Mapping the habitat for the moose population in Northeast China by combining remote sensing products and random forests
Many wildlife species face the risks of habitat loss, habitat fragmentation or local extinction in response to climate change and anthropogenic disturbance. Moose (Alces alces) in Northeast China is on the southernmost edge of the geographical range of Eurasian moose, the distribution of this popula...
Published in: | Global Ecology and Conservation |
---|---|
Main Authors: | , , , , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Elsevier
2022
|
Subjects: | |
Online Access: | https://doi.org/10.1016/j.gecco.2022.e02347 https://doaj.org/article/bb4e7fa44c51407b857e45acf8deff08 |
id |
ftdoajarticles:oai:doaj.org/article:bb4e7fa44c51407b857e45acf8deff08 |
---|---|
record_format |
openpolar |
spelling |
ftdoajarticles:oai:doaj.org/article:bb4e7fa44c51407b857e45acf8deff08 2023-05-15T13:13:07+02:00 Mapping the habitat for the moose population in Northeast China by combining remote sensing products and random forests Xiaoliang Zhi Hairong Du Minghai Zhang Zexu Long Linqiang Zhong Xue Sun 2022-12-01T00:00:00Z https://doi.org/10.1016/j.gecco.2022.e02347 https://doaj.org/article/bb4e7fa44c51407b857e45acf8deff08 EN eng Elsevier http://www.sciencedirect.com/science/article/pii/S2351989422003493 https://doaj.org/toc/2351-9894 2351-9894 doi:10.1016/j.gecco.2022.e02347 https://doaj.org/article/bb4e7fa44c51407b857e45acf8deff08 Global Ecology and Conservation, Vol 40, Iss , Pp e02347- (2022) Alces alces Down-sampling random forests Habitat suitability Remote sensing Geater Khingan Mountains Lesser Khingan Mountains Ecology QH540-549.5 article 2022 ftdoajarticles https://doi.org/10.1016/j.gecco.2022.e02347 2022-12-30T20:07:25Z Many wildlife species face the risks of habitat loss, habitat fragmentation or local extinction in response to climate change and anthropogenic disturbance. Moose (Alces alces) in Northeast China is on the southernmost edge of the geographical range of Eurasian moose, the distribution of this population is retreating, and population number has been declining for the last several decades. However, little is known about its habitat suitability over a large spatial scale, which hinders further effective conservation of the moose in China. It is critical to explore the moose-habitat relationships and habitat suitability to understand moose habitat requirements, potential land use impacts, and effective management. In this paper, we combined remote sensing-derived predictors and machine learning methods (down-sampling random forests) to explore the moose-habitat associations and map moose habitat suitability. Results showed that our model performed well to excellently in terms of three evaluation metrics (AUCROC, AUCPR, CBI), which indicates the advantages of the combination of remote sensing and machine learning methods in predicting moose habitat. We identified the main factors driving moose distribution in Northeast China are the human footprint index, the mean monthly maximum temperature of the late spring, the percentage of coniferous forest, the minimum dynamic habitat index, the minimum temperature of the coldest month, and the distance from town. Moose responds to these variables nonlinearly. Generally, variables related to human disturbance and heat stress are the main drivers of moose occurrence and are negatively associated with moose occurrence probability. High suitability areas are mainly distributed in eastern and northern Greater Khingan Mountains. Highly suitable habitat covers only a small proportion of the study area. We identified 67,400 km2 of suitable habitat covering 13.6% of the study area. Our study can provide critical information for decision-makers when designing conservation and ... Article in Journal/Newspaper Alces alces Directory of Open Access Journals: DOAJ Articles Global Ecology and Conservation 40 e02347 |
institution |
Open Polar |
collection |
Directory of Open Access Journals: DOAJ Articles |
op_collection_id |
ftdoajarticles |
language |
English |
topic |
Alces alces Down-sampling random forests Habitat suitability Remote sensing Geater Khingan Mountains Lesser Khingan Mountains Ecology QH540-549.5 |
spellingShingle |
Alces alces Down-sampling random forests Habitat suitability Remote sensing Geater Khingan Mountains Lesser Khingan Mountains Ecology QH540-549.5 Xiaoliang Zhi Hairong Du Minghai Zhang Zexu Long Linqiang Zhong Xue Sun Mapping the habitat for the moose population in Northeast China by combining remote sensing products and random forests |
topic_facet |
Alces alces Down-sampling random forests Habitat suitability Remote sensing Geater Khingan Mountains Lesser Khingan Mountains Ecology QH540-549.5 |
description |
Many wildlife species face the risks of habitat loss, habitat fragmentation or local extinction in response to climate change and anthropogenic disturbance. Moose (Alces alces) in Northeast China is on the southernmost edge of the geographical range of Eurasian moose, the distribution of this population is retreating, and population number has been declining for the last several decades. However, little is known about its habitat suitability over a large spatial scale, which hinders further effective conservation of the moose in China. It is critical to explore the moose-habitat relationships and habitat suitability to understand moose habitat requirements, potential land use impacts, and effective management. In this paper, we combined remote sensing-derived predictors and machine learning methods (down-sampling random forests) to explore the moose-habitat associations and map moose habitat suitability. Results showed that our model performed well to excellently in terms of three evaluation metrics (AUCROC, AUCPR, CBI), which indicates the advantages of the combination of remote sensing and machine learning methods in predicting moose habitat. We identified the main factors driving moose distribution in Northeast China are the human footprint index, the mean monthly maximum temperature of the late spring, the percentage of coniferous forest, the minimum dynamic habitat index, the minimum temperature of the coldest month, and the distance from town. Moose responds to these variables nonlinearly. Generally, variables related to human disturbance and heat stress are the main drivers of moose occurrence and are negatively associated with moose occurrence probability. High suitability areas are mainly distributed in eastern and northern Greater Khingan Mountains. Highly suitable habitat covers only a small proportion of the study area. We identified 67,400 km2 of suitable habitat covering 13.6% of the study area. Our study can provide critical information for decision-makers when designing conservation and ... |
format |
Article in Journal/Newspaper |
author |
Xiaoliang Zhi Hairong Du Minghai Zhang Zexu Long Linqiang Zhong Xue Sun |
author_facet |
Xiaoliang Zhi Hairong Du Minghai Zhang Zexu Long Linqiang Zhong Xue Sun |
author_sort |
Xiaoliang Zhi |
title |
Mapping the habitat for the moose population in Northeast China by combining remote sensing products and random forests |
title_short |
Mapping the habitat for the moose population in Northeast China by combining remote sensing products and random forests |
title_full |
Mapping the habitat for the moose population in Northeast China by combining remote sensing products and random forests |
title_fullStr |
Mapping the habitat for the moose population in Northeast China by combining remote sensing products and random forests |
title_full_unstemmed |
Mapping the habitat for the moose population in Northeast China by combining remote sensing products and random forests |
title_sort |
mapping the habitat for the moose population in northeast china by combining remote sensing products and random forests |
publisher |
Elsevier |
publishDate |
2022 |
url |
https://doi.org/10.1016/j.gecco.2022.e02347 https://doaj.org/article/bb4e7fa44c51407b857e45acf8deff08 |
genre |
Alces alces |
genre_facet |
Alces alces |
op_source |
Global Ecology and Conservation, Vol 40, Iss , Pp e02347- (2022) |
op_relation |
http://www.sciencedirect.com/science/article/pii/S2351989422003493 https://doaj.org/toc/2351-9894 2351-9894 doi:10.1016/j.gecco.2022.e02347 https://doaj.org/article/bb4e7fa44c51407b857e45acf8deff08 |
op_doi |
https://doi.org/10.1016/j.gecco.2022.e02347 |
container_title |
Global Ecology and Conservation |
container_volume |
40 |
container_start_page |
e02347 |
_version_ |
1766256104999747584 |