TL;DR
This study investigates the increasing problem of arable land abandonment in China's developed Yangtze River Delta region, using Pingyang County as a case study. By analyzing remote sensing data and employing the FLUS model, the research reveals a rapid expansion of abandoned land and predicts its future distribution, highlighting the need for targeted land protection policies.
What Problem Does This Paper Address?
Arable land abandonment is a growing concern globally, particularly in China, posing a significant threat to national food security and social stability. This phenomenon is exacerbated by rapid urbanization and economic development, which often lead to rural depopulation and a shift away from agricultural livelihoods. While previous research has touched upon the causes and current status of land abandonment, there's a need for a deeper understanding of its spatio-temporal dynamics, especially at the landscape patch level, and for accurate predictions of future trends. This paper aims to fill this gap by analyzing the spatial and temporal evolution of abandoned arable land in a developed county and predicting its future distribution to inform effective land management strategies.
Methods
The study employs a multi-faceted approach combining remote sensing, landscape pattern analysis, and a land use simulation model.
- Remote Sensing and GIS: Landsat TM imagery from 2000, 2010, and 2018 was used to identify and map abandoned arable land. This was complemented by land-use change survey data and high-resolution satellite and aerial imagery for accuracy.
- Landscape Pattern Index: To quantify the characteristics of abandoned land, several landscape pattern indices were calculated using Fragstats 4.2 software and interpolated in ArcGIS 10.3. These include:
- Patch Density (PD): Measures the fragmentation of abandoned land.
- Largest Patch Index (LPI): Indicates the dominance of the largest abandoned patch.
- Mean Patch Size (MPS): Represents the average size of abandoned patches.
- Landscape Shape Index (LSI): Quantifies the complexity of the shape of abandoned patches.
- FLUS Model (Future Land Use Simulation): This model was used to predict the spatial and temporal distribution of abandoned arable land in 2028. The FLUS model integrates an artificial neural network (ANN) for suitability probability calculation with an adaptive inertial competition mechanism to simulate land use changes based on historical trends and influencing factors.
Datasets
- Landsat TM Remote Sensing Images: Acquired for the farming seasons (May-September) of 2000, 2010, and 2018, with a spatial resolution of 30m.
- 1:10,000 Land-Use Change Survey Data: Provided by the Pingyang County Bureau of Land and Resources.
- Google Earth Satellite Images: With a spatial resolution of 0.13m.
- Multi-temporal Aerial Images: With a spatial resolution of 1m.
Key Insights and Ideas
- Rapid Expansion: Arable land abandonment isn't a slow creep but a rapid expansion, with the abandoned area in Pingyang County increasing nearly 12-fold in just 18 years. This highlights the urgency of the issue in economically dynamic regions.
- Geographic Concentration: Abandonment isn't uniform. High rates are consistently observed in the western and northwestern mountainous areas, likely due to challenging topography and farming conditions.
- Shifting Landscape Patterns: The "character" of abandoned land is changing. Fragmentation is increasing, particularly in western mountainous areas, while the complexity of abandoned land shapes is also growing, moving from the center towards northern and western regions. This suggests a more irregular and dispersed pattern of abandonment over time.
- Future Hotspots: Predictions for 2028 indicate a further concentration of high abandonment rates in the northwest, with rates potentially reaching 15.76% to 24.89%. This suggests that existing trends will likely continue and intensify in these specific areas.
- Policy Implications: The study's findings directly inform land protection policies by identifying high-risk areas and understanding the spatio-temporal dynamics, enabling more targeted interventions.
Key Results
- Area Increase: The total area of abandoned arable land in Pingyang County grew from 441.58 hm² (0.42% of total arable land) in 2000 to 5797.40 hm² (5.46%) in 2018.
- Abandonment Rate Trend: The abandonment rate increased significantly from 2000 to 2018, with areas of high abandonment rate expanding and becoming more concentrated.
- Landscape Index Changes:
- Patch Density (PD) increased from 0.0038 to 0.6013, indicating increased fragmentation.
- Largest Patch Index (LPI) decreased from 3.2051 to 0.4649, suggesting a shift away from large contiguous abandoned areas.
- Mean Patch Size (MPS) initially decreased but then rebounded, indicating a complex pattern of patch size changes.
- Landscape Shape Index (LSI) increased from 27.5642 to 109.4218, showing a trend towards more complex and irregular shapes of abandoned land.
- FLUS Model Accuracy: The FLUS model achieved an overall accuracy of 0.875 and a Kappa coefficient of 0.71 for predicting 2018 abandonment, validating its reliability.
- 2028 Prediction: The model predicts a further increase in abandoned land area to 7138.80 hm² by 2028, with an abandonment rate of 6.93%. High abandonment rates are expected to be more concentrated in the northwest.
Limitations and Future Work
- FLUS Model Limitations: The paper acknowledges that the FLUS model's reliance on proximity diffusion can lead to radial expansion, which might not perfectly capture real-world abandonment patterns influenced by diverse, unpredictable factors. This can result in some discrepancies between simulated and actual results at local scales.
- Future Work Suggestions: The authors propose developing more targeted countermeasures for arable land protection, considering the specific ecological environment and resource endowments of the study area. This includes strategies like financial subsidies for farmers, promoting large-scale land operations, and developing characteristic agriculture in mountainous regions. The study also suggests that these findings have broader relevance for governing arable land abandonment in other developed regions and for ensuring national and global food security.
Summary updated