Rwanda is a small, densely populated country in central east Africa. The economy is mainly dominated
by agriculture, with more than 90% of the population practicing subsistence agriculture. Consequently,
farming lands are highly fragmented into small farms. Most natural forests have been cleared and the
remnant natural forests are now protected areas. Currently, most forest products are supplied from tiny
private woodlots and trees scattered on farms. Tree resources outside forest (TROF) are becoming more
and more important for sustainable development of the country because they provide a number of
forest goods and services. In fact, TROF reduce pressure on remaining forest resources and therefore
they should be managed in a sustainable manner. However, this can only be possible after assessing
available TROF in terms of quantity, species richness and distribution in different land use systems. This
study aims to assess the spatial distribution of TROF in relation to dominant land use systems in one
administrative sector in Central Rwanda. A subset of an Ikonos multispectral image of July 2007 was used
to determine dominant land use classes using ERDAS IMAGINE 9.2 and a land use map for the study area
was prepared using ArcGIS 10. Field data on tree species diversity and density in different land use types
were collected using a Stratified Systematic Cluster Sampling method with random start to enable
statistical analysis of the results. A one‐way analysis of variance was used to test for significant
differences between selected parameters in studied land use systems. Duncan’s multiple range tests
were used to compare pairs of land use systems in order to establish which of the land use systems were
significantly different. The results showed that the pasture land use system registered the highest TROF
basal area (with an average of 8 m2
/ha), number of trees (average of 123 trees/ha) and species richness
(average of 25 species/ha). The highest number of tree stumps (average of 38 stumps/ha) occurred in
cropland meaning that TROF are more harvested in croplands than in other land uses. Therefore on farm
tree planting schemes along with improved agroforestry technologies and adequate TROF management
interventions should be initiated on croplands and scaled up on other sectors with similar socio‐
economic and ecological conditions.