Rangeland species potential mapping using machine learning algorithms

Sharifipour, Behzad; Gholinejad, Bahram; Shirzadi, Ataollah; Shahabi, Himan; Al‐Ansari, Nadhir; Farajollahi, Asghar; Mansorypour, Fatemeh; Clague, John J.

doi:10.1016/j.ecoleng.2023.106900

Cited by 5 publications

(2 citation statements)

References 88 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…SDMs determine possible distributions and habitat suitability based on plant occurrence information and environmental variables [11,12]. With the increasing availability of various types of environmental big data, SDM studies using different machine learning technologies have been attempted [13][14][15][16][17]. For example, Pouteau et al confirmed the potential habitats of native and endemic plants using support vector machines (SVMs) and suggested the necessity of conservation strategies [14].…”

Section: Introductionmentioning

confidence: 99%

“…Edalat et al derived the importance of 13 environmental variables that affect the distribution of medical plants and compared the results using five machine learning algorithms (a generalized linear model, a generalized boosting model, boosted regression trees, functional discrimination analysis, and mixture discriminant analysis) [13]. Furthermore, Sharifipour et al derived distribution maps of three rangeland plant species using five algorithms (SVM, an artificial neural network, naïve Bayes, Bayes net, and classification and regression tree) and compared their performance [15].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Predicting Changes in and Future Distributions of Plant Habitats of Climate-Sensitive Biological Indicator Species in South Korea

Shin,

Lee

et al. 2024

Sustainability

View full text Add to dashboard Cite

Climate change has been progressing rapidly in recent years; consequently, current plant habitats are expected to change. Therefore, to monitor plant movement caused by changed habitat environments, certain plants are designated as bioindicators and managed accordingly. Monitoring changes in plant habitats is important for protecting vulnerable plant species and establishing suitable measures for vegetation environments with suitable plant species under future climates. As part of this task, South Korea manages climate-sensitive plant species for each biological classification group, including plants. Accordingly, in this study, possible current habitats were identified and future habitats were predicted for nine climate-sensitive species in South Korea under climate change scenarios (representative concentration pathways RCP 4.5 and RCP 8.5) using a species distribution model (SDM) and based on national data acquired through field surveys. The MaxEnt algorithm, with high accuracy, was used for the SDM analysis. The MaxEnt algorithm is a powerful tool that analyzes the effects of environmental variables based on occurrence data and indicates possible habitats. To obtain precise results, environmental variables were utilized by collecting comprehensive climatic and topographic data for South Korea. Based on a current habitat analysis, the model accuracy of nine species yielded a high value of more than 0.9, on average, which indicates the extremely high performance of the model. Under climate change scenarios, evergreen coniferous and deciduous broadleaf plant habitats were predicted to expand inland and to the north of South Korea. The results of this study provide valuable data for establishing future conservation and management strategies for climate-sensitive plant species in South Korea. In addition, the detailed environment variable construction method and SDM analysis method used in this study could be applied to the analysis of changes in comprehensive plant habitats caused by climate change in other countries.

show abstract