*Article* **Changes in Soil Ectomycorrhizal Fungi Community in Oak Forests along the Urban–Rural Gradient**

**Hongyan Shen <sup>1</sup> , Baoshan Yang <sup>1</sup> , Hui Wang 1,\*, Wen Sun <sup>1</sup> , Keqin Jiao <sup>1</sup> and Guanghua Qin <sup>2</sup>**


**Abstract:** The ectomycorrhizal fungi communities of forests are closely correlated with forest health and ecosystem functions. To investigate the structure and composition of ectomycorrhizal fungi communities in oak forest soil and their driving factors along the urban–rural gradient, we set up a *Quercus acutissima* forest transect and collected samples from the center to the edge of Jinan city (urban, suburban, rural). The results showed that the ectomycorrhizal fungal community composition at the phyla level mainly included Basidiomycota and Ascomycota in three sites. At the genus level, the community compositions of ectomycorrhizal fungi, along the urban–rural gradient, exhibited significant differences. *Inocybe*, *Russula*, *Scleroderma*, *Tomentella*, *Amanita* and *Tuber* were the dominant genera in these *Quercus acutissima* forests. Additionally, the diversity of ectomycorrhizal fungi was the highest in rural *Quercus acutissima* forest, followed by urban and suburban areas. Key ectomycorrhizal fungi species, such as *Tuber*, *Russula* and Sordariales, were identified among three forests. We also found that pH, soil organic matter and ammonium nitrogen were the main driving factors of the differences in ectomycorrhizal fungi community composition and diversity along the urban–rural gradient. Overall, the differences in composition and diversity in urban–rural gradient forest were driven by the differences in soil physicochemical properties resulting from the forest location.

**Keywords:** *Quercus acutissima* forest; soil physicochemical properties; community structure; LEfSe analysis; driving factors

## **1. Introduction**

Ectomycorrhizal (ECM) fungi are one of the taxa of fungi that are among the most abundant and widespread in forests [1]. There are about 20,000~25,000 ECM fungi in the world, and they can associate with 6000 species of trees and shrubs mainly distributed in tropical and temperate forests [2–4]. In particular, the tree species from Fagaceae, Betulaceae, Pinaceae and Legumes are the dominant symbiotic plants of ECM fungi [1]. These plants provide carbon synthesized by photosynthesis for the fungi, while ECM fungi supply nutrients to the host plants, such as nitrogen and phosphorus [5,6]. Moreover, ECM fungi play critical roles in global carbon cycles by secreting extracellular enzymes and organic acids to facilitate the decomposition of soil organic matter [7]. In addition, previous studies have confirmed that ECM fungi play a significant role in the enhancement of plant tolerance to adverse environments [8] and the sustainment of forest ecosystem diversity and functions [1,9]. Generally, ECM diversity and community structure are considered to be the important indicators to evaluate the health and stability of forest ecosystems [10]. In this sense, studies on ECM fungi diversity and its driving factors can provide us with a better understanding of ecological processes such as geochemical element cycling, plant nutrient acquisition, sustainable forest development and global environmental change [1,7].

**Citation:** Shen, H.; Yang, B.; Wang, H.; Sun, W.; Jiao, K.; Qin, G. Changes in Soil Ectomycorrhizal Fungi Community in Oak Forests along the Urban–Rural Gradient. *Forests* **2022**, *13*, 675. https:// doi.org/10.3390/f13050675

Academic Editor: Christopher Gough

Received: 21 March 2022 Accepted: 25 April 2022 Published: 27 April 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 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/).

The diversity and richness of ECM fungi are affected by different soil conditions [11,12]. It has been found that higher levels of organic matter and more phosphorus available in soil usually inhibit ECM fungi occurrence [13,14]. Lower pH has been found to be conducive to increases in ECM diversity and richness [15]. ECM communities are significantly different in the cold temperate zone, warm temperate zone and Mediterranean climate zone, indicating that climatic conditions are important driving factors for ECM composition and structure [11]. Furthermore, on the basis of their specificity to host plants, some ECM fungi are able to associate with specific tree species, which is another important factor affecting ECM fungi structure and diversity in different forest ecosystems [1]. Therefore, understanding the influence of environmental factors on the ECM community is critical for the exploitation and protection of ECM fungi function in forest ecosystems.

Oak (*Quercus* Spp.) is one of the dominant deciduous tree genera in boreal forests around the world, and it has great ecological and economic value for environmental protection, seedling establishment and wood yield [16]. Its roots rely on obligate symbiosis with ECM fungi. It cannot survive and establish itself without ECM fungi [17]. Oak roots can improve nutrient absorption and plant adaptability once they are associated with ECM fungi, especially in adverse conditions [16,17]. Thus, the composition and diversity of ECM fungi in oak forests have been paid much attention in previous studies. He et al. [18] observed that the compositions of ECM fungal communities varied with the stand ages of oak forests, but their diversities were broadly similar across the 20-, 30- and 40-yearold *Quercus Mongolica* forests. Furthermore, the ECM fungal community structure of oak forest soil was obviously different in different habitats. The soil fungal biodiversity had a great impact on hosts, especially at the nursery stage, where it could influence the quality of the plating material [19]. Jin et al. [20] found that the ECM fungi in oak nurseries were mainly composed of Boletales while Agaricales dominated in the Guizhou wild oak forest. Moreover, previous studies have shown that ECM fungal diversity and richness in oak forests significantly change with seasonal shifts and that they generally present the highest abundance in summer [20,21]. Although the changes in ECM fungal communities have been extensively studied, there is little information on the composition of ECM fungal communities and its driving factors in *Quercus acutissima* forests along the urban–rural gradient. As an effective ecological research method, the rural–urban gradient method has been widely used to study ecological problems in various regions of the world [22,23]. Here, we investigated the ECM community in *Quercus acutissima* forests along the urban–rural gradient in Jinan city, northern China. The aims were to: (1) explore the changes in ECM community composition and diversity along the urban–rural gradient and (2) reveal the effects of soil physicochemical properties on ECM community compositions and diversity. Our results provide practical guidance for urban and suburban forest ecosystem management.

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

## *2.1. Sample Site*

Jinan city, the capital of Shandong Province, is located between 34◦460–37◦320 N and 116◦130–117◦580 E, in the middle and lower reaches of the Yellow River. The terrain in the south is higher than that in the north in this area. Jinan has a typical temperate monsoon climate characterized by distinct seasons, a mean annual temperature of 13.8 ◦C and mean precipitation of 650–700 mm. Jinan city has abundant forest resources, dominated by *Quercus acutissima*, *Platycladus orientalis*, *Robinia pseudoacacia* and *Pinus thunbergii Parl* [24]. The urbanization level in Jinan reached 73.46% as of 2021 [25]. The typical urban areas, ecologically sensitive areas and southern mountain control areas were designed in Jinan along with the development of the economy. Here, we focused on the *Quercus acutissima* forest, which is a typical type of vegetation in Jinan. The study areas were Quancheng Park (36◦380 N, 117◦380 E) in the urban area, Liubu Forest Farm (36◦270 N, 117◦110 E) in the suburbs and Yaoxiang Forest Park (36◦120 N, 117◦4 0 E) in the countryside. The geographic information for the study areas is shown in Figure 1.

**Figure 1.** Maps of the sampling site and study area location. **Figure 1.** Maps of the sampling site and study area location.

information for the study areas is shown in Figure 1.

#### *2.2. Sample Collection and Treatment Process 2.2. Sample Collection and Treatment Process*

In October 2020, three independent sample plots (20 m × 20 m each; the distances were more than 1000 m from each other) were set up in the urban, suburban and rural *Quercus acutissima* forests of Jinan city, respectively. Five healthy trees with similar tree heights and diameters at breast height were randomly selected in each plot. The distances between the trees were more than 10 m. Humus such as dead branches and leaves was removed from the soil surface, and then four soil cores were collected from the east, west, south and north directions around each tree with 3.5 cm diameter soil augers at 20 cm depths. All the soil samples were evenly combined into one in each plot and a total of nine composite soil samples were thus obtained from these urban, suburban and rural *Quercus acutissima* forests (three replicated samples per forest × three forests). Afterward, the collected samples were stored in cryogenic ice boxes and quickly transported back to the laboratory. The fresh soil samples were sieved through a 2 mm sieve after roots and residues were removed. After being mixed evenly, samples were divided into two parts: one was stored at 4 °C for the determination of soil physicochemical properties and the other was frozen at −80 °C until molecular testing of ECM fungi. In October 2020, three independent sample plots (20 m × 20 m each; the distances were more than 1000 m from each other) were set up in the urban, suburban and rural *Quercus acutissima* forests of Jinan city, respectively. Five healthy trees with similar tree heights and diameters at breast height were randomly selected in each plot. The distances between the trees were more than 10 m. Humus such as dead branches and leaves was removed from the soil surface, and then four soil cores were collected from the east, west, south and north directions around each tree with 3.5 cm diameter soil augers at 20 cm depths. All the soil samples were evenly combined into one in each plot and a total of nine composite soil samples were thus obtained from these urban, suburban and rural *Quercus acutissima* forests (three replicated samples per forest × three forests). Afterward, the collected samples were stored in cryogenic ice boxes and quickly transported back to the laboratory. The fresh soil samples were sieved through a 2 mm sieve after roots and residues were removed. After being mixed evenly, samples were divided into two parts: one was stored at 4 ◦C for the determination of soil physicochemical properties and the other was frozen at −80 ◦C until molecular testing of ECM fungi.

along with the development of the economy. Here, we focused on the *Quercus acutissima* forest, which is a typical type of vegetation in Jinan. The study areas were Quancheng Park (36°38′ N, 117°38′ E) in the urban area, Liubu Forest Farm (36°27′ N, 117°11′ E) in the suburbs and Yaoxiang Forest Park (36°12′ N, 117°4′ E) in the countryside. The geographic

#### *2.3. Soil Physicochemical Properties Determination*

*2.3. Soil Physicochemical Properties Determination*  The soil pH (water: soil = 1: 2.5) was measured using an UltraBasic pH meter (Denver Instruments, UB-10, US) and conductivity (EC) (water: soil = 1:5) was measured using a portable conductivity meter. Microbial biomass carbon (MBC) and microbial biomass nitrogen (MBN) in the soil samples were determined by chloroform fumigation. [26]. Soil organic matter (SOM) was determined using the H2SO4-K2CrO7 oxidation method [27]. Total nitrogen (TN) was analyzed using an automatic Kjeldahl apparatus (KDY-9820, Tongrunyuan, Beijing, China) [27]. Soil ammonium nitrogen (NH4+-N) and nitrate nitrogen (NO3–-N) in the KCl extraction were measured using a UV-visible spectrophotometer (UV-5200PC, Yuanxi, Shanghai, China) [28]. The ratio of C/N in the soil was calculated based on soil organic carbon and total nitrogen contents. Soil available phosphorus (AP) and available potassium (AK) were determined with atomic absorption spectrometers (SHI-MADIV AA-7000, Shimadzu, Tokyo, Japan) after being extracted with 0.5 mol/L NaHCO3 The soil pH (water: soil = 1: 2.5) was measured using an UltraBasic pH meter (Denver Instruments, UB-10, Bohemia, NY, USA) and conductivity (EC) (water: soil = 1:5) was measured using a portable conductivity meter. Microbial biomass carbon (MBC) and microbial biomass nitrogen (MBN) in the soil samples were determined by chloroform fumigation. [26]. Soil organic matter (SOM) was determined using the H2SO4-K2CrO<sup>7</sup> oxidation method [27]. Total nitrogen (TN) was analyzed using an automatic Kjeldahl apparatus (KDY-9820, Tongrunyuan, Beijing, China) [27]. Soil ammonium nitrogen (NH<sup>4</sup> + - N) and nitrate nitrogen (NO<sup>3</sup> −-N) in the KCl extraction were measured using a UV-visible spectrophotometer (UV-5200PC, Yuanxi, Shanghai, China) [28]. The ratio of C/N in the soil was calculated based on soil organic carbon and total nitrogen contents. Soil available phosphorus (AP) and available potassium (AK) were determined with atomic absorption spectrometers (SHIMADIV AA-7000, Shimadzu, Tokyo, Japan) after being extracted with 0.5 mol/L NaHCO<sup>3</sup> and 1 mol/L CH3COONH<sup>4</sup> (pH = 7), respectively [29].

#### and 1 mol/L CH3COONH4 (pH = 7), respectively [29]. *2.4. Soil DNA Extraction, PCR Amplifification and Sequencing*

*2.4. Soil DNA Extraction, PCR Amplifification and Sequencing*  DNA extraction, PCR amplification and high-throughput sequencing of the fungal ITS sequences of the soil samples were completed by Hangzhou LC-Bio Technology Co., Ltd. DNA extraction was performed using an OMEGA Soil DNA Kit (OMEGA, Norcross, GA, USA) according to the manufacturer's instructions. The ITS1 and ITS2 regions were used to identify the fungal species, and the analysis was carried out with ITS1FI2 (50 -GTGARTCATCGAATCTTTG-30 ), ITS2 (50 -TCCTCCGCTTATTGATATGC-30 ), F (50 -GAACCWGCGGARGGATCA-30 ) and R (50 -GCTGCGTTCTTCATCGATGC-30 ) primers.

PCR amplification was performed in a reaction mixture with a total volume of 25 µL containing 25 ng of template DNA, 12.5 µL PCR Premix, 2.5 µL of each primer and PCR-grade water to adjust the volume. The PCR conditions to amplify the ITS fragments consisted of an initial denaturation at 98 ◦C for 30 s; 32 cycles of denaturation at 98 ◦C for 10 s, annealing at 54 ◦C for 30 s and extension at 72 ◦C for 45 s; and then final extension at 72 ◦C for 10 min. The PCR products were confirmed with 2% agarose gel electrophoresis. The PCR products were purified with AMPure XT beads (Beckman Coulter Genomics, Danvers, MA, USA) and quantified with Qubit (Invitrogen, Carlsbad, CA, USA). The amplicon pools were prepared for sequencing and the size and quantity of the amplicon library were assessed with an Agilent 2100 Bioanalyzer (Agilent, Palo Alto, CA, USA) and the Library Quantification Kit for Illumina (Kapa Biosciences, Woburn, MA, USA), respectively. The libraries were sequenced on a NovaSeq PE250 platform. Samples were sequenced on an Illumina NovaSeq platform according to the manufacturer's recommendations, provided by LC-Bio. QIIME software was used to divide the OTU of Illumina Miseq sequencing data. The RDP database and UNITE database (https://unite.ut.ee/ (accessed on 1 March 2021)) were applied to classify and annotate ECM fungi, and Mothur software was used to calculate the Chao1 index, Shannon index and Simpson index.

#### *2.5. Data Analysis*

IBM SPSS Statistics 23.0 (SPSS Inc., Chicago, IL, USA) was used for one-way analysis of variance (ANOVA) and multiple comparison to evaluate the significant differences in the different forest soils (*p* < 0.05). Origin 2021 (OriginLab Inc., Northampton, MA, USA) was used to draw a histogram of community relative abundance at the order and genus levels (others < 0.01). Redundancy analysis (RDA) was performed with CANOCO 5.0 to examine the habitat differences in key ECM fungi species. We utilized linear discriminant analysis effect size (LEfSe) on the website of LC-Bio Technology Co., Ltd., to identify the ECM fungi that might explain the differences among the three forests along the urban–rural gradient (LDA > 4, *p* < 0.05). In addition, Pearson's correlation analysis was applied with Origin 2021 to examine the relative effects of ECM fungi characteristics (diversity and richness) and soil factors (pH, SOM, TN, NH<sup>4</sup> + -N, NO<sup>3</sup> −-N, AP, AK, EC).

#### **3. Results**

#### *3.1. Differences in Soil Physicochemical Properties*

There were significant differences in the soil physicochemical properties among the three sampling areas along the urban–rural gradient in Jinan (Table 1). Specifically, the pH value and the contents of NH<sup>4</sup> + -N, AK and AP in the urban forest soil were significantly higher than those in the suburban and rural forests (*p* < 0.05), whereas TN and NO<sup>3</sup> −-N in the urban forest were clearly lower than those in suburban and rural forests, and the difference between suburban and rural forests was not significant. In addition, the content of SOM in the suburban forest soil was 2.50 and 2.19 times higher than that in urban and rural forests, respectively. No significant differences in C/N among any of the sampling sites were observed.

**Table 1.** Soil physicochemical properties in *Quercus acutissima* forests along the urban–rural gradient.


Different letters in each column indicate significant differences in soil samples (*p* < 0.05). The values in the table represent the mean ± standard error.

#### *3.2. MBC and MBN Contents in the Quercus acutissima Forest Soil*

Both the MBC and MBN in the forest soil exhibited significant differences in *Quercus acutissima* forests along the urban–rural gradient (Figure 2). Generally, the average content **Site pH SOM** 

**Sampling** 

ent.

**(g∙kg<sup>−</sup>1)** 

**TN (g∙kg<sup>−</sup>1)** 

of MBC in urban, suburban and rural soil was 84.21, 53.68 and 205.96 mg·kg−<sup>1</sup> , respectively, showing a descending order of rural > urban > suburban. With the same trend as the MBC, the average content of MBN in urban, suburban and rural forest soil was 18.18, 10.78 and 48.15 mg·kg−<sup>1</sup> , respectively. tively, showing a descending order of rural > urban > suburban. With the same trend as the MBC, the average content of MBN in urban, suburban and rural forest soil was 18.18, 10.78 and 48.15 mg∙kg−1, respectively.

Both the MBC and MBN in the forest soil exhibited significant differences in *Quercus acutissima* forests along the urban–rural gradient (Figure 2). Generally, the average content of MBC in urban, suburban and rural soil was 84.21, 53.68 and 205.96 mg∙kg−1, respec-

**Table 1.** Soil physicochemical properties in *Quercus acutissima* forests along the urban–rural gradi-

Different letters in each column indicate significant differences in soil samples (*p* < 0.05). The values

**AP (g∙kg<sup>−</sup>1)**  **AK** 

**(g∙kg<sup>−</sup>1) C/N** 

**NO3–-N (mg∙kg<sup>−</sup>1)** 

ECM fungi diversity and abundance were significantly different.

*Forests* **2022**, *13*, x FOR PEER REVIEW 5 of 14

**NH4+-N (mg∙kg<sup>−</sup>1)** 

in the table represent the mean ± standard error.

Urban 7.87 ± 0.03 a 10.88 ± 4.13 b 2.67 ± 0.09 b 7.28 ± 0.03 b 11.65 ± 1.59 a 17.92 ± 1.11 a 297.61 ± 9.38 a 3.26 ± 1.28 a Suburban 4.88 ± 0.11 b 27.16 ± 9.88 a 2.92 ± 0.03 a 8.59 ± 0.19 a 3.81 ± 0.68 b 5.22 ± 0.06 b 242.02 ± 15.24 b 5.44 ± 1.05 a Rural 4.50 ± 0.31 b 12.40 ± 8.40 b 2.89 ± 0.01 a 8.81 ± 0.13 a 3.48 ± 0.20 b 4.42 ± 0.10 b 228.14 ± 9.80 b 2.50 ± 1.70 a

*3.2. MBC and MBN Contents in the Quercus acutissima Forest Soil* 

**Figure 2.** MBC and MBN contents in *Quercus acutissima* forest soil along the urban–rural gradient. Different letters in each column indicate significant differences in soil samples (*p* < 0.05). **Figure 2.** MBC and MBN contents in *Quercus acutissima* forest soil along the urban–rural gradient. Different letters in each column indicate significant differences in soil samples (*p* < 0.05).

#### *3.3. The Change in ECM Fungi Diversity*

*3.3. The Change in ECM Fungi Diversity*  All the Good's coverage estimators were greater than 99%, which suggested that there was an overall good sampling (Table 2). The ECM fungal community diversity in rural forest soil was the highest, with a Shannon index of 2.76 and a Simpson index of 0.92, followed by urban forest (2.54 and 0.89) and suburban forest (2.12 and 0.83). However, Chao1 values were broadly similar in the forest soil of urban, suburban and rural forests, which tat indicated habitat changes had little influence on the ECM fungi richness along the urban–rural gradient. This also showed that the effects of the urban–rural gradient on All the Good's coverage estimators were greater than 99%, which suggested that there was an overall good sampling (Table 2). The ECM fungal community diversity in rural forest soil was the highest, with a Shannon index of 2.76 and a Simpson index of 0.92, followed by urban forest (2.54 and 0.89) and suburban forest (2.12 and 0.83). However, Chao1 values were broadly similar in the forest soil of urban, suburban and rural forests, which tat indicated habitat changes had little influence on the ECM fungi richness along the urban–rural gradient. This also showed that the effects of the urban–rural gradient on ECM fungi diversity and abundance were significantly different.

**Table 2.** Community diversity indices of ECM fungi in *Quercus acutissima* forest soil along the urban–rural gradient.


Different letters in each column indicate significant differences in soil samples (*p* < 0.05). The values in the table represent the mean ± standard error.

#### *3.4. ECM Fungal Community Compositions Vary with Urban–Rural Gradient*

A total of 282 ECM fungal OTUs were retrieved from the ITS sequence of soil samples collected from *Quercus acutissima* forests in Jinan along the urban–rural gradient. These ECM fungi were distributed in 3 phyla, 5 classes, 11 orders and 26 genera. At the phyla level, the most dominant were Basidiomycota (81.11%), followed by Ascomycota (13.83%) and Mucoromycota (1.06%). Agaricomycetes were the most dominant guild at the class level, accounting for up to 86.53%, 99.94% and 72.69% of the total community in urban, suburban and rural sampling sites, respectively. Sordariomycetes were the second class, accounting for 0.56%, 0.12% and 27.57% in urban, suburban and rural forests, respectively. Pezizomycetes, Eurotiomycetes and Endogonomycetes were occasionally detected, implying they were the minor classes.

Major differences were observed in ECM fungal community compositions along the urban–rural gradient (Figure 3). At the order level, Agaricales had the highest relative abundance in urban and rural *Quercus acutissima* forests (Figure 3a), which indicates that it can survive in broad habitats and plays crucial roles in sustaining ecosystem balance [30]. Russulales were extensively found in these *Quercus acutissima* forests along the urban–rural gradient, especially in suburban areas where the highest proportion was 56.02%, implying that Russulales may play an important role in the growth and health of oak trees in the sensitive areas disturbed by human activities. The top 13 genera of ECM fungi in *Quercus acutissima* forest soil are shown in Figure 3b. *Inocybe* was the dominant genus in the urban site and its relative abundance was 37.87%, followed by 5.49% for the rural site and 2.84% for the suburban site. However, the relative abundance of *Russula* was highest in the suburban area but lowest in the urban area. High proportions of *Amanita* were observed in the soil samples from the rural area, whereas *Amanita* was almost absent in urban and suburban areas. *Tomentella* was also one of the representative genera in the soil of *Quercus acutissima* forest. The order of its relative abundance in the study sites along the urban– rural gradient was suburban > urban > rural. Overall, the dominant ECM fungi in *Quercus acutissima* forest soil showed remarkable differences along the urban–rural gradient. *Forests* **2022**, *13*, x FOR PEER REVIEW 7 of 14

**Figure 3.** The relative abundance of ECM fungi at the order (**a**) and genus (**b**) levels in the soil of *Quercus acutissima* forests along the urban–rural gradient. **Figure 3.** The relative abundance of ECM fungi at the order (**a**) and genus (**b**) levels in the soil of *Quercus acutissima* forests along the urban–rural gradient.

urban, suburban and rural *Quercus acutissima* forest soils, respectively.

explaining habitat variations that showed significant differences in relative abundance in the urban soil. Meanwhile, *Russula* was the taxon of ECM fungi present in the suburban soil. In addition, there were two characteristic groups of ECM fungi with significant differences in relative abundance in the rural site compared to the urban and suburban sites. They were an unclassified genus from Sordariales and an unclassified genus from Thelephoraceae, respectively. The LDA value also further indicated that *Tuber*, *Russula* and Sordariales were the key groups in responses to more pronounced habitat shifts in the

As shown in Figure 4, remarkable differences in key ECM fungi taxa were observed

*3.5. LEfSe Analysis of ECM Fungi Community* 

#### *3.5. LEfSe Analysis of ECM Fungi Community*

As shown in Figure 4, remarkable differences in key ECM fungi taxa were observed among the urban, suburban and rural *Quercus acutissima* forest soils. Specifically, *Tuber* and an unclassified genus from Thelephoraceae were two distinct ECM fungi members explaining habitat variations that showed significant differences in relative abundance in the urban soil. Meanwhile, *Russula* was the taxon of ECM fungi present in the suburban soil. In addition, there were two characteristic groups of ECM fungi with significant differences in relative abundance in the rural site compared to the urban and suburban sites. They were an unclassified genus from Sordariales and an unclassified genus from Thelephoraceae, respectively. The LDA value also further indicated that *Tuber*, *Russula* and Sordariales were the key groups in responses to more pronounced habitat shifts in the urban, suburban and rural *Quercus acutissima* forest soils, respectively. *Forests* **2022**, *13*, x FOR PEER REVIEW 8 of 14

**Figure 4.** Cladogram (**a**) and LDA distribution histogram (**b**) for ECM fungi in the soil of *Quercus acutissima* forests along the urban–rural gradient. **Figure 4.** Cladogram (**a**) and LDA distribution histogram (**b**) for ECM fungi in the soil of *Quercus acutissima* forests along the urban–rural gradient.

(**b**)

Redundancy analysis confirmed that soil physicochemical properties had the greatest effect on dominant ECM fungi at the genus level (top six relative abundances) (Figure

The dominant *Russula* exhibited a strongly positive relationship with SOC and C/N, while *Scleroderma* was negatively correlated with SOC and C/N, indicating that soil carbon could drive much more variation in *Russula* and *Scleroderma*. *Inocybe* and *Tuber* showed a remarkably positive relationship with NO3–-N, AK, AP and pH, but a strongly negative relationship with NH4+-N and TN. In addition, *Tomentella* weakly positively correlated with SOC, NO3–-N, AK and AP. *Amanita* significantly positively correlated with MBC and MBN, but weakly related with NH4+-N and TN. The three sampling sites for urban, suburban and rural forests were distributed in different quadrants, and the three sampling plots in the same area showed obvious clustering phenomena, implying that the ECM fungal community composition differed significantly in the *Quercus acutissima* forests

*3.6. Driving Factors of ECM Fungal Composition* 

along the urban–rural gradient (Figure 3).

#### *3.6. Driving Factors of ECM Fungal Composition*

Redundancy analysis confirmed that soil physicochemical properties had the greatest effect on dominant ECM fungi at the genus level (top six relative abundances) (Figure 5). The first axis and second axis explain the variations of 54.34% and 27.56%, respectively. The dominant *Russula* exhibited a strongly positive relationship with SOC and C/N, while *Scleroderma* was negatively correlated with SOC and C/N, indicating that soil carbon could drive much more variation in *Russula* and *Scleroderma*. *Inocybe* and *Tuber* showed a remarkably positive relationship with NO<sup>3</sup> −-N, AK, AP and pH, but a strongly negative relationship with NH<sup>4</sup> + -N and TN. In addition, *Tomentella* weakly positively correlated with SOC, NO<sup>3</sup> −-N, AK and AP. *Amanita* significantly positively correlated with MBC and MBN, but weakly related with NH<sup>4</sup> + -N and TN. The three sampling sites for urban, suburban and rural forests were distributed in different quadrants, and the three sampling plots in the same area showed obvious clustering phenomena, implying that the ECM fungal community composition differed significantly in the *Quercus acutissima* forests along the urban–rural gradient (Figure 3). *Forests* **2022**, *13*, x FOR PEER REVIEW 9 of 14

**Figure 5.** RDA of dominant ECM fungi and soil physicochemical properties in *Quercus acutissima*  forests. The solid and dashed arrows represent ECM fungi and soil properties. **Figure 5.** RDA of dominant ECM fungi and soil physicochemical properties in *Quercus acutissima* forests. The solid and dashed arrows represent ECM fungi and soil properties.

#### *3.7. Relationships of ECM Fungal Diversity with Soil Factors*

*3.7. Relationships of ECM Fungal Diversity with Soil Factors*  Among all the soil factors examined (Figure 6), the diversity indices (Shannon index and Simpson index) and richness index (Chao1 index) were strongly positively related to TN, NH4+-N, MBC and MBN, but negatively related to NO3−-N, AP and AK, which was consistent with the research conclusions of Erlandson et al. [11] and He et al. [18]. Moreover, NH4+-N was negatively correlated with AP and AK, while NO3–-N was positively correlated with AP and AK. It is worth noting that SOM and TN showed positive relationships with ECM fungal diversity, but negative relationships with ECM fungal richness, suggesting that the higher SOM and TN contents in soil were probably associated Among all the soil factors examined (Figure 6), the diversity indices (Shannon index and Simpson index) and richness index (Chao1 index) were strongly positively related to TN, NH<sup>4</sup> + -N, MBC and MBN, but negatively related to NO<sup>3</sup> −-N, AP and AK, which was consistent with the research conclusions of Erlandson et al. [11] and He et al. [18]. Moreover, NH<sup>4</sup> + -N was negatively correlated with AP and AK, while NO<sup>3</sup> −-N was positively correlated with AP and AK. It is worth noting that SOM and TN showed positive relationships with ECM fungal diversity, but negative relationships with ECM fungal richness, suggesting that the higher SOM and TN contents in soil were probably associated with the higher ECM fungi diversity.

**Figure 6.** Pearson correlation heat map of soil physicochemical properties and ECM fungal diver-

sity. Red is positive, blue is negative; the darker color means the correlation is stronger.

with the higher ECM fungi diversity.

**Figure 6.** Pearson correlation heat map of soil physicochemical properties and ECM fungal diversity. Red is positive, blue is negative; the darker color means the correlation is stronger. **Figure 6.** Pearson correlation heat map of soil physicochemical properties and ECM fungal diversity. Red is positive, blue is negative; the darker color means the correlation is stronger.

**Figure 5.** RDA of dominant ECM fungi and soil physicochemical properties in *Quercus acutissima* 

Among all the soil factors examined (Figure 6), the diversity indices (Shannon index and Simpson index) and richness index (Chao1 index) were strongly positively related to TN, NH4+-N, MBC and MBN, but negatively related to NO3−-N, AP and AK, which was consistent with the research conclusions of Erlandson et al. [11] and He et al. [18]. Moreover, NH4+-N was negatively correlated with AP and AK, while NO3–-N was positively correlated with AP and AK. It is worth noting that SOM and TN showed positive relationships with ECM fungal diversity, but negative relationships with ECM fungal richness, suggesting that the higher SOM and TN contents in soil were probably associated

forests. The solid and dashed arrows represent ECM fungi and soil properties.

*3.7. Relationships of ECM Fungal Diversity with Soil Factors* 

with the higher ECM fungi diversity.

#### **4. Discussion**

*4.1. Diversity and Composition of ECM Fungal Community in the Quercus acutissima Forest Soil along the Urban–Rural Gradient*

ECM fungi are one of the most important functional phyla in the forest soil ecosystem [31–33]. *Quercus acutissima* is a typical tree species associated with ECM fungi and can form a huge mycelium network to participate in the nutrient cycle and energy metabolism of the host [21]. ECM fungal community composition and diversity are commonly used to evaluate the stability of forest ecosystems because their changes are strongly related to soil health [10,34]. In this study, ECM fungal diversity exhibited more pronounced shifts along the urban–rural gradient. It was obvious that there were minimum levels of MBC and MBN in the suburban forest soil, indicating that microorganism growth was significantly inhibited in this ecotone. The contents of TN, NH<sup>4</sup> + -N and SOM were higher in the soil of the suburban *Quercus acutissima* forest than in urban and rural forests, whereas the Shannon index, Simpson index and MBC and MBN contents of ECM fungi were the lowest, implying that the ECM fungi were not apt to occur in more fertile soil [35]. This finding could have been related to the higher contents of carbon and nitrogen in the suburban forest soil, where ECM fungi were not necessary for the host to obtain more nutrients. Moreover, the decrease in ECM fungi diversity also resulted in a weaker decomposition effect on the soil organic matter or litter [25,26]. Additionally, the ECM fungal diversity and soil microbial biomass in the urban *Quercus acutissima* forest were significantly lower than in rural forest, which implied that human activities in urban areas likely had a negative influence on ECM fungal diversity and even on other kinds of microbial diversity [36]. Previous studies have also shown that lower abundance and diversity of soil microorganisms can result in the limitation of community function. Thus, conservation of ECM diversity is of great significance for maintaining the function of urban forest ecosystems [1].

Key species are generally critical indicators of environment conditions and community functions [36]. This study of ECM fungal community composition in forests could provide a basis for the exploration of the ectomycorrhizal symbiotic mechanism. Earlier studies of ECM fungal communities were only based on morphological and anatomical observations [1,37]. In recent years, use of molecular biology methods has provided new insights into the below-ground ECM fungal community and a more precise approach to the study of ECM fungal diversity [1]. Here, the ECM fungal community compositions in *Quercus acutisana* forest soil were investigated in different habitats (urban, suburban and rural forests) in Jinan city. Some genera, such as *Inocybe*, *Russula* and *Tomentella*, were

found to be dominant, which was similar to the results reported in other oak forests [18]. It was confirmed that these taxa may have important ecological significance in terms of improving the environmental adaptability of trees due to their high genetic diversity and probable functional diversity [38]. *Inocybe*, *Russula*, *Tomentella*, *Amanita* and *Scleroderma* belonging to Basidiomycetes and *Tuber* from Ascomycetes were identified as the key ECM fungi taxa. They were closely associated with the community function; thus, changes in these ECM fungi's relative abundance could potentially alter forest structure and function [20]. Interestingly, *Inocybe* frequently demonstrated a relatively high abundance in urban forest soil, but it was also occasionally observed in suburban and rural forests soil. Similarly, Jin [20] also found that the relative abundance of *Inocybe* in artificial forests was significantly higher in comparison to that growing in natural environment. Our result may be attributable to the presence of simple tree species in urban artificial forests, the conditions of which are favorable for *Inocybe* colonization. *Russula* is a widely distributed ECM fungi genus that can associate and be symbiotic with *Quercus*, *Pine, Abies* and *Spruce* throughout the world, especially in boreal temperate forests [39]. In our study, *Russula* showed a high relative abundance in the *Quercus acutissima* forest ecosystem along the urban–rural gradient, notably accounting for up to 56.01% in the suburbs. *Amanita* was the most abundant ECM fungi in the rural *Quercus acutissima* forest. Previous studies have confirmed that both *Russula* and *Amanita* are late successional species [18,39,40]. In the current study, the suburban and rural *Quercus acutisana* forests were older than the urban forest, which might have been the reason why the relative abundances of *Russula* and *Amanita* were higher in suburban and rural forest ecosystems. Yang et al. [41] found that *Scleroderma* and *Tomentella* had a strong Cd tolerance under the stress of Cd and benomyl. Both were also prevalent in this study, indicating that they play an essential role in resisting environmental stress, as well as maintaining plant growth and metabolism. In addition, some ECM fungi with lower abundances, such as *Elaphomyces*, *Helvella*, *Boletus*, *Amphinema* and *Hydnum*, were only observed in the rural *Quercus acutissima* forest, while *Pisolithus* was occasionally identified in the suburban forest. The functions of these rare or specific ECM fungi in different habitats need to be further studied.

#### *4.2. Influencing Factors of ECM Fungal Community Composition in Quercus acutissima Forests along the Urban–Rural Gradient*

The changes in ECM fungal community composition and diversity are often used to predict ecosystem health [1,20]. Here, there was an obvious correlation between soil physicochemical properties and ECM fungal community structure in the *Quercus acutissima* forests along the urban–rural gradient. Soil physicochemical properties can affect the growth and distribution of EMC fungi, thereby resulting in different EMC fungal communities [1,42]. In particular, pH exerts an important role in soil nutrient transformation and cycling by influencing the microbial community composition and activity [43]. It has been reported that most EMC fungi can survive in weakly acidic conditions, while their growth is inhibited in alkaline conditions [21]. Studies have found that the seedling roots associated with mycorrhizal fungi not only perform normal respiration to produce CO2, but also secrete organic acids and H<sup>+</sup> , leading to soil acidification in forests [44,45]. In this study, the soil pH ranged from 4.50 to 5.00 in the suburban and rural forests, while it was about 7.80 in the urban forest. The ECM fungal Shannon and Simpson indices in the rural forest were higher than those in the urban forest, suggesting that soil pH was significantly negatively correlated with diversity indices. This conclusion is consistent with the previous results obtained by Kyaschenko, who also stated that the presence of a large number of ECM fungi likely reduces soil pH values and that the diversity of a fungal community decreases as the pH values increase [46].

SOM is one of the main sources of metabolic substances and energy for the colonization of mycorrhizal fungal [47]. In turn, the interaction between fungal communities can also regulate the SOM accumulation and soil fertility [47,48]. Remarkably higher SOM content was detected in the suburban *Quercus acutissima* forest soil in Jinan. A possible reason was that the proportion of other mixed tree species was larger in the *Quercus acutissima* forests, which probably enhanced the input of organic matter from the aboveground litter layer into the soil and provided sufficient sources during SOM retention. In contrast, the *Quercus acutissima* forest in the urban area was strongly disturbed by human activities, which resulted in a low level of litter input and affected the formation and accumulation of soil SOM. Nitrogen is widely recognized as a key limiting nutrition in controlling species composition and diversity in forest ecosystems [49]. The soil nitrate nitrogen content in the *Quercus acutissima* forest was significantly higher than that of ammonium nitrogen. The NO<sup>3</sup> <sup>−</sup>-N has a negative charge and the adsorption with soil is weak [50]. Thus, NO<sup>3</sup> −-N is likely to be leached from soil, which is not conducive to fertilizer conservation [50]. However, the content of NH<sup>4</sup> + -N was more than twice that of NO<sup>3</sup> −-N in the suburban and rural *Quercus acutissima* forest soils, which could have effectively prevented soil nitrogen loss and maintained soil fertility [50]. Hao et al. [51] found that most EMC fungi have a preference for NH<sup>4</sup> + -N in the growth process. Similar results were found in this study: ECM fungal diversity was significantly positively correlated with NH<sup>4</sup> + -N and negatively correlated with NO<sup>3</sup> <sup>−</sup>-N, indicating that most ECM fungi also preferred NH<sup>4</sup> + -N in the *Quercus acutissima* forest soil. However, a number of studies have reported that excessive nitrogen deposition (mainly nitrate) also inhibits the growth of EMC fungi in temperate broad-leaved oak forests and leads to a decrease in their diversity [52]. Moreover, EMC fungi have different preferences and adaptability in different soils [34]. In the present study, there were different correlations between soil properties and key ECM fungi taxa. For example, *Russula* and *Scleroderma* were extremely sensitive to SOM content and *Russula* was suitable for growing in suburban forests with high SOM content. However, *Scleroderma* preferred to colonize in urban and rural forests with relatively low SOM content. It also had a wide adaptive capacity in the environment and its abundance was not affected by the contents of nitrogen, phosphorus and potassium in soil. *Scleroderma* was dominant in low-nutrient-level soil, which means that this genus may have a unique function in nutrient uptake. In comparison with *Scleroderma*, *Inocybe* is generally considered as a later successional ECM fungus [53], and it showed a very significant positive correlation with AP, AK and NO<sup>3</sup> −-N contents in the soil, suggesting that the growth of *Inocybe* requires abundant nutrient supply and a stable ecosystem. These results indicate that the occurrence of ECM fungi in soil is affected by different soil physicochemical properties, which may be related to ECM fungal species and their ecological adaptability [31]. In turn, ECM fungi also greatly alter the chemical form and availability of soil nutrients by secreting functional substances, thus playing a regulatory role in the host rhizosphere environment [20]. Hence, ECM fungi in forests along the urban–rural gradient strongly change with the various soil properties. The effects of soil physicochemical properties on soil health and ECM fungi should be considered in the cultivation of *Quercus acutissima* forests. As far as ECM fungi protection is concerned, it is necessary to detect in a timely manner the changes in soil physicochemical properties in forests. In order to obtain detailed information on ECM fungi along the urban–rural gradient, further studies need to be implemented including focuses on sampling sites, frequency and the driving force of human activities.

#### **5. Conclusions**

Soil ECM fungi are fundamental components of forest ecosystems. The three *Quercus acutissima* forests demonstrated significant shifts in ECM fungal communities and diversity in different habitats, suggesting that different forest management strategies may be needed for microbial biodiversity conservation along the urban–rural gradient. Moreover, the distribution of ECM fungi in the oak forest soil implies that the dominance of ECM fungi differs depending on the spatial location. Our results provide basic information on forest protection and management along the urban–rural gradient in Jinan city. More abundant information about the location-specific relationship between environmental factors and microbial community is urgently needed for effective forest soil management.

**Author Contributions:** Conceptualization, H.S. and W.S.; methodology, H.S.; software, H.S. and K.J.; validation, B.Y., G.Q. and H.W.; formal analysis, H.S.; investigation, H.S. and B.Y.; resources, H.S.; data curation, H.S.; writing—original draft preparation, H.S.; writing—review and editing, H.S.; visualization, H.S.; supervision, H.W.; project administration, H.W.; funding acquisition, H.W. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Natural Science Foundation of China (41877424; 31870606), the Research Leader Studio Project (2021GXRC094), the Key R & D project of Shandong Province (2021LZGC005-02-02) and the Fundamental Research Funds for the Central Universities, CHD (300102351505).

**Acknowledgments:** The authors acknowledge Hangzhou LC-Bio Technology Co., Ltd. (Hangzhou, China), for providing the sequencing platform and technical support.

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

#### **References**


**Decheng Liu 1,†, Zongqiang Chang 2,† , Xiaohui Liang <sup>1</sup> and Yuxia Wu 1,\***


**Abstract:** The degrees of adaptive responses of different halophytes to saline–alkali soil vary substantially. *Kalidium* (Amaranthaceae), a genus comprised of six species of succulent euhalophytes with significantly differing distributions in China, provides ideal material for exploring the ecophysiological relationships involved in these variations. Thus, in a large-scale field survey in 2014–2018, samples of soil (at 20 cm depth intervals spanning 0 to 100 cm) and seeds were collected from areas where these six species are naturally distributed. Chemical properties of soils in the areas and germinability of the species' seeds in media with 0–500 mM NaCl and 0–250 mM Na2SO<sup>4</sup> were then analyzed to test effects of salinity-related factors on the species' distributions. The pH of the soil samples mainly ranged between 8.5 and 10.5 and positively correlated with their mean total salt contents. Germination rates of all six species' seeds were negatively correlated with concentrations of NaCl and Na2SO<sup>4</sup> in the media, and their recovery germination rates in distilled water were high (>74%). The results show that the species' distributions and chemical properties of their saline soils are strongly correlated, notably the dominant cation at all sites is Na<sup>+</sup> , but the dominant anions at *K. cuspidatum* and *K. caspicum* sites are Cl− and SO<sup>4</sup> <sup>2</sup>−, respectively. Species-associated variations in concentrations of Ca2+ were also detected. Thus, our results provide clear indications of major pedological determinants of the species' geographic ranges and strong genotype-environment interactions among *Kalidium* species.

**Keywords:** germination percentage; *Kalidium*; halophytes; pH; ion content; total salt contents; saline soil

#### **1. Introduction**

One of the diverse environmental factors that strongly affect terrestrial plants' natural distributions is soil salinity, which has growth-impairing and lethal effects on all plants when it exceeds species-dependent thresholds [1–3]. Further, salinity reportedly reduces crop yields on about a fifth of all irrigated land and, in combination with increasing global scarcity of water resources, salinization of soil and water is seriously threatening crop yields and future food production [4,5]. However, halophytes, comprising about 1% of the world's flora, can grow in saline environments with relatively high concentrations of electrolytes [6]. For example, the growth and development of glycophytes is severely inhibited by exposure to 100–200 mM of NaCl, while halophytes can tolerate and complete their life cycles at substantially higher concentrations [2,4,7]. This is due to anatomical adaptations in halophytes, such as salt bladders, salt hairs, and/or salt glands in the leaves [7,8] and various physiological tolerance mechanisms. For example, excess salt may be excreted through trichomes of halophytic grasses [8], or diluted by increases in the water content and thickness of succulent halophytes' leaves [9,10]. Moreover, different halophytes growing together on the same saline–alkali soil often have substantially differing elemental

**Citation:** Liu, D.; Chang, Z.; Liang, X.; Wu, Y. Soil Chemical Properties Strongly Influence Distributions of Six Kalidium Species in Northwest China. *Forests* **2022**, *13*, 2178. https://doi.org/10.3390/f13122178

Academic Editors: Xiao-Dong Yang and Nai-Cheng Wu

Received: 21 November 2022 Accepted: 16 December 2022 Published: 19 December 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 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/).

concentrations, indicating that their physiological selectivity varies [11]. Generally, the dominant ions in salty habitats are Na<sup>+</sup> and Cl−, but other ions (including Ca2+, Mg2+ , K+, SO<sup>4</sup> <sup>2</sup>−, and CO<sup>3</sup> <sup>2</sup>−) may also be abundant [9,12]. Moreover, both their absolute and relative concentrations may vary, and influence the composition of the associated plant communities [13–15]. fering elemental concentrations, indicating that their physiological selectivity varies [11]. Generally, the dominant ions in salty habitats are Na<sup>+</sup> and Cl<sup>−</sup> , but other ions (including Ca2+, Mg2+, K+, SO<sup>4</sup> 2- , and CO<sup>3</sup> 2- ) may also be abundant [9,12]. Moreover, both their absolute and relative concentrations may vary, and influence the composition of the associated plant communities [13–15].

halophytes growing together on the same saline–alkali soil often have substantially dif-

*Forests* **2022**, *13*, x FOR PEER REVIEW 2 of 14

*Kalidium* (Amaranthaceae) is a genus of succulent halophytes with five species (*Kalidium caspicum* (L.) Ung.-Sternb., *Kalidium cuspidatum* (Ung.-Sternb.) Grubov., *Kalidium foliatum* (Pall.) Moq., *Kalidium gracile* Fenzl, and *Kalidium schrenkianum* Bunge ex Ung.-Sternb.), mainly distributed as shrubs in Northwest Asia and Southeast Europe. Some authorities have also recognized *K. wagenitzii* as an endemic species in Turkey, but others include it in *K. foliatum* [16]. In addition, two varieties of *K. cuspidatum* (var. *cuspidatum* and *sinicum* A. J. Li) have been recognized as separate species [17]. Here these varieties are called species, mainly due to genetic differences in DNA barcodes [17], but partly because our results indicate that they have significant adaptive differences. All six *Kalidium* species (including the two varieties of *K. cuspidatum* as species) naturally grow in deserts in northwest China. These succulent halophytes provide important fodder for livestock in winter, and have high ecological value for soil and water conservation in their semi-arid and arid areas. An extensive field survey showed that their distributions in China significantly differ (Figure 1). *K. foliatum* is the most widespread. Most *K. cuspidatum* sites are in Ningxia, *K. gracile* is mostly located in Gansu, Qinghai and Inner Mongolia. *K. sinicum* is naturally distributed in East Gansu and West Ningxia, while *K. caspicum* and *K. schrenkianum* are mainly restricted to regions north and south of the Tianshan Mountains in Xinjiang, respectively. *Kalidium* (Amaranthaceae) is a genus of succulent halophytes with five species (*Kalidium caspicum* (L.) Ung.-Sternb., *Kalidium cuspidatum* (Ung.-Sternb.) Grubov., *Kalidium foliatum* (Pall.) Moq., *Kalidium gracile* Fenzl, and *Kalidium schrenkianum* Bunge ex Ung.- Sternb.), mainly distributed as shrubs in Northwest Asia and Southeast Europe. Some authorities have also recognized *K. wagenitzii* as an endemic species in Turkey, but others include it in *K. foliatum* [16]. In addition, two varieties of *K. cuspidatum* (var. *cuspidatum* and *sinicum* A. J. Li) have been recognized as separate species [17]. Here these varieties are called species, mainly due to genetic differences in DNA barcodes [17], but partly because our results indicate that they have significant adaptive differences. All six *Kalidium* species (including the two varieties of *K. cuspidatum* as species) naturally grow in deserts in northwest China. These succulent halophytes provide important fodder for livestock in winter, and have high ecological value for soil and water conservation in their semi-arid and arid areas. An extensive field survey showed that their distributions in China significantly differ (Figure 1). *K. foliatum* is the most widespread. Most *K. cuspidatum* sites are in Ningxia, *K. gracile* is mostly located in Gansu, Qinghai and Inner Mongolia. *K. sinicum* is naturally distributed in East Gansu and West Ningxia, while *K. caspicum* and *K. schrenkianum* are mainly restricted to regions north and south of the Tianshan Mountains in Xinjiang, respectively.

**Figure 1.** Map showing soil sampling sites in areas occupied by the six *Kalidium* species. Stars indicate seed sampling sites. **Figure 1.** Map showing soil sampling sites in areas occupied by the six *Kalidium* species. Stars indicate seed sampling sites.

The adaptive evolution and origin of key halotolerance mechanisms have been intensively studied, as reviewed for example by Flowers et al. (2010) and Cheeseman (2013). However, the NaCl concentration is not the sole stressor in saline environments. Thermal and water stresses are also often important [18,19]. Furthermore, salinity may be associated with extreme pH and/or variations in relative proportions of both cations and anions [20]. Effects of these variations have been less extensively studied, so this study focused on their impacts on distributions of the six *Kalidium* species in China. For this purpose, soil samples were collected from sites of the six species, across their ranges in China, then the pH, total salt contents, and ion contents of the soil were assayed at 20 cm depth intervals spanning 0 to 100 cm. In addition, the germinability of seeds of the six species was The adaptive evolution and origin of key halotolerance mechanisms have been intensively studied, as reviewed for example [2,18]. However, the NaCl concentration is not the sole stressor in saline environments. Thermal and water stresses are also often important [19,20]. Furthermore, salinity may be associated with extreme pH and/or variations in relative proportions of both cations and anions [21]. Effects of these variations have been less extensively studied, so this study focused on their impacts on distributions of the six *Kalidium* species in China. For this purpose, soil samples were collected from sites of the six species, across their ranges in China, then the pH, total salt contents, and ion contents of the soil were assayed at 20 cm depth intervals spanning 0 to 100 cm. In addition, the germinability of seeds of the six species was determined under different concentrations of NaCl and Na2SO4. Correlations between these abiotic factors and distributions of the

*Kalidium* species were then examined, to explore mechanisms affecting the relationship between biodiversity and ecosystem functions.

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

#### *2.1. Field Investigation and Sampling*

In a comprehensive field investigation of the areas where *Kalidium* Moq is distributed in northwest China during 2014–2018, we identified 103 representative sites, in total, of the six *Kalidium* species. These sites are mainly located in gravel deserts and/or gravel dunes, according to the FAO soil classification system (FAO 2016). Vegetation at the sites is dominated by *Kalidium* Moq and *Halocnemum strobilaceum* of the Amaranthaceae, *Halogeton arachnoideus* (Amaranthaceae), *Stipa glareosa* (Gramineae) and various other halophytes. Distances between neighboring sites mainly ranged from 150 to 200 km. Soil samples were collected from centers of these areas using a soil auger, with three replications (three-point sampling), at 20 cm depth intervals from 0 to 100 cm. Furthermore, mature seeds were collected from the six *Kalidium* species and stored in a refrigerator at −20 ◦C before the start of experiments. The altitude and geographic coordinates of each site were measured using an Etrex GIS unit (Garmin, Taiwan). Locations of the collection sites are shown in Figure 1.

#### *2.2. Determination of Soil Chemical Properties*

The soil samples were air-dried, passed through a 1 mm sieve, then their pH was measured at a 1:5 soil: water ratio (*w*/*v*) using a PHS-25 pH meter (Shanghai Biocotek Inc., Shanghai, China), and their electrical conductivity (mS/cm) using a DDSJ-318 conductivity meter (Shanghai Biocotek Inc., Shanghai, China). Their contents of eight ions were also analyzed: Na<sup>+</sup> , K<sup>+</sup> , Ca2+, and Mg2+ using a 180-80 Polarized Zeeman atomic absorption spectrophotometer (Hitachi Inc., Tokyo, Japan); CO<sup>3</sup> <sup>2</sup><sup>−</sup> and HCO<sup>3</sup> − by the double indicator titration method; Cl− by silver nitrate titration; and SO<sup>4</sup> <sup>2</sup><sup>−</sup> by the turbidimetric method [22].

#### *2.3. Determination of Seeds' Germinability*

Mean seed mass was calculated by weighing 1000 seeds of each *Kalidium* species with three replications (Table S1). Then germination and recovery experiments were conducted in an LRH 550-G programmed controlled-environment chamber (Shaoguan taihong, China) providing 16 h light/8 h dark cycles with cool white fluorescent lamps 100 µmol m−<sup>2</sup> s −1 (Philips), and 25/19 ◦C day/night temperatures.

Seeds of the six *Kalidium* species were subjected to treatment with NaCl at seven concentrations by placing them on filter paper in Petri dishes (9 cm diameter) moistened with 10 mL of 0, 50, 100, 200, 300, 400, and 500 mM NaCl solution. Other batches were exposed to Na2SO<sup>4</sup> with corresponding Na concentrations (0, 25, 50, 100, 150, 200 and 250 mM). Germination (regarded as emergence of the radicle from the seed by about 1 mm) was scored every day for 7 days. The germination rate at that point was calculated for each species, then ungerminated seeds were transferred to distilled water, and incubated under otherwise identical conditions. Germination of these seeds was scored for a further 7 days, after which the recovery germination percentage was calculated for each species.

#### *2.4. Data Analysis*

The soils' total salt contents were estimated from electrical conductivity measurements, using the following empirically determined linear relationship between NaCl concentration (y) and conductivity (x): y = 0.0159x (R<sup>2</sup> = 0.9811) [23]. Relationships between total salt contents and pH in 20 cm layers of the top 100 cm of soil at sites of the six species were examined by linear regression. The significance of differences in germination rates and recovery germination rates of seeds at different salt concentrations was tested by one-way ANOVA. Values of 18 bioclimatic factors covering most of the distributions of the six species were downloaded from the Global Climate Database (http://www.worldclim.org/bioclim, accessed on 9 November 2021). After excluding redundant bioclimatic factors, by applying a cumulative contribution ratio threshold of 80% [24], eight remained (Table S2). These were

used in combination with altitude and two soil factors (pH and total salt concentration of the soil) in the Principal Component Analysis (PCA) of abiotic factors affecting distributions of each of the six studied *Kalidium* species. For this, SPSS (Version 19.0., Chicago, IL, USA) was used (with the significance threshold set at *p* < 0.05). factors affecting distributions of each of the six studied *Kalidium* species. For this, SPSS (Version 19.0., Chicago, IL, USA) was used (with the significance threshold set at *p* < 0.05). **3. Results**

way ANOVA. Values of 18 bioclimatic factors covering most of the distributions of the six species were downloaded from the Global Climate Database (http://www.worldclim.org/bioclim, accessed on 9 November 2021). After excluding redundant bioclimatic factors, by applying a cumulative contribution ratio threshold of 80% [23], eight remained (Table S2). These were used in combination with altitude and two soil factors (pH and total salt concentration of the soil) in the Principal Component Analysis (PCA) of abiotic

#### **3. Results** *3.1. pH and Total Salt Contents*

#### *3.1. pH and Total Salt Contents* The pH of soils in the areas occupied by the *Kalidium* species mainly ranged from 8.5

*Forests* **2022**, *13*, x FOR PEER REVIEW 4 of 14

The pH of soils in the areas occupied by the *Kalidium* species mainly ranged from 8.5 to 10.5, although the *K. foliatum* sites had a wider range (7.3–11). The mean pH of soil samples from areas occupied by *K. cuspidatum*, *K. gracile*, *K. sinicum*, *K. foliatum*, *K. caspicum,* and *K. schrenkianum* was 9.41, 9.05, 9.13, 9.03, 8.94, and 8.90, respectively (Figure 2). Mean total salt contents in areas occupied by *K. cuspidatum*, *K. gracile*, *K. foliatum*, *K. caspicum* and *K. sinicum* were 22.4, 19.5, 17.6, 16.5 and 14.3 g/kg, respectively (Figure 2), while the highest mean value for any layer in *K. schrenkianum* areas was just 13.0 g/kg (in the 0–20 cm layer) and the overall mean was just 7.08 g/kg (Figures 2 and 3a). Generally, there was a positive correlation between pH values and mean salt contents in the topsoil (0–20 cm), but at >40 cm depths the relationship turned negative in areas occupied by all six species (Figure 3a). However, salt contents of samples from *K. cuspidatum* areas (with the highest mean pH and salt contents) significantly varied with depth while they were much more constant in *K. caspicum* areas (Figure 3). In addition, pH and total salt contents varied much more widely in soil samples from *K. foliatum*, *K. gracile*, and *K. sinicum* areas than in samples from *K. schrenkianum* areas (Figures 2 and 3). to 10.5, although the *K. foliatum* sites had a wider range (7.3–11)*.* The mean pH of soil samples from areas occupied by *K. cuspidatum*, *K. gracile*, *K. sinicum*, *K. foliatum*, *K. caspicum,* and *K. schrenkianum* was 9.41, 9.05, 9.13, 9.03, 8.94, and 8.90, respectively (Figure 2). Mean total salt contents in areas occupied by *K. cuspidatum*, *K. gracile*, *K. foliatum*, *K. caspicum* and *K. sinicum* were 22.4, 19.5, 17.6, 16.5 and 14.3 g/kg, respectively (Figure 2), while the highest mean value for any layer in *K. schrenkianum* areas was just 13.0 g/kg (in the 0– 20 cm layer) and the overall mean was just 7.08 g/kg (Figures 2 and 3a). Generally, there was a positive correlation between pH values and mean salt contents in the topsoil (0–20 cm), but at >40 cm depths the relationship turned negative in areas occupied by all six species (Figure 3a). However, salt contents of samples from *K. cuspidatum* areas (with the highest mean pH and salt contents) significantly varied with depth while they were much more constant in *K. caspicum* areas (Figure 3). In addition, pH and total salt contents varied much more widely in soil samples from *K. foliatum*, *K. gracile*, and *K. sinicum* areas than in samples from *K. schrenkianum* areas (Figures 2 and 3).

**Figure 2.** Scatter plots of total salt contents and pH of soil samples from areas occupied by indicated *Kalidium* species. **Figure 2.** Scatter plots of total salt contents and pH of soil samples from areas occupied by indicated *Kalidium* species.

**Figure 3.** Linear regression plots showing the relationship between total salt contents and pH at indicated depths, and overall, in soil from areas occupied by the six *Kalidium* species. (**a**) total salt contents at different soil depth; (**b**) total salt contents at different species **Figure 3.** Linear regression plots showing the relationship between total salt contents and pH at indicated depths, and overall, in soil from areas occupied by the six *Kalidium* species. (**a**) total salt contents at different soil depth; (**b**) total salt contents at different species.

#### *3.2. Na<sup>+</sup> , Ca2+ , K<sup>+</sup> , and Mg2+ Concentrations 3.2. Na<sup>+</sup> , Ca2+, K<sup>+</sup> , and Mg2+ Concentrations*

Na<sup>+</sup> concentrations in areas occupied by the six species were all high (Figure 4), especially in *K. cuspidatum* areas, where they declined with increases in depth but even at 80– 100 cm exceeded 2.5 g/kg, the upper limit for sensitive crops according to the FAO. In *K. gracile* and *K. schrenkianum* areas, Na<sup>+</sup> concentrations peaked at 20–40 cm depth. Mean Na<sup>+</sup> concentrations in samples from all soil layers in areas occupied by the six species ranged from 0.90 to 4.22 g/kg. As shown in Table 1, Na<sup>+</sup> contents were also significantly positively correlated (*r* > 0.89) with total salt contents in soil from the six species' areas, except for 20–40 cm samples (*r* = 0.632), strongly suggesting that Na<sup>+</sup> made the largest contributions to the total salt contents. Na<sup>+</sup> concentrations in areas occupied by the six species were all high (Figure 4), especially in *K. cuspidatum* areas, where they declined with increases in depth but even at 80–100 cm exceeded 2.5 g/kg, the upper limit for sensitive crops according to the FAO. In *K. gracile* and *K. schrenkianum* areas, Na<sup>+</sup> concentrations peaked at 20–40 cm depth. Mean Na<sup>+</sup> concentrations in samples from all soil layers in areas occupied by the six species ranged from 0.90 to 4.22 g/kg. As shown in Table 1, Na<sup>+</sup> contents were also significantly positively correlated (*r* > 0.89) with total salt contents in soil from the six species' areas, except for 20–40 cm samples (*r* = 0.632), strongly suggesting that Na<sup>+</sup> made the largest contributions to the total salt contents.

**Table 1.** Correlation coefficient (*r*) between total salt contents and concentrations of indicated ions in soil from indicated depths in areas occupied by the six *Kalidium* species. **Table 1.** Correlation coefficient (*r*) between total salt contents and concentrations of indicated ions in soil from indicated depths in areas occupied by the six *Kalidium* species.


\*, *p* < 0.05; \*\*, *p* < 0.01.

Ca2+ was the second most abundant cation in all tested soil samples (Figure 4). In *K. foliatum* and *K. caspicum* areas, mean concentrations increased as soil depth increased, reaching 0.74 and 0.98 g/kg at 80–100 cm, respectively. In contrast, in *K. cuspidatum* and *K. sinicum* areas, mean Ca2+ concentrations declined as depth increased and then increased (Figure 4), to <86 mg/kg at 40–60 cm in *K. cuspidatum* areas (Figure 4). Overall, as shown in Table 1, there was a negative correlation between mean Ca2+ concentrations and total salt contents in the topsoil (*r* = −0.841, *p* < 0.05), but a positive correlation between them at 60–80 cm (*r* = 0.934, *p* < 0.05) and 80–100 cm (*r* = 0.965, *p* < 0.01). Ca2+ was the second most abundant cation in all tested soil samples (Figure 4). In *K. foliatum* and *K. caspicum* areas, mean concentrations increased as soil depth increased, reaching 0.74 and 0.98 g/kg at 80–100 cm, respectively. In contrast, in *K. cuspidatum* and *K. sinicum* areas, mean Ca2+ concentrations declined as depth increased and then increased (Figure 4), to <86 mg/kg at 40–60 cm in *K. cuspidatum* areas (Figure 4). Overall, as shown in Table 1, there was a negative correlation between mean Ca2+ concentrations and total salt contents in the topsoil (*r* = −0.841, *p* < 0.05), but a positive correlation between them at 60–80 cm (*r* = 0.934, *p* < 0.05) and 80–100 cm (*r* = 0.965, *p* < 0.01).

K<sup>+</sup> concentrations did not exceed 260 mg/kg in any tested soil samples. In *K. gracile*  and *K. sinicum* areas, as depth increased they first declined and then increased, and were almost twice as high in the former as the latter at each soil depth (0–60 cm) (Figure 4). K<sup>+</sup> concentrations significantly declined with depth (from 216, and 257 mg/kg, respectively, in topsoil) at all sites occupied by *K. schrenkianum* and *K. cuspidatum* (Figure 4). K + concentrations did not exceed 260 mg/kg in any tested soil samples. In *K. gracile* and *K. sinicum* areas, as depth increased they first declined and then increased, and were almost twice as high in the former as the latter at each soil depth (0–60 cm) (Figure 4). K<sup>+</sup> concentrations significantly declined with depth (from 216, and 257 mg/kg, respectively, in topsoil) at all sites occupied by *K. schrenkianum* and *K. cuspidatum* (Figure 4).

Mg2+ concentrations were also substantially lower than Na+ and Ca2+ concentrations (consistently < 300 mg/kg). In areas occupied by *K. caspicum*, *K. gracile*, and *K. sinicum* they first increased and then declined with increases in soil depth (Figure 4), peaking at 268 mg/kg at 40–60 cm in samples from *K. caspicum* areas (Figure 4). Overall, as shown in Table Mg2+ concentrations were also substantially lower than Na<sup>+</sup> and Ca2+ concentrations (consistently < 300 mg/kg). In areas occupied by *K. caspicum*, *K. gracile*, and *K. sinicum* they first increased and then declined with increases in soil depth (Figure 4), peaking at 268 mg/kg at 40–60 cm in samples from *K. caspicum* areas (Figure 4). Overall, as shown in

Table 1, at sites of all six species, Mg2+ concentrations were positively correlated with total salt contents at 40–60 cm (*r* = 0.980, *p* < 0.01) and 80–100 cm (*r* = 0.991, *p* < 0.01) depths. 1, at sites of all six species, Mg2+ concentrations were positively correlated with total salt contents at 40–60 cm (*r* = 0.980, *p* < 0.01) and 80–100 cm (*r* = 0.991, *p* < 0.01) depths.

**Figure 4.** Concentrations of Na<sup>+</sup> , Ca2+ , K<sup>+</sup> , and Mg2+ in soil from indicated depths in areas occupied by the six *Kalidium* species. The values shown are means with SE (*n* = 3). **Figure 4.** Concentrations of Na<sup>+</sup> , Ca2+, K<sup>+</sup> , and Mg2+ in soil from indicated depths in areas occupied by the six *Kalidium* species. The values shown are means with SE (*n* = 3).

*3.3. Cl<sup>−</sup>*

5, Table 1).

*, SO<sup>4</sup>*

*<sup>2</sup><sup>−</sup> and HCO<sup>3</sup>*

#### *3.3. Cl*−*, SO<sup>4</sup> <sup>2</sup>*<sup>−</sup> *and HCO<sup>3</sup>* − *Concentrations* Mean Cl<sup>−</sup> concentrations were highest in *K. cuspidatum* areas, where they declined

*<sup>−</sup> Concentrations*

*Forests* **2022**, *13*, x FOR PEER REVIEW 7 of 14

Mean Cl− concentrations were highest in *K. cuspidatum* areas, where they declined from 10.62 g/kg in topsoil but still exceeded 2.5 g/kg at 80–100 cm depth, and lowest in *K. caspicum* areas, where the mean topsoil concentration was only 1.53 g/kg (Figure 5). It was also a major anion in *K. sinicum*, *K. foliatum* and *K. gracile* areas, where mean concentrations were 2.6, 2.7 and 3.2 g/kg, respectively (Figure 5). Overall, as shown in Table 1, Cl− concentrations were positively correlated with total salt contents at 0–20 cm (*r* = 0.856, *p* < 0.05), 40–60 cm (*r* = 0.880, *p* < 0.05) and 80–100 cm (*r* = 0.933, *p* < 0.05) depths. from 10.62 g/kg in topsoil but still exceeded 2.5 g/kg at 80–100 cm depth, and lowest in *K. caspicum* areas, where the mean topsoil concentration was only 1.53 g/kg (Figure 5). It was also a major anion in *K. sinicum*, *K. foliatum* and *K. gracile* areas, where mean concentrations were 2.6, 2.7 and 3.2 g/kg, respectively (Figure 5). Overall, as shown in Table 1, Cl<sup>−</sup> concentrations were positively correlated with total salt contents at 0–20 cm (*r* = 0.856, *p* < 0.05), 40–60 cm (*r* = 0.880, *p* < 0.05) and 80–100 cm (*r* = 0.933, *p* < 0.05) depths.

**Figure 5.** Concentrations of Cl<sup>−</sup> , SO<sup>4</sup> 2− , and HCO<sup>3</sup> − in soil from indicated depths in areas occupied by the six *Kalidium* species. The values shown are means with SE (*n* = 3). **Figure 5.** Concentrations of Cl−, SO<sup>4</sup> <sup>2</sup>−, and HCO<sup>3</sup> − in soil from indicated depths in areas occupied by the six *Kalidium* species. The values shown are means with SE (*n* = 3).

In *K. caspicum* areas, SO<sup>4</sup> <sup>2</sup><sup>−</sup> was the main anion, and its mean concentration increased as soil depth increased, reaching 11.78 g/kg at 80–100 cm, while concentrations were much lower in *K. cuspidatum* areas (just 1.52 g/kg at 60–80 cm depth) (Figure 5). In areas occupied by the other four species—*K. gracile*, *K. foliatum*, *K. sinicum*, and *K. schrenkianum*—the mean concentration first declined (from 3.7, 5.4, 3.5 and 3.43 g/kg, respectively, in topsoil) and In *K. caspicum* areas, SO<sup>4</sup> <sup>2</sup><sup>−</sup> was the main anion, and its mean concentration increased as soil depth increased, reaching 11.78 g/kg at 80–100 cm, while concentrations were much lower in *K. cuspidatum* areas (just 1.52 g/kg at 60–80 cm depth) (Figure 5). In areas occupied by the other four species—*K. gracile*, *K. foliatum*, *K. sinicum*, and *K. schrenkianum*—the mean concentration first declined (from 3.7, 5.4, 3.5 and 3.43 g/kg, respectively, in topsoil) and

then increased as depth increased. Overall, it was only correlated with total salt contents at 60–80 cm depth (r = 0.966, *p* < 0.01) in areas occupied by the six *Kalidium* species (Figure then increased as depth increased. Overall, it was only correlated with total salt contents at 60–80 cm depth (r = 0.966, *p* < 0.01) in areas occupied by the six *Kalidium* species (Figure 5, Table 1). HCO<sup>3</sup> <sup>−</sup> concentrations in soil from areas occupied by all six species were very low. Its

*Forests* **2022**, *13*, x FOR PEER REVIEW 8 of 14

HCO<sup>3</sup> − concentrations in soil from areas occupied by all six species were very low. Its mean concentrations first increased then declined as depth increased in *K. schrenkianum*, *K. cuspidatum* and *K. sinicum* areas (Figure 5), and were highest in the 40–60 cm layer of soil in *K. sinicum* areas, at just 14 mg/kg. Moreover, HCO<sup>3</sup> − concentrations were negatively correlated with total salt contents at depths below 20 cm in areas occupied by all species (Figure 5, Table 1). CO<sup>3</sup> <sup>2</sup><sup>−</sup> was undetectable with the applied equipment in most samples. mean concentrations first increased then declined as depth increased in *K. schrenkianum*, *K. cuspidatum* and *K. sinicum* areas (Figure 5), and were highest in the 40–60 cm layer of soil in *K. sinicum* areas, at just 14 mg/kg. Moreover, HCO<sup>3</sup> <sup>−</sup> concentrations were negatively correlated with total salt contents at depths below 20 cm in areas occupied by all species (Figure 5, Table 1). CO<sup>3</sup> <sup>2</sup><sup>−</sup> was undetectable with the applied equipment in most samples.

#### *3.4. Germination Rates 3.4. Germination Rates*  Germination rates of seeds of the six *Kalidium* species were negatively correlated with

Germination rates of seeds of the six *Kalidium* species were negatively correlated with concentrations of both NaCl and Na2SO<sup>4</sup> in the media (*p* < 0.05, Figures 6 and 7). In distilled water their germination rates ranged from 90.8% for *K. sinicum* to 100% for *K. caspicum*. However, less than half of all species' seeds germinated when the concentration exceeded 200 mM NaCl, and no *K. sinicum* seeds germinated at higher concentrations (400 or 500 mM NaCl) (Figure 6). Germination rates also declined with increases in Na2SO<sup>4</sup> concentrations in the medium, and at 150 mM exceeded 50% (56.3%) for seeds of only one species (*K. caspicum*) (Figure 7). concentrations of both NaCl and Na2SO<sup>4</sup> in the media (*p* < 0.05, Figures 6 and 7). In distilled water their germination rates ranged from 90.8% for *K. sinicum* to 100% for *K. caspicum*. However, less than half of all species' seeds germinated when the concentration exceeded 200 mM NaCl, and no *K. sinicum* seeds germinated at higher concentrations (400 or 500 mM NaCl) (Figure 6). Germination rates also declined with increases in Na2SO<sup>4</sup> concentrations in the medium, and at 150 mM exceeded 50% (56.3%) for seeds of only one species (*K. caspicum*) (Figure 7).

**Figure 6.** Total germination percentages after 14 days at the indicated NaCl concentrations in the six studied *Kalidium* species. Germination rates (%, means and standard deviations) of the six *Kalidium* species after treatments with indicated concentrations of NaCl (0, 50, 100, 200, 300, 400, and 500 mM NaCl) after 7 days. The dark bars indicate germination rates after the treatments and light bars the total percentages that germinated during the recovery treatment in distilled water with additional 7 days. Asterisks indicate significant differences in each condition respect to the corresponding control (according to Dunnet test, *p* < 0.05). **Figure 6.** Total germination percentages after 14 days at the indicated NaCl concentrations in the six studied *Kalidium* species. Germination rates (%, means and standard deviations) of the six *Kalidium* species after treatments with indicated concentrations of NaCl (0, 50, 100, 200, 300, 400, and 500 mMNaCl) after 7 days. The dark bars indicate germination rates after the treatments and light bars the total percentages that germinated during the recovery treatment in distilled water with additional 7 days. Asterisks indicate significant differences in each condition respect to the corresponding control (according to Dunnet test, *p* < 0.05).

The recovery germination rates of seeds of all six *Kalidium* species after treatment with both salts were high. The recovery germination rates of *K. gracile* seeds following the NaCl treatment were negatively correlated with the NaCl concentration during the treatment, declining from 92.2 to 74.2% following exposure to 500 and 200 mM NaCl, respectively (Figure 6). However, the recovery germination rate of *K. sinicum* seeds exposed to 500 mM NaCl was high (>80%). Moreover, there were no significant differences in

recovery germination rates of *K. cuspidatum* seeds exposed to different NaCl concentrations (*p* >0.05) and those of the other three species were all above 78% at 500 mM NaCl (Figure 6). Following treatment with Na2SO4, the recovery germination percentages of *K. sinicum* and *K. schrenkianum* seeds were lower than those of the other four species at the same concentrations, but were still high (76.6 and 83.2%, respectively), following exposure to the highest Na2SO<sup>4</sup> concentration, 250 mM (Figure 7). *Forests* **2022**, *13*, x FOR PEER REVIEW 9 of 14

> All six species produce small seeds (<0.5 g/1000 seeds, Table S1), and there was no linear relationship between the mean mass and germination rate of their seeds.

**Figure 7.** Total germination percentages after 14 days at the indicated Na2SO<sup>4</sup> concentrations in the six studied *Kalidium* species. Germination rates (%, means and standard deviations) of the six *Kalidium* species after treatments with indicated concentrations of Na2SO<sup>4</sup> (0, 25, 50, 100, 150, 200, and 250 mM) after 7 days. The dark bars indicate germination rates after the treatments and light bars the total percentages that germinated during the recovery treatment in distilled water with additional 7 days. Asterisks indicate significant differences in each condition respect to the corresponding control (according to Dunnet test, *p* < 0.05). **Figure 7.** Total germination percentages after 14 days at the indicated Na2SO<sup>4</sup> concentrations in the six studied *Kalidium* species. Germination rates (%, means and standard deviations) of the six *Kalidium* species after treatments with indicated concentrations of Na2SO<sup>4</sup> (0, 25, 50, 100, 150, 200, and 250 mM) after 7 days. The dark bars indicate germination rates after the treatments and light bars the total percentages that germinated during the recovery treatment in distilled water with additional 7 days. Asterisks indicate significant differences in each condition respect to the corresponding control (according to Dunnet test, *p* < 0.05).

#### The recovery germination rates of seeds of all six *Kalidium* species after treatment *3.5. Principal Component Analyses*

with both salts were high. The recovery germination rates of *K. gracile* seeds following the NaCl treatment were negatively correlated with the NaCl concentration during the treatment, declining from 92.2 to 74.2% following exposure to 500 and 200 mM NaCl, respectively (Figure 6). However, the recovery germination rate of *K. sinicum* seeds exposed to 500 mM NaCl was high (>80%). Moreover, there were no significant differences in recovery germination rates of *K. cuspidatum* seeds exposed to different NaCl concentrations (*p* >0.05) and those of the other three species were all above 78% at 500 mM NaCl (Figure 6). Following treatment with Na2SO4, the recovery germination percentages of *K. sinicum* and *K. schrenkianum* seeds were lower than those of the other four species at the same concentrations, but were still high (76.6 and 83.2%, respectively), following exposure to the highest Na2SO4 concentration, 250 mM (Figure 7). All six species produce small seeds (<0.5 g/1000 seeds, Table S1), and there was no linear relationship between the mean mass and germination rate of their seeds. *3.5. Principal Component Analyses* Two soil factors (pH and total salt contents), altitude, four temperature factors and two precipitation factors were used in PCA to explore relationships between abiotic factors and spatial distributions of each of the six *Kalidium* species (Figure 8 and Table S2). Principal Component (PC1) explained 40.4–50.5% of the variance and was most strongly influenced by temperature factors (maximum temperature of the warmest month, annual mean temperature, and mean temperature of the warmest quarter; Tables 2 and S2). PC2 explained 20.8–28.4% of the variance and largely reflected effects of precipitation factors on four of the six species. In addition, the two soil factors (total salt contents and pH) strongly contributed to PC3, explaining 9.96–18.4% of the variation in distribution of the six species. In total, PCs 1–3 explained >85% of the variation in the six species' distributions (Figure 8, Tables 2 and S2). The results strongly indicate that the most important ecological variables for adaptation to the species' saline environments are temperature during the driest month and precipitation. They also indicate that the selected ecological variables are strongly associated with the six *Kalidium* species' spatial distributions through their effects on the soil environment (Figure 1).

Two soil factors (pH and total salt contents), altitude, four temperature factors and two precipitation factors were used in PCA to explore relationships between abiotic fac-

Principal Component (PC1) explained 40.4–50.5% of the variance and was most strongly influenced by temperature factors (maximum temperature of the warmest month, annual mean temperature, and mean temperature of the warmest quarter; Tables 2 and S2). PC2 explained 20.8–28.4% of the variance and largely reflected effects of precipitation factors on four of the six species. In addition, the two soil factors (total salt contents and pH)

strongly contributed to PC3, explaining 9.96–18.4% of the variation in distribution of the six species. In total, PCs 1–3 explained >85% of the variation in the six species' distributions (Figure 8, Tables 2 and S2). The results strongly indicate that the most important ecological variables for adaptation to the species' saline environments are temperature during the driest month and precipitation. They also indicate that the selected ecological

through their effects on the soil environment (Figure 1).

**Figure 8.** Score plots obtained from Principal Component Analysis (PCA) of effects of abiotic factors on distributions of the six *Kalidium* species. *x* axis, *y* axis, and *z* axis indicate the first, the second and the third Principle Component, respectively. **Figure 8.** Score plots obtained from Principal Component Analysis (PCA) of effects of abiotic factors on distributions of the six *Kalidium* species. *x* axis, *y* axis, and *z* axis indicate the first, the second and the third Principle Component, respectively.


**Table 2.** Loadings of the main factors influencing the first three Principle Components obtained from Principle Component Analysis of the relationships between abiotic factors and distributions of the six *Kalidium* species.

bio1, annual mean temperature; bio5, max temperature of the warmest month; bio9, mean temperature of the driest quarter; bio12, annual average precipitation; bio14, precipitation of the driest month; TS, total salt contents.

#### **4. Discussion**

The mean pH was high (8.9–9.4) in soil samples from areas occupied by all six of the *Kalidium* species. Mean total salt contents were also high, but covered a substantial range (22.4, 19.5, 17.6, 16.5, 14.3, and 7.08 g/kg in soils from *K. cuspidatum*, *K. gracile*, *K. foliatum*, *K. caspicum*, *K. sinicum*, and *K. schrenkianum* areas, respectively). The indication that *K. schrenkianum* has relatively low salt tolerance, according to the low mean salt content of samples from areas it occupies, may at least partly explain why this species is restricted to a narrow range to the south of the Tianshan Mountains [18]. In areas occupied by all six species, total salt concentrations were positively related to pH at 0–20 cm and 20–40 cm soil depths (*r* = 0.794 and *r* = 0.248, respectively), indicating that salt contents are particularly strongly correlated to pH in topsoil in the study region.

In areas occupied by all the species, Na<sup>+</sup> was the main cation, and its concentrations were significantly correlated with total salt contents at all soil depths except 20–40 cm (*r* = 0.632), showing that they either require high Na<sup>+</sup> concentrations for optimal growth and development or at least tolerate them [6]. Ca2+ was the next most abundant cation at 0–40 cm soil depths, indicating that high concentrations of Ca2+ and Na<sup>+</sup> likely accumulate in all the *Kalidium* species, as previously found in roots and leaves of *K. foliatum*, *K. cuspidatum,* and various other halophytes [25,26]. As a kind of antagonist, absorption of large amounts of Ca2+ by roots could potentially alleviate damage to plants by other ions [27]. Concentrations of the major nutrient K<sup>+</sup> , which is required by all living cells and often deficient in barren soil [28,29], ranged from 28.2 to 256.9 mg/kg in soil from *K. sinicum* and *K. cuspidatum* areas, respectively. Generally, in soils from areas of all six species the K + concentration was much lower than the Na<sup>+</sup> concentration and (hence) the Na+/K<sup>+</sup> ratio was high (>30:1). In addition, there was no significant correlation between the K<sup>+</sup> concentration and total salt contents, indicating that the species' requirements for Na<sup>+</sup> and K + ions significantly differ. Mean Mg2+ concentrations were highest and lowest in soil from *K. gracile* and *K. foliatum* areas (ca. 217 mg/kg and 2.3-fold lower, respectively), and substantially lower than concentrations of the other measured cations in all surveyed areas.

The most abundant anion was Cl− in *K. cuspidatum* areas (where concentrations of both Na<sup>+</sup> and Cl<sup>−</sup> were highest) and SO<sup>4</sup> <sup>2</sup><sup>−</sup> in *K. caspicum* areas (where Cl<sup>−</sup> concentrations were lowest). Thus, anions in soils in these areas were strongly dominated by Cl− and SO<sup>4</sup> <sup>2</sup>−, respectively (mainly balanced in both cases by Na<sup>+</sup> ). These were also the major ions in habitats of the other species, but there was less dominance by Cl<sup>−</sup> or SO<sup>4</sup> <sup>2</sup>−, e.g., mean SO<sup>4</sup> <sup>2</sup>−, Cl<sup>−</sup> and Na<sup>+</sup> concentrations in *K. foliatum* areas were 5.5, 2.6, and 2.8 g/kg, respectively. There were also wide variations in pH and total salt contents in soil samples from *K. foliatum* areas, indicating that the species has strong adaptive ability and, thus, can thrive in relatively diverse habits. Concentrations of CO<sup>3</sup> <sup>2</sup><sup>−</sup> were nondetectable and HCO<sup>3</sup> − concentrations were very low (with no significant differences) in areas of the *Kalidium* species.

Plants must be sufficiently adapted to the salinity of their environments to germinate [19,30] and establish [31–34]. In our assays, the germination rates of seeds of the six *Kalidium* species were all negatively correlated with NaCl and Na2SO<sup>4</sup> contents of the medium. Soil salinity fluctuates with precipitation, and can be alleviated in periods with high precipitation, so high proportions of seeds of many halophytes stored in highly saline soil may germinate during such periods [30,35,36]. Moreover, their recovery germination parameters may be major determinants of their distributions. The recovery germination rates of all six *Kalidium* species exceeded 74% after the NaCl and Na2SO<sup>4</sup> treatments, corroborating the conclusions that the quality of topsoil is the first selective barrier affecting plants' distributions [14,35]. In addition, PCA showed that maximum temperature, summer rainfall and total salt contents of the soil strongly affect geographic distributions of the six *Kalidium* species. Similarly, distributions and yields of wild barley are clearly related to climatic factors, especially precipitation [37], and distributions of *Arabidopsis halleri* and *A. lyrate* are apparently linked to differences in their tolerance of the heavy metals Zn and Cd [38]. Clearly, therefore, ecological factors (and genetically-based adaptations to them) are key determinants of plants' distributions [18,37,38].

#### **5. Conclusions**

Adaptation to topsoil salinity in early stages is a major determinant of the six *Kalidium* species' geographic distributions in the study region. The dominant cation at all sites is Na<sup>+</sup> , but the dominant anions at *K. caspicum* and *K. cuspidatum* sites are SO42<sup>−</sup> and Cl−, respectively. Both salinity and their distributions are affected by numerous interacting factors. Inter alia, temperatures during the driest month and precipitation directly and/or indirectly affect soils' salt contents and pH, which also strongly influence the six *Kalidium* species' distributions. Clearly, high tolerance of salinity stress is a key adaptive trait of halophytes, which has multiple evolutionary origins. Moreover, major changes have occurred in plants' distributions and population sizes during desertification, following which halophytes may occupy extensive semi-arid and arid regions.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/f13122178/s1:, Table S1: Masses of 1000 seeds of the six *Kalidium* species; Table S2: Loading matrix of the Principle Components (PCs).

**Author Contributions:** Y.W. and Z.C. conceived and managed the project; Y.W., D.L. and X.L. performed the experiments; Y.W., Z.C., D.L. and X.L. analyzed the data; Z.C. and D.L. performed statistical analyses; Y.W. and D.L. wrote the original manuscript; and Y.W. and Z.C. reviewed and edited the final manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the Central Government Guides Local Scientific and Technological Development Programs of Gansu Province (Grant Number 22ZY2QG001), National Key Research and Development Program of China (Grant Number 2022YFF1303301), the National Natural Science Foundation of China (NSFC, Grant Number 41871092), and Science and Technology Project of Forestry and Grassland Bureau of Gansu Province (Grant Number 2022kj063).

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** We are grateful to Kuibing Meng and Fengzhu Zhang for collecting samples in the field.

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

## **References**

