Transitions in land use and land cover (LULC) are key indicators of human activities, impacting both the environment and human well-being. Spatiotemporal analysis and the projection of future LULC trends under various scenarios are crucial tools for monitoring past changes, predicting future developments, and guiding management interventions to achieve desired LULC trajectories.

This study used the freely available, open source LULC maps for 2001 (when the
dataset was first available), 2011 and 2021 to perform a spatiotemporal analysis of LULC changes in key areas in Kenya where changes in forest cover could help to prevent land degradation. Three scenarios were generated and used to project LULC in 2050: achieving a 10% forest cover, converting forests to croplands, and restoring degraded forests. The analysis was conducted for four counties that are primarily arid and semiarid lands (ASAL), and in which the Integrated Management of Natural Resources for Resilience in Arid and Semi-Arid Lands (IMARA) program is implemented. These counties are Turkana, West Pokot, Elgeyo Marakwet and Narok.

Based on the findings of this study, the distribution of LULC classes among the four counties varies considerably. This suggests that decisions regarding the selection and implementation of sustainable land management interventions should be tailored to the specific geographical context. For counties with conditions similar to Turkana, the focus should be on increasing tree cover and restoring degraded lands. In contrast, for counties resembling Narok, West Pokot and Elgeyo Marakwet, halting deforestation may be a more effective starting point for environmental conservation.

As a signatory to multilateral agreements addressing climate change and its impacts, the Kenyan government is committed to achieving a 10% forest cover due to its critical role in climate regulation and environmental protection. This study found that adopting interventions to maintain a 10% forest cover (scenario 1) would yield the most favourable environmental outcomes by 2050. In all four counties, this scenario not only increased forest cover but was also associated with an expansion in croplands.

The study modelling also indicates that by 2050, the cropland proportion attained under scenario 1 (achieving 10% forest cover) will not be significantly different from that under scenario 2 (forest-to-cropland conversion). This is attributed to the significant role of healthy forest cover in providing favourable conditions to croplands through climate regulation, conservation of underground water (groundwater resources) and protection of soils from erosion.

The researchers observed the occurrence of croplands along rivers, especially in Turkana and West Pokot counties, which implies that forests are likely to be converted to cropland in these areas due to the availability of water resources for cultivation. Coupled with the conversion of wetlands to croplands, as identified in our analysis, we anticipate possible wetland degradation through over-extraction and pollution of water resources.

Likewise, the observed proximity of farmlands to forest also could promote deforestation through forest-to-cropland conversion. This potential outcome underscores the importance of zoning to clearly demarcate forest lands from cultivable parcels. Such zoning can help prevent encroachment into forests and support sustainable conservation efforts. According to scenario 2, the “Shamba system”, in which farmers are allowed to cultivate within forests, is expected to significantly contribute to the loss of forest cover.

Spatial mapping of LULC also reveals the linear distribution of forest cover along water bodies, especially in Turkana County. Owing to the challenges of invasive species, and the fact that the severity of their implications increases with their spread, it is important for the local communities to understand the long-term implications of invasive species and the approaches to contain their extent to an acceptable threshold. Capacity-building programs aimed at building consensus on the definition, impacts and management of invasive species can be valuable tools for early detection and rapid response to invasions by “alien” or “non-native” species.

Scenario analysis of LULC provides insight into the implications of alternative management strategies. However, while spatiotemporal LULC analysis is important in monitoring the implications of alternative LULC changes, integrating local knowledge and experience can enhance our understanding of the drivers behind the observed and projected land cover transitions. Consequently, land management interventions should be context-specific and evidence-based for them to yield practical and sustainable solutions.