 43" (95 days), "RD 57" (110 days), and "RD 41" (105 days) rice varieties were planted twice a year. The pregerminated rice seeds were sown using the broadcasting method, and chemical fertilizers were applied using 46-0-0 (62.5 kg ha−1) and 16-20-0 (156.3 kg ha−1). Glyphosate (48% *w*/*v* SL) and alachlor (48% *w*/*v* EC) were applied to control weeds, whereas acephate (75% S) and chlorpyrifos (40% EC) were used for disease and insect control. A harvesting machine was used for rice harvesting. Rice straw and stubble were burnt in the field once a year after 20–25 days of sun-drying in August. After 1–3 days of burning, tillage was started for the second rice cultivation, which was harvested in November. As there was not sufficient water for a third rice cultivation, the field was left to fallow for approximately 5 months (December–April) to allow sufficient time for the natural decomposition of rice straw and stubble, and the first rice cultivation was started in May, with harvesting in September.

Five replicated plots were investigated in each field to assess the effects of rice straw and stubble burning on soil properties and bacterial communities, with an area of 5 × 5 m for each plot (Figure 1). Burning was conducted at 10.00–12.00 am and took approximately 98–130 s for each plot (the average burning speed was 6.5 s m−2). Air, fire, and soil temperatures as well as soil moisture were measured in each plot at three periods: before burning (19 August 2021), immediately after burning (19 August 2021), and 1 year after burning (27 August 2022). The fire temperature during burning was measured using an infrared thermometer. The soil temperature and soil moisture were measured at a depth of 5 cm with a Thermocouple Type K and a soil moisture meter, respectively. Each plot was burnt only once, meaning that there was no repeat burning even if rice straw and stubble remained, reflecting the current burning practice.

(**c**)

**Figure 1.** Rice straw and stubble burning practices: (**a**) during burning, (**b**) postburning, and (**c**) one year after burning. Photos were taken by Noppol Arunrat.

#### *2.3. Sample Collection and Analysis*

Soil samples before burning (preburning), immediately after burning (postburning), and 1 year after burning were obtained from five adjacent fields. In each field, five replicated plots at depths of 0–5 cm were investigated. In each plot, the soil samples were collected from five positions (four positions at the four corners of the plot and one in the center). Roots, grasses, stones, and residues were removed manually from the samples, which were then mixed to obtain one composite sample for each field. Approximately 1 kg of soil was placed into a plastic bag for the analysis of the soil physical and chemical properties. Additionally, 100 g samples of soil were placed into zip-lock plastic bags, cooled, and transported to the laboratory for the analysis of bacterial communities. To determine the soil bulk density, a 5.0 × 5.5 cm soil core was taken from each layer, and bulk density was measured after drying at 105 ◦C for 24 h.

Soil texture was determined using a hydrometer. The soil pH was measured with a pH meter using 1:1 suspensions of solids in water [18]. Electrical conductivity (ECe) was measured in saturation paste extracts using an EC meter [19]. The available calcium (Ca), magnesium (Mg), and potassium (K) levels were measured with atomic absorption spectrometry (NH4OAc pH 7.0 extraction) [20]. The available phosphorus (P) concentration was determined using the molybdate blue method (Bray II extraction) [21]. The ammonium nitrogen (NH4-N) and nitrate-nitrogen (NO3-N) levels were measured with the KCl extraction method, and total nitrogen (TN) was measured with the micro-Kjeldahl method. The cation exchange capacity (CEC) was analyzed using the NH4OAc pH 7.0 method. The organic carbon (OC) contents were determined following the method described by Walkley and Black [22]. The SOC stock was estimated using the following equation:

$$\text{SOC stock} = \text{OC} \times BD \times L \times 10,000,\tag{1}$$

where *SOC* is the soil organic carbon stock (Mg C ha<sup>−</sup>1), *OC* is the organic carbon (%), *BD* is the soil bulk density (Mg m<sup>−</sup>3), and *L* is the soil thickness (m).

The STN stock was calculated using the following equation:

$$STN\,stock = TN \times BD \times L \times 10,000,\tag{2}$$

where *STN* is the amount of soil total nitrogen (Mg N ha−1), *TN* is the total nitrogen (%), *BD* is the soil bulk density (Mg m<sup>−</sup>3), and *L* is the soil thickness (m).

#### *2.4. DNA Extraction, Bacterial 16s Amplification, and Sequencing*

The DNA was extracted from approximately 0.25 g of soil using a DNeasy PowerSoil Pro DNA Kit (Qiagen, USA). The hypervariable V3–V4 region was amplified with the 16s rRNA gene using primers 341F (5 -CCTAYGG-GDBGCWSCAG) and 805R (5 -GGACTAC-NVGGGTHTCTAAT-3 ) [23]. Subsequently, the PCR products were sequenced using the Paired-end Illumina Miseq platform (2 × 250 bp) at the Omics Sciences and Bioinformatics Center of Chulalongkorn University (Bangkok, Thailand). All sequencing data associated with this study can be found in the National Center for Biotechnology Information (NCBI) under the BioProject accession number PRJNA881635.

#### *2.5. Bacterial Taxonomic and Functional Identification*

The bioinformatic analysis of the bacterial 16s rRNA gene was conducted on QIIME2 v. 2022.2 [24]. Raw sequence data were quality-filtered and merged, and chimera were removed using the DADA2-plugin [25]. Amplicon sequence variants (ASVs) with less than two sequence reads (singletons) were eliminated. Bacterial taxonomy was assigned using the Silva v. 138 database [26,27], and ASVs that were assigned to mitochondria or chloroplasts were removed. The remaining ASVs were then resampled and normalized to a minimum number of sequences from each sample using the rarefy plugin. This rarefied dataset was functionally assigned using PICRUSt2 [28] to predict the bacterial functions based on marker genes. The gene families for the bacterial sequences were annotated corresponding to the enzyme classification numbers (E.C. numbers). In this study, we highlighted 15 soil enzymes that potentially indicated soil health [29]. The E.C. numbers and names of these enzymes are presented in Supplementary Materials, Table S1.

#### *2.6. Statistical Analysis*

The soil properties before burning, immediately after burning, and after harvesting (1 year after burning) were compared with a one-way ANOVA and post hoc Tukey's HSD tests. The visual graphics were generated with the 'ggplot2' package in the R environment (v. 4.0.2) [30]. The alpha diversity indices, which included the observed richness, Chao-1, Simpson, and Shannon indices, were computed and statistically compared among the sites using the ANOVA (for normal distribution data) or Kruskal–Wallis tests (for nonnormal distribution data). Bacterial community and functional compositions were analyzed and visualized through a principal coordinate analysis (PCoA) based on the Bray–Curtis distance. The differences in compositions were tested using permutational multivariate analyses of variance (PERMANOVAs). A redundancy analysis (RDA) was employed to determine the influence of soil properties on soil bacterial community compositions, and the significance of the correlation between them was confirmed using the Mantel test.

#### **3. Results**

#### *3.1. Soil Physical and Chemical Properties*

No significant differences in soil moisture and soil temperature were found among the three periods (preburning, postburning, and 1 year after burning) (Table 1). At a depth of 5 cm, the soil moisture ranged from 45.1% to 48.4% in the preburning sites, and a slight decrease to 44.5–46.0% was detected in the postburning samples. A rise in soil temperature was measured at 25.9–26.8 ◦C after burning compared with preburning (25.7–26.5 ◦C). During burning, the fire temperature in the litter layer ranged from 415.5 to 469.5 ◦C.

**Table 1.** Soil moisture and soil, air, and fire temperatures at the study sites (minimum–maximum).


a letters denote significant statistical differences (*p* ≤ 0.05).

There were no significant differences of soil physicochemical properties were detected between preburning and 1 year after burning (Table 2). At a depth of 0–5 cm in paddy soils, the bulk density, organic matter (OM), CEC, and soil texture showed no significant changes after burning. The soils had a significantly lower acidity as well as lower NO3-N levels after burning. Conversely, the burned soils showed higher levels of TN, NH4-N, available P, K, Ca, and Mg, and had higher ECe values (Table 2). At 1 year after burning, the soil pH, ECe, CEC, NH4-N, NO3-N, and available P, K, Ca, and Mg were higher than the initial values (preburning), whereas OM and TN were reduced and lower than the preburning values, but the significant differences were not found.

**Table 2.** Physicochemical properties of paddy soil samples before burning, after burning, and 1 year after burning.


a–b letters denote significant statistical differences (*p* ≤ 0.05).

The SOC stock was 17.10 ± 0.3 Mg C ha−<sup>1</sup> in the preburning samples, which increased insignificantly to 17.19 ± 0.1 Mg C ha−<sup>1</sup> after burning. A slightly declined SOC stock was observed at 1 year after burning, with 16.91 ± 0.23 Mg C ha−1, which was slightly lower than the preburning value (Figure 2a). A significantly higher STN stock, with an

increase from 2.03 ± 0.09 Mg N ha−<sup>1</sup> to 2.27 ± 0.081 Mg N ha<sup>−</sup>1, was also identified in the postburning soils. A significant reduction in the STN stock was detected at 1 year after burning (1.85 ± 0.07 Mg N ha<sup>−</sup>1) (Figure 2b).

**Figure 2.** Soil organic carbon (SOC; (**a**) and soil total nitrogen (STN; (**b**) levels in paddy soils. \*\* denotes significant statistical differences (*p* ≤ 0.05).

#### *3.2. Overview of the Sequencing Analysis*

A total of 616,815 (41,121 sequences/sample) clean sequences were obtained in this study. As shown in Figure 3, the rarefaction curves of all samples gradually flattened, indicating that the number of sequences obtained in this study could reflect the bacterial community in the study sites. Here, the sequences were grouped into 18,715 ASVs, which were then classified into 52 phyla, 119 classes, 245 orders, 312 families, and 543 genera.

**Figure 3.** Rarefaction curves of all samples in rice fields at preburning (pre-B), postburning (pos-B), and 1 year after burning (AB).

#### *3.3. Taxonomic Distribution*

As shown in Figure 4, whilst the most abundant taxa in the preburning and postburning samples were Chloroflexi (33%), followed by Actinobacteria (Pre-B = 15%; Pos-B = 18%), and Proteobacteria (Pre-B = 12%; Pos-B = 9%), those in the site 1 year after burning were Actinobacteria (22%), followed closely by Proteobacteria (20%) and Chloroflexi (16%) (Figure 4a). Overall, 11 phyla, 15 order, and 45 genera were indicated as differentially abundant taxa (*p* < 0.05; LDA score > 3). The ANOVA results showed that the relative abundances of 7 out of 10 abundant phyla (average relative abundance > 1%) were notably different among the study samples (Figure 4a). Chloroflexi abundance dramatically decreased by approximately 15–17% at 1 year after burning, whereas the abundance of Planctomycetes increased immediately after burning and then decreased, reaching a level similar to that before burning. The abundances of Actinobacteria, Proteobacteria, Verrucomicrobia, and Gemmatimonadetes significantly increased by 7%, 8%, 3%, and 1%, respectively, at 1 year after burning compared to the pre- and postburning soils. At the order level, significant changes were found to have occurred for several taxa (Figure 4b). The abundances of Ktedonobacterales, Bacillales, and Betaproteobacteria increased slightly immediately after burning, but significantly decreased by 13–14%, 3–5%, and 2–3%, respectively, at 1 year after burning. Myxococcales and Gaiellales abundances significantly increased by 3–4% at 1 year after burning, compared to the two other timepoints.

**Figure 4.** Stacked bar plot showing the relative abundances of the bacterial phyla (**a**) and orders (**b**) in rice fields at preburning (pre-B), postburning (pos-B), and 1 year after burning (AB). Asterisks beside the phylum name indicate statistical significance (*p* < 0.05).

Ten abundant genera (average relative abundance > 1%) were detected in this study. *Bacillus* and HSB OF53-F07 dominated at the pre- and postburning soils, accounting for 5% and 4% in the preburning sites and 6% and 4% in postburning soils. As shown in Figure 5, the abundances of these two taxa increased immediately after burning, but decreased significantly 1 year after burning. This trend was also found for *Conexibacter* and Acidothermus. On the other hand, the abundances of *Anaeromyxobacter* and *Candidatus* Udaeobacter increased significantly 1 year after burning, accounting for 3.1%, 2.7%, and 2.5%, respectively, of the taxa (Figure 5).

#### *3.4. Bacterial Diversity, Community Compositions, and Correlations to Soil Properties*

As shown in Table 3, all alpha diversity indices, including observed richness, Chao-1, Simpson, and Shannon indices, presented similar trends. The diversity indices did not change immediately after burning, but there were significant increases after 1 year. The beta diversity, presented by the PCoA ordination based on the Bray–Curtis distance, overlapped between the pre- and postburning samples, but these two groups were separate from those 1 year after burning (Figure 6a). According to the PERMANOVA results, the bacterial community compositions in the rice fields were significantly differed at 1 year after burning (Figure 6a).

**Table 3.** Alpha diversity indices.


a–b letters denote significant statistical differences (*p* ≤ 0.05).

**Figure 5.** Bar plots for the most abundant genera in this study. Plots with different letters were statistically different. Pre-B—preburning; Pos-B—postburning; AB—1 year after burning.

As shown in Figure 6b, the redundancy analysis (RDA) revealed that the soil properties could explain 67.4% of the total variations in the bacterial community compositions. According to the Mantel test, the TN, STN, sand, clay, soil moisture, and soil temperature were significant soil parameters, shaping the bacterial communities (Table 4). Among these parameters, only clay and soil moisture presented a strong correlation (correlation coefficient > 0.7) (Table 4).

**Table 4.** Correlations and significant values of the bacterial communities and soil properties determined with the Mantel test.


**Table 4.** *Cont.*

**Figure 6.** Bacterial community composition and correlations to soil properties. (**a**) Principal coordinate analysis (PCoA) ordination based on the Bray–Curtis distance, showing the community composition of bacteria detected in the study sites. (**b**) RDA ordination presents soil properties that significantly correlated with community composition. Significant parameters indicated with the Mantel test. Pre-B—preburning; Pos-B—postburning; AB—1 year after burning. \* indicates statistically difference.

#### *3.5. Predictive Functions*

We used PICRUSt2 to predict the functions of the bacterial community based on enzymatic genes. In total, 2372 predictive enzymes were detected across all samples. As shown in Figure 7a, PCoA, which explained 90.1% of the total functional composition, showed that the functional compositions of the bacterial communities found in preburning

and postburning sites were largely similar, although they differed significantly 1 year after burning (PERMANOVA test, *p* > 0.05). In addition, 15 enzymes involved in the carbon, nitrogen, and phosphorus cycles were selected to highlight the potential enzyme activities in soils. The ANOVA results revealed that 10 out of 15 selected enzymes differed significantly among the study sites (Figure 7b). Whilst the abundances of β-glucosidase, chitinase, and urease were higher in pre- and postburning soils compared to the sites 1 year after burning, alpha-N-acetylglucosaminidase and endo-1,4-beta-xylanase presented inconsistent trends. Cellulase and nitrogenase levels increased 1 year after burning compared to the postburning soils. Pectin lyase and nitrate reductase showed similar levels at preburning and 1 year after burning.

**Figure 7.** Bacterial functions predicted using PICRUSt2. (**a**) Principal coordinate analysis (PCoA) ordination, based on the Bray–Curtis distance, shows the functional composition of bacteria. (**b**) Heatmap shows the mean abundances of soil enzymes potentially produced by bacteria detected in the study sites. Pre-B—preburning; Pos-B—postburning; AB—1 year after burning. \* indicates statistically difference.

As shown in Figure 8, Spearman's rank correlation analysis indicated correlations between soil properties and the selected soil enzymes. The parameters OM, OC, TN, STN, soil texture, moisture, and temperature were significantly correlated with several soil enzymes. Positive correlations were found, for example, for OM and amidase, cellulase, and chitinase, for SOC and amidase, cellulase, and TN, and for STN and β-glucosidase, chitinase, and urease. In contrast, negative correlations were found for these enzymes and the clay and soil moisture levels.

**Figure 8.** Spearman's rank correlation between soil properties and soil enzymes predicted with PICRUSt2. X marks indicate insignificant correlation, whereas circles indicate a significant correlation (*p* < 0.05). Circle color corresponds to the correlation coefficient.

#### **4. Discussion**

#### *4.1. Effects of Burning Rice Straw and Stubble on Soil Physicochemical Properties*

Fire can alter the physical and chemical properties of soils through ash deposit. In our study, significant increases in pH, ECe, TN, NH4-N, and soil nutrients (available P, K, Ca, and Mg) were found (Table 2) after fire was used. An elevated soil pH following a fire was also reported in the meta-analysis of Ribeiro Filho et al. [31]. Granged et al. [32] also detected an increase in soil pH from 6.2 to 7.5 immediately after a fire. This increase could be explained by the loss of OH-groups from clay minerals, the formation of oxides, and the incorporation of ash into the soil, leading to an increase in the base cations in soils [33,34]. Kumar et al. [12] and Nigussie and Kissi [35] found increased ECe values after crop residue burning due to increased ash contents in the soil, along with higher available P levels after burning, possibly derived from the available P in ash. Kumar et al. [12] reported that the available P increased from 68.39 to 76.31 kg ha−<sup>1</sup> after the use of fire. Nigussie and Kissi [35] detected an increase in the available P by 73.41% after burning.

Similarly, burning rice straw and stubble provided ash containing high K levels, resulting in an increase in the available K after burning, which was consistent with the results of the study by Gangwar et al. [36]. Burning facilitated the release of nitrogen from rice straw and stubble, increasing the TN content. Pellegrini et al. [37] and Parro et al. [38] also reported an increase in nitrogen after the use of fire. Ammonium (NH4-N) and nitrate (NO3-N) are the plant-available forms of nitrogen, and are generated via the

decomposition of organic N compounds [39]. Previous studies [40–42] have reported high NH4-N concentrations after high-severity burning, whereas NO3-N did not form directly through heating [43], but was produced after burning through the nitrification of NH4-N [44]. This might explain the significant decrease in NO3-N of the paddy soils after burning (Table 2). According to Wan et al. (2001), NH4-N increases immediately after burning and returns to a preburning level after 1 year, whereas NO3-N can recover to preburning levels after 5 years. Covington et al. [45] also found that changes in NO3-N could not be detected immediately after a fire, but the levels increased 1 year after burning, exceeding the preburning levels.

Interestingly, the bulk density, OM, OC, CEC, and soil texture remained unaffected by the burning of rice straw and stubbles (Table 2), most likely because the fire was not severe enough to alter these soil properties. In contrast, Baldock and Smernik [46], Oguntunde et al. [47], and Ayodele et al. [48] detected a decreased bulk density due to the conversion of residues to a char form and the residue remaining from incomplete combustion. Although previous studies reported a decline in OM after burning [7,39,49], this was not the case in the present study. As shown in Figure 1, burning was practiced while the rice stubbles mostly stood in the fields, resulting in a lower impact of the fire on the soil surface. Due to the insignificant changes in OM and clay content after burning, the CEC was largely unaffected, as it is closely related to the OM and clay content. Similarly, Fonseca et al. [50] detected unchanged CEC levels after shrub burning in the northeast of Portugal. As shown in Figure 2, burnt paddy soils contained higher SOC stocks than unburnt paddy soils. The slight increase in the SOC stock after burning may have occurred as a result of the charred material and ash. However, a significant difference was not detected, indicating that burning rice straw and stubble did not alter the SOC stock. We, therefore, hypothesized that the fire was not severe enough to consume the soil carbon. This was in agreement with the findings of Neill et al. [51], who detected no significant changes in soil carbon levels in a prescribed burning in a Cape Cod oak–pine forest. In contrast, the STN stock increased significantly in the postburning soils (Figure 2), which was also observed in previous studies [52,53].

Due to remaining ash in the paddy fields, the soil pH, ECe, CEC, NH4-N, NO3-N, and available P, K, Ca, and Mg levels were higher 1 year after burning than before burning. On the other hand, the OM levels were lower after burning, most likely because of the slow decomposition of residues. A similar trend was observed for TN, which may have occurred due to plant uptake (Table 2). Burning rice straw and stubble resulted in temporary changes in SOC and STN stocks, as well as soil nutrient levels. Several studies have also reported that changes in soil nutrient levels were temporary and that, generally, the levels returned to those measured before burning [54–56].

#### *4.2. Soil Bacterial Community Composition and Diversity Responses Immediately after Burning*

Bacteria represent the most abundant and diversified population of microorganisms worldwide [57], and play an important role in the decomposition of OM and the cycling of nutrients in agricultural ecosystems [58,59]. Fire directly compromises the survival of soil microbial communities through soil heating [9]; in our study, straw burning caused a decline in soil microorganisms at the soil surface due to an increase in soil surface temperatures. Mickovski [10] reported that burning straw and stubble heated the soil temperature in the uppermost 0–3 cm to approximately 50–70 ◦C immediately after a fire. The heat resulted in the mortality of 77% and 9% of the heterotrophic microorganisms in the topsoil (0–5 cm) and the 5–10 cm layer, respectively. Biederbeck et al. [11] investigated the effects of burning cereal straw and reported that repeated burning in the field caused a reduction in the bacterial population by more than 50%. Kumar et al. [12] also reported that the soil temperature increased to 55 ◦C immediately after rice straw burning, resulting in a significant decrease in bacteria, fungi, actinomycetes, phosphate-solubilizing microorganisms, potassium-solubilizing microorganisms, and cellulose degraders after burning. The results of the abovementioned studies were, however, in contrast to our findings, namely, concerning the relative abundances of bacterial phyla, which did not significantly differ (*p* > 0.05) between preburning and postburning sites (Figure 4a). Our sites had a high initial soil moisture; therefore, the burning finished after only a short period of time (6.5 s m−<sup>2</sup> on average), resulting in the fire not being severe enough to kill the soil microbial communities. Although the fire temperatures in the residue layer reached between 298.2 and 603.0 ◦C, the soil temperature after burning only increased by 1.6–3.8 ◦C compared to the preburning samples (Table 1). Moreover, it was hypothesized that whilst the fire could affect the soil microbial communities in the uppermost layer (0–1 cm), no effect was observed in the deeper layer (2–5 cm) because the high soil moisture contents may have had limited the heat transfer to the deeper soil layers. According to Busse et al. [60], heat transfer can decrease when the soil water content exceeds 20%. This was also supported by Whelan et al. [61], who reported that the prevalence of dormant soil microorganisms decreased with low soil moisture contents and high temperatures.

Previous studies reported that fire has a significant effect on the thin topsoil layer, such as the 0–1 cm layer [62], the 0–2.5 cm layer [11], and the 0–3 cm layer [63]. This could explain the insignificant difference between the soil microbial communities of the preburning and postburning sites observed here at a depth of 0–5 cm (Figure 4a). Similarly, Li et al. [64] investigated the responses of soil Acidobacteria to a wildfire disturbance in the topsoil (0–10 cm) and subsoil (10–20 cm), and found no significant differences in Acidobacterial α-diversity between these two soil layers across different fire severities. However, studies on the impacts of rice straw burning on soil microbial dynamics are still limited, since most studies have focused on wildfires or prescribed fires. Kumar et al. [12] investigated the responses of soil microbial communities to paddy straw burning in sandy loam soil. The authors discovered that soil microorganisms and microbial enzymes temporarily decreased after burning, before recovering 30–60 days after burning.

Chloroflexi were the dominant bacteria in both preburning and postburning fields, followed by Actinobacteria, Firmicutes, Planctomycetes, Proteobacteria, and Acidobacteria, although no significant difference was detected among the sites (Figure 4a). Previous studies also reported the dominance of these bacterial phyla in paddy soils [65–68]. Chloroflexi act as primary degraders of polysaccharides under the anaerobic conditions of rice fields, whereas Actinobacteria can also degrade OM under such conditions [69] and produce enzymes involved in carbon cycling for plant residue decomposition and carbon sequestration in soils [70]. Firmicutes comprise a range of thermophilic, antibiotic, and endospore-producing members, many of which are extremely resistant to desiccation, heat, and radiation, and can survive in extreme conditions [71]. Planctomycetes and Proteobacteria are important phyla involved in nitrification and denitrification processes in the soil [72,73]. Trivedi et al. [74] reported that Proteobacteria are classified as "copiotrophs" (R-strategists), which are more abundant in nutrient-rich conditions. Proteobacteria also play a key role in OM decomposition and produce several types of glycosyl hydrolases, thus, promoting plant growth [75]. Acidobacteria use diverse metabolic pathways in ecological processes, including biogeochemical cycles, biopolymer decomposition, and exopolysaccharide secretion [76]. Stinca et al. [77] investigated the soil in a beech forest 2 years after a wildfire, and found that the most abundant phyla (Proteobacteria, Acidobacteria, Bacteroidetes, Planctomycetes, Firmicutes, Gemmatimonadetes, and Chloroflexi) had remained relatively unaffected by the fire, except for Actinobacteria.

At the genus level, *Bacillus* and HSB OF53-F07 were the dominant taxa in both the preburning and postburning sites, but significant differences were not detected. Moreover, *Anaeromyxobacter* was more abundant in the preburning than postburning sites (*p* > 0.05), whereas *Acidothermus* was more abundant in the postburning sites than the preburning ones (*p* > 0.05) (Figure 5); *Acidothermus* is more resistant to higher temperatures than other bacteria. This was consistent with the findings of Mohagheghi et al. [78], who reported that the *Acidothermus* bacteria could remain active at high temperatures, with an optimum growth temperature of 55 ◦C. Furthermore, the *Acidothermus* genome contains genes that encode for thermostable enzymatic cellulose degradation.

Most soil microorganisms die at temperatures exceeding 50 ◦C due to changes in their cells and enzymes [79]. In the present study, β-glucosidase, chitinase, and urease were the dominate soil enzymes produced by the soil bacterial communities. However, the relative abundances of those enzymes did not differ significantly between preburning and postburning soils (Figure 7b). Stott et al. [80] reported that β-glucosidase is an important enzyme for the decomposition of crop residues and provides the energy source for heterotrophic bacteria.

#### *4.3. Changes in Soil Physicochemical Properties and Soil Bacterial Community Composition and Diversity*

Fires can promote the growth of soil bacteria with heat-resistance capacities and enhance those with potential fast-growth strategies [81]. Increased abundances of Actinobacteria, Proteobacteria, Verrucomicrobia, and Gemmatimonadetes were observed 1 year after burning, whereas the abundance of Chloroflexi decreased, and could not reach the initial value (Figure 4a). Although the Planctomycetes abundance increased immediately after burning, the relative abundance decreased and returned to that of the preburning soils. This was consistent with previous studies reporting that changes in soil physical and chemical properties after a fire could indirectly affect soil microbial communities [82–84]. Zhao et al. [17] reported that the relative abundances of Proteobacteria, Betaproteobacteria, and Actinobacteria significantly increased with increased levels of STN, available N, and available P. Our study also pointed out that STN, TN, sand, clay, soil moisture, and soil temperature significantly shaped the bacterial community composition (Table 2). The increased pH after burning (Table 2) was a crucial factor in determining microbial activities, as it led to changes in community composition and diversity.

At the genus level, the abundances of *Bacillus*, HSB OF53-F07, *Conexibacter*, and *Acidothermus* increased immediately after burning and then significantly decreased, whereas *Anaeromyxobacter* and Candidatus *Udaeobacter* dominated 1 year after burning (Figure 5). We hypothesized that *Bacillus*, HSB OF53-F07, *Conexibacter*, and *Acidothermus* were dominant genera in paddy soils, and may be highly resistant to heat, but with low growth rates. In contrast, *Anaeromyxobacter* and Candidatus *Udaeobacter* may have high potential growth rates and mainly occupy areas with increased soil nutrient levels after a fire. However, for a better understanding of the behavior of these bacteria, long-term studies are necessary.

*Bacillus* can promote plant growth through nitrogen fixation, the solubilization and mineralization of phosphorus, zinc, and iron [85]. *Conexibacter* can transform nitrate into nitrite via denitrification under limited oxygen conditions [86]. Seki et al. [87] reported that *Conexibacter* is a slow-growing microorganism and can survive in arid environments. *Acidothermus* is involved in cellulose degradation and enhances plant growth [88], whereas *Anaeromyxobacter* can fix and assimilate N2 to NH4N with the use of nitrogenase, and is involved in the reduction of N2O to N2 [89,90]. This could explain the increase in the abundance of *Anaeromyxobacter* (Figure 5) and in the levels of nitrogenase (Figure 7) 1 year after burning. Moreover, *Anaeromyxobacter* is an iron-reducing bacterium, using OM as the electron donors [91], and can denitrify NO2 <sup>−</sup> to NO through Fe2+ oxidation [92]. Candidatus *Udaeobacter* can secrete antibiotics in the soil and has the potential to remove trace gases, especially H2 [93]. Brewer et al. [94] revealed that Candidatus *Udaeobacter* is an aerobic heterotroph with numerous amino acid and vitamin transporters, minimizing the cellular architecture and sacrificing metabolic versatility to become dominant in the soil.

The cellulase and nitrogenase levels were increased 1 year after burning (Figure 7), most likely because the increased soil nutrient levels accelerated the microbial enzyme production. Cellulase is generally produced for cellulose decomposition mechanisms [95], whereas nitrogenase can reduce N2 to NH3 [96]. Soil enzymes play the most active part in all biochemical processes in soil. In the present study, positive correlations were found between OM and amidase, cellulase, and chitinase, and between SOC and amidase. Moreover, cellulase increased with increasing TN levels, and an increase in STN influenced β-glucosidase, chitinase, and urease (Figure 8). This led us to infer that soil enzymes were

more related to OM levels and directly involved in OM mineralization, thus, affecting carbon and nitrogen cycles. To further reveal the effects of repeated fires on soil bacterial activities and soil properties, long-term studies are needed. Furthermore, it should be noted that the enzymatic results from this study based on the prediction tool, the actual measurements of the enzymatic activities, should be further investigated in the future.

#### **5. Conclusions**

Soil chemical properties (pH, ECe, NH4N, NO3N, TN, and available P, K, Ca, and Mg) of paddy soils significantly increased after burning. However, these values largely returned to prefire levels after 1 year. The most abundant taxa in preburning and postburning soils were Chloroflexi, Actinobacteria, and Proteobacteria. At 1 year after burning, Chloroflexi abundance decreased dramatically, whereas the abundances of Actinobacteria, Proteobacteria, Verrucomicrobia, and Gemmatimonadetes significantly increased. At the genus level, *Bacillus*, HSB OF53-F07, *Conexibacter*, and *Acidothermus* abundances increased immediately after burning, and then significantly declined, with lower levels 1 year after burning. *Anaeromyxobacter* and Candidatus *Udaeobacter* dominated at 1 year after burning. Fires tend to directly and indirectly affect soil microbial communities and functions. Direct effects are exerted through soil heating, killing some bacterial species, whereas indirect effects are due to changes in soil physicochemical properties. In this study, rice straw and standing stubble were burnt under high soil moisture conditions, and burning finished within a very short period of time, indicating that the fire was not severe enough to sufficiently heat the soil and kill the soil microorganisms.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/biology12040501/s1, Table S1: the enzyme classification numbers and their descriptions.

**Author Contributions:** Conceptualization, N.A. and S.S.; methodology, N.A., C.S. and P.K.; investigation, N.A. and S.S.; writing—original draft preparation, N.A., C.S., P.K., S.S. and R.H.; writing—review and editing, N.A., C.S., P.K., S.S. and R.H.; supervision, R.H. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research project was supported by the Mahidol University (Basic Research Fund: fiscal year 2021). We would also like to thank Thailand Science Research and Innovation (TSRI) for their funding of this project, under contract number BRF1-A8/2564.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the institutional review board of the Institute for Population and Social Research, Mahidol University (IPSR-IRB) (COA. no. 2021/01-003; date of approval: 28 January 2021).

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Raw sequence data generated for this study are available in the sequence read archives (SRAs) of the National Center for Biotechnology Information (NCBI) under BioProject accession number PRJNA819169.

**Acknowledgments:** We would like to thank Mahidol University (Basic Research Fund: fiscal year 2021) and Thailand Science Research and Innovation (TSRI) for their funding. Moreover, we would like to express special thanks to all key informants for their information, including the reviewers for their comments and suggestions to improve this paper.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

**Abdulaziz M. Assaeed , Basharat A. Dar , Abdullah A. Al-Doss , Saud L. Al-Rowaily, Jahangir A. Malik and Ahmed M. Abd-ElGawad \***

> Plant Production Department, College of Food and Agriculture Sciences, King Saud University, Riyadh 11451, Saudi Arabia

**\*** Correspondence: aibrahim2@ksu.edu.sa; Tel.: +966-5626-80864

**Simple Summary:** Plants adapt themselves to harsh environmental conditions by changing morphological parameters through phenotypic plasticity. Plants modify their functional traits and allocate biomass to either tolerate or resist the stress caused by their variable habitats. In this study, we observed that *Aeluropus lagopoides,* being among the few halophytic palatable species of salt marshes, adapt to the harsh salt marshes of different eco-regions by significantly modifying its morphological and physiological traits. Due to this structural modification, this plant showed great potential to rehabilitate different inland and coastal saline flat areas (sabkha), taking saline agriculture and soil remediation into consideration.

**Abstract:** Understanding the response variation of morphological parameters and biomass allocation of plants in heterogeneous saline environments is helpful in evaluating the internal correlation between plant phenotypic plasticity mechanism and biomass allocation. The plasticity of plants alters the interaction among individuals and their environment and consequently affects the population dynamics and aspects of community and ecosystem functioning. The current study aimed to assess the plasticity of *Aeluropus lagopoides* traits with variation in saline habitats. Understanding the habitat stress tolerance strategy of *A. lagopoides* is of great significance since it is one of the highly palatable forage grass in the summer period. Five different saline flat regions (coastal and inland) within Saudi Arabia were targeted, and the soil, as well as the morphological and physiological traits of *A. lagopoides,* were assessed. Comprehensive correlation analyses were performed to correlate the traits with soil, region, or among each other. The soil analysis revealed significant variation among the five studied regions for all measured parameters, as well as among the soil layers showing the highest values in the upper layer and decreased with the depth. Significant differences were determined for all tested parameters of the morphological and reproductive traits as well as for the biomass allocation of *A. lagopoides*, except for the leaf thickness. In the highly saline region, Qaseem, *A. lagopoides* showed stunted aerial growth, high root/shoot ratio, improved root development, and high biomass allocation. In contrast, the populations growing in the low saline region (Jizan) showed the opposite trend. Under the more stressful condition, like in Qaseem and Salwa, *A. lagopoides* produce low spikes in biomass and seeds per plant, compared to the lowest saline habitats, such as Jouf. There was no significant difference in physiological parameters except stomatal conductance (*gs*), which is highest in the Jizan region. In conclusion, the population of *A. lagopoides* is tolerant of harsh environments through phenotypic plasticity. This could be a candidate species to rehabilitate the saline habitats, considering saline agriculture and saline soil remediation.

**Keywords:** functional traits; saline flat regions; halophytes; biomass allocation; desalination

#### **1. Introduction**

Different eco-regions with different climatic conditions can alter the available resources and conditions crucial to plant performance. The response of the plants to these

**Citation:** Assaeed, A.M.; Dar, B.A.; Al-Doss, A.A.; Al-Rowaily, S.L.; Malik, J.A.; Abd-ElGawad, A.M. Phenotypic Plasticity Strategy of *Aeluropus lagopoides* Grass in Response to Heterogenous Saline Habitats. *Biology* **2023**, *12*, 553. https://doi.org/10.3390/ biology12040553

Academic Editors: Daniel Puppe, Panayiotis Dimitrakopoulos, Baorong Lu and Caifu Jiang

Received: 23 February 2023 Revised: 28 March 2023 Accepted: 3 April 2023 Published: 5 April 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

environmental changes is through induced phenotypic changes [1]. Plant species with wide distribution among different eco-regions show large intraspecific variations in most functional and phenotypic traits [2,3]. The spatial variation in functional traits and their phenotypic plasticity can help plants persist under global climate change [4,5]. Variations in biotic and abiotic factors in different geographical regions can lead to morphologically and functionally different ecotypes [6]. Plants have to adjust to environmental heterogeneity through the plasticity of adaptive traits and respond to changes in light availability [7–9], water availability [9], nutrients [9,10], salinity [11,12], and temperature to survive and sustain in the soil-plant atmospheric continuum environment [13].

One major dependency for plant species to maintain their populations under variable environmental conditions is phenotypic plasticity [14,15]. Plants persist through this potential mechanism of phenotypic plasticity when faced with faster environmental changes and lead it toward homeostasis levels, thus allowing their proper functioning [16]. In the face of global warming, phenotypic plasticity has become a benchmark for understanding its potential for population persistence and adaptation [17]. Alterations in environmental conditions like light regimes, soil properties, humidity, and rainfall may shift several phenotypic traits [18–20]. Different plant populations exhibit adaptive plasticity in response to spatial variability of environmental conditions, such as climate and edaphic factors [21–26]. In heterogeneous environments, both abiotic and biotic factors can influence plant life-history traits (seed germination, growth, flowering, and reproduction) and adaptation (plasticity or differentiation) [15,27]. Measuring these phenotypic trait variations (both reproductive success and vegetative growth) and investigating the environmental variability of sites in which the population occurs is necessary to assess the adaptability and conservation status of the target species [28,29]. The plasticity of plants has a crucial impact on the community structure and dynamics, where plasticity alters many interactions between organisms and their abiotic and biotic factors of the environments [30]. The plant species characterized by phenotypic plasticity can colonize a wide range of habitats and modify the community structure as they tolerate different environmental factors.

Plants respond to variable environments by adjusting their multiple aspects of allocation and architecture, morphology, and physiology [1,31] to mitigate stress levels and increase the uptake of the limiting resources [32]. Biomass allocation among plant parts is driven by environmental conditions, which define many plant growth processes [33,34] and is related to the phenotypic characteristics of plants. Therefore, plant phenotypic plasticity can be used as a potential covariate for understanding biomass allocation [35,36].

In arid and semi-arid regions, water loss due to evapotranspiration increases the salt concentration in soil components [37], leading to more severe salinization issues. Natural saline habitats vary in salinity levels both spatially and temporally due to topography, soil properties, and micro-climate differences [38]. One such saline habitat is sabkha i.e.., a flat area of clay, silt, or sand with an overlying crust of soil [39]. The salt stress, moisture content, physio-chemical soil characteristics, and other environmental factors in saline areas tend to show relative stability with time [40], which has an extensive influence on community structure, plant morphological structure, and biomass allocation [41]. The biomass allocation of plants represents their growth and metabolism and affects the plant's functional attributes [42].

Most salt marsh plant species are halophytes with a high degree of phenotypic variations, occupying a broad range of environmental conditions and possessing various traits to adapt to saline conditions [43–45]. Halophytic species have developed different mechanisms for regulating growth and development to ensure their survival in highly-saline inland or coastal areas, salt marshes, dunes, and desert habitats [46,47]. The distribution of some halophytic species is best correlated along a gradient of soil variables, such as salinity, moisture content, soil texture, organic matter, and calcium carbonate [48]. Halophytic grasses can tolerate salinity at a species-specific level and vary with the ecotype, region's habitat, and specific environmental factors [14,49]. These halophytic species show adaptive

phenotypic plasticity, enabling them to cope with different saline environments [50], as most traits exhibited considerable plasticity in response to different salinity stresses [51].

*Aeluropus lagopoides* (L.) Thwaites (Poaceae) is a stoloniferous halophytic perennial C4 photosynthetic grass ranging in distribution from Northern Africa (Morocco to Somalia), Italy, and Cyprus, through the deserts of the Middle East to Central Asia, Pakistan, and India [52]. In Saudi Arabia, it is found in different regions of saline coastal environments and inland areas. *A. lagopoides* was recorded from the inland wadi (valley) of Qareenah, Riyadh, saltmarsh areas of Qaseem and Jouf, and coastal zones of Salwa and Jizan region [53–56]. *A. lagopoides* is of economic importance as it is a palatable summer forage in arid areas as well as a sand stabilizer [14] and can be used for landscaping the urban areas of desert regions [57]. The plant withstands high salinity stresses up to 25 dS·m−<sup>1</sup> and can adapt to heterogeneous environments due to structural adaptations and phenotypic trait modifications [58]. There is a considerable variation in environmental conditions of *A. lagopoides* habitats between different coastal and inland regions of Saudi Arabia, with variable effects on water relations, salinity, light, ambient temperature, and edaphic factors [55]. Consequently, the only dependency to maintain its populations under stressful environmental conditions is adaptive plasticity [14,15]. The ability of *A. lagopoides* to grow in different regions provides an excellent opportunity to study its phenotypic trait variations with respect to the regions in which it grows. However, the relationship between biomass allocation and root/shoot morphological strategies of *A. lagopoides* growing in different saline regions is unclear. Therefore, in this study, we aimed to explore the linkage of the variation in the functional traits of vegetative and root parts of *A. lagopoides* (i.e., phenotypic plasticity) with the differences in the habitats (edaphic factors). We aimed to clarify the following questions (1) how do morphological parameters of *A. lagopoides* synergistically change in response to habitat conditions? (2) what biomass allocation strategies did *A. lagopoides* have under different saline regions?

#### **2. Materials and Methods**

#### *2.1. Surveyed Regions and Soil Analysis*

The populations of *A. lagopoides* were studied along Saudi Arabia during the years 2020–2021 and were found in five saline regions (Figure S1) were identified as follows: (1) Salwa (lowland coastal saline flat area), (2) Jizan (southern coastal saline flat area), (3) Qareenah (inland saline flat areas of wadi Hargan, Riyadh region), (4) Qaseem (an inland saline flat area of Al-Aushazia) and (5) Jouf ( an inland saline flat area in Domat Aljandal). The regions' details are presented in Table S1. Each region was geographically different from the others as the shortest point-to-point distance between them was more than 300 km (Table S1 and Figure S2). Within each region, five distinct *A. lagopoides* patches were randomly selected for soil sampling and plant morphological traits measurements. From September to March (when *A. lagopoides* become fully flourished), three random plots (5 × 5 m) were selected within each patch (Figure 1). To assess the relationship between morphological and biomass allocation of the plant and resource allocation of the rhizosphere soil properties, three core soil sampling was selected. At three random points, three core soil samples were collected from three soil layers (0–15 cm, 15–30 cm, and 30–45 cm) within each plot. Each corresponding soil layer of these three soil samples was merged as one composite sample. A total of three composite soil samples represented each plot, and subsequently, a total of 9 samples from each patch. Hence, a total of 225 composite soil samples from all the studied region (5 regions × 5 patches × 3 plots × 3 layers) were collected. For soil moisture content, part of each sample was collected in duly labeled moisture tin, and the moisture content was immediately determined by the weight-loss method for all the samples.

**Figure 1.** Different studied flat saline regions, (**A**) Qareenah, (**B**) Qaseem, (**C**) Salwa, (**D**) Jouf, and (**E**) Jizan. The left is an overview, and the right is a close view of *A. lagopoides*.

For further analysis, all the soil samples were collected in plastic bags, duly labeled, and transferred to Range Science Lab., College of Food and Agriculture Sciences, King Saud

University, Riyadh, Saudi Arabia. All the soil samples were spread over separate plastic sheets, air-dried at room temperature, and sieved through a 2 mm sieve to remove any debris, and the soil samples were analyzed similarly to the previously reported approach of Dar et al. [55]. In brief, the soil texture was determined using the hydrometer method [59]. Soil organic matter was determined by wet combustion with dichromate at 450 ◦C [60]. Soil water extracts (1:5) were prepared for the estimation of soil electrical conductivity (EC) and pH [60]. Soluble inions (Cl− and SO4 <sup>2</sup>−) were determined by the titration method, while soluble cations (Ca2+, Mg2+, Na+, and K+) were determined using a flame photometer according to Rhoades [61].

#### *2.2. Morphological Traits Measurements*

Within each plot, five fully matured individuals were randomly selected for morphological parameters, including both on-field and off-field functional trait measurements were recorded. A total of 375 individuals (5 regions × 5 patches × 3 plots × 5 individuals) were targeted for the measurements. In the field, shoot length, number of tiller/plant, number of leaves/tiller, number of spikes/plant, spike length, top internode length of the main tiller, number of stolon/plant, and top internode stolon length were measured.

On the other hand, after taking the field measurement, the same individuals were excavated and collected in labeled plastic bags. The bags were brought in an ice-cool box to the Range Science Laboratory, College of Food and Agriculture Sciences, King Saud University, Riyadh, Saudi Arabia, for other measurements like leaf area, biomass, and root morphological parameters. Plant samples were separated into root and shoot systems. The leaf area of five leaves of each individual was measured using the WinDIAS system (Delta-T Devices Ltd., Cambridge, UK). Also, the average area and biomass of five spikes of each individual were measured. The root system of all the individuals was thoroughly washed, and their main root length, root hair length, total root area, and root dry matter were measured. Based on these measurements, specific leaf area (SLA) was determined as the ratio of leaf area to leaf dry mass. Leaf dry matter content (LDMC) was calculated as the ratio of leaf dry mass to saturated fresh mass [62]. Leaf thickness was calculated as the ratio (SLA × LDMC−1). For resource allocation, root/shoot ratio, root mass fraction, and shoot mass fraction were calculated. These functional traits were selected to assess the response and plasticity of *A. lagopoides* to the environmental factors within different regions, according to Perez-Harguindeguy et al. [63].

#### *2.3. Determination of Ecophysiological Traits*

Before the targeted plant individuals were excavated, chlorophyll fluorescence, chlorophyll content, leaf temperature, and stomatal conductance were measured. Chlorophyll fluorescence was measured with an Opti-Sciences OS30p+ chlorophyll fluorometer (Opti-Sciences, Hudson, NY, USA). Fluorescence was measured at midday (11.00–13.00 h, solar time). Chlorophyll fluorescence, initial fluorescence (F0), maximum fluorescence (Fm), and variable fluorescence (Fv) were determined, and the ratios of Fv/F0 and Fv/Fm were calculated using MINI-PAM fluorometer (Heinz Walz GmbH, Effeltrich, Germany). The minimum and maximum dark-adapted fluorescence (F0, Fm) and Fv/Fm (where Fv = Fm—F0) were obtained after the leaves of the plants were dark-adapted for at least 20–25 min [64].

In-situ stomatal conductance (gs) was measured using a steady-state diffusion porometer (model SC-1, Decagon Devices, Inc., Pullman). Each day before measurements, the porometer was calibrated, and gs was measured on the adaxial surface of a fully developed penultimate leaf in the afternoon (13:00–15:00 h). The chlorophyll content was measured according to the method of Lichtenthaler and Wellburn [65] with some modifications. About 0.5 g of the fresh plant sample was extracted by acetone, and the content of chlorophyll a (Chl. a), chlorophyll b (Chl. b), and total chlorophyll were determined by measurements of the absorbance at 663 and 645 nm with the UV-VIS spectrophotometer (SHIMADZU, Kyoto, Japan, UV1800).

#### *2.4. Statistical Analysis*

To compare the various traits of *A. lagopoides* and determine the significant variations among regions, the data for the plant functional traits and ecophysiological parameters were analyzed by one-way ANOVA with the region as a factor. However, the soil properties were analyzed for three-way ANOVA with the regions, soil layers, and samples as factors. The ANOVA was performed using Statistix 8.1 software. The soil samples were also used as a factor in the soil analysis to check the homogeneity of soil samples within the studied region. Mean values were compared by the Duncan Multiple Range (DMR) test using SAS 9.1.3. The standard error (SE) was calculated for each mean value.

To correlate the various plant traits (vegetative and reproductive) with each region, a data matrix of the shoot, root, and reproductive traits from the five studied regions was subjected to principal component analysis (PCA). The plant morphological traits were plotted as loading vectors in a bi-plot, while the region was plotted as observations.

On the other side, in order to assess the relationship between the soil parameters of each region and morphological traits, two datasets were prepared; one of the various morphological traits and the second of the soil variables of the studied regions at the three layers (0–15 cm, 15–30 cm, and 30–45 cm). These two datasets were subjected to ordination analysis using canonical correspondence analysis (CCA). Also, the agglomerative hierarchical clustering (AHC) and heatmap were performed based on the data of the top layer soil parameters and the morphological traits of *A. lagopoides* populations within the five studied regions. The AHC was performed based on the Pearson correlation coefficient and weighted pair-group average agglomeration method. PCA and AHC were performed using the XLSTAT software program (version 2018, Addinsoft, NY, USA), while CCA was produced using the MVSP software program, ver. 3.1 [66].

#### **3. Results**

#### *3.1. Soil Layer Variations among the Regions of A. lagopoides*

Soil analysis revealed significant variation (*p* < 0.05) among the five studied regions supporting the growth of *A. lagopoides* for all measured parameters, except for HCO3 as well as among the soil layers (Table 1). The Qaseem, Salwa, and Qareenah regions had the highest and comparable pH values for the top layer soil (0–15 cm), while the Jizan and Jouf regions attained the lowest values of the pH (Table 1). In general, the pH values decreased significantly (*p* = 0.0219) with the soil depth in all studied regions.

The soil salinity is highly significantly varied among soil layers for all regions (*p* = <0.001). The soil of the Qaseem region generally had the highest values of EC (25.95 dS/m for the top layer, 10.28 dS/m for the middle layer, and 6.32 dS/m for the lower layer), cation (Ca2+, Mg2+, and Na+) and anions (Cl−) of all the regions as well as for all the soil layers. However, the values of K+ and SO4 <sup>2</sup><sup>−</sup> varied in significance from layer to layer among the locations, the highest value of K+ for the top layer (20.22 meq/L) being in the Qaseem region and the below two layers (9.95 meq/L for the top layer and 3.34 meq/L for the lower layer in Jouf region (Table 1). The highest value of SO4 <sup>2</sup><sup>−</sup> for the top layer (71.8 meq/L) was recorded in the Jizan region. However, Qaseem attained the highest amount of Cl− (237.60, 73.30, and 51.00 meq/L for the top middle and lower layers, respectively) for all three layers compared to other locations. The cations and anions of the Jouf region soil showed a trend of lower concentration with soil depth. The soil of the Qareenah region attained the highest content of CaCO3 among all regions, and the content increased with the increase in soil depth. On the other hand, the Qaseem region attained the highest organic matter content for the top layer (1.78%), followed by the Qareenah region for all the soil layers (1.63%, 0.98%, and 0.91% for the top, middle, and lower layers, respectively).


**Table 1.** Physical and chemical properties of different soil layers supporting *Aeluropus lagopoides* in different regions.

Different capital letters showed significant variation among regions at *p* < 0.05 (Duncan's test), with df 4 for the region and 4 for soil layers, respectively. Different small letters revealed significant differences among soil layers (0–15 cm, 15–30 cm, and 30–45 cm). Capital letters indicate the significance of regions and small letters soil layers, EC: electrical conductivity, OM: organic matter, MC: moisture content, \**p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001, and "ns" for *p* > 0.05.

Regarding soil texture, the sand content is highest in Salwa for all three layers and the lowest in silt content, while the Qaseem region had the lowest values of sand and the highest of silt for all three soil layers. Clay content was highest in Qaseem for the top layer (16.80%), while it was highest in the Jizan region for the layer of 15–30 cm (13.58%) and the layer of 30–45 cm (12.32%). Moisture content showed a significant difference (*p* < 0.0001) among layers, and it was highest in the Qaseem region for all three layers compared to other regions' respective layers. Overall, the results of soil analyses showed that the highest soil characteristic values trend from top to bottom layer (0–15 > 15–30 > 30–45) for all the regions, with some minor exceptions.

#### *3.2. Morphological Traits Variations among the Studied Regions of A. lagopoides*

By comparing the five studied regions of *A. lagopoides*, significant differences were determined for all tested parameters of the morphological and reproductive traits as well as for the biomass allocation, except for the leaf thickness, where no significant difference was observed (Table 2).

**Table 2.** Single-factor analysis of variance (ANOVA) showing the effect of different saline regions on plant functional traits and biomass allocation of *A. lagopoides* having a degree of freedom of the studied regions.


SS (Sum of Squares), MS (Mean Square), \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001, and "NS" for *p* > 0.05.

#### 3.2.1. Shoot Traits

A highly significant difference in the shoot length and biomass was observed among regions (*p* < 0.001). Moreover, the shoot length of *A. lagopoides* growing in the Jizan and Salwa regions was the highest, while it was lowest in Qareenah and Jouf regions (Figure 2). For shoot biomass (fresh and dry weight), Qassem and Salwa regions attained the highest values, while the lowest values of shoot fresh and dry weight were assessed in the Qareenah region (4.68 g and 3.05 g, respectively). The number of tillers per plant, stolon number per plant, and stolon length showed highly significant differences (*p* < 0.001) among regions of *A. lagopoides.* The number of tillers per plant was higher in the Jouf region, while the number and length of stolon were higher in Jizan and Qaseem regions (Figure 2). Jouf region attained the lowest values of stolon measurements.

**Figure 2.** Comparison of shoot traits of *Aeluropus lagopoides* growing in different saline flat regions of Saudi Arabia. Values are average (*n = 75*), and the bar represents the standard error. Different letters among regions showed significant differences at *p* < 0.05 after Duncan's test. \*\* *p* < 0.01\*\*\* *p* < 0.001.

#### 3.2.2. Root Traits

All studied root traits of the samples showed highly significant differences (*p* < 0.001) among the five studied regions (Figure 3). For root length, Qaseem, Salwa, and Jizan regions attained the highest values, while Qareenah and Jouf had the lowest root length. For biomass, *A. lagopoides* growing in the Qaseem region attained the highest root fresh and dry weight. Moreover, the highest value of the root area was determined for the *A. lagopoides* growing in the Qaseem region (45.27 cm2), followed by Salwa (35.71 cm2) and Jizan (32.31 cm2) regions (Figure 3).

#### 3.2.3. Leaf Traits

All leaf traits of *A. lagopoides* plants showed highly significant variation among regions (*p* < 0.001), except for leaf thickness which did not vary significantly (*p* = 0.29) from one region to another (Figure 4).

The Jizan region attained the highest values of leaf biomass (fresh and dry weight), while the Qareenah region had the lowest values of both leaf fresh and dry weight. The number of leaves per plant was highest for the population of the Jouf region, while the Qareenah region attained the lowest number of leaves. In contrast, the Qareenah region attained the highest values of specific leaf area (0.16 cm−<sup>2</sup> g<sup>−</sup>1), and the Jizan region showed the highest value of leaf dry matter content (90.16%). However, no significant difference in leaf thickness was found in all studied regions.

**Figure 3.** Measured root traits for *Aeluropus lagopoides* growing in different saline flat regions of Saudi Arabia. Values are average (*n = 75*), and the bar represents the standard error. Different letters among regions showed significant differences at *p* < 0.05 after Duncan's test. \*\*\* *p* < 0.001. FW: fresh weight, DW: dry weight.

**Figure 4.** Comparison of leaf traits for *Aeluropus lagopoides* growing in different saline flat regions of Saudi Arabia. Values are average (*n = 10*), and the bar represents standard error. Different letters among regions showed significant differences at *p* < 0.05 after Duncan's test. SLA: specific leaf area, LDMC: leaf dry matter content. FW: fresh weight, DW: dry weight. \*\*\* *p* < 0.001, and "ns" for *p* > 0.05.

#### 3.2.4. Reproductive Traits

All the measured reproductive traits of *A. lagopoides* showed a highly significant difference (*p* < 0.001) among the studied regions (Figure 5). Regarding spike numbers, the populations of the Jouf region showed the maximum production of spikes and seeds per plant, while Qareenah and Jizan attained the lowest values (Figure 5). The average spike length of the Qaseem regions was the highest, while Qareenah and Salwa regions attained the lowest values of the spike length. For spike biomass (fresh and dry weight), the Jizan region attained the highest values (0.09 and 0.06 mm, respectively), while Qareenah and Salwa regions showed the lowest biomass among the studied regions.

#### *3.3. Variation in Ecophysiological Parameters of A. lagopoides among Different Regions*

The photosynthetic pigments (chlorophyll *a,* chlorophyll *b*, total chlorophyll) varied slightly among different regions but had no significance (Figure 6).

Similarly, there is no significant difference in Fv/Fm, and all values were under 0.79. On the other hand, stomatal conductance showed a highly significant difference among the regions (*p* < 0.001), as the Jizan region had the highest stomatal conductance (52.88 mmole m−<sup>2</sup> s<sup>−</sup>1), while the Jouf region had the lowest value (20.46 mmole m−<sup>2</sup> s<sup>−</sup>1). Moreover, a highly significant difference was observed among the studied regions for the *A. lagopoides* leaf temperature.

**Figure 6.** Ecophysiological parameters of *Aeluropus lagopoides* growing in different saline flat regions of Saudi Arabia. Values are average (*n = 75*), and the bar represents the standard error. Different letters among regions showed significant differences at *p* < 0.05 after Duncan's test. \*\*\* *p* < 0.001, and "ns" for *p* > 0.05. FW: fresh weight.

#### *3.4. Biomass Allocation of A. lagopoides in Response to Different Habitats*

The biomass proportion of plant parts in *A. lagopoides* was significantly different (*p* < 0.001) among the studied regions (Figure 7). Concerning root:shoot ratio, the population of the Qaseem region showed the highest value (0.23), followed by Qareenah (0.16), while the population of the Jizan region attained the lowest values of root/shoot ratio (Figure 7). Similarly, the root mass fraction showed the same pattern, where *A. lagopoides* of the Qaseem region showed the highest root mass fraction (0.18), followed by Qareenah (0.13), while *A. lagopoides* of the Jizan region attained the lowest value (0.04). In contrast, the population of the Jizan region showed the highest shoot mass fraction (0.96), while the population of the Qaseem and Qareenah regions attained the lowest values.

**Figure 7.** Biomass allocation among *Aeluropus lagopoides* collected from different saline flat regions of Saudi Arabia, based on dry matter. Values are average (*n = 75*), and the bar represents the standard error. Different letters among regions showed significant differences at *p* < 0.05 after Duncan's test. \*\*\* *p* < 0.05.

#### *3.5. Correlation Analysis among Functional Plant Traits, Regions, and Soil Variables* 3.5.1. Plant Functional Traits-Regions Correlations

The principal component analysis (PCA) revealed the existence of a close correlation between different morphological traits of *A. lagopoides* and the studied regions (Figure 8). The PCA revealed that the number of spikes and number of seeds per plant correlate with the number of tillers and leaves per plant (Figure 8). The *A. lagopoides* growing in Qaseem and Salwa regions are closely correlated and showed a positive correlation with root and shoot biomass. The root traits (root biomass, root area, and root length) are separated on the upper-left side of the PCA biplot, where it showed correlations to each other as well as with shoot biomass. Spike length showed a correlation with the shoot length. The *A. lagopoides* population of the Jizan region showed a substantial correlation with the leaf biomass, spike biomass, and stolon length, where spike biomass revealed a correlation with the leaf biomass as well as the stolon length (Figure 8). The Jouf region showed a significant positive correlation with specific leaf area, number of leaves per plant, number of tillers per plant, and number of seeds per plant (Figure 8). However, leaf dry matter content was the only morphological trait closely correlated to the Qareenah region.

**Figure 8.** Principal component analysis (PCA) of the measured traits (shoot, represented with red arrows, root represented with brown arrows, reproductive traits represented with blue arrows of *Aeluropus lagopoides* within different saline flat regions (represented with yellow circle) of Saudi Arabia.

#### 3.5.2. Correlations among Soil Variables, Plant Functional Traits, and Regions

The data of soil variables of each layer of each region and the functional traits were correlated using CCA (Figure 9). In general, Qassem and Salwa regions show a close correlation to each other, while Jizan is segregated alone on the lower-right side of the CCA biplot. Jouf region was different from other regions for the soil profile of the three layers (upper, middle, and lower) and separated on the lower-left side of the CCA biplot. Finally, the Qareenah region was separated at the center of the CCA biplot, revealing no specific correlation to any parameters.

Regarding the top layer of the soil (0–15 cm), the Qassem and Salwa regions showed a close correlation to most of the soil parameters that are correlated together, such as moisture content, pH, salinity, organic matter, Na, Cl, Mg, K, and Ca. (Figure 9a). Jizan region showed a close correlation to sulfate content, where it showed a correlation to leaf and spike biomass traits. The soil of the top layer in the Jouf region is different and separated on the lower-left side of the CCA biplot, where it showed a correlation to calcium carbonate and sand contents, and it showed a negative correlation with all morphological traits (Figure 9a).

**Figure 9.** Canonical correspondence analysis (CCA) showing the correlations among the soil variables of different layers separately ((**a**): 0–15 cm, (**b**): 15–30 cm, and (**c**): 30–45 cm layers), regions, and morphological traits of *A. lagopoides*. SFW: shoot fresh weight, SDW: shoot dry weight, RFW: root fresh weight, RDW: root dry weight, lvs/*p*; number of leaves per plant, SLA: specific leaf area, Spk/*p*: number of spikes per plant, AvgSpkL: average spike length, RA: root area, LT: leaf thickness, SL: shoot length, RL: root length, Stl/*p*: number of stolon per plant, LDMC: leaf dry matter content, LFW: leaf fresh weight, LFW: leaf fresh weight, LDW: leaf dry weight, SpkFW: spike fresh weight, SpkDW: spike dry weight, ASL: average stolon length.

The heatmap correlation analysis, based on the soil data of the top layer, revealed that root biomass (root fresh weight and root dry weight) has a significant correlation with Ca2+, Na+, Cl−, Clay, and moisture content, while root area showed a significant correlation to only clay content (Figure S3). On the other hand, specific leaf area showed a significant positive correlation with organic matter, while leaf dry matter content showed a correlation to calcium carbonates.

The leaf thickness revealed a significant positive correlation to Na and K contents. The spike biomass showed a significant correlation with sulfate content. However, leaf dry weight showed a significant negative correlation with K, bicarbonates, and organic matter (Figure S3).

For the middle layer of the soil (15–30 cm), the Qaseem and Salwa regions again showed a close correlation with moisture content, salinity, organic matter, Na, and Ca, while the Jizan region showed a correlation to the clay content. However, the Jouf region showed a close correlation to potassium ions but a negative correlation with all studied plant traits (Figure 9b). Pearson's correlation heatmap of the middle layer showed that root biomass significantly correlates with moisture content and sulfate (Figure S3). Moreover, the specific leaf area showed a significant positive correlation with organic matter like the top layer. The number of tillers and seeds per plant showed a significant correlation with the potassium ion. However, leaf biomass revealed a significant negative correlation with organic matter and calcium carbonates (Figure S3).

The PCA analysis of the lower layer of the soil (30–45 cm) revealed a different pattern compared to the upper and middle layers (Figure 9c). In this soil layer, the plant traits did not show a positive correlation to any soil parameters, except for clay, which correlated to the leaf biomass, spike biomass, and average stolon length of the *A. lagopoides* growing in Jizan (Figure 9c). Pearson's correlation heatmap of the lower layer showed a similar pattern to the middle layer (Figure S3).

#### *3.6. Cluster Analysis of Regions Based on Soil and Plant Functional Traits*

The hierarchical clustering for soil variables showed that Qareenah and Salwa regions are quite similar and showed a little pit correlation to the Jouf region (Figure 10A). However, the Jizan region differs in its soil characteristics from other regions. Regarding the plant functional traits, the Qareenah and Jizan regions are closely related, while Qaseem and Salwa regions showed a close correlation to each other (Figure 10C). However, the Jouf region is unique in the functional traits of *A. lagopoides.*

The combination of clustering with heatmap analysis revealed that the *A. lagopoides* populations growing in the Jouf Region showed a negative correlation with all the soil variables except Na+1 and pH (Figure 10B), while the Salwa region showed a positive correlation with organic matter and chloride. On the other hand, the heatmap analysis of morphological traits revealed that *A. lagopoides* populations growing in the Qareenah region showed a negative correlation with spike dry weight and fresh weight. In contrast, the Jizan region was positively correlated with root length and fresh weight. The Salwa region revealed a positive correlation with spike length (Figure 10D).

**Figure 10.** Agglomerative hierarchical clustering (AHC) and heatmaps of the studied parameters within different saline flat regions of *Aeluropus lagopoides*. (**A**) AHC and (**B**) heatmap based on the soil variables, (**C**) AHC and (**D**) heatmap based on the morphological and reproductive traits. EC: electrical conductivity, OM: organic matter, SLA: specific leaf area, LDMC: leaf dry matter content.

#### **4. Discussion**

Desert vegetation faces various ecological constraints like high temperature, soil salinity, and low soil moisture due to low precipitation, making the desert region a challenging environment for plants to grow [67]. Under stressful environments, desert grasses show specific structural and functional modifications in morphological and physiological characteristics to thrive well in such harsh environments [68]. The present study revealed that the various saline flat areas inhabited by *A. lagopoides* differed significantly in soil physicochemical characteristics from region to region as well as with soil depth (up to 30 cm), i.e., among layers (Table 1). These edaphic factors shaped these study sites' community structure and species association [55]. Salt stress can cause a reduction in water potential in soil and can induce osmotic stress in plants [69]. The structural and functional mechanisms of differently adapted populations of a desert halophyte (*A. lagopoides*) were studied for its survival and growth in hyper-arid-saline environments. When species undergo specific drought events and variable edaphic factors like soil salinity and high pH, they restrict their growth by utilizing energy for survival rather than further growth and development [70,71]. The continuity of environmental effects and the distance between the studied regions may support the likelihood for specific characteristics to become fixed in this grass over time.

Soil physical and chemical parameters of the habitats of all these five studied regions were significantly different, indicating the adaptive potential of *A. lagopoides* to cope with variable environmental conditions. Thus *A. lagopoides* plant faces the dual environmental stress of salt and water scarcity. Most of the soil physio-chemical characteristics values like salinity, pH, organic matter, cations, and some anions like Cl, and HCO3, in all three soil layers of the inland saline flat area of the Qaseem region, were high, followed by the inland saline flat area of the Qareenah region (Table 1). The coastal Salwa and Jouf regions were moderately saline, and the least saline was the coastal saline flat area of the Jizan region (Table 1). The soil salinity, pH, cations, and anions were highest in the top soil layer and decreased significantly with depth, where this observation could be attributed to the high evaporation rate [72]. However, the cations and anions increase in value as soil depth increases in the Jouf region which could be ascribed to waterlogging at the site of the Jouf region [73].

Based on these soil characteristic variations, *A. lagopoides* evolved independently to these different salt levels among regions and responded quite differently relating to their morphological and physiological parameters (Figures 2–6). In this study, the highly saline Qaseem region (25.95 dS·m<sup>−</sup>1) has the stunted growth of aerial parts of *A. lagopoides* (shoot length, stolon length, leaf biomass). This could be due to the high salinity level in the soil of this region. The same less shoot height was reported in *Aeluropuslittoralis* under different salinity [74]. This growth restriction of the aerial part of the plant is an essential morphological adaptation because a short-statured plant may conserve the energy required for vital metabolic processes [75]. In contrast, *A. lagopoides* growing in an inland saline flat area of the Qaseem region had improved root development, such as increased root length, high root fresh weight, root dry weight, and more root area than the low saline Jizan region. The investment in root development is an essential line of defense against salt stress [76] and determines the capacity of the plants to obtain water and nutrients [77,78]. Generally, root parameters increase under salinity in most halophytic species [79], while the opposite is true for glycophytic and less salt-tolerant species [80]. *A. lagopoides*, an indicator species of highly saline soils, grow well and uses Na<sup>+</sup> for many physiological processes [81,82]. The well-developed root system of *A. lagopoides* in highly saline habitats may have provided additional benefits to this plant under physiological drought in extracting moisture from the deeper soil layer, a common phenomenon in plants subjected to limited water availability [83]. This observation is supported by data on biomass allocation, where the *A. lagopoides* growing in the Qaseem region attained the highest root/shoot ratio as well as the root mass fraction (Figure 7). This reflects that when *A. lagopoides* is subjected to more salinity, it invests more energy in root development compared to shoot.

On the other side, most aerial parts like shoot length, shoot fresh weight, and shoot dry weight of the Jouf region were stunted, and the soil was moderately less saline than in the Qaseem region (Figure 2). This may be due to the combined effect of salinity and drought (low soil moisture content). Previous studies also reported reduced stem elongation of *Abies alba* [84] under saline and drought-prone environments. The number of tillers in *A. lagopoides* in the Jouf region is significantly more than in other regions. This, again, may be due to the low soil salinity of the Jouf region. The high osmotic stress of the salt outside the roots reduces the formation rate of new leaves and tiller productions [85] in moderate to high saline regions. *A. lagopoides* tends to produce thick leaves with low fresh and dry weight and low LDMC in the Qaseem region (Figure 4). This may be due to the highest soil salt content and the strongest degree of salinization in this habitat. The leaves of *A. lagopoides* became fleshy and developed a lot of water storage palisade tissues and water transport tissues [86]. Firstly, a fleshy leaf structure can also dilute the cell salt ion concentration to avoid its toxic effect. Secondly, it can also increase vacuole concentration and decrease water potential via ion regionalization, thus alleviating the water stress caused by salt stress [87]. Therefore, forming thick leaves in highly saline habitats may be a survival strategy for inland salt marsh plants to adapt to the harsh environment for a longer time.

Regarding reproductive traits, *A. lagopoides* showed a highly significant difference among the studied regions. Under the more stressful condition, like in Qaseem and Salwa, *A. lagopoides* produce more spikes while showing low spikes in biomass and seeds per plant, compared to the lowest saline habitats, such as Jouf (Figure 5). This could be explained

as the plant, under stressful conditions, invests more in seed production and does not make an effort to produce spikes, which is a strategy to maintain more seeds that is more important to the survival of the species in harsh environments [88]. The spike number and inflorescence biomass of *Spartina alterniflora* have been reported to be decreased with increasing salinity [89]. However, the Jouf region showed maximum production of spikes and seeds per plant, while Jizan, in spite of the lowest production, attained the highest value for fresh and dry weight. This could be due to high spike and seed size and mass which the plants tradeoff for spike number and seeds based on the resources available in the habitats they adapt. Plasticity in reproductive investment is also an important trait in varying environments because changes in spike length, seed number, weight, and size directly influence plant fitness [90].

Different levels of salinity adversely influence stomatal conductance (gs). *A. lagopoides* exhibited declined stomatal conductance (gs) with increasing salinity (Figure 6). It depicts that stomatal conductance is an effective strategy to prevent water loss for maintaining the normal function of photosynthetic activity under saline conditions. However, stomatal conductance decreased in the low saline and low moisture content Jouf region. The decrease in stomatal conductance could reduce water loss, which is an adaptation mechanism by plants in dry soil conditions [91]. The highest efficiency of the PSII photochemistry (*Fv/Fm*) method has been extensively used to detect plant stress differences in response to environmental challenges and, consequently, to screen tolerance levels to environmental stress [92]. The *Fv/Fm* of *A. lagopoides* did not show a significant difference among the studied regions, while all values were under 0.79 (Figure 6), meaning that plants are under stress conditions.

Under limited resources, *A. lagopoides* improves its fitness by balancing biomass allocation between aboveground and belowground plant parts and synergistic morphological variation between shoot and root systems. In variable environments, plants' developmental traits and biomass allocation strategies are responses toward morphological characteristics of a plant's location adaptation to resource heterogeneity [93]. Vegetative (especially leaves) and roots are essential for plants to acquire resources. Plant morphologically changes with the environmental gradient to obtain most of the resources and strategies ecologically to adapt to environmental changes [94]. Under both water and salt stress in the Qaseem region, the roots of *A. lagopoides* adopted a strategy of root development and expansion to obtain resources to improve their adaptive ability. Thus *A. lagopoides* formed a good root architecture and produced a well-developed network of fibrous roots by increasing the root area and increasing root biomass. Our results are in agreement with the conclusion that an increase in soil salt concentration increased root development [95].

Overall, the present results demonstrated that the morphological architecture and biomass allocation of *A. lagopoides* are significantly affected in different saline flat area regions based on habitat heterogeneity vis a vis moisture content and salinity as a strategy for adaptation to harsh environments.

#### **5. Conclusions**

The morphological, reproductive, and physiological traits of *A. lagopoides* in the present study show plasticity with the change in the environmental conditions of saline flat area habitats. The Regions with high salinity, such as Qaseem and Salwa, showed the highest values of most of the shoot and root traits. However, the population of *A. lagopoides* in Qaseem and Salwa showed more spikes and lower spikes in biomass and seeds per plant compared to the lowest saline habitats, such as Jouf. Under stressful conditions, i.e., high salinity, the grass produces more seeds instead of spike biomass or other morphological traits. This plasticity reflects the strategy of *A. lagopoides* to cope with the harsh/saline environment. The ability of *A. lagopoides* to change its morphology with the variations in the environmental conditions enables it to colonize, dominate, and shape the community structure within the salt marsh habitat of different regions. The data on biomass allocation in the present study revealed that *A. lagopoides* invests more energy toward roots than shoots under stressful conditions. Due to the extensive fibrous root network, this plant could be a promising candidate as a soil stabilizer in saline flat areas during the summer season. Based on our data, we can conclude that the population of *A. lagopoides* shows great potential to rehabilitate the saline habitats of inland and coastal saline flat regions, taking saline agriculture, saline soil remediation, and stabilization into consideration, particularly this grass flourished in the dry summer season when these habitats are devoid of forage vegetation. Further study is recommended to evaluate the transplantation of this promising forage grass on a large scale in saline rangeland habitats degraded due to heavy grazing of a few palatable halophytic species.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/biology12040553/s1, Table S1: Geographical addresses of distinct patches of *A. lagopoides* populations of the studied regions of Saudi Arabia, along with yearly climatic data. Figure S1: Map of Saudi Arabia showing sampled regions of *Aeluropus lagopoides* populations; Figure S2: Pairwise geographical distance between the studied regions; Figure S3. Pearson's Correlation heatmap between the top, middle, and bottom layer soil parameters and the different morphological and reproductive traits of *Aeluropus lagopoides* within different saline flat regions.

**Author Contributions:** Conceptualization, A.M.A., B.A.D. and A.M.A.-E.; formal analysis, A.M.A., B.A.D. and A.M.A.-E.; investigation, A.M.A., A.M.A.-E., S.L.A.-R., A.A.A.-D., J.A.M. and B.A.D.; writing—original draft preparation, B.A.D. and A.M.A.-E.; writing—review and editing, A.M.A., B.A.D., S.L.A.-R., A.A.A.-D., J.A.M. and A.M.A.-E.; visualization, A.M.A.-E. and B.A.D. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by The Researchers Supporting Project number (RSPD2023R676) King Saud University, Riyadh, Saudi Arabia.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors extend their appreciation to The Researchers Supporting Project number (RSPD2023R676) King Saud University, Riyadh, Saudi Arabia.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
