This study investigates the patterns of land use change, the reasons for those changes and the vulnerability of forests in Nandi North forest zone in Kenya between the years 1986 to 1995 and 1995 to 2006. Landsat TM images of 1986, 1995 and ETM+ of 2006 were used to create single image depicting the sequence of land use change for the period between 1986 and 2006. The spatial relationship between forest cover change and the causes was determined based on the ground truthing and additional literature from the zonal forest officer and the zone management plan where a total of seven classes including dense natural forest, sparse forest, grassland, agricultural land, plantation forest, tea plantation and built up area was used for classification. Both unsupervised and supervised classifications were carried out using a false colour composite image for interpretation. Out of the seven classes, the results indicate an increase in agriculture over the study period while natural forests were reducing. Extreme poverty, planned deforestation for development needs, and unsustainable forest practices such as illegal logging, charcoal burning and encroachment were seen as key drivers of land use change as indicated by the information from the officer. Accessibility to the forest was a key threat to the existence of the forest. This is seen in the result where forest zones with gentle slopes and flat areas, lower elevation, close to roads and settlement had higher probability for conversion to other land uses due to easier access to them compared to steep slopes, with high elevation and which are from roads and settlement. The resulting vulnerability map categorized the forest into four degrees of varying vulnerability namely: highly vulnerable, moderately vulnerable, vulnerable and least vulnerable starting from the easily accessible to least accessible. The resultant map is important in forest conservation by the forest department as areas highly vulnerable can be put into high consideration at the same time involving the community adjacent to be involved in the management process. Key words: Forest cover, LULC, change detection, vulnerability, GIS.
Read full abstract