*Article* **Effects of Landscape Positions and Landscape Types on Soil Properties and Chlorophyll Content of Citrus in a Sloping Orchard in the Three Gorges Reservoir Area, China**

**Siyue Sun <sup>1</sup> , Guolin Zhang <sup>2</sup> , Tieguang He 3,\*, Shufang Song <sup>4</sup> and Xingbiao Chu 1,\***


**Abstract:** In recent years, soil degradation and decreasing orchard productivity in the sloping orchards of the Three Gorges Reservoir Area of China have received considerable attention both inside and outside the country. More studies pay attention to the effects of topography on soil property changes, but less research is conducted from the landscape. Therefore, understanding the effects of landscape positions and landscape types on soil properties and chlorophyll content of citrus in a sloping orchard is of great significance in this area. Our results showed that landscape positions and types had a significant effect on the soil properties and chlorophyll content of citrus. The lowest soil nutrient content was detected in the upper slope position and sloping land, while the highest exists at the footslope and terraces. The chlorophyll content of citrus in the middle and upper landscape position was significantly higher than the footslope. The redundancy analysis showed that the first two ordination axes together accounted for 81.32% of the total variation, which could be explained by the changes of soil total nitrogen, total phosphorus, total potassium, available nitrogen, available potassium, organic matter, pH, and chlorophyll content of the citrus. Overall, this study indicates the significant influence of landscape positions and types on soil properties and chlorophyll content of citrus. Further, this study provides a reference for the determination of targeted land management measures and orchard landscape design so that the soil quality and orchard yield can be improved, and finally, the sustainable development of agriculture and ecology can be realized.

**Keywords:** agriculture landscape; chlorophyll content of citrus; landscape position; soil properties; terraces

## **1. Introduction**

The Three Gorges Reservoir Area of China is one of the most suitable ecological areas for the growth of citrus. The citrus industry has achieved a dominant position in the development of mountainous agriculture and rural economy in this area [1]. Currently, the mountainous agriculture in this area is transforming from conventional farming systems to suburban modern agriculture and leisure and sightseeing agriculture, and the ecosystem service function contained in it has begun to attract attention [2]. Citruses are planted on the sloping land along the Yangtze River and its tributaries because of cultivated land tension in this area. For a long time, orchard managers have been ignoring soil conservation practices, and indiscriminately using large amounts of fertilizers without considering soil differences, which has led to serious problems such as chemical fertilizer pollution, soil degradation, and orchard production reduction, and brought great harm to agricultural production and ecological environment in the area [3–5].

**Citation:** Sun, S.; Zhang, G.; He, T.; Song, S.; Chu, X. Effects of Landscape Positions and Landscape Types on Soil Properties and Chlorophyll Content of Citrus in a Sloping Orchard in the Three Gorges Reservoir Area, China. *Sustainability* **2021**, *13*, 4288. https://doi.org/ 10.3390/su13084288

Academic Editors: Bharat Sharma Acharya and Rajan Ghimire

Received: 11 March 2021 Accepted: 6 April 2021 Published: 12 April 2021

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

**Copyright:** © 2021 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/).

Soil properties, the most important factor determining soil quality, not only affect crop output, but also have a significant impact on the cultivated land use and soil environmental protection [6–8]. Several previous studies have confirmed the effects of terrain, land use, hedgerows, and other environmental factors and management measures on soil quality in the Three Gorges Reservoir Region [9]. Teng Mingjun et al. [10] found that topographical factors are the main factors that cause the spatial heterogeneity of soil organic carbon in the reservoir area. Shen Zhenyao et al. [11] found that load intensities of nitrogen, phosphorus, and other non-point source pollutants were significantly different in soils with different land-use patterns. Xu Feng et al. [12] showed that slope ecological engineering with contour hedgerows could effectively control slope erosion and nutrient loss.

However, in the Three Gorges Reservoir Region, a fragile ecoregion and a developing leisure and sightseeing agricultural area, there are a few studies on the change of soil properties caused by the landscape. Each soil property has a respective spatial distribution in the landscape. The landscape position affects the process of soil formation, so it is considered to be one of the key factors affecting the changes of soil properties [13,14]. Simultaneously, landscape types also cause changes in soil spatial distribution. Arnaz's [15] study found that when sloping land transformed into terraced land, the slope's length and angle decreased significantly, resulting in a decrease in soil erosion and sediment yield. Although soil properties are influenced by many factors, such as climate, parent material, and biological factors, the influence of landscape types and landscape positions cannot be ignored on the regional scale [16]. Therefore, in this area, soil-landscape analysis is crucial to understand the spatial variation law of soil properties for determining targeted land management interventions to improve soil quality, form a charming farmland landscape, and achieve sustainable agricultural development [17].

Additionally, more researchers believe that it is best to combine soil analysis with leaf analysis to comprehensively diagnose the soil quality and citrus nutritional status to guide rational fertilization, improve citrus quality, and increase citrus yield. Mohesh et al. [18,19] Showed that the chlorophyll content of plant leaves determines the photosynthetic capacity and nutritional status of leaves, which can be used as an indicator of plant health. Haboudane et al. [20–22] point out that chlorophyll content in crops plays a key in precision agriculture because it is related to nitrogen concentration in the leaf of a crop. It reflects how the crop responds to nitrogen application, as well as being an important indicator of photosynthetic activity, which determines crop yield. Therefore, it is necessary to assess the comprehensive impact of environmental factors on soil and citrus trees in combination with the changes of the chlorophyll content of citrus (CCC).

The purposes of this study are: (1) to evaluate the effects of landscape position and landscape type on soil properties, including soil total nitrogen, total phosphorus, total potassium, available potassium, available nitrogen, available phosphorus, soil organic matter, and pH; and (2) to evaluate the effects of landscape position and landscape type on CCC. This information provides a reference for knowing how the local ecosystem works and assessing the impact of future landscape changes. It is not only helpful for the determination of targeted orchard land management measures and the formation of a good agricultural landscape but also related to the ecological environment safety and sustainable development of the agricultural economy in the middle and upper reaches of the Yangtze River.

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

#### *2.1. Study Area*

The study area is in the citrus orchard in Guo Jiagou, Fengjie County, Chongqing, China (31◦060 N, 109◦270 E, Figure 1). Fengjie County, with an altitude ranging from 86 to 2123 m above sea level, is a mountainous landform in the eastern Sichuan Basin and the mountainous area accounts for 88.3% of the total (Figure 1D). The Yangtze River runs through the middle of Fengjie County, stretching 41.5 km, with Mei Xi River, Da Xi River, Shi Sun River, Cao Tang River, Zhu Yi River, and other rivers. Fengjie County is

a typical subtropical monsoon climate with four distinct seasons, abundant rainfall, and long sunshine hours. Due to the influence of topography and landform, the vertical change of climate is more obvious and forms a typical three-dimensional climate. The frost-free period is about 287 days; average annual temperature is 16.3◦ , average precipitation is about 1150 mm, and average sunshine duration is 1639 h. subtropical monsoon climate with four distinct seasons, abundant rainfall, and long sunshine hours. Due to the influence of topography and landform, the vertical change of climate is more obvious and forms a typical three-dimensional climate. The frost-free period is about 287 days; average annual temperature is 16.3°, average precipitation is about 1150 mm, and average sunshine duration is 1639 h.

2123 m above sea level, is a mountainous landform in the eastern Sichuan Basin and the mountainous area accounts for 88.3% of the total (Figure 1D). The Yangtze River runs through the middle of Fengjie County, stretching 41.5 km, with Mei Xi River, Da Xi River, Shi Sun River, Cao Tang River, Zhu Yi River, and other rivers. Fengjie County is a typical

*Sustainability* **2021**, *13*, 4288 3 of 14

**Figure 1.** (**A**) The geographical location of the Yangtze watershed in China. (**B**) The land use map of Chongqing province. (**C**) The soil distribution map of study area. (**D**) The geographical structure of study area. (**E**) The geomorphic profile of the study site. **Figure 1.** (**A**) The geographical location of the Yangtze watershed in China. (**B**) The land use map of Chongqing province. (**C**) The soil distribution map of study area. (**D**) The geographical structure of study area. (**E**) The geomorphic profile of the study site.

Fengjie County is located in the core area of the citrus industrial belt in the middle and upper reaches of the Yangtze River. As of March 2021, the citrus planting area of Fengjie County has reached 246.7 km2, the output has reached 370,000 tons, and the comprehensive output value has exceeded 3.5 billion yuan, accounting for about 20% of the total agricultural output value. This has led to the income increase of 0.3 million people in 24 towns and 70,000 households, and the employed population accounts for 28.4% of the total population of Fengjie County. Citrus trees in the study area were planted in 1980; the variety is Feng Yuan 72–1, the row spacing of citrus plants is 4 m × 4 m, and the height of the canopy is 3–4 m. In 1990, part of the sloping land was changed to contour terraces, and the slope of the surface was 5 degrees, showing a trend of high in-Fengjie County is located in the core area of the citrus industrial belt in the middle and upper reaches of the Yangtze River. As of March 2021, the citrus planting area of Fengjie County has reached 246.7 km<sup>2</sup> , the output has reached 370,000 tons, and the comprehensive output value has exceeded 3.5 billion yuan, accounting for about 20% of the total agricultural output value. This has led to the income increase of 0.3 million people in 24 towns and 70,000 households, and the employed population accounts for 28.4% of the total population of Fengjie County. Citrus trees in the study area were planted in 1980; the variety is Feng Yuan 72–1, the row spacing of citrus plants is 4 m × 4 m, and the height of the canopy is 3–4 m. In 1990, part of the sloping land was changed to contour terraces, and the slope of the surface was 5 degrees, showing a trend of high inside and low outside. At present, in the study area the most agricultural landscape is sloping landscape, with a few being terraced landscape.

The soil distribution in Fengjie County is shown in Figure 1C and Table 1. The soil in the study area is mainly yellow soil, and the profile configuration is A–B–C type. The thickness of the soil layer is generally more than 60 cm, and the color is yellow or brownish yellow. The content of silt is 36.9–44.38%, clay is 30%, and the texture is loamy clay. This kind of soil has a deep soil layer, heavy texture, total porosity of about 59%, and strong water and fertility conservation.


**Table 1.** Information about the geological structure, soils, and land use of the analyzed catchment.

The data of soil and topography is from Fengjie County. The data of land use is from Chongqing.

#### *2.2. Experimental Design*

This study was conducted in August 2020 in the citrus orchard operated by Fengjie County Agriculture Development Ecology Co., Ltd., Chongqing, China. Two adjacent slopes with different landscape types were chosen to explore the effects of landscape positions and landscape types on soil properties and CCC in the study area. One of the slopes is a sloping landscape, the other is a contour terrace in the upper slope position, and the middle slope and footslope positions are sloping landscapes. Refer to the research methods of Brubaker et al. to divide slope landscape position [23]. The research is divided into three parts (Figure 1E): the upper landscape position (US), the middle landscape position (MS), and the foot of slope (FS). Landscape types are divided into sloping landscape and terraced landscape. A total of representative 24 plots (4 m × 4 m) were selected based on landscape positions and landscape types: 8 in the US, 8 in the MS, and 8 in the FS. The landscape position, landscape type, and geographical location of each sampling plot were recorded, and slope gradient, slope aspect, and elevation were measured. The basic information of sample plots in different landscape positions is shown in Table 2.



FS = the footslope position, MS = the middle slope position, US = the upper slope position.

#### *2.3. Soil Sampling and Chlorophyll Content of Citrus*

In August 2020, soil samples were collected for 24 plots by the diagonal sampling method. After clearing the top litter, three individual soil samples (0–15 cm, one from the center of the field, two from diagonal corners) were taken from a plot and mixed to get 1 sample. The samples were air-dried and sent to the Guangxi Academy of Agricultural Sciences to determine the contents of soil total nitrogen, total phosphorus, total potassium, available nitrogen, available phosphorus, available potassium, organic matter, and pH.

The CCC, expressed as a chlorophyll content index (CCI) was measured using the CCM-200 plus Chlorophyll Content Meter (OPTI–SCIENCES, Hudson, NH, USA) between 8:00 and 10:00 a.m. on 28 and 29 August 2020 (sunny cloudless). The measurement method was to select well-developed and fully developed leaves from the upper, middle, and lower locations of the south of a citrus tree. After measuring each leaf four times, the mean value of chlorophyll content in the upper, middle, and lower tree locations was taken.

#### *2.4. Data Analysis*

#### 2.4.1. Statistics Analysis

Statistical analyses were conducted using Excel 2016 (Microsoft, Redmond, WA, USA) and SPSS 24.0 (IBM, Armonk, NY, USA). A descriptive statistic was performed to describe the soil properties and CCC. Then, Pearson correlation analysis was used to show correlations between soil properties. One-way ANOVA was used to examine the effects of landscape position and landscape type on soil properties and CCC, and the least significant difference method (LSD) was used to compare the mean values.

#### 2.4.2. Redundancy Analysis Method

To find the most important environmental factors that affect the soil properties and CCC in the experimental area, redundancy analysis (RDA) was carried for a constrained ranking analysis based on the experimental data. A Monte Carlo permutation test was used to find the relative importance of each environmental factors in explaining changes in soil properties. Redundancy analysis is a direct gradient analysis technique. Through community ranking, the community sample plots (species) investigated in an area were arranged according to their similarity to analyze the relationships between various species and the environment and effectively evaluate the impact of environmental variables on species [24]. In this research, RDA was applied to find the relationship between species variables (soil properties and CCC) and environmental variables (landscape positions and landscape types, slope surface, gradient, and aspect).

It is worth noting that the gradient length should be measured by detrended correspondence analysis (DCA) on the sample before constraint analysis. Since the first gradient length was 0.6 < 3.0, RDA is the most appropriate method [25]. Before RDA analysis, two data matrices (species data and environmental data) were built, and the environmental data were encoded. In this study, landscape types were divided into two types: 1 represents the sloping landscape and 2 the terraced landscape. Landscape position can be divided into FS, MS, and US, represented by 1, 2, and 3, respectively. The experimental slope surface was divided into the terraced field and sloping field, represented by 1 and 2, respectively. The actual measured values were used to represent the slope and aspect.

#### **3. Results and Discussion**

#### *3.1. The Description of Soil Properties and CCC*

Table 3 shows the statistical variables of soil properties and CCC under different landscape positions and landscape types. The content ranges of soil properties and CCC were as follows: CCC: 66.03–163.40, total nitrogen: 0.55–2.70 g/kg, total phosphorus: 0.19–1.75 g/kg, total potassium: 11.76–35.11 g/kg, available nitrogen: 30.60–185.50 mg/kg, available phosphorus: 0.30–50.70 mg/kg, available potassium: 50.50–273.40 mg/kg, soil organic matter: 9.30–51.00 g/kg, and soil pH: 5.73–7.54. The average content of soil properties followed a decreasing order: available potassium > available nitrogen > available phosphorus > soil organic matter > total potassium > total nitrogen > total phosphorus. The coefficient of variation (CV) of soil properties and CCC ranged from 27.60% to 83.82% (Table 3), showing medium variation (10% ≤ CV ≤ 100%), indicating that the soil properties and CCC varied greatly in different landscape positions and types.


**Table 3.** Descriptive statistics for soil properties and the chlorophyll content of citrus (CCC).

SD = standard deviation, CV = coefficient of variation, TN = total nitrogen, TP = total phosphorus, TK = total potassium, AN = available nitrogen, AP = available phosphorus, AK = available potassium, SOM = soil organic matter, CCC = chlorophyll content of citrus.

#### *3.2. Soil Properties Correlations*

A significant correlation was found between many soil properties measurements (Table 4). The content of total nitrogen (TN) was significantly positively correlated with total phosphorus (TP), available nitrogen (AN), available phosphorus (AP), available potassium (AK), and soil organic matter (SOM). SOM was significantly positively correlated with AN, TP, AP (*p* < 0.01), and AK (*p* < 0.05). The increasing SOM can promote the activity of soil enzymes, and promote the decomposition of plant and animal residues and humus, thus releasing nitrogen and potassium [26]. Moreover, SOM has a significant promoting effect on soil nutrient retention and nutrient supply capacity, which also suggests that the future fertilization should be based on organic fertilizer, supplemented by fertilizer, to maintain anthropogenic mellowing of soil, to meet the nutritional requirements of stable yield, high yield, and good quality of citrus. Soil pH was negatively correlated with TN, AN, AK, and SOM, and it was an important index reflecting soil parent material properties and weathering and leaching conditions [27].



\* = 0.01 < *p* < 0.05, \*\* = 0.001 < *p* < 0.01.

#### *3.3. The Effect of Landscape Positions on Soil Properties and CCC*

Most soil properties and CCC were significantly different in landscape positions (Table 5). The results of multiple comparisons showed that the CCC among the three landscape positions showed the order of MS > US > FS, and the CCC at the FS was significantly different from the US and MS (*p* ≤ 0.001, Figure 2). Vladimir et al. [28] showed that under the same photosynthetic photon flux density, the fluorescence intensity excited by blue light is 2.5 to 3 times that of red light. Blue light significantly promoted the formation and accumulation of chlorophyll [29,30]. In the mountainous environment, the landscape position directly affects the illumination conditions, and the solar radiation at the FS is lower than the MS and US. Therefore, the CCC at the US and MS is extremely significantly higher than that the FS. At the same time, sunlight may also affect the formation of local microclimates, and differences in microclimates can affect the distribution of plant communities, which in turn affects the growth and development of citrus trees [31]. Additionally, a study by Qiang Fu et al. [32,33] showed that in the soil at a depth of 0–20 cm, the higher the landscape position, the greater the soil moisture content, and chlorophyll content is positively related to soil moisture. Therefore, the significant differences of CCC in landscape positions may be caused by many reasons.

**LS CCC**

**(g/kg)**

**TP (g/kg)**

**Table 4.** Pearson correlation coefficients between soil properties.

AP 0.744 \*\* 0.876 \*\* 0.089 0.737 \*\*

*3.3. The Effect of Landscape Positions on Soil Properties and CCC*

AK 0.607 \*\* 0.458 \* −0.094 0.660 \*\* 0.653 \*\* SOM 0.954 \*\* 0.699 \*\* −0.120 0.876 \*\* 0.744 \*\* 0.475 \*

ferences of CCC in landscape positions may be caused by many reasons.

**AN (mg/kg)**

**Soil Properties TN PH**

**AP (mg/kg)**

**AK (mg/kg)**

**SOM (g/kg)**

**Table 5.** Mean comparisons of soil properties by landscape positions of the study site.

FS 85.80 1.62 1.06 17.25 94.93 27.25 204.30 25.40 6.10 MS 150.88 1.53 0.91 16.96 98.08 19.85 244.40 21.73 6.51 US 141.96 0.70 0.40 14.64 33.25 10.18 121.85 11.05 7.27

\* =  0.01  <  *p*  <  0.05, \*\* =  0.001  < *p* <  0.01, \*\*\* =  *p* <  0.001. Abbreviations as in Table 3.

**TK (g/kg)**

TP 0.687 \*\*

TK −0.058 0.096

\* =  0.01  < *p*  <  0.05, \*\* =  0.001  < *p*  <  0.01.

AN 0.930 \*\* 0.584 \*\* −0.075

**TN TP TK AN AP AK SOM**

pH −0.537 \*\* −0.297 0.089 −0.582 \*\* 0.539 \* −0.611 \*\* −0.420 \*

Most soil properties and CCC were significantly different in landscape positions (Table 5). The results of multiple comparisons showed that the CCC among the three landscape positions showed the order of MS > US > FS, and the CCC at the FS was significantly different from the US and MS (*p* ≤ 0.001, Figure 2). Vladimir et al. [28] showed that under the same photosynthetic photon flux density, the fluorescence intensity excited by blue light is 2.5 to 3 times that of red light. Blue light significantly promoted the formation and accumulation of chlorophyll [29,30]. In the mountainous environment, the landscape position directly affects the illumination conditions, and the solar radiation at the FS is lower than the MS and US. Therefore, the CCC at the US and MS is extremely significantly higher than that the FS. At the same time, sunlight may also affect the formation of local microclimates, and differences in microclimates can affect the distribution of plant communities, which in turn affects the growth and development of citrus trees [31]. Additionally, a study by Qiang Fu et al. [32,33] showed that in the soil at a depth of 0–20 cm, the higher the landscape position, the greater the soil moisture content, and chlorophyll content is positively related to soil moisture. Therefore, the significant dif-

**Figure 2.** Variations of soil properties and CCC in different landscape positions. (**a**) Changes of citrus chlorophyll content; (**b**) Changes of total nitrogen content; (**c**) Changes of total phosphorus content; (**d**) Changes of total potassium content; (**e**) Changes of available nitrogen content; (**f**) Changes of available phosphorus content; (**g**) Changes of available potassium content; (**h**) Changes of soil organic matter content; (**i**) Changes of pH. **Figure 2.** Variations of soil properties and CCC in different landscape positions. (**a**) Changes of citrus chlorophyll content; (**b**) Changes of total nitrogen content; (**c**) Changes of total phosphorus content; (**d**) Changes of total potassium content; (**e**) Changes of available nitrogen content; (**f**) Changes of available phosphorus content; (**g**) Changes of available potassium content; (**h**) Changes of soil organic matter content; (**i**) Changes of pH.

(**i**)


**Table 5.** Mean comparisons of soil properties by landscape positions of the study site.

\* = 0.01 < *p* < 0.05, \*\* = 0.001 < *p* < 0.01, \*\*\* = *p* < 0.001. Abbreviations as in Table 3.

The soil pH value is extremely significantly correlated with the landscape position (*p* < 0.01), manifested as a decrease along the downslope. Due to the citrus orchard being located on the slope of the Three Gorges Reservoir area, the topography has an impact on soil nutrient loss. The leaching or loss of base ions, especially calcium ions, makes the soil acidified [34]. On the contrary, the TN and TK accumulate along the downslope. The multiple comparison results showed that the contents of total nitrogen and total potassium were the lowest in the US position and tended to be the highest at FS position. The content in the US is significantly lower than the FS (*p* < 0.05). This result is consistent with the research conclusion of Hu Chenxia et al. and may be due to the nutrient deposition in the US and the production and residue of plants [35]. The contents of AN and AK were the highest content in the MS, and significantly higher than those in the US (*p* < 0.05, *p* < 0.01, respectively). It could be due to better temperature and moisture in the MS than in the US, or better light conditions than at the FS [36]. Although the difference of TP, AP, and SOM among landscape positions was not significant (*p* > 0.05), they generally tended to accumulate along the downslope, which was consistent with the fact that TN, AN, TK, and AK had lower values on the US and higher values at the FS. Zhang Jianhui et al. [37] showed that the main erosion process in a medium-long slope (40–110 m) was water erosion, and soil nutrients were carried from the upper slope to the foot slope by surface runoff, leading to a decrease in soil quality in the US position.

In general, most soil properties and CCC in orchards have significant correlations among landscape positions. Specifically, most soil properties show a higher value in the MS and FS and lower in the US [38]. The lowest levels of other soil properties, such as soil organic carbon, were also usually found in the US position [39]. The soil properties of the US were worse than those of the MS and FS position. However, unsustainable land use in the upper landscape position has an impact on the lower slope area. The soil nutrient loss is the main reason for the decline of soil quality and non-point source pollution. Therefore, the balance and improvement of soil quality is the critical factor to achieve sustainable agricultural development [40–42]. Xue Zhijing et al. [43] suggested that the high vegetation cover can reduce runoff and soil loss, and then maintain soil nutrients better. Additionally, the effects of tillage erosion cannot be ignored. One solution is to reduce tillage on uphill sites. The other solution is more appropriate to this area to improve the soil fertility by intercropping and applying green manure or organic manure in the US. Applying biological organic fertilizer instead of chemical fertilizer can reduce the emission of nitrogen dioxide and thus improve the acidification and eutrophication of surface water [44]. Therefore, the effects of landscape positions on soil properties and CCC should be further studied.

#### *3.4. The Effect of Landscape Types on Soil Properties and CCC*

Some soil properties of different landscape types in the same landscape position (US) in the orchard were significantly different (Table 6). The mean value of TN and TK ranged from 0.70–1.08 g/kg and 14.64–17.68 g/kg, respectively. Multiple comparison results showed that the TN and TK contents in terraced fields were significantly higher than in sloping fields (*p* < 0.05). The contents of AN, AP, and other nutrients that could be directly absorbed by crops in terraced fields were significantly higher than those in sloping (*p* ≤ 0.001), and the contents of AN and AP in terraced fields were 77.08% and 217.58% higher than those in sloping, respectively. Although there was no significant difference in TP, AK, SOM, and pH among different landscape types (*p* < 0.05), further observation showed that their contents in terraced fields were still higher than those in sloping, indicating that the difference of landscape types did lead to the changes of soil properties.

The research conclusion is consistent with Fu Bojie et al. [45]. The soil properties of sloping land being lower than terraces may be the combined action of water erosion and tillage erosion. Xu Chang et al. [46,47] showed that rainfall and surface runoff were the impetus of soil nutrient loss. Compared with the sloping field, the slope length and angle of the terraced orchard were significantly reduced, which resulted in the reduction of soil erosion caused by the topography and the better preservation of soil nutrients. At the same time, it reduces water body pollution caused by soil particles, nitrogen, phosphorus, and other elements in farmland under the scouring effect of rainwater and runoff. Furthermore, effects of contour tillage on soil movement (translocation and erosion) were examined by Zhang Jianhui et al. in the steep hillslopes of the Sichuan basin using a physical tracer method. The results showed that tillage significantly affects soil migration on sloping land. The tillage erosion rate under contour tillage was 77% lower than that under downslope tillage [48]. In the future farming of sloping, especially at the top and upper slopes with high soil property variability, it is necessary to consider the cultivation method of contour terraced fields, which is conducive to soil conservation and has a charming farmland landscape [49]. Bo Sun et al. [50,51] indicated that excessive application of nitrogen fertilizer and pesticides is considered to cause water pollution in the Yangtze River basin. The solution is to reduce the loss of soil nutrients and maintain a good farmland ecosystem. Therefore, compared with slope land, terrace farming with good ecological, landscape, and economic benefits is a more suitable farming method for farmland in the Yangtze River Basin. At that time, the charming terraced orchard and the beautiful Yangtze River will become a brilliant and unique scene, which will make the leisure and sightseeing agriculture in this region develop better. Moreover, Shimbahri et al. [52] conducted a study on the effect of terraces on soil water content in an arid area of Ethiopia. The results showed that terraces do have good performance in soil and water conservation. Thus, we should also pay attention to the impact of landscape types on soil water content in future research, which is very important for arid and semi-arid regions in the world.


**Table 6.** Mean comparisons of soil properties by landscape types.

\* = 0.01 < *p* < 0.05, \*\*\* = *p* < 0.001.

#### *3.5. Redundancy Analysis (RDA)*

Through redundancy analysis, we obtained the relationship among soil properties and CCC and environmental variables. The significance of the constraint ordination was tested by the Monte Carlo permutation test (499 permutations were performed). The results showed that the tests on the first and second constraint axes were obvious (*p* = 0.028 and 0.002, respectively). The first and second constraint axes together explain 81.32% of the relationship between the species variables and environmental variables. Therefore, we chose the first two constraint axes with high and significant eigenvalues to draw a biplot for observation and then tried to explain it.

The RDA ordination diagram is explained below: environmental variables (explanatory variables) in red arrows indicate species variables (response variables) with blue. The angle between the arrow of the environmental variables and response variables reflects

the correlation (but not the meaning of the angle between the response variable): when the angle is acute, the correlation is positive; correlation is negative when the angle is greater than 90 degrees. The length of the line between the red arrow (environmental variables) and the origin is directly proportional to the degree of correlation between an environmental factor and the distribution of community and species. The angle between the arrow of the environmental variable and the constraint axis represents correlation. The smaller the angle is, the greater the correlation will be. If it is orthogonal, it will be irrelevant. The blue arrow (species variable) points from the origin to the corresponding coordinate of the species score. The direction the arrow points to indicates the direction in which the abundance of the species has increased. The correlation between species and environmental variables was displayed by a perpendicular projection of the species arrow-tips onto the line overlaying the environmental arrow. The longer the projection, the higher the correlation. *Sustainability* **2021**, *13*, 4288 11 of 14 The smaller the angle is, the greater the correlation will be. If it is orthogonal, it will be irrelevant. The blue arrow (species variable) points from the origin to the corresponding coordinate of the species score. The direction the arrow points to indicates the direction in which the abundance of the species has increased. The correlation between species and environmental variables was displayed by a perpendicular projection of the species arrow-tips onto the line overlaying the environmental arrow. The longer the projection, the higher the correlation.

The results of the RDA ordination diagram (Figure 3) show that Axis1 is positively correlated with slope gradient and aspect. The first constraint axis is mainly interpreted as slope gradient and aspect because of the length of the arrow. The second ordination axis (Axis2) has a great negative correlation with landscape position. Therefore, the second ordination axis is mainly interpreted as landscape position. The CCC is highly correlated with landscape position, followed by slope gradient, and negatively with aspect. The highest negative correlation with landscape positions is TN, followed by TK, AN, SOM, and TP. Although most soil properties are significantly affected by landscape positions, the influence of slope and aspect cannot be ignored. The results of the RDA ordination diagram (Figure 3) show that Axis1 is positively correlated with slope gradient and aspect. The first constraint axis is mainly interpreted as slope gradient and aspect because of the length of the arrow. The second ordination axis (Axis2) has a great negative correlation with landscape position. Therefore, the second ordination axis is mainly interpreted as landscape position. The CCC is highly correlated with landscape position, followed by slope gradient, and negatively with aspect. The highest negative correlation with landscape positions is TN, followed by TK, AN, SOM, and TP. Although most soil properties are significantly affected by landscape positions, the influence of slope and aspect cannot be ignored.

In the RDA ordination diagram, slope gradient and aspect are the main determinants of Axis 1 (Figure 3). TN, TP, AN, AK, and SOM are positively correlated with slope gradient and negatively with aspect, which is in keeping with the research conclusions of Holden et al. [53]. Although the heterogeneity of soil properties is influenced by many factors, such as climate, soil parent material, and biology, many soil properties changes can be attributed to topography [54]. The movement and accumulation of soil solutions are significantly affected by slope gradient and aspect, leading to spatial differences in soil properties [55]. In the RDA ordination diagram, slope gradient and aspect are the main determinants of Axis 1 (Figure 3). TN, TP, AN, AK, and SOM are positively correlated with slope gradient and negatively with aspect, which is in keeping with the research conclusions of Holden et al. [53]. Although the heterogeneity of soil properties is influenced by many factors, such as climate, soil parent material, and biology, many soil properties changes can be attributed to topography [54]. The movement and accumulation of soil solutions are significantly affected by slope gradient and aspect, leading to spatial differences in soil properties [55].

**Figure 3.** Ordination biplot of redundancy analysis (RDA) displaying the effects of the selected environmental variables on soil properties and CCC. LT = landscape type, LP = landscape position, **Figure 3.** Ordination biplot of redundancy analysis (RDA) displaying the effects of the selected environmental variables on soil properties and CCC. LT = landscape type, LP = landscape position, SG = slope gradient, SA = slope aspect, SS = slope surface, CCC = chlorophyll content of citrus.

SG = slope gradient, SA = slope aspect, SS = slope surface, CCC = chlorophyll content of citrus. The correlation between CCC and slope gradient and aspect may be related to solar radiation. The slope aspect, a topographic factor that changes regional microclimate, determines the solar radiation amount received by the slope surface [56]. Smith's [57] re-The correlation between CCC and slope gradient and aspect may be related to solar radiation. The slope aspect, a topographic factor that changes regional microclimate, determines the solar radiation amount received by the slope surface [56]. Smith's [57] research shows that many environmental changes are related to solar radiation. Sun Ying et al.

search shows that many environmental changes are related to solar radiation. Sun Ying et

amount of solar radiation probably changes accordingly, which could lead to the CCC

changes.

showed that the chlorophyll content in plants also had a relationship with light intensity [58,59]. Therefore, in this study, with the change of slope gradient and aspect, the amount of solar radiation probably changes accordingly, which could lead to the CCC changes.

#### **4. Conclusions**

In conclusion, our study clearly shows significant changes in soil properties and CCC among different landscape positions and landscape types in the Three Gorges Reservoir Area. Most soil properties showed the highest content in the footslope and terraced landscape and the lowest in the upper slope and sloping. The CCC in the footslope was significantly less than in the MS and US location. In addition to the strong effect of landscape, the well-known principle that spatial heterogeneity of soil properties is affected by topographic factors such as slope gradient and aspect was also confirmed in this research. These results indicated that the changes of soil properties and CCC in this area were mainly affected by landscape position, landscape type, and topography. For Fengjie County to develop suburban modern agriculture and sightseeing agriculture, determining the targeted land management measures of orchards to change the farmland landscape and orchard planting layout is more in line with the requirements of regional development [60]. It can not only improve the soil quality and yield of orchards and reduce the unnecessary nutrient waste and non-point source pollution caused by orchards, but also provide ideas for the landscape design of orchards to realize the sustainable development of agriculture and ecology in this area.

It is necessary to carry out similar and larger scale research in other catchments such as the Yellow River basin or citrus planting areas such as Southern Jiangxi, China to determine the complex influence law between soil–landscape–crop under different soil parent materials, climate, geographical conditions, etc., and formulate the applicable regional eco-agricultural transformation scheme. In addition, this study has the same reference value for the development of sustainable agriculture in other tropical and subtropical countries such as India and Nigeria. However, our research method has some limitations, such as the research on the spatial distribution of soil properties being weak. Therefore, we summarized several lessons for future researchers to conduct further study. 1. More soil physical properties, such as soil bulk density, aeration, permeability, and adhesion, can be included in soil analysis. 2. More response variables can be included in leaf analysis, such as chlorophyll a, chlorophyll b, carotenoids, leaf area, specific leaf index, etc.

**Author Contributions:** Conceptualization and methodology, G.Z. and S.S. (Siyue Sun); software, validation, investigation and formal analysis, S.S. (Siyue Sun) and S.S. (Shufang Song); resources and data curation, T.H.; writing—original draft preparation, S.S. (Siyue Sun); writing—review and editing, and supervision, G.Z. and X.C.; visualization, S.S. (Siyue Sun); project administration, X.C.; funding acquisition, T.H., G.Z., and S.S. (Shufang Song). All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Natural Science Foundation of China, grant number: 51808215, Advantage Research Team Project of Guangxi Academy of Agricultural Sciences, grant number: 2021YT037, and Basic Ability Improvement Project for Young and Middle-aged Teachers in Guangxi Universities and Research and Development Fund for Young Teachers in Guangxi University of Finance and Economics of China, grant numbers: 2018KY0505, 2018QNB10.

**Acknowledgments:** We sincerely thank Fengjie County Agricultural Development Ecology Co., Ltd. for providing us with experimental base and some help. We are also grateful to the journal editors and anonymous reviewers for their constructive comments.

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

#### **References**

1. Wen, Z.F. Thoughts on the Construction of Chongqing Citrus Industry Development Technology System. *Southwest Hortic.* **2002**, *30*, 15–17+20.


**Roshan Babu Ojha 1,† , Sujata Manandhar 2,†, Avishesh Neupane <sup>3</sup> , Dinesh Panday 3,\* and Achyut Tiwari 4,†**


**Abstract:** Mustang valley in the central Himalaya of Nepal is a unique landscape formed by massive soil mass during a glacial period, which is attributed to a mix of vegetations and long agricultural history. Soil nutrients and their sourcing is highly important to understand the vegetation assemblage and land productivity in this arid zone. Twenty soil samples (from 0 to 20 cm depth) were collected from three landscape positions in Mustang district: valley, ridge, and midslope. We explored nutrient sourcing using natural abundance carbon (δ <sup>13</sup>C) and nitrogen isotope (δ <sup>15</sup>N) employing isotope ratio mass spectrophotometry. The results showed that the total soil carbon (TC) and total nitrogen (TN) ranged from 0.3 to 10.5% and 0.3 to 0.7%, respectively. Similarly, the CN ratio ranged from 0.75 to 15.6, whereas soil pH ranged from 6.5 to 7.5. Valley soil showed higher values of TN, CN, and soil pH than the ridge and midslope soils. The valleys had more positive δ <sup>15</sup>N signatures than ridge and midslope, which indicates higher inorganic and organic N fertilizer inputs in the valley bottom than in the midslope and ridge. This suggests that a higher nutrient content in the valley bottom likely results from agro-inputs management and the transport of nutrients from the ridge and midslope. Soil pH and CN ratio were a non-limiting factor of nutrient availability in the study regions. These findings are crucial in understanding the nutrient dynamics and management in relation to vegetation and agricultural farming in this unique topography of the Trans-Himalayan zone of Mustang in central Nepal.

**Keywords:** carbon; isotopic signature; Mustang; natural abundance; nitrogen; nutrient sourcing

#### **1. Introduction**

Nepal exhibits unique topographic features with a great variation in climate and biodiversity observed in every five kilometers across the longitude [1]. The origin of the Nepal Himalaya started from the Miocene period (50 million years ago), throughout which a constant weathering of soil parent materials occurred [2,3]. Mustang geology is believed to have originated around the Plio-Pleistocenous age and is well known as a Thakkhola formation [4]. The evolution of the Thakkhola formation aligns with major Himalayan uplift events that are set on unique geomorphic and climate patterns of Mustang compared to other parts of Nepal [4].

Mustang is located in the high mountains, where weathering is mainly constrained by climate. Overall, the climate of Mustang is characterized by low temperatures and dry seasons with high wind speed. Specifically, the northern part of Mustang represents the rain shadow area of Nepal [5], locally referred to as the Trans-Himalayan zone. The

**Citation:** Ojha, R.B.; Manandhar, S.; Neupane, A.; Panday, D.; Tiwari, A. Carbon and Nitrogen Sourcing in High Elevation Landscapes of Mustang in Central Nepal. *Sustainability* **2021**, *13*, 6171. https:// doi.org/10.3390/su13116171

Academic Editors: Bharat Sharma Acharya and Rajan Ghimire

Received: 11 April 2021 Accepted: 27 May 2021 Published: 30 May 2021

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

**Copyright:** © 2021 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/).

southern part is relatively more humid than the northern part and is covered with forest area which is only 3.3% (12,324 ha) of total landmass [6]. Low temperature and scant precipitation decelerate the weathering process [7] and result in fragile, weak aggregates, and shallow soil mass (i.e., skeletal soil) in Mustang [8]. High wind speed, however, accelerates the physical weathering of rocks and minerals. Therefore, Mustang exhibits unique topographic and climatic features, and its soil behaves differently compared to other parts of Nepal.

The altitude of Mustang district ranges from 2010 to 8167 m above sea level (masl). It is covered by 57.7% barren land, 30.3% grassland, 5.6% forest and bushes, 2.7% sand and cliffs, 2.1% water bodies, and 1.6% cultivated areas [8]. The geomorphology of Mustang is composed of high peaks, ridges, midslopes and valley bottoms attributing to different landscape positions [5]. The following three landscape positions are found in the hillslope. The ridge is the peak of the hill of the sloped land, midslope is in the middle part between the ridge and the valley bottom, and the basal part of the hill is valley bottom which is generally flat land located near the river channels. These landscape positions are characterized by their own specific micro-climate, micro-relief, aspect, and soil type. Generally, the valley bottoms are relatively warmer, moist, fertile, and have a lower slope than the midslope and ridge.

Most of the cultivated area in Mustang is occupied with apple (*Malus domestica*) orchards, one of the major income sources of Mustang residents. It covers around 72% of the district's total fruit production and is mainly dominant in the lower part of Mustang [9]. Besides apples, crops such as maize (*Zea mays* L.), wheat (*Triticum aestivum* L.), buckwheat (*Fagopyrum esculentum*), barley (*Hordeum vulgare*), naked barley (*Hordeum vulgare* ssp. Vulgare), pea (*Pisum sativum*), mustard (*Brassica* sp.), potato (*Solanum tuberosum*), vegetables, and other temperate fruits (apricot and walnut) are grown in Mustang. Generally, orchards are planted in the ridge and midslope areas and crops/vegetables are grown in the valley bottoms. Most parts of the central and southern Mustang and a few villages of the northern Mustang harvest two crops each year.

Carbon (C) and nitrogen (N) are two fundamental nutrients that serve as key soil fertility indicators [10]. The availability of nutrients for crops is governed by soil pH [11] and the Carbon–Nitrogen (CN) ratio [12]. The stable isotopes (natural abundance) of carbon (δ <sup>13</sup>C) and nitrogen (δ <sup>15</sup>N) and the CN ratio have been increasingly used to identify organic matter origin, mixing, and transformations in soil and their cycling in the atmosphere [13–16]. A lower CN ratio indicates a higher mineralization of organic matter and vice versa [17]. The signature of δ <sup>13</sup>C differs with different vegetation assemblage; higher in C<sup>4</sup> vegetation (−9 to −17‰) and lower in C<sup>3</sup> vegetation (−23 to −30‰) [18]. Measuring the natural abundance of δ <sup>13</sup>C provides information about the sourcing and migratory nature of soil organic carbon (SOC) in soil, while the natural abundance of δ <sup>15</sup>N tells us about the biological tracking of the nutrients [16,19]. The isotopic signature of δ <sup>13</sup>C and δ <sup>15</sup>N and the CN ratio can be used to track the processes and mechanisms related to organic matter origin, formation, and turnover [17,20].

The availability of plant nutrients in higher elevation soil is mostly linked to the vegetation types and their photosynthetic pathways. As the elevation gradient changes, a shift in the isotopic signature of C and N occurs, and the change in vegetation types governs these dynamics [21]. For instance, the natural abundance of C (δ <sup>13</sup>C) increases in the foliage of plants along with the increase in elevation [21] but there is a large variation in soil's δ <sup>13</sup>C values [22,23]. Plant leaves or litters are one of the major contributors to soil organic matter formation through microbial decomposition [24,25]. However, the mixing and fractionation of stable isotopes in the soil during the decomposition process results in a larger variation of the isotopic composition [26]. The elevation and the rate of litterfall in natural ecosystems strongly influence the SOC sourcing and mixing [21]. Similarly, the natural abundance of nitrogen (δ <sup>15</sup>N) isotope is enriched in an intensively managed environment. The relationship between the δ <sup>13</sup>C and δ <sup>15</sup>N with soil nutrients provides

important information about the sourcing of nutrients in the higher elevation soils and such information is very limited in the mountainous country of Nepal.

Plant nutrients in the cultivated soils of Mustang are typically low [27], which is attributed to a low mineralization rate, low mobility, and low exchange potential. With an increase in elevation, the availability of plant nutrients are limited due to reduced mineralization rates [21,26]. Nutrient availability in the higher elevation is mostly constrained by the climate, vegetation type, and input management [21,23]. In Mustang, animal husbandry is closely linked with agriculture [28]. There is a significant transfer of biomass from the forest and rangeland to the cropland as fodder and roughages for livestock, ultimately ending in cropland as manure [29]. In addition, as livestock graze, there is also a reciprocal exchange of nutrients from crop residues back to the rangeland and forest through their excrement. This favours the nutrient source mixing in between cropping land and nearby native vegetation coupled with erosion-led nutrient transport [30]. The soils in the agricultural land of the Mustang region are poorly investigated and the existing plant nutrients status is not well known. Furthermore, soil test-based plant nutrient management is barely practised at the local level. The identification of the source, status, dynamism, and retention of those nutrients in the upper elevation soils is critically important to know the managerial aspects of the nutrients in a sustainable manner. We aimed to explore the TC, TN, CN ratio, and soil pH along the transect of different landscape positions (ridge, midslope, and valley) where cropping or orchard plantation is common in Mustang with surrounding natural vegetation. We further analysed the isotopic signature of the natural abundance δ <sup>13</sup>C and δ <sup>15</sup>N to identify the source of C and N where the nutrients source mixing is common in higher elevations of Mustang, Nepal.

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

#### *2.1. Location and Climate*

Mustang is one of the mountainous districts in central Nepal. It is located in the rain shadow of the world's 7th and 10th highest mountains (Dhaulagiri and Annapurna standing 8168 and 8137 masl, respectively) and receives on an average <400 mm annual rain with relatively higher rainfall in the southern part of the district. It presents a diversity of climates ranging from tundra, arid types in the higher elevations above 4500 masl, to alpine and cold temperate in 3000 to 4500 masl and 2000 to 3000 masl, respectively [31]. Furthermore, it is a deeply incised valley of the Kali Gandaki river with an arid valley bottom and characteristic diurnal wind system. It is divided into upper Mustang (above 3800 m) and lower Mustang (below 3800 m), the two divisions differing from each other with respect to the prevailing climatic conditions.

#### *2.2. Soil Sampling Point Determination*

We selected 20 sampling points along the Kaligandaki corridor from the southern (Tukuche) to northern part (Korala) of the Mustang district, considering a vertical transect to capture the best possible landscape positions (Figure 1 and Table 1). Out of 20 points, we collected four samples from the midslope, five samples from the ridge, and eleven samples from the valley in October 2011. Difficulties in accessing the varied topographic or landscape positions resulted in uneven sampling points. The details of sampling points are given in Table 1.

*Sustainability* **2021**, *13*, 6171


**Table 1.** Detailed information on sampling points in the Mustang district of Nepal.

20 29.30347 83.96836 4612 North of Chonup Ridge Orchard Grasses

**Sampling** 

**Figure 1.** Distribution of soil sampling points showing in the digital elevation model (DEM) map across the transects in the study area, the Mustang district of Nepal. **Figure 1.** Distribution of soil sampling points showing in the digital elevation model (DEM) map across the transects in the study area, the Mustang district of Nepal.

#### **Table 1.** Detailed information on sampling points in the Mustang district of Nepal. *2.3. Soil Sampling and Laboratory Analysis*

**Points Latitude Longitude Elevation, masl Location Micro-Relief Site Characteristics Nearby Dominant Vegetation**  1 28.71225 83.64908 2628 Tukuche Midslope Orchard Juniper, Pine 2 28.83692 83.78242 2837 Kagbeni Valley Cropped land Juniper 3 28.80392 83.77322 2852 Between Kagbeni and Lupra Valley Cropped land Juniper We collected soil samples with the help of auger from 0 to 20 cm depth from selected points. A composite sample was taken, which was then spread in a sample box. Roots, undecomposed plant debris, and gravel were removed in the field. Upon returning to the lab, the soil was air dried, ground in mortar and pestle to break aggregates, and sieved to a 2 mm mesh size, which was then subjected to lab analysis.

4 28.80389 83.77322 2852 Between Kagbeni and Lupra Valley Cropped land Juniper 5 28.92494 83.82758 2963 Tsungsang Valley Cropped land Juniper shrub, grasses 6 28.80244 83.79028 2997 Lupra Midslope Orchard Juniper shrub, grasses 7 28.80211 83.78958 3017 Lupra Midslope Apple orchard Juniper shrub, grasses 8 28.88406 83.80836 3092 Thangbe Valley Cropped land Juniper shrub 9 28.96161 83.80847 3447 East of Samar Valley Cropped land Pine, Juniper 10 28.81758 83.84944 3524 Jharkot Ridge Orchard - 11 28.94964 83.80181 3560 South of Samar Midslope Orchard Pine, Juniper The total soil carbon (TC) and total nitrogen (TN) were analysed using the dry combustion method for which soil was ground to 0.5 mm in size. The soil pH was determined in a 1:5 soil to water ratio with a digital soil pH meter. The isotopic signature of carbon (δ <sup>13</sup>C) and nitrogen (15N) were obtained from isotopic ratio mass spectrometry (Thermo Fisher Scientific, Analyzer: FLAS 2000-Conflo IV-Delta V Advantage) for which soil was ground to 0.1 mm in size. The soil sample was replicated twice to determine each of the carbon and nitrogen signatures and the average value of the replicates was reported. Soil standards and reference samples were placed after every 12 samples. The standard error of soil standards/reference sample was <0.23‰.

#### 12 29.06139 83.87169 3579 Ghami Valley Cropped land Planted Populas 13 28.96169 83.80142 3606 Samar Ridge Orchard Pine, Juniper *2.4. Data Analysis*

14 28.99114 83.83819 3778 Syanboche Valley Cropped land - 15 29.18361 83.95714 3823 Lomanthang Valley Cropped land Planted Populas 16 29.18272 83.95711 3825 Lomanthang Valley Cropped land Planted Populas 17 29.25469 83.96025 4027 South of Chonup, North of Lomanthang Valley Cropped land Grasses 18 29.30347 83.96836 4612 North of Chonup Ridge Orchard Juniper 19 29.30347 83.96836 4612 North of Chonup Ridge Orchard Juniper and grasses 20 29.30347 83.96836 4612 North of Chonup Ridge Orchard Grasses Initially, data were entered in MS Excel and then imported to R studio for descriptive analysis. Standard least square analysis of variance (ANOVA) models were used to investigate the effects of landscape position on soil C and N. The data were tested for ANOVA assumptions prior to the analysis, and they met these assumptions. We employed log and square root transformations to fit data into a normal distribution curve before subjecting them to ANOVA. Tukey means separation tests were used for post-hoc comparisons of the soil parameters amongst landscape positions. A correlation between the variables was

calculated at 5% level of significance. Graphs of variables (TC, TN, CN ratio, pH, δ <sup>13</sup>C, and δ <sup>15</sup>N) were prepared in R-studio [32] using ggplot2 package [33]. The relationship between different variables were performed in ggpairs function which is an extension of the ggplot2 package [33]. The statistical significance was determined as *p* < 0.05, unless otherwise noted. Analyses were conducted using 15.0.0 JMP SAS software. To elucidate the relationship between the geographical coordinates and the soil properties, a multiple linear regression model was built where soil properties were kept as a dependent factor and geographical coordinates as independent factors.

#### **3. Results**

The TC was significantly higher in the valley (6.2%) than in the ridge (2.8%) while the midslope had the intermediate values (4.7%) (Figure 2). The maximum and minimum TC in the valley was 10.5 and 4.0%, the midslope was 5.8 and 2.4%, and the ridge was 6.6 and 0.3%, respectively. Although there was no significant difference in the TN contents between these slope locations, the average TN content showed a decreasing trend from the valley (0.8%) to the midslope (0.6%) and ridge (0.5%). The maximum and minimum TN in the valley was 0.7 and 0.3%, the midslope was 0.9 and 0.4%, and the ridge was 0.7 and 0.3%, respectively. *Sustainability* **2021**, *13*, x FOR PEER REVIEW 6 of 12

**Figure 2.** Total soil carbon (TC) and total nitrogen (TN) variation (mean ± SE) at different landscape positions in Mustang district of Nepal. Significant differences between the landscape positions are shown with differing letters. **Figure 2.** Total soil carbon (TC) and total nitrogen (TN) variation (mean ± SE) at different landscape positions in Mustang district of Nepal. Significant differences between the landscape positions are shown with differing letters.

The δ15N values in soil were significantly different across the landscape positions (Figure 3). On average, the valley soil contained the positive δ15N values, whereas the midslope and ridge soils contained negative δ15N values. However, the range of δ15N signature showed both positive and negative values at all landscape positions. The maximum and minimum natural abundance of δ15N in the valley was +7.0 and −15.2‰, the midslope was +7.3 and −16.1‰, and the ridge was +5.4 and −14.9‰, respectively. The soil δ13C value did not differ significantly between the landscape positions, although the average value showed a slightly increasing trend from the valley to the ridge (Figure 3). The maximum and minimum natural abundance of δ13C in the valley was −6.6 and −23.6‰, the midslope was −7.6 and −19.4‰, and the ridge was +1.1 and −22.0‰, respectively. The δ <sup>15</sup>N values in soil were significantly different across the landscape positions (Figure 3). On average, the valley soil contained the positive δ <sup>15</sup>N values, whereas the midslope and ridge soils contained negative δ <sup>15</sup>N values. However, the range of δ <sup>15</sup>N signature showed both positive and negative values at all landscape positions. The maximum and minimum natural abundance of δ <sup>15</sup>N in the valley was +7.0 and <sup>−</sup>15.2‰, the midslope was +7.3 and −16.1‰, and the ridge was +5.4 and −14.9‰, respectively. The soil δ <sup>13</sup>C value did not differ significantly between the landscape positions, although the average value showed a slightly increasing trend from the valley to the ridge (Figure 3). The maximum and minimum natural abundance of δ <sup>13</sup>C in the valley was <sup>−</sup>6.6 and <sup>−</sup>23.6‰, the midslope was −7.6 and −19.4‰, and the ridge was +1.1 and −22.0‰, respectively.

The CN ratio did not differ significantly between the landscape positions, although there was an indication of a higher CN ratio in the valley than in the midslope and the ridge (Figure 4). A significantly higher soil pH was observed in the valley than in the ridge, while the midslope had intermediate values (Figure 4). The maximum and minimum CN ratio in the valley was 12.0 and 6.0, the midslope was 10.0 and 6.0, and the ridge was 16.0 and 1.0, respectively. Similarly, in the valley and midslope, we found the same

**Figure 3.** δ13C and δ15N variation (mean ± SE) in soil at different landscape positions in Mustang district of Nepal. Signifi-

soil pH range (7.0 to 7.5) and in the ridge soil the pH range was 7.0 to 6.5.

cant differences between the landscape positions are shown with differing letters.

**Figure 2.** Total soil carbon (TC) and total nitrogen (TN) variation (mean ± SE) at different landscape positions in Mustang

The δ15N values in soil were significantly different across the landscape positions (Figure 3). On average, the valley soil contained the positive δ15N values, whereas the midslope and ridge soils contained negative δ15N values. However, the range of δ15N signature showed both positive and negative values at all landscape positions. The maximum and minimum natural abundance of δ15N in the valley was +7.0 and −15.2‰, the midslope was +7.3 and −16.1‰, and the ridge was +5.4 and −14.9‰, respectively. The soil δ13C value did not differ significantly between the landscape positions, although the average value showed a slightly increasing trend from the valley to the ridge (Figure 3). The maximum and minimum natural abundance of δ13C in the valley was −6.6 and −23.6‰, the midslope was −7.6 and −19.4‰, and the ridge was +1.1 and −22.0‰, respectively.

district of Nepal. Significant differences between the landscape positions are shown with differing letters.

**Figure 3.** δ13C and δ15N variation (mean ± SE) in soil at different landscape positions in Mustang district of Nepal. Significant differences between the landscape positions are shown with differing letters. **Figure 3.** δ <sup>13</sup>C and δ <sup>15</sup>N variation (mean <sup>±</sup> SE) in soil at different landscape positions in Mustang district of Nepal. Significant differences between the landscape positions are shown with differing letters.

The CN ratio did not differ significantly between the landscape positions, although there was an indication of a higher CN ratio in the valley than in the midslope and the ridge (Figure 4). A significantly higher soil pH was observed in the valley than in the ridge, while the midslope had intermediate values (Figure 4). The maximum and minimum CN ratio in the valley was 12.0 and 6.0, the midslope was 10.0 and 6.0, and the ridge was 16.0 and 1.0, respectively. Similarly, in the valley and midslope, we found the same soil pH range (7.0 to 7.5) and in the ridge soil the pH range was 7.0 to 6.5. The CN ratio did not differ significantly between the landscape positions, although there was an indication of a higher CN ratio in the valley than in the midslope and the ridge (Figure 4). A significantly higher soil pH was observed in the valley than in the ridge, while the midslope had intermediate values (Figure 4). The maximum and minimum CN ratio in the valley was 12.0 and 6.0, the midslope was 10.0 and 6.0, and the ridge was 16.0 and 1.0, respectively. Similarly, in the valley and midslope, we found the same soil pH range (7.0 to 7.5) and in the ridge soil the pH range was 7.0 to 6.5. *Sustainability* **2021**, *13*, x FOR PEER REVIEW 7 of 12

**Figure 4.** The CN ratio and soil pH variation (mean ± SE) at different landscape positions in Mustang district of Nepal. Significant differences between the landscape positions are shown with differing letters. **Figure 4.** The CN ratio and soil pH variation (mean ± SE) at different landscape positions in Mustang district of Nepal. Significant differences between the landscape positions are shown with differing letters.

Both a positive and a negative correlation was observed between the soil parameters (Figure 5). A significant positive correlation was found between the TC and TN (r = 0.7, *p* < 0.001); the TC and CN ratio (r = 0.6, *p* < 0.001); the TC and δ15N (r = 0.7, *p* < 0.001); the TN and δ15N (r = 0.5, *p* = 0.019); and a significant negative correlation was found between the TN and the δ13C (r = −0.6, *p* = 0.004). A detailed correlation matrix including all variables is provided in Appendix Figure A1. We did not find any significant correlation between geographical co-ordinates (latitude, longitude, and elevation) and soil parameters (Appendix Figure A1). Furthermore, there was no significant effect of latitude, longitude, and elevation on soil properties as suggested by multiple linear regression models (results not shown), which is in compliance with correlation results. Both a positive and a negative correlation was observed between the soil parameters(Figure 5). A significant positive correlation was found between the TC and TN (r = 0.7, *p* < 0.001); the TC and CN ratio (r = 0.6, *p* < 0.001); the TC and δ <sup>15</sup>N (r = 0.7, *p* < 0.001); the TN and δ <sup>15</sup>N (r = 0.5, *p* = 0.019); and a significant negative correlation was found between the TN and the δ <sup>13</sup>C (r = <sup>−</sup>0.6, *<sup>p</sup>* = 0.004). A detailed correlation matrix including all variables is provided in Appendix A Figure A1. We did not find any significant correlation between geographical co-ordinates (latitude, longitude, and elevation) and soil parameters (Appendix A Figure A1). Furthermore, there was no significant effect of latitude, longitude, and elevation on soil properties as suggested by multiple linear regression models (results not shown), which is in compliance with correlation results.

**Figure 5.** A correlation matrix between the selected variables, scatter points on the lower panel, and correlation values and

their 5% significance level on the upper panel. \*\*\* *p*-value < 0.001; \*\* *p*-value < 0.01, \* *p*-value < 0.05.

shown), which is in compliance with correlation results.

**Figure 4.** The CN ratio and soil pH variation (mean ± SE) at different landscape positions in Mustang district of Nepal.

Both a positive and a negative correlation was observed between the soil parameters (Figure 5). A significant positive correlation was found between the TC and TN (r = 0.7, *p* < 0.001); the TC and CN ratio (r = 0.6, *p* < 0.001); the TC and δ15N (r = 0.7, *p* < 0.001); the TN and δ15N (r = 0.5, *p* = 0.019); and a significant negative correlation was found between the TN and the δ13C (r = −0.6, *p* = 0.004). A detailed correlation matrix including all variables is provided in Appendix Figure A1. We did not find any significant correlation between geographical co-ordinates (latitude, longitude, and elevation) and soil parameters (Appendix Figure A1). Furthermore, there was no significant effect of latitude, longitude, and elevation on soil properties as suggested by multiple linear regression models (results not

Significant differences between the landscape positions are shown with differing letters.

**Figure 5.** A correlation matrix between the selected variables, scatter points on the lower panel, and correlation values and their 5% significance level on the upper panel. \*\*\* *p*-value < 0.001; \*\* *p*-value < 0.01, \* *p*-value < 0.05. **Figure 5.** A correlation matrix between the selected variables, scatter points on the lower panel, and correlation values and their 5% significance level on the upper panel. \*\*\* *p*-value < 0.001; \*\* *p*-value < 0.01, \* *p*-value < 0.05.

#### **4. Discussion**

Our current study showed that the valley bottom had a higher TC and TN concentration compared to the midslope and ridge. There is a progressive increment in the TC and TN concentration from the ridge to the valley bottom. This implies that the valley bottom is more fertile than the ridge and the midslope. The transport of the nutrients from the upper slope (ridge and midslope) towards the lower slope (valley bottom) due to soil erosion may have resulted higher concentration of the TC and TN [34] in the valley bottom. The deposition of sediment in the valley bottom and nutrient transport from the upper slope and forest litter [30] coupled with heavy textured soil [27] resulted in higher TC and TN. Lü et al. [35] reviewed the nutrient transport process associated with rainfall-runoff events and reported their results in terms of factors, forms, carriers, and sources of nutrient transport. Lü et al. [35] concluded that during the erosion process water is a carrier of soil nutrients in the soluble form. The dissolved organic C and available form of N resulting from litter decomposition from the forest is the primary source of nutrient transport, along with eroding water and sediments from the upper slope to the lower slope [36,37]. The lower slope, where the deposition of sediments occurs, are generally high in soil carbon and nitrogen compared to the eroding upper slope [38–40]. Thus, in the current study area, this erosion-led nutrient transport is dominant where the nutrients from the upper slope of the ridge and the midslope are deposited in the valley bottom.

The low CN ratio in the soil that we observed in the current study might be due to presence of inorganic carbon (natural carbonates) or inorganic nitrogen (input management) or both. The presence of inorganic forms of C and N alters the CN ratio [41]. The ratio of total organic C to N is the indicator of the decomposition rate of organic matter with an inverse relationship [42]. The CN ratio at any landscape position does not affect the nutrient availability [43] in the Mustang soil. Similarly, the average soil pH of the study area was 6.9 ± 0.5 units, which is the most favourable range for the nutrient availability. Many of the nutrients essential for plants are available in the pH range of 6.5 to 7.5 [44]. There, the CN ratio and soil pH are found to be non-limiting factors for nutrient availability in the study area.

The range of δ <sup>13</sup>C value from <sup>−</sup>23.6 to +1.1‰ in the current study indicates the presence of both inorganic and organic C in soil. The C content of few soil samples, particularly from the ridge, was dominated by inorganic C (i.e., carbonate mineral). Plant photosynthesis discriminates against the heavier C isotope, and the degree of discrimination mainly varies with the type of photosynthetic pathways (C3, C4, CAM). The δ <sup>13</sup>C values of C<sup>3</sup> and C<sup>4</sup> plants generally lie between −20 to −40‰, and −9 to −17‰, respectively, while δ <sup>13</sup>C values in CAM typically lie in between −10 to −20‰ [45]. Garzione et al. [46] reported that C<sup>3</sup> plants, dominantly trees, shrubs, and cool-growing-season grasses in the region produce soil respiration with δ <sup>13</sup>C values of about <sup>−</sup>22 to <sup>−</sup>32‰, whereas C<sup>4</sup> plants, dominantly warm-growing-season grasses, produce soil respiration with δ <sup>13</sup>C values between <sup>−</sup><sup>10</sup> and −15‰. The average observed range of soil carbonate formed in equilibrium with C3-respired CO<sup>2</sup> is δ <sup>13</sup>C = <sup>−</sup>13 to <sup>−</sup>9‰, whereas soil carbonates formed in the presence of C<sup>4</sup> plants have δ <sup>13</sup>C = +1 to +3‰ [47].

Galy et al. [48] reported the presence of δ <sup>13</sup>C value of <sup>−</sup>0.4 to +1.9‰ in the carbonates of the bedload sediment around the Lomangthang and Kagbeni regions of the Mustang district. The presence of carbonates in the Tethyan sedimentary series, which also includes the Mustang district, comprises of Paleozoic–Mesozoic carbonates and clastic sediments that have δ <sup>13</sup>C values ranging from <sup>−</sup>2.5 to 0‰ [49]. The <sup>δ</sup> <sup>13</sup>C value in Paleosol carbonates of the Thakkhola formation is between −5.6‰ and +3.5‰ with a mixed C<sup>3</sup> and C<sup>4</sup> plant species, but predominantly C<sup>4</sup> species [46]. An arid environment like Mustang is likely to have higher δ <sup>13</sup>C values resulting from a low respiration rate or the dominance of C<sup>4</sup> plants. Garzione et al. [46] collected different species of grass in between 3000 and 4000 masl and found δ <sup>13</sup>C values from <sup>−</sup>12.3 to <sup>−</sup>12.8‰ in these grass species. They concluded that the higher value of δ <sup>13</sup>C in the valley floor of the Thakkhola formation deposition is from paleosol carbonates with the presence of both C<sup>3</sup> and C<sup>4</sup> vegetation. However, the presence of only C<sup>3</sup> vegetation in the lower elevations (Tetang formation), yielded δ <sup>13</sup>C values from <sup>−</sup>21.9 to <sup>−</sup>26.5‰. Szpak et al. [50] demonstrated that foliar <sup>13</sup>C values increased with a site's altitude, which is in agreement with our data trend of greater soil <sup>13</sup>C enrichment in the ridge compared to the valley. Therefore, the sourcing of soil C can be attributed to the mixture of soil organic matter and paleosol carbonates with the increasing influence of organic matter (mostly from C3 vegetation) as we move from the ridge towards the valley floor of the Mustang district.

The natural abundance of δ <sup>15</sup>N in soil represents an integrated signal of the ecosystem's N processes that help constrain N budgets, identify sources, and their fates. The range of δ <sup>15</sup>N values in our study sites is between <sup>−</sup>16.1 to +7.0‰. Since Mustang is located in an arid climate, the δ <sup>15</sup>N enrichment in soil and plants is expected. The loss of <sup>14</sup>N and enrichment of δ <sup>15</sup>N values in soil and vegetation samples in dry regions have been reported previously [51,52]. Zhou et al. [53] reported increasing δ <sup>15</sup>N values with decreasing rainfall in the Qinghai–Tibetan Plateau, a region similar to our site. The authors suggested that the precipitation and temperatures that influence the CN content and ratio are also the primary factors determining the patterns of soil δ <sup>15</sup>N on a regional scale [53]. In a landscape scale similar to our site, variability in δ <sup>15</sup>N was positively related to moisture availability, soil fertility, and vegetation cover [54]. Szpak et al. [50] demonstrated that foliar δ <sup>15</sup>N values decreased with an increase in altitude, similar to our trend observed in soil δ <sup>15</sup>N values. The higher value of δ <sup>15</sup>N in valleys in our sites could also be associated with the use of organic fertilizer (δ <sup>15</sup>N value of around 2 to 30‰) and synthetic N fertilizer (δ <sup>15</sup>N value of around −4 to 4‰) [55]. Regmi et al. [56] reported the use of synthetic N (23 to 280 kg urea ha−<sup>1</sup> ) and farmyard manure (2.1 to 5.3 t ha−<sup>1</sup> ) in apple orchards of the Mustang district. Hence, the enrichment of δ <sup>15</sup>N in the Mustang district and the variations we observed between various sites might come from the agro-inputs or enrichment due to micro and macro scale topographic and climatic variations in these sites.

Furthermore, the higher positive correlation of the TC with the TN, δ <sup>15</sup>N, and CN ratio; δ <sup>13</sup>C with CN ratio; TN with δ <sup>15</sup>N; and the higher negative correlation of the TN with δ <sup>13</sup>C (Figure 5) indicates the interdependence of sourcing between the TC and the TN. This

indicates that the source mixing between soil organic matter is C<sup>4</sup> and C<sup>3</sup> vegetation. The nutrient source mixing is further augmented by the transfer of forest litter as manure to farmland by the farmers of the Mustang valley [34]. The use of fresh animal manure in the valley bottom might result in higher δ <sup>15</sup>N [57]. There is a mixing of isotopically distinctive carbon and nitrogen in the study area as a weak and negative correlation between δ <sup>13</sup>C and δ <sup>15</sup>N (Figure 5) [58]. Hence, the isotopic techniques are useful in organic matter source identification, their mixing, and stability in the soil.

#### **5. Conclusions**

Our study discloses the sourcing and availability of C and N in higher elevation soils of the Mustang District, Nepal using stable isotope techniques, CN ratio, and soil pH. The findings suggest that the landscape positions strongly influence the nutrient sourcing and mixing in higher elevation soils. As our data is limited to cover the broader geographic range, we do not find a relation of C and N with longitude, latitude, and elevation. C and N sourcing are specific to different landscape positions in Mustang. The ridge and midslopes are dominant with the litter decomposition, either of the forest trees species or of fruit orchards. Valley slopes are mostly dominant with the fresh organic and inorganic substrates through agro-input management by farmers, along with the source mixing of C and N transfer associated with erosion-led nutrient transport from the ridge and midslopes. Hence, we emphasize that the cultivated croplands in the valley or orchards in the midslope and the ridge should be managed according to the nutrient sourcing in the region for sustainable land management. It is important to consider landscape position, broad geographic co-ordinates, and micro-relief to study C and N sourcing in future studies. Further research is necessary to study the micro-climate, decomposition rate constant, and the microbial diversity to understand the cycling of C and N in the higher landscapes.

**Author Contributions:** Conceptualization—S.M.; funding acquisition—D.P.; methodology—S.M., R.B.O., and A.T.; investigation—S.M.; data curation—R.B.O. and A.N.; formal analysis—A.N. and R.B.O.; visualization—R.B.O., A.N., and D.P.; software—A.N. and R.B.O.; validation—R.B.O., A.N., D.P., S.M., and A.T.; writing—original draft preparation—R.B.O.; writing—review and editing— A.N., A.T., R.B.O., D.P., and S.M. All authors have read and agreed to the published version of the manuscript.

**Funding:** The MEXT and the Global COE Program at the University of Yamanashi (UoY), Japan supported/funded this study.

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

**Informed Consent Statement:** Not applicable.

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

**Acknowledgments:** The authors would like to sincerely thank Futaba Kazama for her support and acknowledge the MEXT and the Global COE Program at the University of Yamanashi (UoY), Japan for funding this study. We are thankful to Takashi Nakamura from Interdisciplinary Centre for River Basin Environment, UoY, Japan for his support in the lab analysis. We also extend our thanks to Devraj Chalise for the study area map. We would like to thank three anonymous reviewers and an academic editor for their valuable comments and suggestions, which helped us in improving this paper.

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

**Sample Availability:** Not applicable.
