What question did this study set out to answer?

To evaluate how different land-use systems affect soil organic carbon storage and enzyme activity in the Indo-Gangetic Plains.

March 21, 2026

Impact of Forest Land Conversion on Soil Organic Carbon and Enzyme Activity Across Different Land‐Use Systems in the Indo‐Gangetic Plains, India: Assessment Through Machine Learning

Key Points

To evaluate how different land-use systems affect soil organic carbon storage and enzyme activity in the Indo-Gangetic Plains.
Soil samples were collected from 0 to 100 cm depths across ten different land use practices.
Soil organic carbon distribution, carbon lability, and enzymatic activities were analyzed.
A Classification and Regression Tree model was used to identify drivers of carbon management index.
Structural equation modeling assessed the direct and indirect influences of land use on carbon dynamics.
Grassland showed the highest total soil organic carbon stocks, comparable to natural forest and agro-forestry.
Active carbon pools decreased with soil depth, whereas passive pools increased.
Grassland had the highest carbon management index value at the surface, dropping significantly below 30 cm depth.
Enzymatic activities decreased with soil depth, with natural forest showing the highest dehydrogenase activity.

Abstract

ABSTRACT Understanding soil organic carbon (OC) storage in deep soil layers across various land use and management practices (LUMP) is vital for evaluating carbon sequestration and human‐induced land use impacts, especially in the north‐western Indo‐Gangetic Plains where intensive cultivation, soil disturbance and contrasting land‐use systems can markedly alter deep‐soil carbon stocks. Therefore, soil samples from 0 to 100 cm depths across ten land use practices—natural forest, grassland, agro‐forestry, horticulture‐based, river bed, rice‐based, reclaimed alkali, maize‐based, sewage‐irrigated and stabilised dumping sites were collected. These were analysed for organic carbon distribution, carbon lability and enzymatic activities. Total soil organic carbon stocks (CSTOCK) were strongly influenced by both land use types and soil depth. Grassland soil showed the highest CSTOCK, comparable to natural forest and agro‐forestry especially within the 0–15 cm and 15–30 cm layers, while stabilised dumping sites had the lowest. Active carbon pools decreased with soil depth (57% in agro‐forestry to 32% in river bed), while passive pools increased with depth. Carbon management index (CMI) values were highest in grassland at the surface but dropped significantly below 30 cm suggesting that beyond 30 cm depth, CMI poorly reflects carbon dynamics. Enzymatic activities (dehydrogenase, β‐glucosidase) decreased with depth, with natural forest showing high dehydrogenase activity. A Classification and Regression Tree model pinpointed CSTOCK as the primary CMI driver, with a threshold of > 13.9 Mg/ha for optimal carbon sequestration. Structural equation modelling revealed land use influences CMI directly through labile carbon or indirectly via CSTOCK. Sustainable practices like natural forests, agro‐forestry and horticulture which prioritise deep‐rooted systems, enhance carbon sequestration and stabilise deep soil carbon, necessitating supportive policies and further research. Future work also could focus on development of predictive models for different land‐use and management practices, allowing these findings to be generalised to similar ecosystems elsewhere.

Mark Helpful

Bookmark

Relay