*Article* **Scale Effects on the Relationship between Plant Diversity and Ecosystem Multifunctionality in Arid Desert Areas**

**Jiaxin Liu <sup>1</sup> , Dong Hu <sup>2</sup> , Hengfang Wang <sup>1</sup> , Lamei Jiang <sup>1</sup> and Guanghui Lv 1,\***

<sup>2</sup> College of Life Science, Northwestern University, Xi'an 710069, China

**\*** Correspondence: guanghui\_xju@sina.com; Tel.: +0991-2111427

**Abstract:** Understanding the relationship between biodiversity and ecosystem multifunctionality is popular topic in ecological research. Although scale is an important factor driving changes in biodiversity and ecosystem multifunctionality, we still know little about the scale effects of the relationship between the different dimensions of biodiversity and ecosystem multifunctionality. Using plant communities in the northwest of the Qira Desert Ecosystem National Field Research Station of the Chinese Academy of Sciences in Qira County, Xinjiang, as the study object, we explored the scale effects of plant diversity and ecosystem multifunctionality at different sampling scales (5 m × 5 m, 20 m × 20 m, and 50 m × 50 m) and the relative contribution of different dimensions of diversity (species diversity, functional diversity, and phylogenetic diversity) to variation in ecosystem multifunctionality. At different scales, a significant scale effect was observed in the relationship between plant diversity and ecosystem multifunctionality. Species diversity dominated ecosystem multifunctionality at large scales (50 m × 50 m), and species diversity and ecosystem multifunctionality varied linearly between scales. Functional diversity made the greatest contribution in small scales (5 m × 5 m), and the relationship between phylogenetic diversity and ecosystem multifunctionality tended to show a single-peaked variation between scales, with a dominant effect on multifunctionality at the mesoscale (20 m × 20 m). The results of the study deepen the understanding of the scale effect of the relationship between plant diversity and ecosystem multifunctionality in arid desert areas, and help to further conserve plant diversity and maintain ecosystem multifunctionality.

**Keywords:** desert ecosystems; scale; plant diversity; ecosystem multifunctionality

#### **1. Introduction**

Global climate change and habitat fragmentation play a significant negative role in biodiversity conservation and the sustainability of ecosystem functions. [1,2]. The study of the relationship between biodiversity and ecosystem function is important to enhance biodiversity and restore ecosystem function. Biodiversity includes species diversity, functional diversity, and phylogenetic diversity [3]. Early studies on the relationship between biodiversity and ecosystem function mostly considered the relationship between species diversity and single ecosystem function [4,5]. With the progress in research, scholars have found that functional diversity and phylogenetic diversity have significant effects on ecosystem function and cannot be replaced by species diversity, and considering only single ecosystem functions may underestimate the role of biodiversity in ecosystem function [6,7]. Therefore, the study of the relationship between multidimensional biodiversity and ecosystem multifunctionality contributes to a deeper understanding of the biodiversity maintenance mechanisms.

Studies on biodiversity in China and elsewhere tended to focus on different environmental gradients, disturbance levels and successional stages [8,9], with less attention paid to the scale dependence of biodiversity. However, biodiversity depends on the

**Citation:** Liu, J.; Hu, D.; Wang, H.; Jiang, L.; Lv, G. Scale Effects on the Relationship between Plant Diversity and Ecosystem Multifunctionality in Arid Desert Areas. *Forests* **2022**, *13*, 1505. https://doi.org/10.3390/ f13091505

Academic Editor: Takuo Nagaike

Received: 27 August 2022 Accepted: 14 September 2022 Published: 16 September 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/).

<sup>1</sup> College of the Ecology and Environment, Xinjiang University, Urumqi 830017, China

number, composition and distribution of community species, and these factors are scaledependent. [10,11]. The earliest studies on the relationship between plant diversity and ecosystem multifunctionality at different scales showed that alpha diversity was significantly positively correlated with ecosystem multifunctionality, with alpha diversity playing a dominant role [12]. Studies in subtropical regions further demonstrated that the positive correlation also showed a trend of rapid increase followed by a gentle increase [13,14], but the effect of α-diversity on single ecosystem function is not significant in forest ecosystem studies [15]. Functional and phylogenetic diversity is also being studied in greater depth by researchers. A study by Dang et al. (2018) [16] in desert ecosystems has shown that functional diversity indices vary significantly between scales and shape different community structures. Changes in sampling scale have a significant effect on the divergence and aggregation of genealogical structure and the level of genealogical diversity [17]. Studies on cave plants have also shown significant differences in genealogical diversity between large and small scales [18]. Ecosystem multifunctionality and biodiversity interactions are also scale-dependent [19]. First, different species play different roles between scales, allowing inter-scale differences in ecosystem multifunctionality [20,21]. Second, as the scale increases, community differences lead to a constant exchange of materials and energy flows between communities, which affects ecosystem multifunctionality [22]. Finally, the composition and distribution of functional traits among species vary by scale. As scale changes, functional traits segregating or overlapping in trait space as scale changes, making ecosystem multifunctionality change in response [19].

Arid desert ecosystems are sensitive areas of global change and priority areas for biodiversity conservation. As an important part of terrestrial ecosystems, arid zones have distinctive climatic environments, geographical locations, and resource distribution patterns that make them unique in terms of biodiversity and ecosystem multifunctionality [23].

Located at the southern edge of the Taklamakan Desert, Xinjiang's Qira County has a dry climate with little rain and wind, and its ecosystem type is a typical temperate desert ecosystem. Due to its geographical location and topographical constraints, the region is ecologically fragile, and desertification is severe, which has led to a reduction in biodiversity and diminished ecosystem function services [24]. Considering the scale dependence of biodiversity and ecosystem multifunctionality, this study explored the relationship between multidimensional biodiversity and ecosystem multifunctionality based on three sampling scales (5 m × 5 m, 20 m × 20 m, and 50 m × 50 m), aiming to address the following scientific questions: (1) How do plant diversity and ecosystem multifunctionality relate to scale? (2) How do the relative contributions of species diversity, functional diversity, and phylogenetic diversity to ecosystem multifunctionality vary at different scales?

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

#### *2.1. Overview of Experimental Area*

The study area is located on the northern foot of the Kunlun Mountains and on the periphery of the oasis at the southern edge of the Tarim Basin (80◦3701200 E, 37◦2 00 00 N). The climate in the reserve is extremely arid, and water resources are scarce, with an average annual precipitation of 35.1 mm and a potential annual evaporation of 2595.3 mm [25]. The main types of soil are gray-brown desert soil, gray desert soil, and wind-sand soil, with a high degree of soil salinity [26]. The natural vegetation is dominated by perennial desert plants, with the main species including *Populus euphratica*, *Tamarix chinensis*, *Alhagi sparsifolia*, *Salsola collina*, and *Hexinia polydichotoma.*

#### *2.2. Research Method*

#### 2.2.1. Sample Setting

A 100 m × 100 m sample plot was set up in the northwest of the Qira Desert Ecosystem National Field Research Station of the Chinese Academy of Sciences in July 2019, where quadrats of three scales (50 m × 50 m, 20 m × 20 m, and 5 m × 5 m) were set up. Using the 5 m × 5 m quadrat as the basic unit, 100 quadrats were randomly selected at each scale

within the sample plots. Where each 20 m × 20 m quadrat contains 16 quadrats of 5 m × 5 m and each 50 m × 50 m quadrat contains 100 quadrats of 5 m × 5 m (Figure 1). Qira. m and each 50 m × 50 m quadrat contains 100 quadrats of 5 m × 5 m (Figure 1). Qira.

quadrats of three scales (50 m × 50 m, 20 m × 20 m, and 5 m × 5 m) were set up. Using the 5 m × 5 m quadrat as the basic unit, 100 quadrats were randomly selected at each scale within the sample plots. Where each 20 m × 20 m quadrat contains 16 quadrats of 5 m × 5

**Figure 1.** Location of the study area and the investigated plots. **Figure 1.** Location of the study area and the investigated plots.

*Forests* **2022**, *13*, x FOR PEER REVIEW 3 of 14

#### 2.2.2. Collection of Plant Samples

2.2.2. Collection of Plant Samples Plant species information, abundance, and plant height were recorded in the field at a minimum sampling scale of 5 m × 5 m. Three 1 m2 samples were selected on the diagonal of each 5 m × 5 m square to record and calculate the herbaceous abundance within the 5 m × 5 m square. In a 5 m × 5 m sample, approximately 30 mature leaves were collected from each plant species, and three were selected to measure leaf length (LL), leaf width (LW), leaf thickness (LT), and fresh leaf weight. All leaves were taken back to the laboratory to be dried, ground, and used for the measurement of leaf dry matter content Plant species information, abundance, and plant height were recorded in the field at a minimum sampling scale of 5 m <sup>×</sup> 5 m. Three 1 m<sup>2</sup> samples were selected on the diagonal of each 5 m × 5 m square to record and calculate the herbaceous abundance within the 5 m × 5 m square. In a 5 m × 5 m sample, approximately 30 mature leaves were collected from each plant species, and three were selected to measure leaf length (LL), leaf width (LW), leaf thickness (LT), and fresh leaf weight. All leaves were taken back to the laboratory to be dried, ground, and used for the measurement of leaf dry matter content (LDMC), leaf carbon content (LC), leaf nitrogen content (LN), and leaf phosphorus content (LP) indexes.

#### (LP) indexes. 2.2.3. Collection of Soil Samples

2.2.3. Collection of Soil Samples The soil was sampled in 5 m × 5 m units, with the diagonal method of taking a 0–20 cm surface layer of soil at the center, using an aluminum box to store the soil and calculate the soil water content, and then taking a sample in a sealing bag for the determination of The soil was sampled in 5 m × 5 m units, with the diagonal method of taking a 0–20 cm surface layer of soil at the center, using an aluminum box to store the soil and calculate the soil water content, and then taking a sample in a sealing bag for the determination of other soil physical and chemical properties. The method of determination [27] is shown in Table 1.

(LDMC), leaf carbon content (LC), leaf nitrogen content (LN), and leaf phosphorus content

other soil physical and chemical properties. The method of determination [27] is shown in Table 1. **Table 1.** Soil index and determination method.


#### Total nitrogen Kjeldahl method Nitrate nitrogen UV spectrophotometry *2.3. Data Calculation and Analysis*

Ammonium nitrogen UV spectrophotometry 2.3.1. Calculation of Plant Diversity

In this study, we selected species diversity indices, namely Shannon–Wiener diversity index, Simpson diversity index, Margalef richness index, and Pielou evenness index ([28]; functional diversity indices, including FRic richness index, FEve evenness index, FDiv divergence index, and RaoQ quadratic entropy index [29,30]; and phylogenetic diversity indices, including the mean interspecific distance index (MPD), mean nearest interspecific distance (MNTD), and Faith diversity index (PD) (for calculation methods, see Supplementary Table S1) [31,32].

#### 2.3.2. Calculation of Ecosystem Multifunctionality

In this study, soil environmental factors (total nitrogen, total phosphorus, ammonium nitrogen, nitrate nitrogen, fast-acting phosphorus, and organic matter) were used as indicators of ecosystem multifunctionality, and the "Z-score" mean method was used to calculate ecosystem multifunctionality, represented by the formula [33,34]:

$$MF\_a = \sum\_{i}^{F} \frac{\mathcal{g}\left(r\_{i\mid}f\_{i\mid}\right)}{F}$$

.

In the above equation, *MF<sup>a</sup>* represents ecosystem multifunctionality, *f<sup>i</sup>* represents the measured value of function *i*, *r<sup>i</sup>* is the mathematical function that converts *f<sup>i</sup>* into a positive value, *g* represents the normalization of all measured values, and *F* represents the number of functions measured.

#### 2.3.3. Data Analysis

Excel 2019 was used for the initial processing and calculation of the data. Differences in plant diversity and ecosystem multifunctionality between the three scales (5 m × 5 m, 20 m × 20 m, 50 m × 50 m) were analyzed in SPSS 26.0 using one-way analysis of variance (ANOVA). When the variance was equal, the least significant difference (LSD) method was used for the results of multiple comparisons; when the variance was not uniform, the results of multiple comparisons were tested using a non-parametric test. The Kolmogorov– Smirnov test (K–S test) were used to test the normality of ecosystem multifunctionality. Random sampling of the sample plot is done in R4.1.3. A community phylogenetic tree was created in R4.1.3 using the "V.PhyloMaker" package [35]. The species diversity index, functional diversity index, and phylogenetic diversity index were calculated using the "vegan", "FD", and "picante" packages, respectively.

The model was selected in R4.1.3 using the function "dredge" from the "MuMin" package [36], based on the corrected Akaike's information criterion (AICc; ∆AICc < 2) [37]. A selection procedure was used to select the best predictor of ecosystem multifunctionality, and when multiple models were selected, model averaging was performed based on AICc weights. The model residuals were inspected for constant variance and normality. All predictors and response variables were standardized before the model was constructed. Predictors were log-transformed as necessary before analysis to meet the assumptions. The model calculated relative explanatory rates for each diversity index and compared them with the total explanatory rates for all diversity indicators in the model, after which the explanatory rates for the indices in the model were categorized and summed by species diversity, functional diversity, and phylogenetic diversity to obtain the relative importance of different diversity dimensions (species diversity, functional diversity, and phylogenetic diversity) as drivers of ecosystem multifunctionality.

#### **3. Results**

#### *3.1. Characteristics of Plant Diversity*

Among the species diversity indices, the Shannon diversity index, Simpson diversity index, and Margalef richness index tended to increase with scale and showed significant differences between scales (*p* < 0.05), and the Pielou evenness at large scales (50 m × 50 m) was significantly smaller than at small (5 m × 5 m) and medium scales (20 m × 20 m). For the functional diversity index, the RaoQ index showed an increasing trend from small to large scales. The FRic richness index, FEve evenness index, and FDiv divergence index were significantly different between small and large scales. The MNTD index of phylogenetic diversity showed a decreasing trend with increasing scale. The PD index showed an

**3. Results** 

**3. Results** 

*3.1. Characteristics of Plant Diversity* 

*3.1. Characteristics of Plant Diversity* 

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

not significantly different between scales (*p* > 0.05) (Figure 2).

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

opposite trend and was significantly different between scales, while the MPD index was not significantly different between scales (*p* > 0.05) (Figure 2). not significantly different between scales (*p* > 0.05) (Figure 2).

Among the species diversity indices, the Shannon diversity index, Simpson diversity index, and Margalef richness index tended to increase with scale and showed significant differences between scales (*p* < 0.05), and the Pielou evenness at large scales (50 m × 50 m) was significantly smaller than at small (5 m × 5 m) and medium scales (20 m × 20 m). For the functional diversity index, the RaoQ index showed an increasing trend from small to large scales. The FRic richness index, FEve evenness index, and FDiv divergence index were significantly different between small and large scales. The MNTD index of phylogenetic diversity showed a decreasing trend with increasing scale. The PD index showed an opposite trend and was significantly different between scales, while the MPD index was

Among the species diversity indices, the Shannon diversity index, Simpson diversity

index, and Margalef richness index tended to increase with scale and showed significant differences between scales (*p* < 0.05), and the Pielou evenness at large scales (50 m × 50 m) was significantly smaller than at small (5 m × 5 m) and medium scales (20 m × 20 m). For the functional diversity index, the RaoQ index showed an increasing trend from small to large scales. The FRic richness index, FEve evenness index, and FDiv divergence index were significantly different between small and large scales. The MNTD index of phylogenetic diversity showed a decreasing trend with increasing scale. The PD index showed an opposite trend and was significantly different between scales, while the MPD index was

**Figure 2.** Characteristics and differences in plant diversity between scales (Mean ± SE)*.* Note: error lines are standard errors; different lowercase letters on the error line for the same diversity index indicate highly significant differences between data (*p* < 0.05). **Figure 2.** Characteristics and differences in plant diversity between scales (Mean ± SE). Note: error lines are standard errors; different lowercase letters on the error line for the same diversity index indicate highly significant differences between data (*p* < 0.05). Using the K–S test and Raincloud plot (Figure 3), the ecosystem multifunctionality index calculated by the mean method was distributed normally at all three scales, with

#### *3.2. Characteristics of Ecosystem Multifunctionality* the values of the multifunctionality index varying from −1.085 to 1.129 for small-scale

*3.2. Characteristics of Ecosystem Multifunctionality*  Using the K–S test and Raincloud plot (Figure 3), the ecosystem multifunctionality index calculated by the mean method was distributed normally at all three scales, with the values of the multifunctionality index varying from −1.085 to 1.129 for small-scale samples, −0.931 to 0.973 for medium-scale samples, and −0.484 to 0.718 for large-scale sam-Using the K–S test and Raincloud plot (Figure 3), the ecosystem multifunctionality index calculated by the mean method was distributed normally at all three scales, with the values of the multifunctionality index varying from −1.085 to 1.129 for small-scale samples, −0.931 to 0.973 for medium-scale samples, and −0.484 to 0.718 for large-scale samples, but the ecosystem multifunctionality index did not vary significantly between scales (*p* > 0.05). samples, −0.931 to 0.973 for medium-scale samples, and −0.484 to 0.718 for large-scale samples, but the ecosystem multifunctionality index did not vary significantly between scales (*p* > 0.05).

#### ror lines are standard errors; same lowercase letters on the error line indicate no significant differences between data (*p* > 0.05). *3.3. Characterization of the Relationship between Diversity and Multifunctionality*

**Figure 3.** Characteristics and differences in ecosystem multifunctionality between scales. Note: Error lines are standard errors; same lowercase letters on the error line indicate no significant differences between data (*p* > 0.05). The results of the variance decomposition showed that there was a scale effect on the contribution of different dimensions of diversity to ecosystem multifunctionality. The FRic, FEve, and RaoQ indexes of plant functional diversity explained 56% of the variation in multifunctionality together on small scales, and they were the main factors driving ecosystem multifunctionality. The FRic index was significantly correlated with ecosystem multifunctionality.

However, at the mesoscale, phylogenetic diversity was the main factor shaping multifunctionality, with the MPD, MNTD, and PD indices accounting for 53% of multifunctionality. The MPD index was significantly correlated with multifunctionality, while the functional diversity of plants explained less of multifunctionality, but the FRic index

was still significantly correlated with multifunctionality. Species diversity had a stronger influence on multifunctionality, with the Margalef index being significantly correlated with multifunctionality. fluence on multifunctionality, with the Margalef index being significantly correlated with multifunctionality. The contribution of species diversity to multifunctionality reached a maximum of

The results of the variance decomposition showed that there was a scale effect on the contribution of different dimensions of diversity to ecosystem multifunctionality. The FRic, FEve, and RaoQ indexes of plant functional diversity explained 56% of the variation in multifunctionality together on small scales, and they were the main factors driving ecosystem multifunctionality. The FRic index was significantly correlated with ecosystem

However, at the mesoscale, phylogenetic diversity was the main factor shaping multifunctionality, with the MPD, MNTD, and PD indices accounting for 53% of multifunctionality. The MPD index was significantly correlated with multifunctionality, while the functional diversity of plants explained less of multifunctionality, but the FRic index was still significantly correlated with multifunctionality. Species diversity had a stronger in-

*Forests* **2022**, *13*, x FOR PEER REVIEW 6 of 14

multifunctionality.

*3.3. Characterization of the Relationship between Diversity and Multifunctionality* 

The contribution of species diversity to multifunctionality reached a maximum of 75% on large scales, and it was the main explanatory factor for ecosystem multifunctionality. The Margalef and Pielou indexes were significantly correlated with multifunctionality. Among the functional diversity indices, the FRic, FDiv, and RaoQ indexes together contributed to 13% of the variation in multifunctionality and were all significantly correlated with multifunctionality. Among the phylogenetic diversity indices, the MNTD index was significantly correlated with multifunctionality (Figure 4). 75% on large scales, and it was the main explanatory factor for ecosystem multifunctionality. The Margalef and Pielou indexes were significantly correlated with multifunctionality. Among the functional diversity indices, the FRic, FDiv, and RaoQ indexes together contributed to 13% of the variation in multifunctionality and were all significantly correlated with multifunctionality. Among the phylogenetic diversity indices, the MNTD index was significantly correlated with multifunctionality (Figure 4).

**Figure 4.** Relationship between plant diversity and ecosystem multifunction at different scales. Note: \*, \*\* and \*\*\* represent *p* < 0.05, *p* < 0.01 and *p* < 0.001.

#### **4. Discussion**

*4.1. Plant Diversity at Different Scales*

Plants are driven by interspecific interaction and environmental influence, resulting in certain spatial distribution patterns [38]. When spatial scales change, plant community structure and diversity characteristics also change. Exploring the relationship between sampling scale and plant diversity can contribute to a more comprehensive understanding of community diversity trends and species coexistence mechanisms [39]. Our study found that species diversity was strongly scale-dependent [40]. The Margalef richness index

showed an increasing trend with increasing scale. This is because the study site is highly windy and sandy, and water resources are scarce; thus, plants are clumped and aggregated to avoid wind and sand attacks and water scarcity [40], which limits the number of plant species that can be accommodated at small scales. As the sampling scale increases, the number of plant species increases, the composition of dominant species at each level becomes more diverse, and the Margalef index also increases. The Shannon diversity index and Simpson diversity index also increased with expansion in scale because, as the scale increases, habitat heterogeneity, the number of plants that can be accommodated, the number of species, and the level of species increases, and the structure of plant communities becomes more integrated and complex [41]. In addition, the Shannon diversity, Simpson diversity, and Margalef richness indexes showed a rapid increase and then a steady climb with the expansion of the scale, which is similar to the results of Deng et al. (2015) [14] in mixed coniferous forests of *Pinus radiata*. This indicates that the diversity of desert plant species increases with scale and plateaus after reaching a certain threshold [42].

Functional diversity is an extremely important part of plant diversity and plays an irreplaceable role in shaping the structure of plant communities and altering ecosystem functions [43]. Previous studies have shown that functional diversity varies with scale due to phenotypic plasticity [44–46]. Similar results were obtained in our study. The FEve evenness index and the FRic richness index both showed a decreasing trend with increasing scale and were opposite the trend in species richness, with significant differences between scales (*p* < 0.05); this indicates that species diversity and functional diversity have relatively independent trends [47]. This may be attributed to the obvious environmental filtering effect of the arid zone, where the functional composition of species is restricted to a certain range of functional traits, resulting in a more homogeneous pool of functional traits in this study area and an increase in species richness, leading to a more refined division of ecological niches rather than greater functional diversity [48], thus producing a different trend in functional diversity from that of species diversity [49,50].

The FDiv divergence index shows the degree of overlap in ecological niches between species within a community; that is, the heterogeneity of community character values [51], and a higher FDiv index indicates a high degree of ecological niche differentiation and higher resource use [52,53]. In our study, the FDiv index was significantly greater at large scales than at small scales, probably because, as scale increases, plant competition for the same or several habitat-specific resources diminishes, and ecological niches diverge further; thus, the FDiv index increases. The results of this study showed that the FRic richness index tended to decrease with increasing scale and was negatively correlated with the Shannon diversity index. This may be because functional richness is influenced not only by the functional ecological niche of the species but also by the range of functional trait values [54]. To overcome extreme drought conditions, functional traits of species in the study area are prioritized in response to selection pressure to adapt to drought [48,55], and functional traits tend to develop homogeneously, with increasing scale leading to a continuous increase in species richness followed by a deepening of functional redundancy [56,57], the FRic richness index declines, and the results of previous studies in Pinus oak forests in the Qinling Mountains are consistent with our study [58]. The negative correlation between species diversity and functional diversity because of scale expansion suggests that species diversity alone should not be considered when extrapolating functional diversity but also species differences and functional redundancy between scales [58].

In our study, the PD and MNTD indices varied significantly with scale, with the PD index increasing and the MNTD index decreasing with scale. This is probably because there is a significant correlation between the PD index and the Margalef index [59], with species richness increasing with scale, which is conducive to the maintenance of genealogical diversity at large scales. Genealogical structure is one of the most important expressions of community structure, as it reflects the process of community construction and evolution. Numerous studies have shown that the divergence or aggregation of genealogical structure is related to scale size [60–62]. The MNTD index in our study decreases with increasing

scale and is significantly different between the three scales, suggesting that the community genealogical structure gradually moves from divergence to aggregation with increasing scale [63]. Therefore, the community is composed of more distantly related species on small scales and more closely related species on large scales. This is probably because of the low variation in habitat conditions on a small scale, combined with the harsh environmental conditions in the study area, where competition for a particular resource between species makes competitive exclusion the dominant community-building process. Therefore, the genealogical structure tends to diverge. At large scales, habitat heterogeneity increases, habitat-filtering ecological processes begin to dominate [64], and the genealogical structure gradually tends to agglomerate. Studies in the evergreen broadleaf forests of the Gutian Mountains and in the tropical rainforests of Panama [61,65] have also shown that competitive exclusion at small scales has a negative effect on the coexistence of closely related species, and that there is a tendency for the genealogical structure of communities to change from divergence to aggregation with increasing scale, but when the spatial scale exceeds a certain area, there is no correlation between genealogical structure and scale, and the genealogical structure becomes aggregated [66].

#### *4.2. Ecosystem Multifunctionality at Different Scales*

Early studies of ecosystem multifunctionality focused on the effects of species diversity on a single ecosystem function on the same scale [50,67–70]. As research has progressed, researchers have realized that differences in the choice of scale of study influence the expression of ecosystem multifunctionality [20]. However, the results of this study showed that ecosystem multifunctionality did not diverge significantly among the three scales. This is probably because variations in ecosystem multifunctionality in this study area are more related to changes in species composition, environmental conditions, or temporal scales than to spatial scales [33,71,72]. Our study area is located in an arid desert region where plant species composition is highly dependent on soil water and salinity conditions, which has led to a high degree of similarity in the overall community species composition of the region; however, significant differences in ecosystem multifunctionality between scales depend not only on significant differences in species diversity between scales but also on diverse community structure [73]. In addition, this study did not consider climatic conditions and time scales when discussing ecosystem multifunctionality, but numerous studies have shown that ecosystem multifunctionality varies significantly depending on the climatic conditions or time span of the study [12,74]. Finally, although three scales—large, medium, and small—were chosen for this study, the largest scale was only 50 m × 50 m. There is still much potential to expand the area of the sampling scale; thus, work on larger scales should be carried out in the future.

#### *4.3. Relationship between Plant Diversity and Ecosystem Multifunctionality at Different Scales*

The relationship between plant diversity and ecosystem multifunctionality varies with scale. Our study found that the highest levels of species diversity at large scales explained the greatest amount of variation in ecosystem multifunctionality, suggesting that the maintenance of multiple ecosystem functions simultaneously in this study area required a greater number of species and higher levels of species diversity to support them and that high species diversity could effectively support the maintenance of ecosystem multifunctionality. This is probably due to ecological niche differences on large scales when different species or functional groups coexist in resource-limited communities [75], where larger scales have more species and therefore have complementary advantages in resource use and, thus, the greatest explanatory power for ecosystem multifunctionality. The relationship between species diversity indices and multifunctionality also varied between scales, with the Margalef index being significantly correlated with ecosystem multifunctionality at the meso- and macro-scales but not at the small-scale, and the Pielou index was significantly correlated with ecosystem multifunctionality at the macro-scale. This demonstrated that under the dimension of species diversity, species richness and species evenness played a

dominant role in driving multifunctionality. This likely results from the fact that the maintenance of ecosystem multifunctionality requires a high level of species diversity [72,76], and a smaller quantity of species cannot support all ecosystem functions [77]. Domestic and international studies have shown that larger vegetation communities and a larger number of species increase the use of habitat resources by plants and play an important role in maintaining high levels of ecosystem multifunctionality [69,78–81]. Species are relatively sparse in desert areas, and when the scale is small, the number of species in the sample is relatively infrequent, the Margalef index is comparatively low, the level of species diversity is not high, and the impact on ecosystem multifunctionality is limited. When the scale increases, the spatial ecosystem heterogeneity of the sample increases, it can accommodate a larger number and more species of plant samples, the Margalef index and Pielou indexes increase, and the explanatory power of species diversity on multifunctionality is strengthened, which is consistent with the hypothesis that species diversity is positively correlated with ecosystem multifunctionality [82].

Studies have shown that species diversity is no substitute for functional diversity to simply quantify multifunctionality [70], and in recent years, many studies have demonstrated that functional diversity has a stronger explanatory role for multifunctionality [83–85]. In our study, functional diversity contributes most to multifunctionality at a small scale, which is determined by a combination of species composition and structure. At small scales, species richness is low to allow for an increase in ecological niche space, which facilitates the expansion of the range of functional plant traits [86], so functional richness is at a higher level at small scales, which indicates a higher proportion of available resources [79], thus allowing functional diversity to dominate variations in ecosystem multifunctionality at small scales. In addition, the FRic index of functional diversity was significantly correlated with ecosystem multifunctionality at all three scales (*p* < 0.05), and the FDiv index was highly significantly correlated with ecosystem multifunctionality at large scales. The FRic index is a dominant driver of ecosystem multifunctionality at different scales, probably because plant species are limited in desert areas, and changes in species community composition and structure directly alter functional richness [87], which in turn affects single ecosystem functions and thus ecosystem multifunctionality. In addition, functional dispersion is significantly correlated with ecosystem multifunctionality at large scales, probably because as habitat heterogeneity increases on larger scales, variation in functional traits increases, competition among species decreases, complementarity of functional traits for resource use increases, overlap of species' ecological niches decreases, and the overall degree of resource use within scales increases, which also has a positive effect on the maintenance of ecosystem multifunctionality [86]. The results are also similar to the findings of Huang et al. (2019) [79] in Yunnan.

Genealogical diversity reflects the historical course of species evolution [88], can express community information not represented by species diversity and functional diversity, and is an important component in the study of the relationship between plant diversity and ecosystem multifunctionality [89–91]. In our study, genealogical diversity played a dominant role in ecosystem multifunctionality at the mesoscale, probably because habitat heterogeneity has an important influence on the maintenance of genealogical diversity. The results of the study on typical areas of karst landscapes showed that phylogenetic diversity had a single-peaked change with increasing habitat heterogeneity and was greatest in areas of moderate habitat heterogeneity [59]. In addition, a study by Li et al. (2021) [92] in the Daiyun Mountains also showed that phylogenetic diversity showed an intermediate peak with altitude, which was similar to the results of our study. At the mesoscale, habitat heterogeneity is at a moderate level compared to the large and small scales, which contributes more to genealogical diversity, and higher levels of genealogical diversity increase the rate of explanation for ecosystem multifunctionality. Our study also showed that the MPD index was significantly correlated with ecosystem multifunctionality on the large scale, and the PD index was not significantly correlated with ecosystem multifunctionality on any scale, indicating that the proximity of affinities among species plays a major role

in maintaining ecosystem multifunctionality at different scales. Although the PD index increases with the number of species, the total number of plant species in the desert area is limited, and when the PD index reaches a certain value, it will not continue to promote the maintenance of ecosystem multifunctionality. In our study area, species relatedness tended to change from distant to close, based on scale expansion. This is likely because the habitat filtering effect in the study area increases with scale, the extreme arid habitat conditions make it easier for species with similar life history strategies to survive [93,94], and similar survival strategies and close affinities use resources in similar ways, allowing limited resources to be fully utilized [95]. Genealogical structure has a strong influence on ecosystem multifunctionality, so the correlation between MNTD and MPD indexes and ecosystem multifunctionality is also stronger than that between PD indices.

#### **5. Conclusions**

We uncovered a significant scale effect in the relationship between plant diversity and ecosystem multifunctionality. Differences in sampling scales altered the composition and distribution of species in plant communities. The relative contribution of plant diversity to ecosystem multifunctionality also changed consequently. This highlights the close coupling between scale effects and ecosystem multifunctionality in the plant communities of this study area. Therefore, we will continue to expand the sampling scale in the next research work, aiming to better enhance plant diversity and maintain ecosystem function.

**Supplementary Materials:** The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/f13091505/s1, Table S1: Calculation formula of plant diversity indexes.

**Author Contributions:** Conceptualization, J.L. and G.L.; methodology, J.L.; software, J.L., H.W. and D.H.; writing—original draft preparation, J.L.; writing—review and editing, J.L., L.J. and D.H.; supervision, G.L. funding acquisition, G.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by the National Natural Science Foundation of China (42171026) and Xinjiang Uygur Autonomous Region innovation environment Construction special project & Science and technology innovation base construction project (PT2107).

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

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not available.

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

#### **References**


**Yan Luo <sup>1</sup> and Yanming Gong 2,3,\***


**Abstract:** In the past 30 years, Northwest China has experienced a warm and humid climate increase trend. How this climate change will affect the species diversity of plant communities is a hot issue in ecological research. In this study, four α diversity indexes were applied in 29 shrub communities at desert sites in Xinjiang, including the Margalef index, Simpson index, Shannon–Wiener index, and Pielou index, to explore the relationship between the α diversity of the desert shrub communities and climate factors (mean annual temperature (MAT) and mean annual precipitation (MAP)). The species diversity indexes varied across these different desert shrub communities. *Tamarix ramosissima* communities had the highest Margalef index, while the *Krascheninnikovia ewersmannia* communities had the lowest Margalef index; *T. ramosissima* communities also showed the highest Simpson index and Shannon–Wiener index, but *Alhagi sparsifolia* communities showed the lowest Simpson index and Shannon–Wiener index. The *Ephedra przewalskii* communities and *Karelinia caspica* communities showed the highest and the lowest Pielou index, respectively. The α diversity indexes (except the Pielou index) of desert shrub communities had a significantly positive correlation with MAP (*p* < 0.05) but a non-significantly correlation with MAT (*p* > 0.05). These results indicate that, compared with temperature, water conditions are still a more vital climatic factor affecting the species diversity of desert shrub communities in Xinjiang, and thus, the recent "warm and humid" climate trend in Xinjiang affects the α diversity of desert shrub communities.

**Keywords:** desert plants; shrub community; biodiversity; climate change; arid region

## **1. Introduction**

As the basis of ecosystem biodiversity, plant diversity plays an important role in maintaining ecosystem versatility, productivity, stability, and anti-interference ability [1,2]. Climate change has a significant impact on the water and heat dynamics of ecosystems, resulting in significant changes in the species composition and community structure of ecosystems [3]. Some research has found that climate change restricts the growth and distribution of plants, drives changes in plant diversity, affects interspecific relationships and community productivity, and seriously threatens the biodiversity of desert ecosystems [4,5]. Plant diversity will also alleviate the impact of climate change [4]. Therefore, the impact of climate change on plant diversity is a mutual process. In addition, to explain the distribution pattern of species diversity, a variety of hypotheses have been proposed, the most discussed of which is the energy hypothesis, which argues that changes in species diversity are controlled by energy according to the different forms of energy and their impact on the species diversity mechanism. The water-energy dynamic hypothesis holds that the large-scale pattern of species diversity is determined by water and energy. In this hypothesis, energy refers to thermal kinetic energy (or thermal energy), usually expressed in terms of potential evapotranspiration or temperature [5]. In addition, some scholars

**Citation:** Luo, Y.; Gong, Y. α Diversity of Desert Shrub Communities and Its Relationship with Climatic Factors in Xinjiang. *Forests* **2023**, *14*, 178. https:// doi.org/10.3390/f14020178

Academic Editor: Cate Macinnis-Ng

Received: 22 November 2022 Revised: 14 January 2023 Accepted: 17 January 2023 Published: 18 January 2023

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

have studied the species richness model of 3637 vascular plants in arid areas of northern China. It was found that water had a greater impact on species richness than energy, which indicated that the species richness pattern of plants in arid areas was mainly limited by water [6]. Therefore, under extreme climatic conditions, the influence of limiting factors on species diversity should be considered first. Desert ecosystems are the most sensitive areas to climate change. Desert plant diversity is one of the key factors determining the structure and functional diversity of desert plant communities [4,7]. Under the background of the rapid loss of biodiversity caused by global change, it is of great significance to study the maintenance mechanism of desert plant diversity for the maintenance and management of biodiversity in this area and the protection of biodiversity in the future.

Due to different regions and ecological types, the effects of temperature and precipitation on plant diversity will also show different response results. On a global scale, rising surface temperatures increase plant growth in the mid- to high latitudes of the Southern and Northern Hemispheres [8]; in inland Asia, temperature is also a major environmental factor for vegetation growth [9]. Other studies have shown that warming reduces the diversity and richness of plant species. For example, in the Arctic tundra ecosystem, short-term warming reduces community species diversity [10]. In the Tibetan Plateau, the increase in temperature accelerates the decrease in species richness [11]. However, studies from Inner Mongolia grassland found that warming had no significant effect on plant diversity [12]. In addition to temperature, the precipitation distribution pattern also affects plant species diversity. Studies have shown that a warm and humid climate is conducive to the increase in perennial plant diversity, while a warm and dry climate is conducive to the improvement of annual plant diversity [13]. In the warm and humid climate, the vegetation coverage in most arid areas of inland China increased significantly [14]. With the increase in annual precipitation, plant diversity changes along the steppe types (from desert to desert steppe-typical to steppe-meadow steppe) [15], which is consistent with Lopez-Angulo research results on the correlation between vegetation species richness and annual precipitation in different grassland types in Xinjiang. However, studies have shown that increased annual precipitation has no significant effect on plant diversity [16,17]. In addition, under the influence of climate change, species composition and vegetation patterns are changing. Kazakis et al. [18] and Pickering et al. [19] believed that climate warming caused the distribution of thermophilic shrubs, herbs, and invasive weeds to tend to occur at higher altitudes. The increase in temperature caused the dominant dwarf shrub, *Dryas octopetala*, in the alpine zone of southern Norway to be replaced by herbs, and the response of shrubs to global changes was more intense [20]. However, there are still few studies on the diversity of desert shrubs.

In recent years, against the background of global warming, the climate in Xinjiang has become more unstable, precipitation variability has increased, and the characteristics of "warm and humid" have become more obvious. Strong human activities have greatly increased the possibility of sudden climate change in Xinjiang, and extreme climate events have increased [21]. This makes the response of the Xinjiang desert ecosystem to global climate change unique and complex. Desert shrubland plants are an important part of biodiversity in the desert ecosystem of Xinjiang. Climate change has reduced plant diversity in originally extremely fragile desert ecosystems. The effects of climate change on the maintenance mechanism of plant diversity and other important scientific questions need to be answered. The study of species diversity is a key issue in desert ecosystem ecology, because the research content involves many aspects, such as the sustainable development of desert ecosystems and the protection and reconstruction of ecosystems. Species diversity is the degree of diversity at the level of species organization within a biological community. Margalef index, Simpson index, Shannon–Wiener index, and Pielou index are mostly used for community species diversity research. In this paper, the shrub community of the desert ecosystem in Xinjiang was taken as the research object. The four diversity indexes of the shrub community of desert plants were systematically studied, and the correlation between species diversity and climatic factors (temperature and precipitation) was discussed to

provide a theoretical basis for the optimal management of desert ecosystems and the protection of plant community species diversity.

correlation between species diversity and climatic factors (temperature and precipitation)

*Forests* **2023**, *13*, x FOR PEER REVIEW 3 of 14

#### **2. Materials and Methods** was discussed to provide a theoretical basis for the optimal management of desert ecosystems and the protection of plant community species diversity.

#### *2.1. Study Area*

Xinjiang is located in the core of the arid region of Central Asia, deep inland, and is the farthest province from the sea in China. The region has a typical warm temperate continental climate, with dry summers, less rain, cold winters, and large temperature differences between day and night. Sandstorm disasters are serious, precipitation is scarce, and evaporation is strong. The study area spans 80◦390–91◦190 E and 40◦140–46◦140 N. The average annual temperature is 9.37 ◦C, and the average annual precipitation is 107.93 mm (Figure 1; Table 1). **2. Materials and Methods**  *2.1. Study Area*  Xinjiang is located in the core of the arid region of Central Asia, deep inland, and is the farthest province from the sea in China. The region has a typical warm temperate continental climate, with dry summers, less rain, cold winters, and large temperature differences between day and night. Sandstorm disasters are serious, precipitation is scarce, and evaporation is strong. The study area spans 80°39′–91°19′ E and 40°14′–46°14′ N. The average annual temperature is 9.37 °C, and the average annual precipitation is 107.93 mm (Figure 1; Table 1).

**Figure 1.** Locations of the 29 sampling sites in Xinjiang, China. The figure was drawn based on the map of Xinjiang at a scale of 1:5,500,000 (Xinjiang Bureau of Surveying Mapping and Geoinformation, No. Xin S (2016) 250). **Table 1.** Characteristics of sites and climatic factors. **Figure 1.** Locations of the 29 sampling sites in Xinjiang, China. The figure was drawn based on the map of Xinjiang at a scale of 1:5,500,000 (Xinjiang Bureau of Surveying Mapping and Geoinformation, No. Xin S (2016) 250).

**Altitude** 

**MAT** 

**MAP** 




**Table 1.** *Cont.*

## *2.2. Field Sampling*

Sampling was conducted in the Xinjiang desert area in August 2018. According to the vegetation characteristics and environmental characteristics, a total of 29 survey plots were selected in the study area. The typical plot method was used to investigate 29 plant communities in Xinjiang desert areas. The basic information of 39 desert species was investigated. The sample design area of the plot is as follows: the shrub sample size is <sup>15</sup> <sup>×</sup> 15 m<sup>2</sup> . In each quadrat, plant species and coverage, plant height, number of plants, and other information of each community were counted to calculate community species diversity, abundance, frequency, dominance, and other indicators.

#### *2.3. Meteorological Data Acquisition*

In this study, annual average temperature (MAT, ◦C) and annual precipitation (MAP, mm) were selected as climatic factors. Annual mean temperature and annual precipitation data are from WorldClim Version 2.0 (http://worldclim.org/version2 (accessed on 7 March 2020)), Fick& Hijmans, 2017). Selecting these two indicators, we can investigate the changes in the diversity characteristics of desert shrub communities with temperature and water factors.

#### *2.4. Data Processing*

SPSS (Version 26.0, Armonk, NY, USA) and Origin (Version 2019, Northampton, MA, USA) were used to analyze the experimental data. One-way ANOVA was used to test the significance of the diversity indexes of different plant communities, and the difference was significant at the 0.05 level. The one-way ANOVA method was used to analyze the variance of the Margalef index, Simpson index, Shannon–Wiener index, and Pielou index of different plant communities. Using linear regression model to analyze the relationship between Margalef index, Simpson index, Shannon–Wiener index, and Pielou index and climatic factors.

The following four diversity indexes were selected as community diversity indexes. Margalef index represents the richness of species; the Simpson index represents the sum of probabilities of all species in a community or sample plot, the Shannon–Wiener index indicates the number of species in a community or sample plot and the evenness of individual distribution among species, and the Pielou index reflects the evenness of individual number distribution of each species, the calculation formula of each index is:

$$\text{Simpson index } (\mathbb{C}): \ C = \sum\_{i=1}^{n} N\_i^2 \tag{1}$$

$$\text{Shannon-Wiener index } (H): \ H' = -\sum\_{i=1}^{n} Pi \ln Pi \tag{2}$$

Margalef index (*Ma*) *Ma* = (*S* − 1)/ln*N* (3)

$$\text{Pieluu index} \left( E \right) E = H' / \ln(S) \tag{4}$$

In the above formula, *S* is the total number of species in the quadrat, *Pi* is the proportion of the abundance of the *i*-th species in the total abundance, and *N* is the total abundance. Gleason index is adopted for plant abundance (*D*). The calculation formula is:

$$D = \mathcal{S} / \ln A \tag{5}$$

In the above formula, *S* and *A* are the total number of plant species and total area (m<sup>2</sup> ) in the sample plot.

#### **3. Results and Analysis**

*3.1. Species Composition and Structural Characteristics of the Desert Shrub Community*

Through the investigation of the study area, a total of 39 species, 27 genera, and 12 families were recorded, represented by Chenopodiaceae, Polygonaceae, Leguminosae, Tamaricaceae, and Zygophyllaceae. Common species include *Tamarix ramosissima*, *Haloxylon ammodendron*, *Halostachys caspica*, *Calligonum mongolicum*, *Alhagi sparsifolia*, *Reaumuria soongarica*, *Ephedra przewalskii*, etc. (Table 2).

**Table 2.** Species classification and characteristics of the desert shrub community.


The plant composition and structural characteristics of 29 plant communities were investigated (Table 3). The overall plant with extremely low coverage (ranging from 6% to 26%), most of the communities contain shrub and subshrub layers, indicating that the distribution of plants is relatively scattered and the coverage is relatively low. Most of them are *T. ramosissima*, *H. ammodendron*, and *A. sparsifolia*, which can grow under extremely dry conditions. In 29 shrub communities, *T. ramosissima* (in S4, S5, S21) and *H. amodendron* (in S2, S25, S26, S27, S29) has the highest frequency with 0.38 and 0.34, respectively. The plant height ranges from 0.07 m to 6.43 m, and the crown width ranges from 0.11 m to 6.48 m in different habitats vary greatly.


**Table 3.** Plant composition and structural characteristics of different community types.

#### **Table 3.** *Cont.*



#### **Table 3.** *Cont.*

Note: S1 to S29 indicated sites with serial numbers from 1 to 29.

*3.2. Relationship between Diversity Index Characteristics of the Desert Shrub Community and Climatic Factors*

The Margalef index is related to the number of species and the total number of plants in the sample community. The Shannon–Wiener index is an important index reflecting the diversity of the community, which is used to explain the richness of species in the sample community. The Simpson index is also called the Simpson dominance index, which reflects the change in the number of species in the community. The larger the Pielou evenness index, the more uniform the distribution of individual species in the community. Table 4 shows the different variation characteristics among the species diversity indexes of the different plant communities. The Margalef species richness index ranged from 0.21 to 1.13. Among them, the *T. ramosissima* community (in S5) was the highest, and the *K. wersmannia* community (in S23) was the lowest, indicating that the plant species in the *T. ramosissima* community were the most species number and *K. wersmannia* community was the least species number. The Shannon–Wiener index ranged from 0.53 to 1.86, among which the *T. ramosissima* community (in S5) was the highest and the *A. sparssifolia* community (in S14) was the lowest, indicating that the *T. ramosissima* community contained a large amount of plant information and that the complexity of the community was higher than that in the other communities. The Simpson index ranged from 0.28 to 0.84 in this study, with the highest in the *T*. *ramosissima* community (in S5) and the lowest in *the A. sparssifolia* community (in S14), indicating that *T. ramosissima* plays an important role in Xinjiang desert ecosystem. The Pielou index ranged from 0.27 to 0.99, with the highest in the *E. przewalskii* community (in S12) and the lowest in the *K. caspica* community (in S19), indicating that the distribution of *E. przewalskii* community was the most uniform.

The relationship between the diversity characteristics of desert shrub communities and climatic factors was obtained by linear regression model analysis. Figure 2 shows that the Margalef index, Simpson index, Shannon–Wiener index, and Pielou index were negatively correlated with MAT but did not reach a significant level. Except for the Pielou index, the Margalef index (R<sup>2</sup> = 0.18, *p* < 0.05), Simpson index (R<sup>2</sup> = 0.19, *p* < 0.05), and Shannon–Wiener index (R<sup>2</sup> = 0.21, *p* < 0.05) were significantly positively correlated with MAP. This indicates that community characteristics are closely related to precipitation conditions and that precipitation conditions have a significant effect on the structure and composition of flora.

**Figure 2.** Relationship between plant community diversity index and climate factors. Note: (**a**,**c**,**e**,**g**) diagrams represent the relationship between Margalef index, Simpson index, Shannon–Wiener index, Pielou index and MAT respectively; (**b**,**d**,**f**,**h**) diagrams represent the relationship between Margalef index, Simpson index, Shannon–Wiener index, Pielou index and MAP respectively. **Figure 2.** Relationship between plant community diversity index and climate factors. Note: (**a**,**c**,**e**,**g**) diagrams represent the relationship between Margalef index, Simpson index, Shannon–Wiener index, Pielou index and MAT respectively; (**b**,**d**,**f**,**h**) diagrams represent the relationship between Margalef index, Simpson index, Shannon–Wiener index, Pielou index and MAP respectively.

tively small, dominated by xerophytic, halophytic, or ultra-xerophytic small shrubs and perennial herbs, reflecting the characteristics of desert, semi-desert, and steppe desert plant communities [22,23]. In this study, the dominant layer of the community is the shrub layer, and its species composition is dominated by dwarf subshrubs. The dominant plants are mainly shrubs or subshrubs of Chenopodiaceae, Tamaricaceae, and Zygophyllaceae.

*4.1. Diversity Characteristics of the Desert Shrub Community* 

**4. Discussion** 


**Table 4.** Diversity index in different plant communities.

Note: S1 to S29 indicated sites with serial numbers from 1 to 29.

#### **4. Discussion**

#### *4.1. Diversity Characteristics of the Desert Shrub Community*

The species and number of natural vegetation in desert shrub communities are relatively small, dominated by xerophytic, halophytic, or ultra-xerophytic small shrubs and perennial herbs, reflecting the characteristics of desert, semi-desert, and steppe desert plant communities [22,23]. In this study, the dominant layer of the community is the shrub layer, and its species composition is dominated by dwarf subshrubs. The dominant plants are mainly shrubs or subshrubs of Chenopodiaceae, Tamaricaceae, and Zygophyllaceae. The proportion of herbaceous plants is lower than that of shrubs and subshrubs, reflecting the harsh desert environment in Xinjiang. Under the conditions of a low degree of heterogeneity, from the perspective of layers, the life form of vegetation composition species gradually tends to be simple or even single.

The community species diversity index can reflect community structure characteristics. The larger the value of the Margalef species richness index, the higher the species number reflecting the community or habitat, and the greater the number of species [22–24]. Among the 29 communities in the Xinjiang desert, the Margalef index is the highest in the *T. ramosissima* community. *T. ramosissima*, as the constructive species, was widely distributed in Xinjiang. Because of its large number of species, large size, and strong ability to adapt to barren habitats, *T. ramosissima* has significant control over the formation of its community structure and community environment. The constructive species of *T. ramosissima* is more likely to form tall shrubs than other species in the process of the barren desert environment, but it will also form significant differences in height, coverage, and other aspects due to environmental changes [25]. The Shannon–Wiener diversity index is used to describe the amount of information contained in a community. The larger the index, the higher the

complexity of the community [26]. In this study, the Shannon–Wiener index was the highest in the *T. ramosissima* community. *T. ramosissima* is the most salt-tolerant species in *Tamarix*, which has the characteristics of strong adaptability, resistance to wind, salt and alkali, drought and moisture, barren soil, sand burial, and developed roots [27]. Therefore, the associated species of the *T. ramosissima* community are abundant. The Simpson index, also known as the ecological dominance index, comprehensively reflects the richness of species in the community. It is one of the more commonly used diversity indexes. The higher the index is, the higher the ecological dominance of dominant species and the higher the probability of species occurrence [24]. In different communities, the Simpson index was still the highest in the *T. ramosissima* community, indicating that the *T. ramosissima* community plays an important role in the desert ecosystem, has good adaptability to the extreme environment, and has a great impact on the improvement of the desert environment. The Pielou index is an index of species evenness, which is generally used to characterize the uniformity of the spatial distribution of species in a community. The larger the value of the index is, the more uniform the distribution of plants [24]. In this study, the Pielou index was the highest in the *E. pseudacorus* community and the lowest in the *K. caspica* community, indicating that the species in the *E. pseudacorus* community had the highest uniformity of spatial distribution, while the *K. caspica* community had the lowest. In desert ecosystems, water, salinity, and other factors have certain effects on the distribution of species, so they will also affect the diversity characteristics of plant communities. However, more samples need to be collected to confirm the specific impact.

#### *4.2. Relationship between Desert Shrub Community Diversity and Climatic Factors*

The results of this study confirmed that climatic factors, especially water conditions, were the main factors leading to differences in plant community types and species composition in the Xinjiang desert. The Margalef index, Simpson index, Shannon–Wiener index, and Pielou index had no significant correlation with MAT but had a significant positive correlation with MAP. This shows that water is still the main limiting factor affecting the diversity of desert plant communities in the Xinjiang desert ecosystem. The frequent occurrence of extreme events in the context of global warming has a particularly significant impact on Xinjiang, which is located in arid and semi-arid regions, resulting in increased risks of extreme precipitation events, storm floods, and snowmelt floods. The increase in precipitation will help to improve the α diversity characteristics of desert shrub communities.

In this study, most diversity indexes had a significant positive correlation with the MAP. This is because the Xinjiang desert area is located in the temperate desert area of China. The environment is harsh, the soil is barren, the community type is poor, and the species' life type is single. It is mainly composed of temperate desert shrubs and semi-shrub communities, and temperate shrubs and meadows are formed in some areas [22,23,26]. In terms of community species composition, shrubs and subshrubs account for a higher proportion, less precipitation in the habitat, simpler community composition, and higher dominance of dominant communities. Environmental factors have important effects on the distribution and diversity of plant species [26]. Extremely arid climatic conditions in the study area are the basis for the formation of desert plants in Xinjiang. At the regional scale, temperature and precipitation are the determinants of vegetation type distribution and species life form and are the basis for the formation of zonal vegetation [16,28]. The growth and development of various plants require precipitation. Precipitation will affect the distribution of different plant species and the species diversity of communities. Precipitation is a comprehensive environmental factor affected by many factors. Especially in recent years, due to the over-exploitation of the earth's resources and the rapid development of industrialization it has accelerated the process of global warming and has also had a huge impact on global plant species diversity [28]. Global warming can not only affect plant growth and productivity by prolonging the seasonal cycle and changing phenological conditions but also directly affect plant photosynthesis by changing precipitation [29]. In

addition, precipitation will also affect the change in soil moisture. The water absorbed by plants mainly comes from soil water, while soil water mainly comes from precipitation, so most of the water available to vegetation comes from natural precipitation [28,29].

Plants absorb water mainly from soil moisture, and soil moisture mainly comes from precipitation, so most of the water that vegetation can use comes from natural precipitation [28]. The effect of precipitation on plant community growth is a cumulative effect, and the plant community structure will change with changes in precipitation [27–29]. Precipitation changes have a significant impact on ecosystem structure and processes, such as community composition and dynamics, species diversity, and species competition. In the desert area of northwest China, water is the main limiting factor controlling plant growth and plant community diversity [29]. The results of this study show that precipitation can significantly improve the Margalef index, Simpson index, and Shannon–Wiener index. Yuan et al. found that the species diversity of alpine grassland in the northern Tibetan Plateau showed an exponential growth relationship with precipitation, and the increase in precipitation led to the optimization of community structure, which was significantly promoted [30]. Soliveres et al. [31] found that there is a good linear relationship between species richness and seasonal precipitation changes. Precipitation pattern changes directly affect species diversity and ecosystem versatility. Therefore, in the future, this paper also needs to consider investigating more data and further revealing the impact of rainfall changes on plant community diversity on a longer time scale.

#### **5. Conclusions**

In a word, this study analyzed the characteristics of different diversity indexes among desert plant communities and their responses to climate factors. Our results show that the *T. ramosissima* community had the highest Margalef index, and the *K. ewersmannia* community had the lowest. The Simpson index and Shannon–Wiener index were the highest in *T. ramosissima* and the lowest in *A. sparsifolia*. The Pielou index was the highest in the *E. przewalskii* community and the lowest in the *K. caspica* community. These results showed that there are significant differences among different plant communities in different diversity indexes. In addition, the α diversity index (except the Pielou index) of desert shrub community species was significantly positively correlated with the annual average precipitation but not significantly negatively correlated with the annual average temperature. These results indicated that water conditions are still an important factor affecting the species diversity of desert shrub communities in Xinjiang. These results provide an important reference for understanding the characteristics of plant community diversity in desert ecosystems and the relationship between desert plants and climate change.

**Author Contributions:** Y.L. carried out the fieldwork and wrote the first draft of the manuscript and Y.G. assisted with revising and editing the draft manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** Third Xinjiang Scientific Expedition Program (Grant No. 2022xjkk040301), China Postdoctoral Science Foundation (Grant No. 2022M722667), and the Department of Education of Xinjiang Uygur Autonomous Region, Dr. Tianchi Program Project (Grant No. TCBS202123).

**Data Availability Statement:** Anyone who needs available data can directly consult and contact the first author of this article.

**Conflicts of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

#### **References**


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