Scholarlink

<p>Land use and land cover change (LULCC) and ecological fire dynamics pose major challenges to the sustainability of Tanzania’s miombo woodlands. This study analysed spatial and temporal patterns of LULCC over a 28-year period (1992–2020) using ESA CCI/C3S land-cover data and MODIS fire hotspots. Offering vital insight informing sustainable management and conservation. A forest-based random forest model was used to examine spatial relationships between ecological fire occurrence and LULCC variables. Deforestation peaked at–0.11 % yr⁻¹ (1992–2006), afforestation at +0.09 % yr⁻¹ (2006–2020), grassland rose 6.9 %, settlement grew 5.4‑fold, reflecting growing human and livestock pressure. Model diagnostics showed moderate performance, with the 1992–2006 model achieving the best predictive accuracy (MSE =223.676; R² = 0.21). Proximity to roads, railways, and protected areas emerged as the strongest predictors of fire occurrence, while direct land-cover conversion played a secondary role. These findings highlight the need for coordinated land governance and integrated fire management. We recommend ecosystem-based land use planning (EB-LUP), landscape-scale conservation, and stakeholder collaboration to support woodlands sustainable management and conservation.</p> Graphical Abstract <p></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

Land use and land cover changes spatial relationships with ecological fires and their implications for sustainable management and conservation of Tanzania's Miombo Woodlands

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月25日

查看全文
  • Isaack S. Baltazary,
  • Barnabas P. Malila,
  • Paulo J. Lyimo,
  • Deo D. Shirima,
  • Charles J. Kilawe,
  • Emmanuel F. Nzunda

摘要

Land use and land cover change (LULCC) and ecological fire dynamics pose major challenges to the sustainability of Tanzania’s miombo woodlands. This study analysed spatial and temporal patterns of LULCC over a 28-year period (1992–2020) using ESA CCI/C3S land-cover data and MODIS fire hotspots. Offering vital insight informing sustainable management and conservation. A forest-based random forest model was used to examine spatial relationships between ecological fire occurrence and LULCC variables. Deforestation peaked at–0.11 % yr⁻¹ (1992–2006), afforestation at +0.09 % yr⁻¹ (2006–2020), grassland rose 6.9 %, settlement grew 5.4‑fold, reflecting growing human and livestock pressure. Model diagnostics showed moderate performance, with the 1992–2006 model achieving the best predictive accuracy (MSE =223.676; R² = 0.21). Proximity to roads, railways, and protected areas emerged as the strongest predictors of fire occurrence, while direct land-cover conversion played a secondary role. These findings highlight the need for coordinated land governance and integrated fire management. We recommend ecosystem-based land use planning (EB-LUP), landscape-scale conservation, and stakeholder collaboration to support woodlands sustainable management and conservation.

Graphical Abstract