*2.3. Methods*

more attention.

#### *2.2. Data* 2.3.1. Construction of the PLES Classification System

The data of administrative districts of the study area were derived from the Remote Sensing Monitoring Institute of Chongqing Planning and Natural Resources Survey and Monitoring Institute. The land-use data used for three phases from 2000 to 2020 were obtained from the Data Center for Resources and Environment Science and Data Center, Chinese Academy of Sciences (http://www.resdc.cn/ (accessed on 9 July 2022)). The resolution of the land-use data is 30 m × 30 m. The types of land-use include six types of the first order: cultivated land, woodland, grassland, water area, residential land, and unused land, and 25 types of the second order. The social–economic data mainly come from the Chongqing Statistical Yearbook (2021). The PLES has the characteristics of complex spatial functions, differences in spatial scales, and the heterogeneity of spatial land use [33]. Its classification is exceptionally complicated. Land resources are the lifeblood of promoting economic development, the source of all productive activities, and the interrelated and unified complex of the PLES. Land-use type has multiple functions, e.g., the construction land has two functions of production and living, and the cultivated land has two functions of ecological and production. Combined with the theory of the PLES and the previous research results [34–36], each land-use category was linked with the leading function of the PLES. Referring to relevant studies [37,38], the PLES classification system was constructed according to the study area's actual situation, as shown in Table 1.

#### 2.3.2. Dynamic of Land–Use Change

*2.3. Methods* 2.3.1. Construction of the PLES Classification System The conversion between land-use types is mainly realized by the land-use transfer matrix, and the dynamic change process of land-use types is mainly expressed by single land-use dynamic and bidirectional land-use dynamic [38].

The PLES has the characteristics of complex spatial functions, differences in spatial scales, and the heterogeneity of spatial land use [Error! Reference source not found.]. Its classification is exceptionally complicated. Land resources are the lifeblood of promoting

unified complex of the PLES. Land-use type has multiple functions, e.g., the construction land has two functions of production and living, and the cultivated land has two functions of ecological and production. Combined with the theory of the PLES and the previous research results [34–36], each land–use category was linked with the leading function of


**Table 1.** The PLES classification system.

The single land-use dynamic (*LCD I<sup>i</sup>* ) refers to the ratio of the area change of land-use types in the region to the study period, which mainly reflect the rate of change of a single land-use type in a certain period. The bidirectional land-use dynamic (*BLCD Ii*) is a further supplement to *LCD I<sup>i</sup>* [38]. It can better describe the change process and the direction of a certain kind of land use, which mainly reflects the transfer intensity between land-use types. The formulas are below:

$$LCDI\_l = \frac{(S\_{lb} - S\_{ia})}{S\_{ia}} \times \frac{1}{T} \times 100\% \tag{1}$$

$$BLCDI\_i = \frac{\left(\sum S\_{ij} + \sum S\_{ji}\right)}{S\_i} \times \frac{1}{T} \times 100\% \tag{2}$$

where *LCD I<sup>i</sup>* is the single land-use dynamic of *i*–th PLES type; *Sia* is the area of *i*–th PLES type at the beginning; *Sib* is the area of *i*–th PLES type at the end; T is the study time interval; *BLCD I<sup>i</sup>* is the bidirectional land-use dynamic of *i*–th PLES type; ∑ *Sij* is the sum of the area of *i*–th PLES type changing into other PLES types; ∑ *Sji* is the sum of the area of other PLES types changing into *i*–th PLES type; *S<sup>i</sup>* is the area of *i*–th PLES type at the beginning.

#### 2.3.3. Construction of LUCs Assessment Model

Land-use systems are dynamic, fragile, and complex [39]. To avoid the fragmentation of regional spatial units, we consider the research scope, scale, spatial resolution, spatial patch status, and data types. The maximum horizontal distance of the study area between east and west is about 74 km, and the maximum vertical distance between north and south is about 112 km. To ensure that there are multiple spatial types in an assessment unit, this paper choose a 4 km × 4 km grid as the evaluation unit. The study area was divided into a total of 397 assessment units. We chose the method of the LER assessment to measure the intensity of LUCs because we consider that the LUCs and landscape ecological risk assessment are highly related (Figure 2).

The spatial complexity index (*CI*), spatial vulnerability index (*FI*), and spatial stability index (*SI*) were chosen to construct the LUCs assessment model to measure the conflict level of spatial units in a region from the ecological point of view. Referring to previous studies [37,40,41], the LUCs assessment model can be expressed as:

$$\text{SCCI} = \text{CI} + \text{FI} - \text{SI} \tag{3}$$

where *SCCI* is the LUCs level; *CI* is the landscape complexity index; *FI* is the landscape vulnerability index; SI is the landscape stability index. uation unit. The study area was divided into a total of 397 assessment units. We chose the method of the LER assessment to measure the intensity of LUCs because we consider that the LUCs and landscape ecological risk assessment are highly related (Figure 2).

is the bidirectional land–use dynamic of *i*–th PLES type; ∑Sij is the sum of the area

Land–use systems are dynamic, fragile, and complex [Error! Reference source not found.]. To avoid the fragmentation of regional spatial units, we consider the research scope, scale, spatial resolution, spatial patch status, and data types. The maximum horizontal distance of the study area between east and west is about 74 km, and the maximum vertical distance between north and south is about 112 km. To ensure that there are multiple spatial types in an assessment unit, this paper choose a 4 km × 4 km grid as the eval-

is the area of *i*–th PLES type at the beginning.

of *i*–th PLES type changing into other PLES types; ∑Sji is the sum of the area of other PLES

*Land* **2022**, *11*, x FOR PEER REVIEW 6 of 19

types changing into *i*–th PLES type; S<sup>i</sup>

2.3.3. Construction of LUCs Assessment Model

BLCDI<sup>i</sup>

**Figure 2.** Conceptual framework of coupled LUCs and LER assessment. **Figure 2.** Conceptual framework of coupled LUCs and LER assessment.

The spatial complexity index (*CI*), spatial vulnerability index (*FI*), and spatial stability (1) Landscape complexity index (*CI*)

index (*SI*) were chosen to construct the LUCs assessment model to measure the conflict level of spatial units in a region from the ecological point of view. Referring to previous studies [37,40,41], the LUCs assessment model can be expressed as: *SCCI = CI + FI − SI* (3) *CI* is mainly due to the low efficiency of land use caused by rapid urbanization, which leads to the negative effects of system fragmentation and complexity [40]. *AWMPFD* index is called the area–weighted average patch fractal dimension in landscape ecology. The higher its value, the greater the degree of human disturbance, and vice versa. In this paper, the *AWMPFD* index is used to express *CI*. The formula is below:

$$A\%MPFD = \sum\_{i=1}^{m} \sum\_{j=1}^{n} \left[ \frac{2\ln(0.25P\_{ij})}{\ln(a\_{ij})} \left(\frac{a\_{ij}}{A}\right) \right] \tag{4}$$

*CI* is mainly due to the low efficiency of land use caused by rapid urbanization, which leads to the negative effects of system fragmentation and complexity [**Error! Reference source not found.**0]. *AWMPFD* index is called the area–weighted average patch fractal dimension in landscape ecology. The higher its value, the greater the degree of human where *Pij* is the perimeter of the patch; *aij* is the area of the patch; *A* is the total area of spatial units in the landscape; *m* is the total number of evaluation units in the study area; *n* is the number of three spatial types.

(2) Landscape vulnerability index (*FI*)

disturbance, and vice versa. In this paper, the *AWMPFD* index is used to express *CI*. The formula is below: = ∑∑[ 2 ( 0.25) ( ) ( )] =1 =1 (4) where *Pij* is the perimeter of the patch; *aij* is the area of the patch; *A* is the total area of spatial units in the landscape; *m* is the total number of evaluation units in the study area; *FI* is mainly due to the vulnerability of a land-use system under the interference of external pressure, which will cause significant damage to the system. With time and space changes in land use, there are significant differences in maintaining ecosystem stability, protecting biodiversity, and improving system functions; that is to say, the various landscape elements have different responses to spatial conflicts, which are related to the stages in the natural succession process. In different locations, the ability of land-use types to resist external disturbance is other. *FI* is used to express it. The formula is below:

$$FI = \sum\_{i=1}^{n} F\_i \times \frac{a\_i}{S} (n=4) \tag{5}$$

where *FI* is the landscape vulnerability index; *a<sup>i</sup>* is the area of various landscapes in the unit; *S* is the total area of units. Referring to the existing research [42], the order of landscape fragility of *FI* from strong to weak is LPS = 4; PES = 3; EPS = 2; ES = 1.

(3) Landscape stability index (*SI*)

*SI* refers to the phenomenon that landscape patches are fragmented under the interference of external pressure. The more fragmented the spatial form of the land-use unit, the stronger the dynamic, the worse the stability, and the stronger the conflict effect. The formula is below:

$$PD = \frac{n\_{\text{i}}}{A}^{\prime} \text{ } SI = 1 - PD \tag{6}$$

where *n<sup>i</sup>* represents the number of *i*–th PLES in the spatial unit; *A* is the space unit area; *PD* is the density of plaque. The larger the *PD* value, the higher the fragmentation degree, the worse the stability, and the worse the anti-interference ability of the unit. For the convenience of calculation, the numerical values in Formulas (3)–(6) are linearly standardized to (0, 1) by using Formula (7) to calculate the conflict in the later period. The standardized formula is below:

$$S = \frac{X - X\_{\text{min}}}{X\_{\text{max}} - X\_{\text{min}}} \tag{7}$$

where *X* is the value in (3), (4), (5), and (6), *X*min is the minimum value, and *X*max is the maximum value according to the existing research [43], based on the inverted "U" model. The LUCs of the study area were divided into four stages [9,23,44]: the stable control stage [0.0, 0.35), the basic control stage [0.35, 0.7), the basic out-of-control stage [0.7, 0.9), and the serious out-of-control stage [0.9, 1). In the stable control stage, the conflict has not yet formed or is in the potential stage, and will have no negative impact on the regional land use. In the basic control stage, conflict begins to form and gradually emerge, but is mostly constructive rather than destructive [45,46]. Appropriate measures should be taken to regulate and avoid or minimize the negative effects of conflict. In the basic out-ofcontrol stage, the conflict broke out gradually, and the direction of land-use transformation gradually lost control. Effective measures must be taken to curb the conflict. Otherwise, the regional land use will gradually be unbalanced. In the serious out-of-control stage, the conflict completely breaks out, which requires the intervention of various administrative, economic, and legal measures. Otherwise, it may evolve from LUCs to a conflict of a social nature [47].

#### 2.3.4. Spatial Relationship of LUCs

The land-use patches are often significantly affected by the land-use patterns around them, which is manifested in two aspects, including spatial agglomeration and spatial adjacency. On one side, the cold spots and hot spots are statistically significant spatial clusters of high values and low values, respectively [48], reflecting the spatial agglomeration relationship and the active degree of the LUCs zones. On another, if there is a more significant conflict difference between adjacent units, the interference of land-use patterns in its adjacent units will be stronger, and the possibility of causing LUCs will be higher [23]. Thus, the methods of cold- and hot-spot analysis and neighborhood analysis were used to reflect the spatial agglomeration relationship and potential risk of LUCs in the UCC. Taking the "3 × 3" rectangular component as the range, the standard deviation of the central element can be obtained through the neighborhood analysis function, to judge the influence degree of the surrounding units on the main unit [23,47]. The potential land-use conflicts risk index (PLUCRI) was constructed to analyze the potential LUCs. The specific formula is below:

$$L\_i = \frac{\sum \left| G\_{nbr\_{ij}} - G\_{scr\_i} \right|}{N\_{nbr\_i}} \tag{8}$$

where *L<sup>i</sup>* is the PLUCRI of the *i*-th evaluation unit, and the larger the value is, the greater the possibility of conflict in the evaluation unit would be. *Gscr<sup>i</sup>* denotes the conflict intensity of the *i*-th evaluation unit, and it is represented by eight LUCs zones (for quantitative calculation, I–VIII is represented by 1, 2, . . . 8, respectively). *Gnbrij* is the conflict intensity of the *j*-th neighborhood unit's *i*-th evaluation unit. *Nnbr<sup>i</sup>* is the number of neighborhood units of the *i*-th evaluation unit [23].

#### **3. Results**
