Scholarlink

Context <p>A long-term and highly dynamic understanding of terrestrial habitats is of extreme importance for enhancing wildlife conservation, habitat management, human-wildlife conflict mitigation in a timely manner. Currently, however, fine-resolution, multi-decade, and inter-annual habitat mapping remains rare due to high costs and data scarcity of wildlife presence/absence data. Moreover, the existing land cover datasets used for habitat mapping have high uncertainty in classification systems, which tend to lead to uneven spatial and discontinuous temporal domains.</p> Objectives <p>The main objectives of this study aim to explore an ensemble method of long-term annual habitat mapping by integrating yearly land cover maps and wildlife habitat preferences derived by a meta-analysis, and to reveal interannual changes in habitat of China’s wild elephant in the context of rapid economic development. Our work innovatively integrates long-term multi-temporal land use/cover data with multi-source environmental factors and applies meta-analysis techniques to overcome habitat mapping and change detection challenges.</p> Methods <p>Random forest (RF) models and hidden Markov model (HMM) were used to produce a highly-accurate annual land cover dataset based on Landsat time series images from 1988 to 2020. Combined with the meta-analysis-derived habitat preferences of wild elephants and multi-temporal remote sensed environmental factors, yearly habitat maps were generated by calculating the habitat suitability index (HSI).</p> Results <p>The results indicated that the elephant habitat decreased significantly (α = 0.05), by 4135.93 km<sup>2</sup> in their current range from 1988‒2020. The area of habitat loss reached approximately 24.32% (1533.33 km<sup>2</sup>) and 19.04% (2602.20 km<sup>2</sup>) in the elephants’ native (in the late 1980s) and newly expanded ranges, respectively. Moreover, the degree of habitat fragmentation was higher in the newly expanded range than in the native range. Over the past 33&#xa0;years, habitat degradation mainly occurred before 2008 in the native range and after 2008 in the newly expanded range.</p> Conclusions <p>These findings demonstrated that the degree and process of wild elephants’ habitat reduction were distinctly different between the native and newly expanded ranges; this highlights the need for different habitat conservation and management policies for different regions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Interannual dynamics of China’s wild elephant habitat revealed by Landsat-based time series land cover and meta-analysis derived habitat preferences

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月1日

查看全文
  • Fei Xie,
  • Xin Zhang,
  • Jiejing Zhang,
  • Fei Chen,
  • Hui Fan

摘要

Context

A long-term and highly dynamic understanding of terrestrial habitats is of extreme importance for enhancing wildlife conservation, habitat management, human-wildlife conflict mitigation in a timely manner. Currently, however, fine-resolution, multi-decade, and inter-annual habitat mapping remains rare due to high costs and data scarcity of wildlife presence/absence data. Moreover, the existing land cover datasets used for habitat mapping have high uncertainty in classification systems, which tend to lead to uneven spatial and discontinuous temporal domains.

Objectives

The main objectives of this study aim to explore an ensemble method of long-term annual habitat mapping by integrating yearly land cover maps and wildlife habitat preferences derived by a meta-analysis, and to reveal interannual changes in habitat of China’s wild elephant in the context of rapid economic development. Our work innovatively integrates long-term multi-temporal land use/cover data with multi-source environmental factors and applies meta-analysis techniques to overcome habitat mapping and change detection challenges.

Methods

Random forest (RF) models and hidden Markov model (HMM) were used to produce a highly-accurate annual land cover dataset based on Landsat time series images from 1988 to 2020. Combined with the meta-analysis-derived habitat preferences of wild elephants and multi-temporal remote sensed environmental factors, yearly habitat maps were generated by calculating the habitat suitability index (HSI).

Results

The results indicated that the elephant habitat decreased significantly (α = 0.05), by 4135.93 km2 in their current range from 1988‒2020. The area of habitat loss reached approximately 24.32% (1533.33 km2) and 19.04% (2602.20 km2) in the elephants’ native (in the late 1980s) and newly expanded ranges, respectively. Moreover, the degree of habitat fragmentation was higher in the newly expanded range than in the native range. Over the past 33 years, habitat degradation mainly occurred before 2008 in the native range and after 2008 in the newly expanded range.

Conclusions

These findings demonstrated that the degree and process of wild elephants’ habitat reduction were distinctly different between the native and newly expanded ranges; this highlights the need for different habitat conservation and management policies for different regions.