Quantifying pasture expansion and associated deforestation: a case study in Ethiopia and Kenya

Organised by Laboratory of Geo-information Science and Remote Sensing

Fri 8 April 2016 08:30 to 09:00

Venue Gaia, gebouwnummer 101

By Yuting Lin (Taiwan)

Deforestation is a result of a number of drivers, one of the main drivers is agriculture, including pasture expansion. In order to understand the spatial relationship between pasture expansion and deforestation, we used FRA RSS datasets with a visual interpretation analysis to maps of deforestation driven by pasture. We selected factors which are associated with pasture expansion, including those related to soil, geomorphology, climate, agricultural environment and socio-economic status, and applied a correlation analysis and multiple logistic regression to identify their relationship with the presence of pasture as a follow up land use after deforestation. These findings were used to predict the risk of pasture expansion in Ethiopia and Kenya. The results show that all factors tested have a significant relationships with the presence or absence of pasture as a follow-up land use. However, the extent to which the factors are related differ in Ethiopia and Kenya. We found that agricultural environment aspects such as the length of the growing period, pasture suitability and livestock density, as well as population density and distance to roads are more related to pasture expansion in Ethiopia, while soil conditions, geomorphology such as elevation and slope, temperature, distance to urban settlements and rivers are more related in Kenya. Our prediction of pasture expansion shows that less dense forests as well as range land use such as shrub-brush and grass land have the highest pasture expansion risk in both countries. These results provide understanding of the factors related to pasture expansion associated with deforestation, and help in identify areas where the risk of pasture expansion into forest is particularly high.

Keywords: Deforestation; Pasture expansion; Visual interpretation; Multiple logistic regression; Kenya; Ethiopia