Preamble

ARID LAND GEOGRAPHY Vol. 48 No. 7 Jul. 2025
Spatiotemporal Variation and Prediction of Habitat Quality in the Tongwei Landslide Area of Gansu Province Based on the PLUS-InVEST Model

ZHANG Xiaoming¹,², SU Xing², ZHANG Jun¹, JIA Jing¹,²
(1. College of Resources and Environmental Science, Gansu Agricultural University, Lanzhou 730070, Gansu, China;
2. Institute of Geological Natural Disaster Prevention and Control, Gansu Academy of Sciences, Lanzhou 730000, Gansu, China)

Abstract: The study of the spatial and temporal evolution of habitat quality in the landslide area of Tongwei County, Gansu Province, is crucial for the ecological sustainable development of Tongwei County and similar regions in northwest China. First, the patch-based land use simulation model was employed to predict land use types for the year 2035. Subsequently, the integrated valuation of ecosystem services and tradeoffs model was utilized to analyze and predict ecosystem quality as well as the temporal and spatial evolution characteristics of habitat quality in Tongwei County and the landslide area from 1985 to 2020 and projected to 2035. Finally, the geodetector model was applied to identify the driving factors behind changes in habitat quality in the landslide area of Tongwei County. The results showed the following. (1) From 2020 to 2035, all land use types in Tongwei County are expected to expand and shift, with the most significant transition occurring from arable land to grassland. Arable land and grassland will continue to dominate the land use types, whereas other types will represent a relatively small proportion. (2) Spatially, habitat quality in Tongwei County and the landslide area exhibited an increasing trend from south to north, remaining predominantly low and lower grades. Habitat degradation in the landslide area was dominated by moderate and high degradation, with decreasing degradation from south to north. (3) Temporally, from 1985 to 2035, habitat quality in Tongwei County and the landslide area followed a pattern of decreasing, then increasing, and subsequently decreasing again. In Tongwei County, the linear trend of the average habitat quality showed a decline, whereas in the landslide area, it showed an increase. The average habitat degradation followed a pattern of decreasing followed by increasing, with a linear fit indicating an upward trend. (4) The normalized vegetation index (NDVI) is the most significant factor affecting spatial differentiation of habitat quality in landslide areas. Interactions among factors are predominantly characterized by nonlinear enhancement, with the interaction between NDVI and average annual precipitation being the strongest.

Keywords: landslide area; PLUS model; InVEST model; habitat quality; geographical detector; Tongwei County

1.1 Study Area Overview

Tongwei County is located in central Gansu Province, between 104°57′~105°38′E and 34°55′~35°29′N, covering a total area of 2907 km². The terrain slopes from high in the northwest to low in the southeast, with elevations ranging from 1376 m to 2475 m and an average elevation of 1970.5 m. The area features typical loess hilly and gully landforms. Situated at the intersection of China's east-west and north-south seismic belts, seismic activity directly influences the occurrence of earthquake-induced landslides. These landslides are predominantly medium to large in scale. Land use in the landslide area is primarily arable land and grassland, with some construction land distributed throughout the region [FIGURE:1].

1.2 Data Sources

Land use data (1985–2020) were obtained from the China Land Cover and Dynamic Dataset (https://zenodo.org/records/8176941). Climate data, including annual average temperature and annual average precipitation, were sourced from the Qinghai-Tibet Plateau Data Center (http://www.gscloud.cn) at 1 km resolution. Digital Elevation Model (DEM) data were obtained from the Geospatial Data Cloud. Evapotranspiration data were obtained from NASA (https://www.nasa.gov/). Normalized Difference Vegetation Index (NDVI) data were derived from MODIS products. Slope and aspect data were extracted using GIS techniques.

Socioeconomic data: Population density data were obtained from the University of Florida Geography Department and Emerging Pathogens Institute (https://www.worldpop.org/). GDP data were obtained from the Chinese Academy of Sciences Resource and Environmental Science Data Center (https://www.resdc.cn/). Distance to roads (including secondary, tertiary, and quaternary roads; distances to national highways, provincial roads, county roads, and township roads) and distance to administrative centers (towns and county government) were obtained from the Geographic Remote Sensing Ecology Network (https://www.gisrs.cn). All data were uniformly projected to WGS_1984_UTM_zone_48N with a 30 m spatial resolution.

To investigate the driving forces of habitat quality changes in the landslide area, eleven factors were selected: annual average temperature (X₁), annual average precipitation (X₂), slope (X₃), aspect (X₄), evapotranspiration (X₅), NDVI (X₆), DEM (X₇), distance to rivers (X₈), population density (X₉), GDP (X₁₀), and distance to roads (X₁₁).

1.3.1 PLUS Model

To provide data support for future habitat quality assessment in landslide areas, the patch-based land use simulation (PLUS) model was employed to predict land use changes from 2020 to 2035. The PLUS model integrates a rule mining method based on land expansion analysis with a cellular automata model based on a multi-type random seed mechanism. The land expansion analysis strategy extracts the expansion portions of various land use types and employs the random forest algorithm to mine the relationship between each land use type expansion and contributing factors, obtaining development probabilities and influence weights of driving factors for each land use type.

The formula for the rule mining method based on land expansion analysis is:
$$P_{d}^{k}(x) = \sum_{n=1}^{N} \frac{1}{M} I(h_n(X) = k)$$

Where: $P_{d}^{k}(x)$ is the probability of spatial unit i converting to land use type k; d=0 indicates the plot cannot be converted to land use type k, d=1 indicates the plot can be converted to land use type k; X represents driving factors; $h_n(X)$ is the predicted land use type calculated when the decision tree is n; I is the indicator function; M is the total number of decision trees.

Based on the development probabilities of various land use types, the cellular automata model with multi-type random seed mechanism and threshold decreasing mechanism is used to dynamically simulate future land use scenarios. The formula for the cellular automata model is:
$$P_{k}^{t} = P_{d}^{k}(X) \times \Omega_{k}^{t} \times D_{k}^{t}$$

Where: $P_{k}^{t}$ is the overall conversion probability of spatial unit i converting to land use type k at time t; $P_{d}^{k}(X)$ is the probability of land use type k growth at spatial unit i when d=1; $\Omega_{k}^{t}$ is the neighborhood weight of land use type k at spatial unit i at time t; $D_{k}^{t}$ is the adaptive driving coefficient for future demand of land use type k, depending on the gap between current land quantity and target demand.

Neighborhood weight reflects the expansion capacity of various land use types, with parameters in [0, 1]. This study uses 0.5 to calculate expansion capacity, where larger values indicate stronger expansion capacity of the land use type.

1.3.2 InVEST Model

The InVEST model was used to analyze habitat quality and habitat degradation in Tongwei County and its landslide area. The model establishes connections between different land use types and threat sources, using landslide area land use type data to define response levels of different land use types to threat sources, simulating habitat quality and degradation patterns. The model assumes habitat quality is proportional to ecosystem stability.

Habitat quality formula:
$$Q_{xj} = H_j \times \left(1 - \left(\frac{D_{xj}^z}{D_{xj}^z + k^z}\right)\right)$$

Where: $Q_{xj}$ is the habitat quality index of grid x in habitat type j, ranging from 0 to 1; $H_j$ is habitat suitability of habitat type j; k is the half-saturation constant, defaulting to 0.5, representing half of the maximum habitat degradation value in Tongwei County; $D_{xj}$ is the habitat degradation index of grid x in habitat type j; z is the model default parameter.

Habitat degradation index formula:
$$D_{xj} = \sum_{r=1}^{R} \sum_{y=1}^{Y_r} \left(\frac{W_r}{\sum_{r=1}^{R} W_r}\right) \times r_y \times i_{rxy} \times \beta_x \times S_{jr}$$

Where: $D_{xj}$ is the habitat degradation index of grid x in habitat type j; $Y_r$ is a set of grids on the threat raster layer; $r_y$ indicates whether grid y provides threat source r; $W_r$ is the threat weight of threat factor r; $\beta_x$ is the resistance of grid x; $S_{jr}$ is the sensitivity of habitat j to threat factor r; $i_{rxy}$ is the impact distance of threat factor r on habitat in grid x to grid y.

Due to different threat factors, $i_{rxy}$ has linear and exponential decay in space. Formulas:
Linear decay: $i_{rxy} = 1 - \frac{d_{xy}}{d_r}$
Exponential decay: $i_{rxy} = \exp\left(-\frac{2.99d_{xy}}{d_r}\right)$

Where: $d_{xy}$ is the linear distance between grid x and grid y; $d_r$ is the maximum impact distance of threat factor r.

Based on the InVEST model user manual and existing research, arable land, construction land, and bare land with strong human disturbance were set as threat factors to habitat quality. Impact weights, maximum impact distances, and impact types were determined according to relevant research, and habitat suitability and relative sensitivity of different land use types to threat factors were established [TABLE:1] and [TABLE:2].

1.3.3 Geodetector

Habitat quality changes are related to natural and socioeconomic factors. Land use change under the combined influence of natural environmental changes and human activities is the known driving force affecting habitat quality. To explore driving factors of habitat quality changes in landslide areas, the geodetector model was used. This model considers spatial heterogeneity among factors and is widely applied in ecological environment management and land use research.

The formula is:
$$q = 1 - \frac{\sum_{h=1}^{L} N_h \sigma_h^2}{N \sigma^2}$$

Where: q is the detection result of habitat quality impact factors, ranging from 0 to 1; h is the classification number of factors; L is the total number of evaluation units; $N_h$ and $\sigma_h^2$ are the variance of each zone; N is the total number of habitat quality units; $\sigma^2$ is the total variance of habitat quality in the region. Factor interaction detection examines the explanatory power of pairwise factor interactions on habitat quality, divided into 5 types [TABLE:3].

2.1.1 Land Use Expansion Analysis

Combining natural environment and socioeconomic conditions of Tongwei County, eleven driving factors were selected: annual average temperature, annual average precipitation, slope, aspect, evapotranspiration, NDVI, DEM, distance to rivers, population, GDP, distance to roads, and distance to towns. Using the PLUS model and random forest algorithm, contribution degrees of different factors to land type expansion were obtained.

Research found that slope contributed most to arable land expansion, followed by aspect and distance to secondary roads, all exceeding 0.15. NDVI and distance to secondary roads contributed most to forest land expansion, exceeding 0.10. Slope contributed most to grassland expansion, followed by aspect and distance to secondary roads, exceeding 0.15. Distance to county government contributed most to construction land expansion, followed by distance to towns and distance to quaternary roads, exceeding 0.10 [FIGURE:2].

2.1.2 Accuracy Verification

Based on the PLUS model, land use types in Tongwei County for 2020 were simulated and compared with actual 2020 land use types for accuracy verification. The Kappa coefficient was 0.82, FoM was 0.21, and accuracy was 0.91, indicating high reliability. Therefore, land use types in 2035 under natural development scenarios can be simulated based on 2020 land use types [FIGURE:3].

2.1.3 Land Use Change Analysis

From 2020 to 2035, land use types in Tongwei County will undergo mutual expansion and transition, with the most significant conversion from arable land to grassland, while grassland to arable land conversion is minimal. Forest land will slightly expand to arable land. Water bodies and unused land will remain basically unchanged. Land use types will still be dominated by arable land and grassland, with smaller proportions of forest land, water bodies, bare land, and construction land [FIGURE:4].

2.2 Habitat Quality in Tongwei County

To more accurately characterize spatial evolution patterns of habitat quality, based on landslide area conditions and existing research, the equal interval method was used to divide Tongwei County habitat quality into five grades: low (0–0.2), lower (0.2–0.4), medium (0.4–0.6), higher (0.6–0.8), and high (0.8–1.0) [FIGURE:5].

Spatially, low-grade habitat quality is mainly concentrated in southern and central Tongwei County, where human activities are frequent and land use is dominated by arable land. Higher and high-grade habitat quality is mainly concentrated in northern Tongwei County, where the Huajialing National Nature Reserve is located and human disturbance is minimal. Overall, habitat quality in Tongwei County shows a general increasing trend from south to north, dominated by low and lower grades [FIGURE:5].

Temporally, from 1985 to 2035, habitat quality in Tongwei County is dominated by low and lower grades, with significant area changes among grades. The proportion of low, medium, and high-grade habitat quality showed an increasing linear trend, while lower and higher grades showed a decreasing linear trend [FIGURE:6]. The average habitat quality index ranged from 0.269 to 0.293, showing an overall pattern of decreasing, then increasing, then decreasing, with a declining linear trend. The average habitat quality index is projected to decrease to 0.269 in 2035 [FIGURE:7].

2.3 Habitat Quality in Landslide Area

Spatially, low and lower-grade habitat quality in landslide areas is mainly concentrated in central and southern Tongwei County landslide zones. Medium-grade habitat quality is mainly concentrated in northern and western landslide zones. Higher-grade habitat quality is mainly concentrated in northern landslide zones. Low and lower-grade habitat quality accounts for a large proportion, with wide distribution and obvious fragmentation [FIGURE:8].

Temporally, from 1985 to 2035, high-grade habitat quality is absent in landslide areas, indicating lower habitat quality than in Tongwei County overall. Landslide area habitat quality is dominated by low and lower grades, with significant proportion changes. Low-grade habitat quality proportion increased from 33.36% to 33.20% (linear increase), lower-grade decreased from 65.70% to 62.92% (linear decrease), medium-grade increased from 3.60% to 3.05% (linear increase), higher-grade decreased from 0.22% to 0.12% (linear decrease), and high-grade decreased from 0.04% to 0.04% (linear decrease). The average habitat quality index ranged from 0.239 to 0.254, showing an overall linear increase, projected to increase to 0.254 in 2035, following a pattern of decreasing, then increasing, then decreasing [FIGURE:10].

2.4 Habitat Degradation in Landslide Area

Landslide area habitat degradation was classified into: mild degradation (0–0.24), moderate degradation (0.24–0.38), high degradation (0.38–0.48), and extreme degradation (0.48–0.70] [FIGURE:11].

Spatially, habitat degradation in Tongwei County landslide areas shows a decreasing trend from south to north, with higher degradation in southern landslide zones and lower degradation in northern zones [FIGURE:11]. Temporally, from 1985 to 2035, moderate and high degradation accounted for over 60.00% of the landslide area. Mild and moderate degradation proportions showed a decreasing linear trend, while high and extreme degradation showed an increasing linear trend. The average habitat degradation index showed an increasing linear trend. In 2035, extreme degradation proportion is projected to increase while other degradation types decrease, with the average degradation index continuing to rise [FIGURE:12].

2.5 Driving Forces of Habitat Quality in Landslide Area

Single factor detection revealed significant differences among driving factors in explaining habitat quality in landslide areas. The explanatory power (q values) from largest to smallest were: NDVI (0.31), average annual precipitation (0.21), average annual temperature (0.19), DEM (0.16), distance to rivers (0.12), GDP (0.11), and distance to roads (0.10) [TABLE:4]. NDVI is the most critical factor affecting spatial differentiation of habitat quality in landslide areas.

Interaction detection showed that pairwise factor interactions had q values greater than single factors, with interactions showing nonlinear enhancement or dual-factor enhancement, predominantly nonlinear enhancement. The interaction between NDVI and average annual precipitation was strongest (q=0.52), followed by NDVI and average annual temperature (0.49), and average annual precipitation and average annual temperature (0.45). NDVI and GDP interaction was also significant (q=0.41) [FIGURE:14]. Natural factors dominate habitat quality changes, while socioeconomic factors also play a role [TABLE:5].

3 Discussion

Accurately predicting future land use structure is challenging. Although the PLUS model shows good simulation performance, subjective factors affecting predictions cannot be avoided, particularly policy impacts on land use. Therefore, improving model prediction accuracy requires full consideration of actual land use conditions. Field investigation revealed that Tongwei County actively implemented the Grain for Green Program in 2000, with large areas of returning farmland to forest. The survival rate of afforestation gradually increased to a peak in 2010, with ecological benefits becoming apparent, while the second round of Grain for Green began showing effects only after 2015.

According to Tongwei County's actual situation, this study used 5-year intervals for predictions, fully considering policy impacts on land use structure. Comparison of major contributing expansion factors showed consistency with previous research. Tongwei County's habitat quality changed significantly in 1985–1990, closely related to policy factors. After five consecutive "No. 1 Central Documents" clarified the household contract responsibility system, Tongwei County's arable land expanded, and landslide area land use types shifted toward arable land. This study shows that the average habitat quality index decreased from 1985–2000, then increased with the Grain for Green policy implementation after 2000. When setting threat factor parameters in the InVEST model, field surveys should be strengthened to improve simulation accuracy while drawing on previous results.

Tongwei County is located in a semi-arid region of northwest China, significantly affected by average annual temperature and precipitation, with few high habitat quality areas. Meanwhile, individual landslide zones are small and don't intersect with high habitat quality areas. Dingxi City and other areas in central Gansu have high population density, low vegetation coverage, and arable land-dominated landscapes, with habitat quality means around 0.25, consistent with this study.

Habitat quality change results from the combined effects of natural and socioeconomic factors. This study used the geodetector model to analyze impacts on spatial evolution patterns of habitat quality and identify dominant controlling factors. However, data acquisition and quantification of other socioeconomic factors remain challenging, and socioeconomic factor support is somewhat insufficient, requiring further breakthroughs in future research.

4 Conclusion

(1) From 2020 to 2035, land use types in Tongwei County will undergo mutual conversion, with significant conversion from arable land to grassland and minimal conversion from grassland to arable land. Land use will remain dominated by arable land and grassland, with smaller proportions of other types.

(2) Spatially, from 1985 to 2035, habitat quality in Tongwei County and its landslide area shows an increasing trend from south to north, with overall low habitat quality dominated by low and lower grades. Habitat degradation in landslide areas is dominated by moderate and high degradation, decreasing from south to north.

(3) Temporally, from 1985 to 2035, the average habitat quality index in Tongwei County follows a decreasing-increasing-decreasing pattern, with a declining linear trend, projected to decrease to 0.269 in 2035. In contrast, the landslide area average habitat quality index follows the same pattern but with an increasing linear trend, projected to increase to 0.254 in 2035. The average habitat degradation index follows a decreasing then increasing pattern, projected to continue increasing in 2035.

(4) NDVI is the most critical factor affecting spatial differentiation of habitat quality in landslide areas. Interactions between any two factors have greater impact than single factors, with NDVI and average annual precipitation interaction being strongest. Factor interactions predominantly show nonlinear enhancement or dual-factor enhancement, mainly nonlinear enhancement.

References

[1] Lie R. Drylands extent and environmental issues: A global approach[J]. Earth Science Reviews, 2016, 161: 259-278.

[2] Zhang Yan, Liang Jinyi, Wei Lingxia, et al. Change in spatial temporal patterns of eco-environment quality and its driving forces for the Lanzhou City[J]. Journal of Arid Land Resources and Environment, 2023, 37(10): 65-74.

[3] Li Xuchun, Bo Yongming, Wang Lijie, et al. Development process and prospects of soil and water conservation research in Dingxi[J]. Soil and Water Conservation in China, 2023(11): 1-4.

[4] Zhang Nan, Zhang Yan, Wang Jiaxi, et al. Quantity and morphological parameters of gullies in small watersheds in the hilly gully Loess Plateau[J]. Journal of Soil and Water Conservation, 2023, 37(3): 109-115.

[5] Newbold T, Hudson L N, Hill S L L, et al. Global effects of land use on local terrestrial biodiversity[J]. Nature, 2015, 520(7545): 45-50.

[6] Zheng Jiancheng, Xie Binggeng, You Xibin, et al. Evolution of habitat quality and its influencing factors in the different terrain gradient of the Guangdong-Hong Kong-Macao Greater Bay Area from 1980 to 2020[J]. Economic Geography, 2022, 42(8): 41-50.

[7] Zhang Ting, Hu Yuzhu, Hu Haihui, et al. Prediction of land use and habitat quality in Harbin City based on the PLUS-InVEST model[J]. Environmental Science, 2024, 45(8): 4709-4721.

[8] Wu Aibin, Chen Fuguo, Zhao Yanxia, et al. Multi-scenario simulation of construction land expansion and its impact on ecosystem carbon storage in Beijing-Tianjin-Hebei urban agglomeration[J]. Environmental Science, 2024, 45(5): 2828-2839.

[9] Zhao Jiyu, Cui Liu, Wang Jia, et al. Multi-scenario simulation of urban green space development and construction timeline based on PLUS model analysis: A case study of the central Zhanjiang City[J]. Acta Ecologica Sinica, 2023, 43(15): 6307-6320.

[10] Hu Jixi, Le Xianwen, Wang Weilin, et al. Temporal and spatial evolution and prediction of ecosystem carbon storage in Jiangxi Province based on PLUS-InVEST model[J]. Environmental Science, 2024, 45(6): 3284-3296.

[11] Ouyang Xiao, He Qingyun, Zhu Xiang, et al. Simulation of impacts of urban agglomeration land use change on ecosystem services value under multi-scenarios: Case study in Changsha-Zhuzhou-Xiangtan urban agglomeration[J]. Economic Geography, 2020, 40(1): 93-102.

[12] Wang Jinfeng, Xu Chengdong. Geodetectors: Principles and prospects[J]. Acta Geographica Sinica, 2017, 72(1): 116-134.

[13] Yuan Yuan, Sheng Yan, Liu Linfu, et al. Spatial-temporal evolution characteristics and driving influence mechanism of habitat quality in Kuye River Basin[J]. Forest and Grassland Resources Research, 2023(6): 67-74.

[14] An Wenju, Yu Yifan, Hao Shanshan, et al. Evolution and response analysis of habitat quality in more sediments and coarse sediments region of northern Shaanxi based on land use change[J]. Arid Land Geography, 2024, 47(3): 474-484.

[15] Luo Shuqi, Hu Xiaomeng, Sun Yuan, et al. Multi-scenario land use change and its impact on carbon storage based on coupled PLUS-InVEST model[J]. Chinese Journal of Eco-Agriculture, 2023, 31(2): 300-314.

[16] Wu Yanxia, Liu Fangnan, Chen Baotong, et al. Spatial and temporal evolution and drivers of habitat quality of urban agglomeration in lower Yellow River Basin[J]. Bulletin of Soil and Water Conservation, 2023, 43(4): 396-404.

[17] Cheng Jing, Wang Peng, Chen Hongxiang, et al. Spatiotemporal evolution of habitat quality in the Weihe River Basin and its topographic gradient effects and influencing factors[J]. Arid Land Geography, 2023, 46(3): 481-491.

[18] Ji Yongfan, Jia Lujing, Yang Lianan, et al. Spatio-temporal evolution and prediction analysis of habitat quality in Yulin City coupled with InVEST-PLUS model[J]. Journal of Soil and Water Conservation, 2023, 37(1): 123-132.

[19] Xie Yuchu, Gong Jie, Zhang Suxin, et al. Spatiotemporal change of landscape biodiversity based on InVEST model and remote sensing technology in the Bailong River Watershed[J]. Scientia Geographica Sinica, 2018, 38(6): 979-986.

[20] Gong J, Xie Y C, Cao E J, et al. Integration of InVEST habitat quality model with landscape pattern indexes to assess mountain plant biodiversity change: A case study of Bailongjiang Watershed in Gansu Province[J]. Journal of Geographical Science, 2019, 29(7): 1193-1210.

[21] Su Junde, Zhao Xiaojiong, Li Guoxia, et al. Spatial-temporal characteristics and driving factors of habitat quality in Qilian Mountain National Nature Reserve[J]. China Environmental Science, 2024, 44(5): 2595-2605.

[22] Yan Li, Cao Guangchao, Kang Ligang, et al. Analysis of spatial-temporal changes in habitat quality and driving factors in Gonghe County using the InVEST model[J]. Arid Zone Research, 2024, 41(2): 314-325.

[23] Zhu Zengyun, Kasimu Alimujiang. Spatial-temporal evolution of habitat quality and its influencing factors in Yili Valley based on geographical detector[J]. Chinese Journal of Ecology, 2020, 39(10): 3408-3420.

[24] Wang Juan, Zhang Huabing, Li Yufeng, et al. Assessment on overwintering habitat quality of red-crowned cranes in Yellow Sea wetlands of Yancheng and its management strategies[J]. Wetland Science, 2022, 20(3): 334-340.

[25] Zhao Xiaojiong, Wang Jian, Su Junde, et al. Assessment of habitat quality and degradation degree based on InVEST model and Moran index in Gansu Province, China[J]. Transactions of the Chinese Society of Agricultural Engineering, 2020, 36(18): 301-308.

[26] Gao Liping. Preliminary results analysis of the implementation of Grain for Green Project and subsequent industrial projects in Tongwei County[J]. Agricultural Science Technology and Information, 2010(12): 43-44.

[27] Zhou Rong, Yan Jin, Zhou Xiaolei, et al. Quantitative analysis of afforestation effectiveness in demonstration site of returning farmland to forest in Tongwei County, Gansu Province[J]. Gansu Science and Technology, 2005(5): 22-23.

[28] Zhou Yeze, Qin Tian, Pei Yang, et al. Land use simulation and multi-scenario prediction based on PLUS model and ground substrate[J]. Geological Review, 2024, 70(Suppl. 1): 315-318.

[29] Xiong Lianghua, Chen Jin. Household contract responsibility system: Farmers spontaneously contract production to households disputes resolved by Central Committee of CPC to promote rural reform[J]. Xinxiang Review, 2020(13): 49-53.