Sachin Timilsina
Research Subjects
Sachin Timilsina has a total of 76 co-authors. Here are the top co-authors by number of shared publications:
- Mahamad Sayab Miya
Systematics and Evolution Laboratory, Department of Biological Sciences Western Kentucky University Bowling Green Kentucky USA
6 shared publications - Deepak Gautam
School of Ecology and Nature Conservation, Beijing Forestry University, Beijing 100083, China
3 shared publications Anna Karina Zimmermann
2 shared publications- Hari Adhikari
Department of Geosciences and Geography, University of Helsinki, Helsinki, Finland
2 shared publications - Kishor Prasad Bhatta
Faculty of Forest Science and Forest Ecology, Georg-August-Universität, Goettingen, Germany
2 shared publications Nathan Le Boursicaud
2 shared publications
Latest Publications
- Open Access
- 21 September 2026
- Published by Springer Nature in Discover Forests
- Vol. 2 (1) , 69
Stand structure drives vegetation carbon storage while soil texture regulates soil organic carbon in Nepalese mid-hill community forests
- 21 September 2026
- Published by Springer Nature in Discover Forests
- Vol. 2 (1) , 69
Open AccessStand structure drives vegetation carbon storage while soil texture regulates soil organic carbon in Nepalese mid-hill community forests
- Open Access
- 27 August 2026
- Published by Elsevier in SSRN Electronic Journal
Multi-Scale LULC Mapping of Nepal: An Optimized Machine Learning Framework in Google Earth Engine
Accurate land use and land cover (LULC) mapping supports environmental monitoring, resource management, and policy planning, especially in Nepal's complex terrain. This study followed a two-phase framework in Google Earth Engine (GEE). In Phase 1, algorithm benchmarking, three Machine learning (ML) algorithms, Random Forest (RF), Gradient Tree Boosting (GTB), and Support Vector Machine (SVM), were tuned across cluster and non-cluster sampling and three satellite sensors. Training sizes from 500 to 100 points were then assessed, followed by scenarios with and without elevation metrics. In Phase 2, multi-scale mapping, the best configuration was applied to generate LULC maps from the national to the ward level. RF outperformed GTB and SVM across most combinations, with the largest gains occurring when elevation, slope, and aspect were added as predictors. The best configuration, RF applied to Landsat 8 imagery with elevation covariates under cluster sampling, achieved an overall accuracy (OA) of 88.4% and a Kappa coefficient (Kc) of 85.2%, with consistent performance across Nepal's twelve physiographic zones, where accuracy ranged from 74.7% to 96%. Forest was the dominant class nationally at 47% of Nepal's area, followed by barren/rocky land (14.8%) and agriculture (13.4%). Snow/glacier, grassland, human buildup, and water bodies covered the remaining area, from 2.5% to 10.8%. A north-south gradient emerged: agriculture and settlement were concentrated in the Terai lowlands, while Snow/glacier and barren/rocky terrain dominated the high-Himalayan regions. These findings show that RF is a reliable method for LULC mapping at multiple scales, supporting land management and policy planning across Nepal.
- 27 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessMulti-Scale LULC Mapping of Nepal: An Optimized Machine Learning Framework in Google Earth Engine
Accurate land use and land cover (LULC) mapping supports environmental monitoring, resource management, and policy planning, especially in Nepal's complex terrain. This study followed a two-phase framework in Google Earth Engine (GEE). In Phase 1, algorithm benchmarking, three Machine learning (ML) algorithms, Random Forest (RF), Gradient Tree Boosting (GTB), and Support Vector Machine (SVM), were tuned across cluster and non-cluster sampling and three satellite sensors. Training sizes from 500 to 100 points were then assessed, followed by scenarios with and without elevation metrics. In Phase 2, multi-scale mapping, the best configuration was applied to generate LULC maps from the national to the ward level. RF outperformed GTB and SVM across most combinations, with the largest gains occurring when elevation, slope, and aspect were added as predictors. The best configuration, RF applied to Landsat 8 imagery with elevation covariates under cluster sampling, achieved an overall accuracy (OA) of 88.4% and a Kappa coefficient (Kc) of 85.2%, with consistent performance across Nepal's twelve physiographic zones, where accuracy ranged from 74.7% to 96%. Forest was the dominant class nationally at 47% of Nepal's area, followed by barren/rocky land (14.8%) and agriculture (13.4%). Snow/glacier, grassland, human buildup, and water bodies covered the remaining area, from 2.5% to 10.8%. A north-south gradient emerged: agriculture and settlement were concentrated in the Terai lowlands, while Snow/glacier and barren/rocky terrain dominated the high-Himalayan regions. These findings show that RF is a reliable method for LULC mapping at multiple scales, supporting land management and policy planning across Nepal.
- Open Access
- 10 June 2026
- Published by Nepal Journals Online (JOL) in Journal of Agriculture and Forestry University
- Vol. 7 (1) , 305-324
Factors Predicting Habitat Suitability for the Occurrence of Gaur Using Ensemble Modeling in Chitwan National Park of Nepal
Gaur, a large herbivore native to South and Southeast Asia, supports ecosystem health by enhancing soil sustainability, promoting plant growth, and aiding plant succession. In Nepal, their presence is limited to the protected habitats of two lowland (Terai) protected areas: Chitwan National Park (CNP) and Parsa National Park (PNP). However, information on the habitat factors influencing Gaur occurrence is limited, yet is essential for improving the existing habitats of this keystone species and prioritizing conservation efforts within the protected area. To address this gap, our study used an ensemble modelling approach to predict Gaur habitat suitability and identify key factors influencing their occurrence in CNP, Nepal. To gather data on Gaur's presence points, field-based surveys were conducted in April and May 2021, including periodic censuses and annual monitoring. Their presence was verified through direct observations, identification of pellet droppings, and hoofmarks found within the park. Ensemble modeling identified approximately 183.53 km2 within the CNP as highly suitable habitat for Gaur, representing 2.1% of the buffer zone and 97.9% of the core area. Of the 33 variables used to construct the habitat suitability model, human population density, slope, and livestock density were the most influencing factors. For long-term Gaur conservation, our study recommends prioritizing habitat management by strictly limiting human activities and livestock grazing in the core of CNP.
- 10 June 2026
- Published by Nepal Journals Online (JOL) in Journal of Agriculture and Forestry University
- Vol. 7 (1) , 305-324
Open AccessFactors Predicting Habitat Suitability for the Occurrence of Gaur Using Ensemble Modeling in Chitwan National Park of Nepal
Gaur, a large herbivore native to South and Southeast Asia, supports ecosystem health by enhancing soil sustainability, promoting plant growth, and aiding plant succession. In Nepal, their presence is limited to the protected habitats of two lowland (Terai) protected areas: Chitwan National Park (CNP) and Parsa National Park (PNP). However, information on the habitat factors influencing Gaur occurrence is limited, yet is essential for improving the existing habitats of this keystone species and prioritizing conservation efforts within the protected area. To address this gap, our study used an ensemble modelling approach to predict Gaur habitat suitability and identify key factors influencing their occurrence in CNP, Nepal. To gather data on Gaur's presence points, field-based surveys were conducted in April and May 2021, including periodic censuses and annual monitoring. Their presence was verified through direct observations, identification of pellet droppings, and hoofmarks found within the park. Ensemble modeling identified approximately 183.53 km2 within the CNP as highly suitable habitat for Gaur, representing 2.1% of the buffer zone and 97.9% of the core area. Of the 33 variables used to construct the habitat suitability model, human population density, slope, and livestock density were the most influencing factors. For long-term Gaur conservation, our study recommends prioritizing habitat management by strictly limiting human activities and livestock grazing in the core of CNP.
- Open Access
- 27 May 2026
- Published by Elsevier in Forest Ecology and Management
- Vol. 617
Passive versus active afforestation of agricultural land: Two decades of impacts on tree biomass and ecosystem carbon storage
- 27 May 2026
- Published by Elsevier in Forest Ecology and Management
- Vol. 617
Open AccessPassive versus active afforestation of agricultural land: Two decades of impacts on tree biomass and ecosystem carbon storage
- Open Access
- 4 January 2026
- Published by Wiley in Ecology and Evolution
- Vol. 16 (1)
Effects of the Irregular Shelterwood System on Regeneration Dynamics of Shorea robusta Gaertn. f. in Baijalpur Community Forest, Nepal
The Irregular Shelterwood System (ISS) has recently been introduced in Nepalese community forests to promote sustainable regeneration of Shorea robusta (Sal). However, limited research exists on its ecological impacts and regeneration dynamics. This study examined regeneration patterns, soil properties, and environmental factors influencing S. robusta regeneration under the ISS in Baijalpur Community Forest, Kapilvastu District, Nepal. Using a stratified systematic random sampling design, data were collected from treated and untreated plots. Results showed a significantly higher density of seedlings (16,800 ha −1 ) and saplings (4693 ha −1 ) in treated plots compared to untreated areas (11,960 and 2688 ha −1 , respectively), indicating the effectiveness of ISS in enhancing regeneration. Most regeneration originated from seedlings rather than coppice shoots. Soil assessments revealed slightly acidic to neutral sandy loam soils, with higher organic matter (OM) and improved nutrient profiles in treated plots. Crown cover and organic carbon were negatively correlated with regeneration, suggesting that moderate canopy openings favor S. robusta establishment. These findings underscore the ecological importance of adaptive silvicultural practices for improving regeneration success, forest resilience, and sustainable community forestry in Nepal.
- 4 January 2026
- Published by Wiley in Ecology and Evolution
- Vol. 16 (1)
Open AccessEffects of the Irregular Shelterwood System on Regeneration Dynamics of Shorea robusta Gaertn. f. in Baijalpur Community Forest, Nepal
The Irregular Shelterwood System (ISS) has recently been introduced in Nepalese community forests to promote sustainable regeneration of Shorea robusta (Sal). However, limited research exists on its ecological impacts and regeneration dynamics. This study examined regeneration patterns, soil properties, and environmental factors influencing S. robusta regeneration under the ISS in Baijalpur Community Forest, Kapilvastu District, Nepal. Using a stratified systematic random sampling design, data were collected from treated and untreated plots. Results showed a significantly higher density of seedlings (16,800 ha −1 ) and saplings (4693 ha −1 ) in treated plots compared to untreated areas (11,960 and 2688 ha −1 , respectively), indicating the effectiveness of ISS in enhancing regeneration. Most regeneration originated from seedlings rather than coppice shoots. Soil assessments revealed slightly acidic to neutral sandy loam soils, with higher organic matter (OM) and improved nutrient profiles in treated plots. Crown cover and organic carbon were negatively correlated with regeneration, suggesting that moderate canopy openings favor S. robusta establishment. These findings underscore the ecological importance of adaptive silvicultural practices for improving regeneration success, forest resilience, and sustainable community forestry in Nepal.
- Open Access
- 15 November 2025
- Published by Elsevier in Next Research
- Vol. 2 (4)
Potential distribution of endangered medicinal plant Nardostachys Jatamansi (D.Don) DC. in Shey Phoksundo National Park, Nepal
- 15 November 2025
- Published by Elsevier in Next Research
- Vol. 2 (4)
Open AccessPotential distribution of endangered medicinal plant Nardostachys Jatamansi (D.Don) DC. in Shey Phoksundo National Park, Nepal
- Open Access
- 11 April 2024
- Published by Sri Lanka Journals Online in Ceylon Journal of Science
- Vol. 53 (2) , 253-265
Ethnobotanical assessment of medicinal plants in Trans-Himalaya of Nepal
Nepal harbors many ethnomedicinal plants distributed throughout the country’s diverse geography. It is crucial to document plants’ ethnobotanical knowledge for further research and conservation. This study aimed to conduct a quantitative ethnobotanical analysis of the medicinal plants used in the Varagung Muktichhetra Rural Municipality of Mustang District, Nepal. Data were collected through a household survey (N = 101), using snowball sampling with 40% sampling intensity, 12 Key Informant Interviews (KIIs), and two Focus Group Discussions (FGDs). The data was analyzed using different quantitative tools: Informant consensus factor (Fic), Fidelity Level (FL), and Use Values (UV). A total of 54 medicinal plants belonging to 32 families were documented, which were used to treat 40 different ailments. The family Asteraceae represented the maximum number of plants being used. Herbs were most frequently used (59%), and root (n =10) was the most used plant part. Neopicrorhiza scrophulariiflora was the most frequently used and preferred species (FL = 65% and UV = 0.62). Fever (Fic = 0.88) and gastrointestinal disorder (Fic = 0.86) were the major ailment categories treated. Oral (76%) was the main route of administration of medicines in the body, and the most used form of medicine preparation was paste (46%). Unsustainable harvesting, illegal collection, and forest fire were the major threats to medicinal plants and were highly significant after applying the non-parametric Friedman test. Enactment of a policy that focuses on conservation, sustainable harvesting, and domestication of highly valued medicinal plants is recommended to policymakers.
- 11 April 2024
- Published by Sri Lanka Journals Online in Ceylon Journal of Science
- Vol. 53 (2) , 253-265
Open AccessEthnobotanical assessment of medicinal plants in Trans-Himalaya of Nepal
Nepal harbors many ethnomedicinal plants distributed throughout the country’s diverse geography. It is crucial to document plants’ ethnobotanical knowledge for further research and conservation. This study aimed to conduct a quantitative ethnobotanical analysis of the medicinal plants used in the Varagung Muktichhetra Rural Municipality of Mustang District, Nepal. Data were collected through a household survey (N = 101), using snowball sampling with 40% sampling intensity, 12 Key Informant Interviews (KIIs), and two Focus Group Discussions (FGDs). The data was analyzed using different quantitative tools: Informant consensus factor (Fic), Fidelity Level (FL), and Use Values (UV). A total of 54 medicinal plants belonging to 32 families were documented, which were used to treat 40 different ailments. The family Asteraceae represented the maximum number of plants being used. Herbs were most frequently used (59%), and root (n =10) was the most used plant part. Neopicrorhiza scrophulariiflora was the most frequently used and preferred species (FL = 65% and UV = 0.62). Fever (Fic = 0.88) and gastrointestinal disorder (Fic = 0.86) were the major ailment categories treated. Oral (76%) was the main route of administration of medicines in the body, and the most used form of medicine preparation was paste (46%). Unsustainable harvesting, illegal collection, and forest fire were the major threats to medicinal plants and were highly significant after applying the non-parametric Friedman test. Enactment of a policy that focuses on conservation, sustainable harvesting, and domestication of highly valued medicinal plants is recommended to policymakers.
- Open Access
- 1 January 2024
- Published by Wiley in Scientifica
- Vol. 2024 (1) , 1739115
Assessment of Soil Carbon Stock and Soil Quality in Different Forest Stands and Management Regimes in Terai Region of Nepal
Assessment of soil organic carbon (SOC) stock and soil quality for informed forest management is hindered by inadequate data across different forest stand types and management regimes. Therefore, this study was conducted to assess SOC stock and soil quality in two forest stand types, i.e., Shorea robusta (Sal) forest and Terai mixed hardwood (TMH) forest, and selected forest management regimes (leasehold forest, community forest, government‐managed forest, and forest area under protected area) in Terai region of Nepal. Stratified random sampling method was adopted for soil sample collection across terai region following Forest Resource Assessment, Nepal. Altogether, 62 composite soil samples from 30 cm depth were taken from the entire Terai region which included these two forest stand types and four management regimes. Different physical (soil texture and bulk density) and chemical (pH, SOC (%), total nitrogen, available phosphorus, and available potassium) properties were analyzed to calculate SOC stock and soil quality. Our result found no significant differences in SOC stock among two forest stand types (p > 0.05). Unexpectedly, leasehold forest had significantly (p < 0.05) higher SOC stock than other forest management regimes. In terms of soil quality, among two forest stand types, Sal forest (0.50) was found to be superior compared with TMH forest (0.46). Similarly, community forest had superior soil quality (0.50) than government‐managed forest (0.47), protected area (0.47), and leasehold forest (0.45). A longitudinal study approach is recommended to observe changes in soil properties over time due to climate change and human activities, offering valuable insights into their dynamics.
- 1 January 2024
- Published by Wiley in Scientifica
- Vol. 2024 (1) , 1739115
Open AccessAssessment of Soil Carbon Stock and Soil Quality in Different Forest Stands and Management Regimes in Terai Region of Nepal
Assessment of soil organic carbon (SOC) stock and soil quality for informed forest management is hindered by inadequate data across different forest stand types and management regimes. Therefore, this study was conducted to assess SOC stock and soil quality in two forest stand types, i.e., Shorea robusta (Sal) forest and Terai mixed hardwood (TMH) forest, and selected forest management regimes (leasehold forest, community forest, government‐managed forest, and forest area under protected area) in Terai region of Nepal. Stratified random sampling method was adopted for soil sample collection across terai region following Forest Resource Assessment, Nepal. Altogether, 62 composite soil samples from 30 cm depth were taken from the entire Terai region which included these two forest stand types and four management regimes. Different physical (soil texture and bulk density) and chemical (pH, SOC (%), total nitrogen, available phosphorus, and available potassium) properties were analyzed to calculate SOC stock and soil quality. Our result found no significant differences in SOC stock among two forest stand types (p > 0.05). Unexpectedly, leasehold forest had significantly (p < 0.05) higher SOC stock than other forest management regimes. In terms of soil quality, among two forest stand types, Sal forest (0.50) was found to be superior compared with TMH forest (0.46). Similarly, community forest had superior soil quality (0.50) than government‐managed forest (0.47), protected area (0.47), and leasehold forest (0.45). A longitudinal study approach is recommended to observe changes in soil properties over time due to climate change and human activities, offering valuable insights into their dynamics.