Application of Geospatial Technology and R for the Wildlife Habitat Analysis in Mae Ping National Park, Thailand DOI

Baromasak Klanreungsang

SSRN Electronic Journal, Journal Year: 2022, Volume and Issue: unknown

Published: Jan. 1, 2022

Wildlife habitat characteristics are critical tool in efficient resource management. At present, Geospatial Technology (GT) and R have been academically proven for evaluating wildlife habitats. This study focuses on analyzing habitats Mae Ping National Park (MPNP), Thailand. GT was used to gather information the physical parameters of train data by a machine learning algorithm. The Landsat 8 set with supervised classification. It found that over 95 percent MPNP is covered forests water resources appropriate Most trees appeared be deciduous dipterocarp forests, followed dry evergreen small amount mixed highlands. data, acquired SMART Patrol Monitoring Center Forest Conservation Area 16 (Chiang Mai), revealed MPNP’s clustered, particularly area’s central-eastern section, density 1.80 animals per square kilometer. With regard species, it wild boars most prevalent, muntjac sambar deer. chi-square test analyze existence causal association between environmental conditions animal distribution. results show distances from resources, altitude slope, saltlicks, roads, tourist attractions all significant relationship at statistical significance p < 0.05. Furthermore, cluster analysis k-medoids algorithm suggested could grouped into three distinct clusters.

Language: Английский

Efficacy of species distribution models (SDMs) for ecological realms to ascertain biological conservation and practices DOI

Mahima Kanwar Rathore,

Laxmi Kant Sharma

Biodiversity and Conservation, Journal Year: 2023, Volume and Issue: 32(10), P. 3053 - 3087

Published: June 20, 2023

Language: Английский

Citations

27
Explainable Boosting Machines for Slope Failure Spatial Predictive Modeling DOI Creative Commons
Aaron E. Maxwell, Maneesh Sharma,

Kurt Donaldson

et al.

Remote Sensing, Journal Year: 2021, Volume and Issue: 13(24), P. 4991 - 4991

Published: Dec. 8, 2021

Machine learning (ML) methods, such as artificial neural networks (ANN), k-nearest neighbors (kNN), random forests (RF), support vector machines (SVM), and boosted decision trees (DTs), may offer stronger predictive performance than more traditional, parametric linear regression, multiple logistic regression (LR), for specific mapping modeling tasks. However, this increased is often accompanied by model complexity decreased interpretability, resulting in critiques of their “black box” nature, which highlights the need algorithms that can both strong interpretability. This especially true when global predictions data points to be explainable order use. Explainable boosting (EBM), an augmentation refinement generalize additive models (GAMs), has been proposed empirical method offers interpretable results performance. The trained graphically summarized a set functions relating each predictor variable dependent along with heat maps representing interactions between selected pairs variables. In study, we assess EBMs predicting likelihood or probability slope failure occurrence based on digital terrain characteristics four separate Major Land Resource Areas (MLRAs) state West Virginia, USA compare those obtained LR, kNN, RF, SVM. EBM provided accuracies comparable RF SVM better LR kNN. generated visualizations included variables, estimation importance average mean absolute scores, scores new add but additional work needed quantify how these outputs impacted correlation, inclusion interaction terms, large feature spaces. Further exploration merited geohazard particular spatial general, value use would greatly enhanced improved interpretability globally availability prediction explanations at cell aggregating unit within mapped modeled extent.

Language: Английский

Citations

40
Assessment of Habitat Suitability and Potential Corridors for Bengal Tiger (Panthera tigris tigris) in Valmiki Tiger Reserve, India, Using MaxEnt Model and Least-Cost Modeling Approach DOI

Roshani Singh,

Md Hibjur Rahaman, Md Masroor

et al.

Environmental Modeling & Assessment, Journal Year: 2024, Volume and Issue: 29(2), P. 405 - 422

Published: March 22, 2024

Language: Английский

Citations

5
Modeling Proboscis monkey conservation sites on Borneo using ensemble machine learning DOI Creative Commons
Anjar Dimara Sakti,

Kurnia Putri Adillah,

Cokro Santoso

et al.

Global Ecology and Conservation, Journal Year: 2024, Volume and Issue: 54, P. e03101 - e03101

Published: July 22, 2024

This study aimed to analyze the habitat suitability of endangered Proboscis monkey (Nasalis larvatus) on Borneo using a multi-machine-learning approach. integrated physical, vegetational, meteorological, and human activity data develop comprehensive model. Four machine-learning algorithms, namely, maximum entropy (MaxEnt), random forest (RF), support vector machine (SVM), gradient tree boosting (GTB), classification regression trees (CART), were employed model index. A total 1943 sample points divided into training (70 %) validation (30 sets for analysis. included three main stages: geospatial database creation, spatial modeling evaluation. In addition, pressure from development index was analyzed. identified high level habitats in nearshore areas. The monkeys observed be 11.54 %, as evidenced by consensus MaxEnt value four algorithms. Conversely, minimum recorded at 13.27 indicated disagreement among all AUC values models ranged 74 % 90 indicating moderate predictive performance. provides valuable insights formulation well-planned programs monkeys. results this will contribute accurate identification potential habitats, thereby providing conservation efforts safeguarding species.

Language: Английский

Citations

4
Machine learning-driven habitat suitability modeling of Suaeda aegyptiaca for sustainable industrial cultivation in saline regions DOI Creative Commons

Sara Edrisnia,

Mohammad Etemadi, Hamid Reza Pourghasemi

et al.

Industrial Crops and Products, Journal Year: 2025, Volume and Issue: 225, P. 120427 - 120427

Published: Jan. 11, 2025

Language: Английский

Citations

0
Use of Google Earth Engine in predicting future giant cane (Arundinaria gigantea (Walter) Muhl.) restoration sites DOI Creative Commons
Sanjeev Sharma

Advances in Bamboo Science, Journal Year: 2025, Volume and Issue: unknown, P. 100164 - 100164

Published: May 1, 2025

Language: Английский

Citations

0
Machine Learning for Criteria Weighting in GIS-Based Multi-Criteria Evaluation: A Case Study of Urban Suitability Analysis DOI Creative Commons

Lan Qing Zhao,

Alysha van Duynhoven, Suzana Dragićević

et al.

Land, Journal Year: 2024, Volume and Issue: 13(8), P. 1288 - 1288

Published: Aug. 15, 2024

Geographic Information System-based Multi-Criteria Evaluation (GIS-MCE) methods are designed to assist in various spatial decision-making problems using data. Deriving criteria weights is an important component of GIS-MCE, typically relying on stakeholders’ opinions or mathematical methods. These approaches can be costly, time-consuming, and prone subjectivity bias. Therefore, the main objective this study investigate use Machine Learning (ML) techniques support weight derivation within GIS-MCE. The proposed ML-MCE method explored a case urban development suitability analysis City Kelowna, Canada. Feature importance values drawn from three ML techniques–Random Forest (RF), Extreme Gradient Boosting (XGB), Support Vector (SVM)–are used derive weights. scores obtained methodology compared with Equal-Weights (EW) Analytical Hierarchy Process (AHP) approach for weighting. results indicate that ML-derived where RF XGB provide more similar than those derived SVM. similarities differences confirmed Kappa indices comparing pairs maps. new processes land-use planning.

Language: Английский

Citations

2
Habitat occupancy of sloth bear Melursus ursinus in Chitwan National Park, Nepal DOI
Rajan Prasad Paudel, Rabin Kadariya, Babu Ram Lamichhane

et al.

Ecology and Evolution, Journal Year: 2022, Volume and Issue: 12(3)

Published: March 1, 2022

Mammals have experienced a massive decline in their populations and geographic ranges worldwide. The sloth bear, Melursus ursinus (Shaw, 1791), is one of many species facing conservation threats. Despite being endangered Nepal, decades inattention to the situation hindered management. We assessed distribution patterns habitat use by bears Chitwan National Park (CNP), Nepal. conducted sign surveys from March June, 2020, 4 × km grids (n = 45). collected detection/non-detection data along 4-km trail that was divided into 20 continuous segments 200 m each. obtained environmental, ecological, anthropogenic covariates understand determinants bear occupancy. were analyzed using single-species single-season occupancy method, with spatially correlated detection. Using repeated observations, these models accounted for imperfect detectability provide robust estimates model-averaged estimate 69% detection probability 0.25. increased presence termites fruits rugged, dry, open, undisturbed habitats. Our results indicate elusive, functionally unique, widespread CNP. Future interventions action plans aimed at management must adequately consider requirements.

Language: Английский

Citations

10
A machine learning model and biometric transformations to facilitate European oyster monitoring DOI Creative Commons
Santiago E A Pineda-Metz, Verena Merk, Bernadette Pogoda

et al.

Aquatic Conservation Marine and Freshwater Ecosystems, Journal Year: 2023, Volume and Issue: 33(7), P. 708 - 720

Published: Jan. 9, 2023

Abstract Ecosystem monitoring, especially in the context of marine conservation and management requires abundance biomass metrics, condition indices, measures ecosystem services key species, all which can be calculated using biometric transformation factors. Following restoration North Sea north‐east Atlantic waters, European oyster ( Ostrea edulis ) its monitoring have substantially increased over past decade. Restoration activities are implemented by diverse approaches practitioners ranging from governmental agencies, research institutions non‐governmental to regional groups, including citizen science projects. Thus, tools for facilitating data acquisition estimation with non‐destructive techniques support quantitatively qualitatively. Weight‐to‐weight factors calculating dry weight O. wet measurements presented. Another important tool is only size measurements. The classical approach achieve these construction allometric models, which, however, greatly vary among regions between years, making them extremely location/season specific. Alternative more flexible models constructed random forests proposed. This algorithm a machine learning technique that increasingly used ecology, has been proven outperform other predictive models. From variable 1,401 individuals, were estimate total, shell body weights, compare 15 forest In general, outperformed ones, lower error when estimating weight. developed thus provide increasing without need sacrificing individuals. Their improvement imply implementation efforts throughout Europe.

Language: Английский

Citations

5
Structural complexity characterizes fine‐scale forest conditions used by Pacific martens DOI Creative Commons
Matthew S. Delheimer, Katie M. Moriarty, Holly L. Munro

et al.

Journal of Wildlife Management, Journal Year: 2023, Volume and Issue: 87(4)

Published: March 13, 2023

Abstract When wildlife species exhibit unexpected associations with vegetation, replication of studies in different locales can illuminate whether patterns use are consistent or divergent. Our objective was to describe fine‐scale forest conditions used by Pacific martens ( Martes caurina ) at 2 study sites northern California that differed composition and past timber harvest. We identified denning resting locations radio‐marked sampled structure‐ plot‐level vegetation using standardized inventory methods between 2009–2021. Woody structures were significantly larger than randomly available across types (e.g., live tree, snag, log) both sites. Den rest occurred areas characterized higher numbers logs snags, lower trees stumps, diameter logs, greater variation tree log diameter. Features largely generally representative heterogeneity increased structural complexity, have been widely associated broader spatial scales (i.e., home range landscape). The occurrence may indicate complexity facilitates marten foraging while reducing predation risk. work offers timely directed information guide management the context landscape change.

Language: Английский

Citations

5