Long-Term Habitat Dynamics in the Greater Kafue Ecosystem of Zambia Using Multi-Temporal Landsat Data (1984–2024)

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

The Greater Kafue Ecosystem (GKE) in central-western Zambia, comprising Kafue National Park and nine surrounding Game Management Areas (GMAs), covers approximately 66,000 km² and ranks among southern Africa’s largest conservation landscapes. Despite its ecological significance, long-term ecosystem-wide assessments of cumulative habitat transformation remain limited. This study employed multi-temporal Landsat imagery (1984–2024), processed in ArcGIS Pro and classified using Random Forest with object-based refinement, to quantify four decades of land-cover change. Classification performance was robust, with overall accuracies of 82–90% (κ = 0.73–0.86), macro-averaged F1-scores of 81.4–90.0%, and a 2024 overall accuracy confidence interval of ± 2.6% (95% CI, n = 600). Natural habitat declined from 98.31% (9.65 million ha) in 1984 to 83.96% (8.24 million ha) in 2024, representing a net loss of approximately 1.41 million ha. Agricultural expansion was the dominant driver, with cropland increasing from 0.86% to 14.53%, accounting for 83% of total conversion. Conversion rates accelerated from 10,209 ha yr⁻¹ (1984–1994) to 57,269 ha yr⁻¹ (2014–2024). Spatial patterns were highly heterogeneous: Kafue National Park retained 98.87%, compared with 84.08% across GMAs, with retention ranging from 96.62% (Lunga Luswishi) to 37.77% (Bilili). Habitat loss exhibited strong spatial structuring, with edge effects extending up to 10 km beyond park boundaries and a 14.4 percentage-point gradient associated with proximity to roads. These findings underscore the effectiveness of strict protection while identifying road-accessible buffer zones as priorities for spatially targeted conservation planning.

Article activity feed