Conservation or Development? Scenario-Driven Forecasts of Land-Use and Land-Cover Change in Kazimzumbwi Forest Reserve (2024–2054)
Keywords:
Conservation, Forests, Land, TanzaniaAbstract
Land-use and land-cover (LULC) changes pose significant threats to peri-urban forest reserves in sub-Saharan Africa, particularly in Tanzania's Kazimzumbwi Forest Reserve (KFR). This study employs a scenario-based modelling approach to project LULC dynamics from 2024 to 2054, analysing three scenarios: Business-as-Usual (BAU), Conservation-Oriented Scenario (COS), and Development-Intensive Scenario (DIS). Descriptive analyses indicate stark differences in landscape trajectories; under BAU, closed forest cover is projected to decline by approximately 39% by 2054, primarily due to conversions to grassland and shrubland. The DIS scenario predicts even more severe outcomes, with forest losses exceeding 45% in certain areas, driven by urban expansion. In contrast, the COS scenario suggests a more stable landscape, with forest persistence probabilities above 0.70 and evidence of regeneration from shrubland and grassland back to forest. Inferential statistics reinforce these findings, with chi-square tests revealing significant deviations in forest persistence under BAU and DIS (p < 0.01). Likelihood-ratio tests confirm that transition probabilities under COS differ markedly from BAU and DIS (p < 0.001), particularly regarding forest persistence and regeneration. Comparisons between DIS and BAU show significantly higher conversion probabilities under development-intensive conditions (p < 0.001) and weak regeneration pathways. Overall, the study reveals that LULC outcomes in KFR depend significantly on policy decisions. Conservation strategies can stabilise forest cover, whereas development-focused approaches risk increased deforestation. The findings provide quantitative support for sustainable land-use planning and forest governance in Tanzania's peri-urban areas.
Published
How to Cite
Issue
Section

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.


