TL;DR
This paper provides a comprehensive national-scale analysis of land-use and land-cover change (LULCC) in India from 1985-2005, integrating high-resolution satellite data with detailed village-level socioeconomic data. It reveals that while cropland-fallow land conversions are driven by factors like labor shortages and land fragmentation, forest loss is primarily linked to agricultural expansion in low-productivity areas, industrial development, and over-reliance on forest resources.
What Problem Does This Paper Address?
India faces immense pressure on its land resources due to a large population, high human and animal density, and rising living standards. Understanding the dynamics and drivers of land-use and land-cover change (LULCC) is crucial for effective land-use planning, enhancing food production, and mitigating environmental degradation. Previous studies often lacked national-scale coverage, relied on coarser data, or focused solely on land cover mapping without deeply integrating socioeconomic determinants at a granular level. This paper aims to fill this gap by providing a detailed, village-level analysis of LULCC drivers across India over two decades.
Methods
The study employs a multi-pronged approach:
- Land Cover Mapping: A wall-to-wall analysis of Landsat satellite imagery at approximately 30m resolution was conducted for three decadal periods: 1985-1995 and 1995-2005. This involved visual interpretation and geometric correction to create consistent land cover maps.
- Spatial Determinants Analysis: To understand the drivers of LULCC, the researchers used spatial logistic regression models. These models link land cover conversion estimates (the dependent variable) with hypothesized socioeconomic and biophysical factors (independent variables) at a 1 km resolution grid cell. A "fractional" binomial logit model was used to handle aggregated LULCC data, and an elastic net penalty was employed for variable selection. Bootstrap resampling was used to account for spatial autocorrelation and provide confidence intervals.
- Synthesis of Case Studies: To complement the quantitative analysis and provide qualitative insights into causality, the researchers synthesized findings from 102 existing case studies on LULCC in India. This involved a systematic literature review to identify common causes and reinforce the modeled results.
Datasets
- Land Cover Data: Decadal land cover maps (1985-1995-2005) derived from Landsat MSS/TM imagery at approximately 30m resolution. This data was further aggregated to 1 km resolution for the regression analysis.
- Socioeconomic Data: A spatiotemporal socioeconomic database compiled from the Indian census (1991 and 2001) at the village/town level, linked with administrative boundaries. This database included over 200 socioeconomic variables.
- Case Study Literature: 102 published case studies on LULCC in India were synthesized to provide qualitative insights into causal factors.
Key Insights and Ideas
- Granularity Matters: The study highlights the critical importance of using village-level socioeconomic data. Coarser administrative units mask the fine-grained socioeconomic variations that are crucial for explaining localized land-use changes.
- Cropland-Fallow Dynamics are Complex: Conversions between cropland and fallow land are not simply about land abandonment. They are strongly linked to labor availability (shortages driving fallow land, gender-biased labor markets influencing it), land fragmentation, and access to irrigation and capital.
- Forest Loss is Multifaceted: While agricultural expansion into low-productivity areas is a major driver of forest loss, industrial development, mining, and a village's economic dependence on forest resources also play significant roles.
- Policy Implementation Challenges: Despite national policies aimed at forest protection, the study indicates that forest conversion to cropland continues, suggesting weak implementation of these policies.
- Complementary Insights: The integration of quantitative spatial analysis with qualitative synthesis from case studies provides a more robust understanding of LULCC drivers, where quantitative methods identify associations and case studies offer insights into causality.
Key Results
- Substantial Land Change: Approximately 35% (1985-1995) and 46% (1995-2005) of land cover conversions in India involved shifts between cropland and fallow land.
- Forest Loss Trends: Gross forest loss was substantial, increasing from ~23,810 km² (1985-1995) to ~25,770 km² (1995-2005). Overall, India experienced a net forest loss of ~18,000 km² during both decades.
- Cropland-Fallow Drivers: Key drivers include lack of irrigation and capital, male agricultural labor shortage, and land fragmentation.
- Forest Loss Drivers: Major drivers include cropland expansion in low-productivity areas, industrial development, mining, and excessive economic dependence on forest resources.
- Forest Gain Drivers: Reversion of cropland and shrubland to forest, often linked to abandonment of less productive land, and the presence of protected areas and community involvement (e.g., sacred groves, joint forest management) contribute to forest gain.
Limitations and Future Work
- Decadal Resolution: The study's decadal resolution means it captures only end-point changes and can mask within-decade variations or intermediary land uses. It also cannot distinguish between fallowing for land restoration versus other reasons.
- Data Limitations: The analysis is limited to data available up to 2005. Extending the analysis to more recent years is a logical next step. The availability of consistent village-level socioeconomic data beyond 2001 is also a limitation.
- Causality vs. Association: While the study identifies strong associations, it acknowledges that determinants do not always imply direct causality. The synthesis of case studies helps bridge this gap.
- Future Work: The authors suggest extending the LULCC analysis to 2015 and leveraging the 2011 socioeconomic census data. Further research could explore the impact of climate variability on fallow land and the effectiveness of different land management strategies.
Summary updated