Prioritizing Abandoned Mine Lands Rehabilitation: Combining Landscape Connectivity and Pattern Indices with Scenario Analysis Using Land-Use Modeling

Liping Zhang, Shiwen Zhang, Yajie Huang, An Xing, Zhiqing Zhuo, Zhongxiang Sun, Zhen Li, Meng Cao, Yuanfang HuangISPRS Int. J. Geo Inf 2018 · DOI

TL;DR

This paper introduces a novel approach to prioritize the rehabilitation of abandoned mine lands (AML) by combining landscape connectivity and pattern indices with land-use modeling. By simulating different reclamation scenarios, the study identifies that prioritizing forestation of AML significantly enhances landscape connectivity, offering a data-driven strategy for sustainable land management.

What Problem Does This Paper Address?

Abandoned mine lands (AML) pose a significant challenge for land management and ecological restoration due to their degraded state and often dispersed nature. Effective rehabilitation requires prioritizing which AML patches to address first and understanding the long-term impact of different land-use choices on the surrounding landscape. Existing approaches often lack quantitative methods for assessing landscape connectivity, especially in the context of post-mining restoration where significant land-use changes are anticipated. This paper aims to fill this gap by developing a systematic framework to prioritize AML reclamation based on landscape connectivity and to evaluate the effectiveness of different reclamation strategies.

Methods

The study employs a multi-faceted approach:

  • Patch-Scale Connectivity: It first quantifies the connectivity of existing AML patches using a combination of patch size and a proximity index. The proximity index, adapted from Gustafson and Parker and calculated using FRAGSTATS, measures how close and contiguous AML patches are to each other within a defined search radius. This helps identify isolated patches versus those that are part of larger, more connected areas.
  • Prioritization: Based on the patch-scale connectivity analysis, AML patches are prioritized for reclamation, with a focus on larger and more connected patches to meet predetermined reclamation targets.
  • Scenario Analysis: Three distinct land-use reclamation scenarios are developed:
  • Fertility-dependent cultivation or forestry: Prioritizes cultivated land based on soil fertility, with remaining areas converted to forest.
  • Elevation-dependent cultivation or forestry: Prioritizes forest land for higher elevations and cultivated land for lower elevations with good soil attributes.
  • Urbanization, cultivation and forestry: Focuses on transforming areas into built-up land for industrial development, with scattered areas for cultivation and the rest for forest.
  • Land-Use Modeling: The Conversion of Land Use and its Effects at the Small Regional Extent (CLUE-S) model is used to predict future land-use changes (excluding AML) in the study area by 2020. This model incorporates ten driving variables (e.g., elevation, slope, distance to infrastructure, soil organic matter, socioeconomic factors) to simulate land-use transitions.
  • Landscape-Scale Connectivity: After simulating future land-use changes and applying the reclamation scenarios to AML, several landscape-scale graph metrics are used to assess the overall connectivity of the post-mining landscape. These include Mean Patch Size (MPS), Number of Patches (NP), Contagion (CONTAG), Aggregation Index (AI), Shape Index (SHAPE), and the Integral Index of Connectivity (IIC).

Datasets

  • Land Use Map (2007): Provided by the Beijing Municipal Bureau of Land and Resources, detailing eight land-use categories.
  • Soil Map: Used in conjunction with the land-use map for scenario 1 evaluation.
  • Digital Elevation Model (DEM): Used to derive elevation and slope data.
  • Road Network, River Network, Town and Rural Residential Site Data: Used to calculate distances for transportation accessibility.
  • Soil Organic Matter Data: Obtained from the Beijing Digital Soil System.
  • Statistical Yearbooks: Provided socioeconomic data (population density, per capita income).

The study uses existing data for the Mentougou District in Beijing, China, and does not introduce new benchmark datasets.

Key Insights and Ideas

  • Connectivity is Key for Prioritization: The study demonstrates that simply looking at patch size isn't enough; proximity and aggregation of AML patches are crucial for effective prioritization. Patches that are larger and closer to other AML patches offer a better starting point for integrated reclamation.
  • Forestation Enhances Connectivity: The most significant finding is that the scenario prioritizing forestation (Scenario 2) resulted in the highest landscape connectivity, as measured by multiple indices, particularly IIC. This suggests that converting AML into forest land is a highly effective strategy for ecological restoration.
  • Scenario Modeling is Essential: Simulating different land-use futures is critical. The CLUE-S model provides a realistic projection of surrounding land-use changes, allowing for a more accurate assessment of how AML reclamation will impact the broader landscape.
  • Proximity Index Nuances: The paper highlights a limitation of the proximity index when dealing with vastly different patch sizes, where large patches can disproportionately influence the connectivity scores of smaller, nearby patches. The proposed method addresses this by considering both size and proximity.

Key Results

  • Prioritization: Six "huge patches" (over 100 ha) were identified as initial priorities for reclamation.
  • Scenario Performance:
    • Scenario 1 (Fertility-dependent): Reclaimed 2092 ha as cultivated land (58.55%), 1084 ha as forest (30.34%), and 397 ha as construction land (11.11%).
  • Scenario 2 (Elevation-dependent): Reclaimed 312 ha as cultivated land (8.73%), 3078 ha as forest (86.15%), and 183 ha as construction land (5.12%). This scenario resulted in the largest area of forest reclamation.
  • Scenario 3 (Urbanization): Reclaimed 281 ha as cultivated land (7.86%), 721 ha as forest (20.18%), and 2571 ha as construction land (71.96%).
  • Connectivity Metrics: Scenario 2 (forested) consistently showed the highest landscape connectivity across several metrics, including the largest MPS, lowest NP, high CONTAG and AI, and the highest IIC (0.6644), indicating greater aggregation and improved habitat connectivity.

Limitations and Future Work

  • Generalizability: The authors acknowledge that the study is specific to the Mentougou District and its unique characteristics. The results and chosen indicators may not be directly generalizable to other regions without further adaptation.
  • Indicator Selection: Future research should explore a wider range of indicators to better describe and characterize landscape connectivity, especially in diverse regional contexts.
  • Suitability Factors: Mined land reclamation is influenced by various factors beyond connectivity, such as environmental hazards, economics, and local suitability. Future studies should integrate these aspects to develop more comprehensive optimal land reclamation scenarios.
  • Dynamic Modeling: While the CLUE-S model predicts future land use, the authors suggest that a more dynamic and iterative approach to connectivity modeling and scenario planning could further enhance the understanding of long-term impacts.

Summary updated

BibTeX
@article{zhang2018prioritizing,
  author = {Liping Zhang and Shiwen Zhang and Yajie Huang and An Xing and Zhiqing Zhuo and Zhongxiang Sun and Zhen Li and Meng Cao and Yuanfang Huang},
  title = {{Prioritizing Abandoned Mine Lands Rehabilitation: Combining Landscape Connectivity and Pattern Indices with Scenario Analysis Using Land-Use Modeling}},
  journal = {ISPRS Int. J. Geo Inf.},
  year = {2018},
  doi = {10.3390/ijgi7080305},
}
Figure