*Article* **The Quality of** *Scutellaria baicalensis* **Georgi Is Effectively Affected by Lithology and Soil's Rare Earth Elements (REEs) Concentration**

**Zijian Sun 1,2, Wei Shen 1, Weixuan Fang 3,\*, Huiqiong Zhang 2, Ziran Chen 2, Lianghui Xiong <sup>2</sup> and Tianhao An <sup>2</sup>**


**\*** Correspondence: weixuanfang123@163.com

**Abstract:** The top-geoherb "Rehe *Scutellaria baicalensis*" was naturally distributed on Yanshan Mountain in Chengde city, Hebei Province, China. Exploring the influences of parent materials on the quality of the top-geoherbs in terms of micronutrient elements is of great significance for the protection of origin and for optimizing replanting patterns of *Scutellaria baicalensis*. In this study, three habitats of *Scutellaria baicalensis* with contrasting geopedological conditions, i.e., naturally grown habitats (NGHs), artificial planting habitats (APHs), and biomimetic cultivation habitats (BCHs), are taken as objects to probe the influences of parent materials on the quality of *Scutellaria baicalensis* in terms of rare earth elements (REEs) by testing on REEs concentrations in the weathering profiles, rhizosphere soil and growing *Scutellaria baicalensis*, as well as their flavonoid compound contents. Hornblende-gneiss was the parent rock in NGHs, whose protolith was femic volcanic rock. Loess was the parent rock in APHs and BCHs. REEs were more abundant in hornblende-gneiss than loess, and therefore, soils developed in NGHs contained higher REE concentrations than those in APHs, which was lower than BCHs after REE-rich micro-fertilizers application. The coefficient of variation (CV) of REEs concentrations in the rhizosphere soils of hornblende-gneiss was higher than that in loess. It possibly was attributed to the complicated minerals compositions and various minerals' grain sizes of hornblende-gneiss, resulting in the variety of weathering intensity involving eluviation, leaching, adsorption, etc., as well as weathering productions, dominated by clay minerals and Fe-(hydro)oxide, and ultimately the remarkable differences in the migrations, enrichments and fractionations within REEs. The biological absorption coefficients (BACs) of REEs for *Scutellaria baicalensis* decreased in the order of NGHs > APHs > BCHs. Roots of *Scutellaria baicalensis* contained similar ΣREE in NGHs (2.02 mg·kg−1) and BCHs (2.04 mg·kg−1), which were higher than that in APHs (1.78 mg·kg−1). Soils developed in hornblende-gneiss were characterized by lower clay fraction content and overall alkalinity with a pH value of 8.06. The absorption and utilization efficiency of REEs for *Scutellaria baicalensis* in NGHs was higher than in APHs and BCHs. Flavonoid compounds, effective constituents of *Scutelleria baicalensis*, showed more accumulations in NGHs than APHs and BCHs, implying their optimal quality of *Scutellaria baicalensis* in NGHs. Flavonoid compounds were remarkably correlated with REEs in the roots, suggesting the influence of REEs concentrations on the quality of *Scutellaria baicalensis*. It can be concluded that high REEs and micronutrient element concentrations of hornblende-gneiss favored the synthesis and accumulation of flavonoid compounds in *Scutellaria baicalensis* after the activation of endocytosis induced by REEs.

**Keywords:** parent materials; weathering; geoherbs; gneiss; flavonoid compounds

## **1. Introduction**

Geoherb is a term used by ancient Chinese to describe the inter-species variation of Chinese medicines relevant to the geographical variation, and top-geoherb (Dao-di Herbs) is the population of a geoherb growing in habitats with natural conditions and ecological

**Citation:** Sun, Z.; Shen, W.; Fang, W.; Zhang, H.; Chen, Z.; Xiong, L.; An, T. The Quality of *Scutellaria baicalensis* Georgi Is Effectively Affected by Lithology and Soil's Rare Earth Elements (REEs) Concentration. *Appl. Sci.* **2023**, *13*, 3086. https://doi.org/ 10.3390/app13053086

Academic Editor: José A. González-Pérez

Received: 9 December 2022 Revised: 21 January 2023 Accepted: 1 February 2023 Published: 27 February 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/).

environment, featuring proverbial superior qualities, better curative effect and popularly used in traditional Chinese medicine clinical practice [1–3], such as *Lycium barbarum* L. from Ningxia, *Radix Angelicae* from Hangzhou, *Radix Rehmanniae* from Henan and *Scutellaria baicalensis* from Rehe in China [4]. Biological factors, e.g., species and intrinsic nature, and geo-environmental factors, e.g., physical and chemical characteristics of parent rock and weathering productions, jointly exert a crucial control on the formation of top-geoherbs' traits. As one of the critical material bases of traditional Chinese medicines, micronutrient elements significantly regulated and controlled the activity of many biological molecules (proteins, enzymes, and hormones) by taking part in enzymatic structures, constructing electron transport systems as carriers, and participating in the synthesis of hormones and vitamins, which ultimately influenced the metabolic activities and accumulation of effective constituents in medicines [5,6]. Therefore, micronutrient element assemblages played a decisive role in forming top-geoherbs' traits [5–8].

Rare earth elements (REEs) are a group of chemically similar elements behaving coherently in nature and consist of lanthanide elements from La to Lu (International Union of Pure and Applied Chemistry, 2005) [9]. REEs can be divided into light rare earth elements (LREEs; La–Eu) and heavy rare earth elements (HREEs; Gd–Lu). Although Y did not belong to REEs, it was studied together because of its chemically similar behavior to REEs. As significant components of micronutrient elements for plant growth, REEs can activate the endocytosis of plant cells [10]. It was demonstrated in previous studies that during the cultivation of Chinese medicines, e.g., *Ginseng, Coix lachryma-jobi, and Eucommia Land wolfberry,* adding moderate concentrations of rare earth fertilizer notably improved yield, quality and the accumulation of effective constituents [11–14]. These discoveries have led to the large-scale application of rare earth micro-fertilizers to medicines production. In order to enable the development of more sustainable rare earth utilization practices, it is significant to conduct research on their sources, migrations, enrichments, fractionations and transformations in the soil–root environments. Bedrock and its weathering production were the parent materials of soil on which plants survived, which was the natural mineral–nutrition elements pools and primary source of REEs for plants [5,15]. In general, the micronutrient element concentrations in the soils that developed in the mountain terrain were primarily inherited from parent materials. Therefore, the latter controlled the background characteristics of micronutrient elements concentrations in the former [16–18] and, ultimately, the adsorption and enrichments of micronutrient elements in Chinese medicines.

The root of the labiate *Scutellaria baicalensis* was initially documented in *Shennong's Classic of Materia Medica* and in *Collective Notes to the Canon of Materia Medica* as traditional Chinese medicines, which were characterized by a cold property and bitter taste. It is a commonly used Chinese medicine recorded in *The Pharmacopoeia of the People's Republic of China* (2015 version) [19], which is beneficial for clearing away hot and toxic substances, promoting diuresis, cooling blood, preventing miscarriage, and relieving cough [20,21]. Chengde city, Hebei Province, China, has been verified to be one of the top-geoherbs habitats for *Scutellaria baicalensis*. It was claimed in *Differentiation of drug production* that *Scutellaria baicalensis* has been widespread around Zhili and Rehe in Hebei Province for a long time, featuring a thick and long root, solid texture, golden-yellow rind, and extra low impurity, and it has been entitled "Rehe *Scutellaria baicalensis*" [22]. According to a field survey, gneiss was confirmed to be the dominant rock type inhabiting Rehe *Scutellaria baicalensis*. *Scutellaria baicalensis* was also observed being planted in loess locations, and REE-rich micro-fertilizers applications were conducted in parts of these habitats. In order to explore the effect of parent materials on the geoherbalism and quality of *Scutellaria baicalensis* in terms of REEs, three habitats of *Scutellaria baicalensis* with contrasting geo-pedological conditions, i.e., naturally grown habitat (NGHs), artificial planting habitats (APHs) without fertilization, and biomimetic cultivation habitats (BCHs) under field management and fertilization were studied. This study focused on the enrichments and fractionations of REEs during bedrock weathering and micronutrients adsorption by *Scutellaria baicalensis*

and further on the controlling mechanism based on field eco-geological investigation and systematical sampling on roots of *Scutellaria baicalensis* and corresponding rhizosphere soil, and in weathering crust from the bottom of bedrock up to soil. REEs test for all samples and flavonoid compound tests for roots of *Scutellaria baicalensis* were conducted.

## **2. Physical Geography and Distribution of** *Scutellaria baicalensis*

The study area is located in Yanshan Mountain, geographically in Chengde city, Hebei Province, China, where Rehe *Scutellaria baicalensis* is regionally widespread. A warm temperate semi-humid continental monsoon climate characterizes the region, with annual precipitation from 500 to 800 mm, frost-free days from 92 to 180 d, an annual effective accumulative temperature from 2800 to 3980 ◦C, and annual sunshine duration from 2500 to 3100 h [23]. The characteristics of the three comparable habitats of *Scutellaria baicalensis* (NGHs, APHs, and BCHs) were as follows (Figure 1).

**Figure 1.** (**a**) Geographical location map; (**b**) Distribution map of sampling in three habitats; (**c**) Distribution map of sampling in biomimetic cultivation habitats (BCHs); (**d**) Distribution map of sampling in artificial planting habitats (APHs).

(1) It was documented that NGHs were concentrated in southern Chengde, such as Luanping County, Xinglong County and Kuanchen County, China. Based on our field survey, NGHs in Hongqi, Luanping County, were selected for sampling (Figure 2a–c). The lithology was dominated by hornblende-gneiss, which were strongly weathered and had intensive fractures along weathering profiles, resulting in thick saprolite and soil with a thickness of 1 m.


**Figure 2.** Field photos of *Scutellaria baicalensis* in various habitats; (**a**–**c**) Natural grown habitats (NGHs); (**d**,**e**) Artificial planting habitats (APHs); (**f**) Biomimetic cultivation habitats (BCHs).

## **3. Materials and Methods**

## *3.1. Sampling Strategy*

For research on the migrations and enrichments of REEs in the soil–roots environment, the sampling numbers of rhizosphere soils and corresponding roots of *Scutellaria baicalensis* in NGHs, APHs, and BCHs were 11, 10 and 8, respectively. Soil samples were collected in the topsoil layer extending 20 cm down from the surface. In addition, two weathering profiles in the close vicinity of field plots (HQ2007 and HQ2008) in NGHs were systemically sampled from the fresh parent rock (bedrocks) through the semi-weathered and highly weathered horizons (regoliths) to the fully weathered horizons (soils). Both profiles were distributed in the second stage of Fenghuangzui formation in the Neoarcjean Dantazi group, whose rock type was dominated by migmatite hornblende-gneiss, amphibolite and biotite-plagioclase-fels. In our study, the rock type of both profiles was hornblende-gneiss; however, their mineralogical compositions were slightly different: the bedrock in profile 1 was mainly composed of medium- and coarse-grained hornblende, plagioclase, quartz and a small amount of fine-grained biotite, while in profile 2, the bedrock was mainly composed of fine-grained plagioclase, biotite, quartz, and a small amount of hornblende. Weathering extended to a depth of 3.6 m in profile 1. The soil layer was 20 cm thick, and its texture was dominated by gravel. The regolith layer extended for a further 340 cm below soil layers with the chemical index of alteration (CIA) values ranging from 65.0 to 71.1. Bedrock extended to a depth of 5.6 m. Weathering extended to a depth of 5.8 m in profile 2. The soil layer was 50 cm thick, with the humus horizon being 10 cm. The soil texture was dominated by gravel. The regolith layer extended for a further 530 cm below

soil layers with the CIA values ranging from 51.1 to 67.2. Bedrock extended to a depth of 7.6 m in which alteration zones, e.g., chloritization, occurred, extending 1 m upwards from the bottom. About 1–2 samples were collected per subhorizon. A total of 8 samples from profile 1 and 10 samples from profile 2 were collected, and each one soil profile was collected per habitats in APHs and BCHs in the close vicinity of field plots (LD2004 and LD2014), respectively. The topsoil layer extending 20 cm downwards from the surface and the loess layer extending for a further 20 cm were sampled, and a total of 2 samples from profile 3 and also 2 samples from profile 4 were collected.

## *3.2. Analytical Methods*

For REE analyses, the prepared samples were first dried at 70 ◦C, and then both the dried soil, regolith and fresh bedrock were crushed, pulverized by an agate mortar and sieved through a 200 mesh into powder. The powder was baked at 105 ◦C to remove adsorbed water before analysis. REEs were measured using inductively coupled plasmamass spectrometry (ICP–MS). For soil samples, approximately 50 mg of soil sample was digested with 0.5 mL HNO3 and 1 mL HF in screw-top, PTFE-lined stainless steel bombs at 185 ◦C for 24 h. The inner Teflon was removed from the hot plate after cooling, and the solution was evaporated to dryness. The solution was then drained and evaporated to dryness with 0.5 mL HNO3. This procedure was repeated twice. The final residue was redissolved by adding 1.5 mL HNO3 and 1.5 mL deionized water. Subsequently, the bomb was resealed and heated at 130 ◦C for 3 h. After cooling to room temperature, the final solution was diluted to 50 mL by adding distilled deionized water. For regolith and bedrock samples, approximately 50 mg of powdered samples was digested with 100 mL APS, which was followed by mechanical shaking for 2 h and standing for 30 min. After filtration at medium speed, 2 mL of filterable solution was diluted to 100 mL by adding distilled deionized water. Subsequently, 10 mL of the solution was diluted to 100 mL by adding 1 mL indium (In) solution, 1 mL HNO3 and deionized water. The *Scutellaria baicalensis* samples were washed with tap water and deionized water successively and then oven-dried at 75 ◦C until the dry weight reached a constant value. Due to the dominant distributions of flavonoid compounds in roots, roots were used for an REE and flavonoid compound test. The dry root samples were ground to a fine powder. Approximately 200 mg of powdered sample was digested with 5 mL HNO3 in a high-pressure digestion tank for 1 h, which was followed by cooling and removing HNO3 at 140 ◦C. After cooling to room temperature, the solution was diluted to 10 mL by adding deionized water. The reagent blanks of soil, regolith, bedrock, and roots were treated following the same procedures as the corresponding samples. The total analytical errors for REEs in this study were within ±6.

High-Performance Liquid Chromatography (HPLC–DAD) was applied to test the six flavonoid compounds, including baicalin, oroxylin A glycoside, wogonoside, baicalein, wogonin, and oroxylin A. First, the reference substances of the six flavonoid compounds were dissolved in chromatographic methanol and shaken to prepare the reference solution. Second, root powder samples were placed in a volumetric flask with 70% ethanol and shaken. The supernatant was extracted for filtration through a 0.22 μm filter membrane after ultrasonic extraction for 40 min and cooling to room temperature. Finally, a high-performance liquid chromatograph Water E2695 was utilized to test the flavonoid compounds of the reference solutions and our specimens. Methanol-0.1% phosphoric acid solution was applied to gradient elute the mobile phase. The accuracy was controlled by adding 10% of blank samples and parallel samples during testing according to the specification requirement.

## *3.3. Parameters on the REE Distribution Characteristics*

ΣREE, ΣLREE and ΣHREE were calculated as the sum of REEs, LREEs and HREEs, respectively. LREE/HREE was the ratio of ΣLREE to ΣHREE and La*N*/Yb*<sup>N</sup>* was the ratio

of La*<sup>N</sup>* and Yb*N*, where La*<sup>N</sup>* and Yb*<sup>N</sup>* represent the chondrite normalized values of La and Yb.

Eu anomaly values are quantifed as (1):

$$\delta \mathbf{E} \mathbf{u} = \frac{\mathbf{E} \mathbf{u}\_{\mathbf{N}}}{\sqrt{\mathbf{S} \mathbf{m}\_{\mathbf{N}} \times \mathbf{G} \mathbf{d}\_{\mathbf{N}}}},\tag{1}$$

where Eu*N*, Sm*N*, and Gd*<sup>N</sup>* represent the chondrite normalized values of Eu, Sm, and Gd, respectively.

Ce anomalies were calculated by formula (2):

$$\delta \text{Ce} = \frac{\text{Ce}\_{\text{N}}}{\sqrt{\text{LaN} \times \text{Pr}\_{\text{N}}}} \text{ } \tag{2}$$

where Ce*N*, La*N*, and Pr*<sup>N</sup>* represent the chondrite normalized values of Ce, La, and Pr, respectively.

The normality of the parameters in various habitats was confirmed using the Shapiro– Wilkes test, and probability (*p* value) lower than 0.05 showed their abnormal distribution. The normal distribution test was conducted in SPSS 26.0.

## **4. Results**

## *4.1. REE Concentrations in the Weathering Crust*

#### 4.1.1. REE Concentrations of Rhizosphere Soils

The concentrations of REEs and several parameters on the REEs distribution characteristics of rhizosphere soils in various habitats are listed in Table 1. According to the Shapiro–Wilkes test, ΣREE, ΣLREE and ΣHREE in the rhizosphere soils were normally distributed in NGHs and APHs (*p* > 0.05) but not in BCHs (*p* < 0.05), and therefore, the median represents the statistical characteristic of REE concentrations in various habitats. Rhizosphere soils in NGHs contained REE concentrations with the median ΣREE values being 173 μg·g−<sup>1</sup> and the median ΣLREE values being 157 μg·g−1, which were higher than that in APHs with the median ΣREE values being 149 μg·g−<sup>1</sup> and the median ΣLREE values being 133 μg·g<sup>−</sup>1. However, ΣHREE exhibited the opposite feature that rhizosphere soils in APHs with higher HREE concentrations. The median LREE/HREE and La*N*/Yb*<sup>N</sup>* of rhizosphere soils in NGHs were 9.76 and 7.86, respectively, which were higher than that in APHs with a median of 11.4 and 8.82, respectively, suggesting greater fractionations with LREEs and HREEs in NGHs. After micro-fertilizers application, it exhibited remarkable enrichments of REEs in the rhizosphere soils of BCHs with the median ΣREE, ΣLREE and ΣHREE values being 187 μg·g<sup>−</sup>1, 166 μg·g−<sup>1</sup> and 20.2 μg·g<sup>−</sup>1, respectively, which were also higher than that in NGHs. Additionally, LREE/HREE and La*N*/Yb*<sup>N</sup>* in the rhizosphere soils of BCHs increased, with the median LREE/HREE and La*N*/Yb*<sup>N</sup>* being 8.36 and 9.53, respectively. The chondrite-normalized REE patterns exhibited an obvious right-inclined style that all the rhizosphere soils were enriched in LREEs (Figure 3a,c,e) with moderately to slightly negative Eu anomalies, especially in APHs and BCHs. However, the Ce anomaly was not apparent. The coefficients of variation (CVs) of ΣREE, ΣLREE and ΣHREE of rhizosphere soils in APHs were 6.84%, 6.88% and 8.84%, respectively, which were much lower than those in NGHs and BCHs, indicating that the REEs concentrations of soils developed in loess without fertilization were relatively homogeneous.

## 4.1.2. REE Concentrations in the Weathering Profile

Two weathering profiles were distributed in hornblende-gneiss locations. The ΣREE values of bedrock were 233 μg·g−<sup>1</sup> and 264 μg·g<sup>−</sup>1, respectively, in profiles 1 and 2, indicating their similar REEs concentrations. Both weathering profiles in NGHs showed similar chondrite-normalized REE patterns from the bottom of bedrock up to regolith, and onto the soil, with enrichment of LREEs and moderately negative Eu anomalies (Table 2, Figure 4), which is in accordance with the weathering characteristics in mountain terrain that bedrock was in-situ weathered to the soil, and the latter has inherited REE concentrations from the

former. However, different enrichments and fractionations within REEs have occurred during weathering. As shown in profile 1 (Figure 4a), the ΣREE was overall higher in the regoliths in which HREEs were deficient, but LREEs were enriched as compared to bedrock, while topsoil was deficient in both LREEs and HREEs. Regoliths contained higher REE concentrations with ΣREE values ranging 242–289 μg·g<sup>−</sup>1, and that of soil was 169 μg·g<sup>−</sup>1. LREE/HREE varied in the range of 8.28–11.2 and La*N*/Yb*<sup>N</sup>* varied in the range of 9.24– 13.8, which were slightly different with bedrock (LREE/HREE = 8.56, La*N*/Yb*<sup>N</sup>* = 9.48). However, profile 2 exhibited that the ΣREE was overall lower in the regoliths and soils, and that LREEs were more deficient than HREEs in both regolith and soil as compared to bedrock, while topsoil was more enriched in REEs than regolith (Figure 4b). Regoliths contained lower REE concentrations with ΣREE values ranging 142–238 μg·g−1, and that of topsoil was 238 μg·g−1. LREE/HREE varied in the range of 6.08–10.7 and La*N*/Yb*<sup>N</sup>* varied in the range of 5.83–14.3, which were different with bedrock (LREE/HREE = 20.6, La*N*/Yb*<sup>N</sup>* = 35.8).

**Figure 3.** Chondrite-normalized REE patterns of rhizosphere soils and roots of *Scutellaria baicalensis*. (**a**) Rhizosphere soils in NGHs; (**b**) Roots in NGHs; (**c**) Rhizosphere soils in APHs; (**d**) Roots in APHs; (**e**) Rhizosphere soils in BCHs; (**f**) Roots in BCHs.


**Table 1.** Parameters on REE distribution characteristics of rhizosphere soils in various habitats, Hebei Province.

**Figure 4.** Chondrite-normalized REE patterns of layers along profiles. (**a**) Chondrite-normalized REE patterns along profile 1 in NGHs; (**b**) Chondrite-normalized REE patterns along profile 2 in NGHs; (**c**) Chondrite-normalized REE patterns along profile 3 in APHs; (**d**) Chondrite-normalized REE patterns along profile 4 in BCHs.

Soils developed in loess also showed chondrite-normalized REE patterns similar to those of parent materials with enrichment of LREEs and lightly negative Eu anomalies (Figure 4c,d). It also exhibited the depletion of REEs in topsoil as compared to parent material in APHs. The ΣREE values of parents materials (loess) were similar, 139 and 133 μg·g−1, whose corresponding soils contained total REE concentrations being 119 and 136 μg·g<sup>−</sup>1. LREE/HREE was, respectively, 8.43 and 8.17, which was similar to their parent materials with LREE/HREE being 7.90 and 8.11. La*N*/Yb*<sup>N</sup>* was, respectively, 9.87 and 9.23, which was similar to their parent materials, with La*N*/Yb*<sup>N</sup>* being 9.32 and 9.26. In general, the REEs of soils and parent material horizons in hornblende-gneiss were more abundant, and they had more substantial variation between LREE and HREE than those in loess.

**Table 2.** Parameters on REEs distribution characteristics along profiles in various habitats, Hebei Province.


## *4.2. REE Concentrations in the Roots of Scutellaria baicalensis*

The concentrations of REEs and several parameters on the REE distribution characteristics for roots of *Scutellaria baicalensis* in various habitats are listed in Table 3. The median was also utilized to represent the statistical characteristic of REE concentrations in the roots of various habitats. Roots of *Scutellaria baicalensis* contained similar ΣREE and ΣLREE in NGHs (ΣREE = 2.02 mg·kg−1, ΣLREE = 1.78 mg·kg−1) and BCHs (ΣREE = 2.04 mg·kg−1, ΣLREE = 1.82 mg·kg−1), which are higher than that in APHs (ΣREE = 1.78 mg·kg−1, ΣLREE = 1.60 mg·kg−1). However, roots in NGHs contained the highest HREE concentrations, with the median ΣHREE being 0.212 mg·kg−1, and roots in BCHs contained the lowest HREE concentrations, with the median ΣHREE being 0.183 mg·kg<sup>−</sup>1, which was exactly opposite to the ΣHREE in the soils. LREE/HREE and La*N*/Yb*<sup>N</sup>* decreased following the order of BCHs > APHs > NGHs, which was inconsistent with the order in the soils. The considerable variance in REEs adsorption by roots from the rhizosphere soil between various habitats can be speculated. The concentrations of REEs with even atomic numbers were higher than those with odd atomic numbers at HREEs, which conformed to the rule of Oddo-Harkins [24,25]. In general, the chondrite-normalized REE patterns presented an obvious right-inclined style with the enrichment of LREEs (Figure 3b,d,f). Positive Eu anomalies were observed for some samples in all habitats, indicating the adsorption of Eu from soil to some extent. However, the negative Ce anomaly of roots was stronger than that of soil, especially in NGHs, since Ce4<sup>+</sup> was hard to be adsorbed by roots.


**Table 3.** Parameters of the REE distribution characteristics of *Scutellaria baicalensis* in various habitats, Hebei Province.

## *4.3. Accumulation of REE in the Roots of Scutellaria baicalensis*

The biological absorption coefficient (BAC) of REEs provides an estimate of the individual availability of REEs to the plant [26–28]. This was adopted to quantify the natural process of element transfer from the soil to the roots of *Scutellaria baicalensis*. The BAC is defined as follows (3):

$$\text{BAC}\_{\text{i}} = \frac{\text{CM}\_{\text{i}}}{\text{CS}\_{\text{i}}},\tag{3}$$

where BAC*<sup>i</sup>* is the migration and accumulation rate of element i, and CM*<sup>i</sup>* and CS*<sup>i</sup>* are the concentrations of element i in the plant and soil, respectively. The results were categorized into five groups based on the magnitude of the coefficient: BAC > 3—very strongly accumulated; from 1.5 to 3.0—strongly accumulated; from 0.5 to 1.5—moderate absorption; from 0.1 to 0.5—weak absorption, and BAC < 0.1—very weak absorption [29]. Under normal circumstances, the ratio of the concentrations of REEs in plants to that in the soil is less than 1, even as low as 0.02 [30]. In the study area, all REEs in *Scutellaria baicalensis* had low BACs: less than 0.1.

One-way ANOVA and LSD tests were applied to analyze whether there were remarkable differences in the BACs of REEs between various habitats after the test for normal distribution and homogeneity of variance. In addition, permutation tests, t tests and nonparametric tests were utilized for the BACs that were not normally distributed or had heterogeneity of variance. The statistical analyses were conducted in the R program. In the box plots, various letters (e.g., a and b) imply that there exist statistically remarkable differences between BACs in various habitats (*p* < 0.05); otherwise, differences are not statistically remarkable(e.g., ab and a, ab and b). Almost all HREEs and Y, except for Gd and

Tb, showed statistically remarkable differences in BACs between NGHs with both BCHs and APHs. According to the median (black line in the box), the BACs of all REEs decreased in the order of NGHs > APHs > BCHs (Figure 5), suggesting higher utilization efficiency for REEs with natural source from bedrock, which was attributed to their self-adaption to the environment for a long time in a complicated open system [31]. On the contrary, the utilization efficiency for anthropogenic REEs was relatively low.

**Figure 5.** Box plot of biological absorption coefficient (BACs) of REEs Y from soils into roots. (**a**) La; (**b**) Ce; (**c**) Pr; (**d**) Nd; (**e**) Sm; (**f**) Eu; (**g**) Gd; (**h**) Tb; (**i**) Dy; (**j**) Ho; (**k**) Er; (**l**) Tm; (**m**) Yb; (**n**) Lu; (**o**) Y. \* *p* < 0.1.

## *4.4. Effective Constituent Content of Scutellaria baicalensis*

Flavonoid compounds are the dominant effective constituents of *Scutellaria baicalensis*, which are composed of two polyhydroxy phenolic benzene rings interconnected by three carbon atoms. Flavonoid compounds included baicalin, oroxylin A glycoside, wogonoside, baicalein, wogonin, and oroxylin A, in which baicalin was recognized as the main criterion of quality [32]. Contents of baicalin ranged from 12.8 to 51.7 mg·g−<sup>1</sup> with an average of 27.5 mg·g−<sup>1</sup> in the NGHs, which were higher than that in APHs (4.22–13.2 mg·g−1) and BCHs (4.04–17.1 mg·g<sup>−</sup>1). Almost all the flavonoid compounds were normally distributed (*p* > 0.05) except for oroxylin A in BCHs. According to the average, baicalin and wogonoside decreased in the order of NGHs > BCHs > APHs, and oroxylin A glycoside, baicalein, wogonin, and oroxylin A followed the order of NGHs > APHs > BCHs (Table 4).



**Table 4.** *Cont.*

## **5. Discussions**

## *5.1. Lithological Influences on the Enrichments and Fractionations of REEs in the Soils*

Micronutrient element concentrations in the soils developed in mountain terrain were primarily inherited from parent materials [16–18]. As mentioned, REE distribution patterns of soils developed in different habitats presented a genetic relationship with their corresponding parent materials. According to 1:250,000 reports of regional geologic survey and our field survey, the protolith of hornblende-gneiss in NGHs was femic volcanic rock, whose REEs concentrations were higher than those that developed in loess; therefore, soils developed from hornblende-gneiss contained higher REEs concentrations (ΣREE = 174 ug·g<sup>−</sup>1) than that from loess (ΣREE = 149 ug·g<sup>−</sup>1), which is in agreement with the expectations from the different lithologies that soils originating from volcanic rock tend to have higher REE concentrations than soils developed from loess [28,33,34].

Weathering crust in hornblende-gneiss showed great variation of REEs concentrations than loess, resulting in relatively homogeneous REEs concentrations in the soils developed from loess (CV = 6.84%) and notable heterogeneity of REEs concentrations in the soils developed from hornblende-gneiss (CV = 33.9%). The weathering of parent materials profoundly impacted the migrations and enrichments of REEs. The REE concentrations distribution pattern in soils of loess was significantly similar to their parent materials since their weak weathering and indistinctive movement of clay particles. However, the complex mineral compositions and grain sizes in hornblende-gneiss significantly influenced the weathering process, resulting in remarkable discrepancies in migrations and enrichments of REEs during weathering [35,36]. Two weathering profiles in hornblende-gneiss locations exhibited different evolution of REEs. In profile 1, REE concentrations decreased in the order of soils < bedrock < regoliths, while in profile 2, REE concentrations decreased followed the order regoliths < topsoil < bedrock (Figure 6a,b). It can be speculated that in profile 1, as eluviation proceeded, soluble components were greatly leached while sparingly soluble components, including REEs, showed relative enrichments, resulting in the overall REE enrichments of the regoliths [37,38]. REEs in the topsoil penetrated downward to the lower part of the profile, which was accompanied with REE-containing clay minerals [39,40], resulting in depletion in the topsoil compared with the lower parts. However, the minerals grain size in the bedrock of profile 2 was much smaller than that in profile 1. As the mineral grain size decreases and specific surface area increases, fine-grained rock offers a potentially increased area over which water penetration may occur [41,42]. Therefore, the reaction with water in the regolith and soils causes releasing of REE from REE-containing minerals and the leaching of REE, especially LREE. It was observed in the field that the humus horizon was 10 cm thick with a higher content of organic materials in profile 2; therefore, topsoil contained higher REE concentrations since organic matter has a fixed effect on REEs [43–45].

**Figure 6.** Bedrock-normalized REE patterns in profiles in NGH. (**a**) Profile 1; (**b**) Profile 2.

The clays and iron (hydro) oxide that formed during chemical weathering in hornblendegneiss locations were notably different, resulting in apparent different fractionations within REEs [46–50], since the mixed speciation of the REEs in clays and Fe-Mn oxyhydroxides is typical for REEs in soils [38,51–53]. Under natural pH, clay minerals with a negative layer charge are effective in the adsorption of REEs through ion exchange, surface complexation, electrostatic attraction and migration into clay structures [54]. The adsorption capacity is largely determined by the surface structure, surface charge of clay minerals and composition [55,56]. In our study area, under alkaline conditions (pH > 7) with abundant Fe2<sup>+</sup>, Mg2<sup>+</sup>, and K+, clay minerals were dominated by smectite, illite or chlorite, especially in profile 2 [46,50,57,58]. Smectite and illite may intrinsically retain HREE more efficiently than LREE [46,58]. As shown in Figure 7a, LREE/HREE had a remarkable positive relationship with SiO2/(Al2O3+Fe2O3) in profile 2 according to Pearson correlation analysis. As weathering proceeded, clay minerals were gradually generated from aluminosilicate minerals with the enrichment of Al2O3 and depletion of SiO2. In addition, enhanced oxidation contributed to the enrichment of Fe2O3. Therefore, the decrease of the SiO2/(Al2O3+Fe2O3) ratio showed the enrichment of clay mineral (smectite and illite) and iron (hydro) oxide, favoring the preferential adsorption of HREE and fractionation within REEs [59]. However, according to CIA, the chemical weathering in profile 1 (CIA = 65.0–71.1) was generally stronger than that in profile 2 (CIA = 51.1–67.2), and a small amount of kaolinite and boehmite with higher maturity was formed from illite/smectite [60]. Therefore, LREEs were preferentially adsorbed by kaolinite in profile 1. In addition, HREEs have a smaller ionic radius and stronger hydrolysis ability than LREEs and therefore have a higher affinity toward iron (hydro) oxide. Previous studies also showed that the chondrite-normalized iron (hydro) oxide fractions are slightly enriched in HREEs via inner-sphere complexation and may play an important role in redistributing HREEs in the weathering crust [61,62]. As shown in Figure 7b, LREE/HREE had a remarkable negative relationship with Fe2O3 in profile 2, suggesting the critical control of iron (hydro) oxide on the REE fractionation. In the study area, profile 2 contained more abundant Fe2O3 (4.92–9.49%, 7.95%) than those in profile 1 (Fe2O3 = 4.40–7.47%, 6.22% ). Therefore, in profile 1, HREEs are predominantly dissolved and migrate as bicarbonate and organic complexes in solution in case of the low contents of iron (hydro) oxide, leading to the fractionation of REEs among weathering crust [63–69]. However, the preferential scavenging of HREEs during the precipitation of pedogenetic iron (hydro) oxide resulted in the HREE enrichment in profile 2 instead of being dissolved and migrating as bicarbonate and organic complexes, which is consistent with the conclusion of Land (1999) [51].

**Figure 7.** Scatter diagram of (**a**) SiO2/(Al2O3+Fe2O3) versus LREE/HREE and (**b**) Fe2O3 versus LREE/HREE in profile 2. \*\* *p* < 0.01.

#### *5.2. Enrichments and Fractionations of REE in the Roots of Scutellaria baicalensis*

The soil mineral particles present in the close vicinity of the roots were the primary source of REEs for plants. It exhibited a remarkable positive correlation between ΣREE in the roots and rhizosphere soils with a coefficient of 0.479 (*p* < 0.01), especially the correlation coefficient between ΣLREE in the roots and rhizosphere soils being 0.511 (*p* < 0.01) (Figure 8). Previous research on the distribution of *Scutellaria baicalensis* demonstrated that they were naturally widespread in the fertile sand or loam underlying the humus horizon, especially chestnut soil or sandy loam, with pH values ranging from 5 to 8 [70]. For a type of species, the REEs concentrations in plants are influenced by some complex related factors, including total REEs concentrations and their occurrences in the rhizosphere soils, as well as pH, Eh, clay fraction contents, nutrient features, etc., in the soil–root environment [71–75]. In our study area, the clay fraction contents of soil developed in NGHs range from 18.8 to 66.9 g/kg, and the sand fraction contents range from 884.1 to 940.9 g/kg (our unpublished data), suggesting that soils were classified into sand according to texture classification. Their pH value varied from 7.31 to 8.57, with an average of 8.06. Therefore, NGHs were optimal for the inhabitation of *Scutellaria baicalensis*, and the absence of clay fraction favored the migrations of REEs into plants, resulting in higher biological absorption coefficients (BACs) of REEs. On the contrary, loess in APHs was generally deemed to comprise higher clay fractions than other lithologies. REEs may be preferentially involved in the crystal lattice of clay minerals as isomorphisms or hosted into REE-rich minerals, e.g., Ti oxides or phosphorite, and therefore inhibit the absorption of REEs into plants [37,76], leading to lower BACs of REEs. However, after rare earth micro-fertilizers application, BACs of all REEs in BCHs were still lower than that in APHs, despite both rhizosphere soils and roots in BCH comprised higher REE concentrations. Previous studies have demonstrated that once exogenous REEs enter the soil, more than 99.5% of them are absorbed by the solid phase of the soil, and only a small amount is dissolved in the water present in the soil [10,77]. This means that within a short period, REE fertilization would not notably increase the concentrations of REEs with high bioavailability [78–82], e.g., water-soluble fractions and iron-exchangeable fractions. In contrast, REEs accumulate in the soil as residue fractions, which are difficult for the plant to adsorb [83]. We can conclude that REE fertilization in BCHs contributes to improving REE concentrations in roots to some extent but that utilization efficiency for anthropogenic REEs was relatively low.

The BCAs of LREEs for *Scutellaria baicalensis* were overall higher than those of HREEs. Previous studies showed that the coprecipitation of rare earth ion salts (mostly in the form of insoluble oxalates or phosphates) and the selective absorption of root cell walls (in the form of trivalent cations) were the main mechanisms through which plant roots fix REEs [10,84,85]. In general, the dominant speciation of the LREEs is as free ions, whereas the HREEs are mainly present as dissolved complexes [38,86,87]. Diffusion through ion channels, i.e., passive diffusion, was the dominant movement mechanism of REEs from the soil into the roots [88–90], and LREEs were dominantly adsorbed into root cells in the form of trivalent cations. In this study area, the BACs distribution patterns of REEs in various

habitats exhibited an overall right-inclined style to different degrees, with a high biological absorption coefficient of Eu (Figure 9), indicating that *Scutellaria baicalensis* preferentially adsorbed LREEs and thus resulted in fractionation within REEs.

**Figure 8.** Scatter diagram of (**a**) ΣREE in the rhizosphere soils versus roots and (**b**) ΣLREE in the rhizosphere soils versus roots. \*\* *p* < 0.01.

**Figure 9.** BACs distribution patterns of REEs in roots of *Scutellaria baicalensis*. (**a**) NGHs; (**b**) APHs; (**c**) BCHs.

## *5.3. Relationships between REE and Effective Constituents of Scutellaria baicalensis*

Flavonoid compounds are effective constituents and secondary metabolites of *Scutellaria baicalensis*, which are controlled by the growing conditions, including illumination, temperature, moisture, and nutrients [91]. In our study area, flavonoid compounds were more abundant in NGHs than APHs and BCHs. Previous data have demonstrated that mineral nutrient elements (e.g., REE3+) control the synthesis and accumulation of flavonoid compounds, either as catalysts for secondary compound metabolism [92–96], by involving in their functional structure [97–100] or by contributing to the growth of cells [101–105]. Pearson correlation analysis was applied to check out further the correlations between micronutrient elements with flavonoid compounds in the roots. The results exhibited that micronutrient elements, including REEs, Cu, Zn, Sr, Ge, and Se, had a significant positive

correlation with flavonoid compounds to some extent (*p* < 0.05 or *p* < 0.01) (Figure 10), indicating that the micronutrient elements of roots have a significant influence on the medicinal composition of *Scutellaria baicalensis*.

**Figure 10.** Pearson correlation analysis between flavonoid compounds and micronutrient elements in the roots of *Scutellaria baicalensis*.

It was widely acknowledged that the activation of endocytosis is the primary response of plant cells to REEs [10,106–108], which induces a series of physiological and biochemical responses, affecting the activation of various enzymes, substance synthesis and cell growth [104,107,109,110]. Previous studies have demonstrated that exposure to REEs can affect the absorption of mineral elements by plants [111,112]. In our study area, roots of *Scutellaria baicalensis* in NGHs contained higher micronutrient elements concentrations than BCHs and APHs, e.g., Cu, Zn, Sr, Ge and Se (Table 5). Therefore, high REEs and micronutrient elements concentrations of hornblende-gneiss favored the synthesis and accumulation of flavonoid compounds in Scutellaria baicalensis after the activation of endocytosis induced by REEs [100,113]. In addition, trivalent lanthanum (La(III)) is similar to Ca2<sup>+</sup> in terms of properties and structures as mentioned above. Therefore, La(III) may substitute for Ca2<sup>+</sup> and present similar effects as Ca2<sup>+</sup> in biological systems [105,114,115]. For instance, La(III) may bind to Ca-binding sites in the CaM molecule by electrostatic attraction or coordination [106,116] based on the concentrations of REEs. Hence, the CAM expression level in plants and its molecular structure was therefore affected by REE concentrations, affecting the accumulation of effective constituents [117].

**Table 5.** Parameters on micronutrient elements distribution characteristics of roots in various habitats, Hebei Province.


As mentioned above, we can conclude that parent materials significantly impact the quality of Scutellaria baicalensis in terms of micronutrient elements involving REEs. Some strategies for management of the habitats of *Scutellaria baicalensis* are as follows:


## **6. Conclusions**

This study presented the influences of parent materials on the inhabitation and quality of top-geoherb *Scutellaria baicalensis* in Chengde City, Hebei province, in terms of REEs. It was greatly significant for protecting the origin and optimizing replanting patterns of *Scutellaria baicalensis*. The migrations, enrichments, fractionations and transformations of REEs in the bedrock–regolith–soil–root continuum were studied in three habitats of *Scutellaria baicalensis* with contrasting geopedological conditions. Parent materials and soils in the hornblende-gneiss locations contained higher REE concentrations than loess locations. REE concentrations in loess soils were relatively homogeneous, while various mineral compositions and mineral grain sizes of the hornblende-gneiss resulted in the heterogeneity of REEs concentration in rhizosphere soils, with a coefficient of variation (CV) being 33.9% as weathering proceeded. Weathering, involving eluviation, leaching, absorption, etc., influenced the migrations and enrichments of REEs in weathering crust in hornblendegneiss, and weathering productions, dominated by clay minerals and iron (hydro) oxide, controlled the fractionations within REEs. Roots of *Scutellaria baicalensis* contained similar ΣREE in NGHs (2.02 mg·kg−1) and BCHs (2.04 mg·kg−1), which are higher than that in APHs (1.78 mg·kg−1). It exhibited a remarkable positive correlation between REE concentrations in the roots and rhizosphere soils with a coefficient of 0.479 (*p* < 0.01). The biological absorption coefficients (BACs) of REEs for *Scutellaria baicalensis* decreased in the order of NGHs > APHs > BCHs. Soils developed in hornblende-gneiss were characterized by high REE concentrations, lower content of clay fraction and overall alkaline with a pH value of 8.06, favoring the inhabitation of Rehe *Scutellaria baicalensis* and adsorption for RREs. Micronutrient elements in the roots, e.g., REEs, Cu, Zn, Sr, Ge and Se, were remarkably correlated with flavonoid compound contents, suggesting their significant impact on the quality of *Scutellaria baicalensis*. The activation of endocytosis induced by REEs favored the adsorption of micronutrient elements and together improved the quality of *Scutellaria baicalensis*. Therefore, *Scutellaria baicalensis* in NGHs, featuring high REEs and other micronutrient elements concentrations, contained higher flavonoid compound content.

**Author Contributions:** Conceptualization, W.F. and Z.S.; methodology, Z.C. and Z.S.; software, T.A.; validation, L.X.; formal analysis, Z.S.; investigation, Z.C. and Z.S.; resources, H.Z.; data curation, L.X.; writing—original draft preparation, Z.S.; writing—review and editing, Z.S., W.F.; visualization, Z.S.; supervision, W.S. and H.Z.; project administration, Z.C. and H.Z. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the China Geological Survey Program (Grant No. DD20190822).

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

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** We would like to thank Wei Liu and Xiaoying Cui from Hebei Huakan Resource Environmental Survey Co., Ltd. for the test of REEs and flavonoid compounds.We thank Xia Li form China Institute of Geo-Environment Monitoring for financial support and Shoulin Zhang from Beijing Institute of Geology for Mineral Resources CO., LTD for valuable suggestions. We also thank editor and three anonymous reviewers for their valuable comments and suggestions

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

## **References**


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