Agroforestry Systems

, Volume 82, Issue 3, pp 303-330

First online:

Tree species selection for land rehabilitation in Ethiopia: from fragmented knowledge to an integrated multi-criteria decision approach

  • Bert ReubensAffiliated withDivision Forest, Nature and Landscape, Katholieke Universiteit LeuvenInstitute for Agricultural and Fisheries Research (ILVO), Plant Sciences Unit—Crop Husbandry and Environment Email author 
  • , Clara MoeremansAffiliated withDivision Forest, Nature and Landscape, Katholieke Universiteit Leuven
  • , Jean PoesenAffiliated withDivision Physical and Regional Geography, Katholieke Universiteit Leuven
  • , Jan NyssenAffiliated withDepartment of Geography, Ghent University
  • , Sarah TewoldeberhanAffiliated withLand Resources Management and Environmental Protection Department, Mekelle University
  • , Steve FranzelAffiliated withInternational Centre for Research in Agroforestry, United Nations Avenue
  • , Jozef DeckersAffiliated withDivision Soil and Water Management, Katholieke Universiteit Leuven
  • , Caleb OrwaAffiliated withInternational Centre for Research in Agroforestry, United Nations Avenue
  • , Bart MuysAffiliated withDivision Forest, Nature and Landscape, Katholieke Universiteit Leuven

Rent the article at a discount

Rent now

* Final gross prices may vary according to local VAT.

Get Access


Dryland regions worldwide are increasingly suffering from losses of soil and biodiversity as a consequence of land degradation. Integrated conservation, rehabilitation and community-based management of natural resources are therefore of vital importance. Local planting efforts should focus on species performing a wide range of functions. Too often however, unsuitable tree species are planted when both ecological suitability for the targeted area or preferences of local stakeholders are not properly taken into account during selection. To develop a decision support tool for multi-purpose species selection, first information needs to be pooled on species-specific ranges, characteristics and functions for a set of potentially valuable species. In this study such database has been developed for the highly degraded northern Ethiopian highlands, using a unique combination of information sources, and with particular attention for local ecological knowledge and preferences. A set of candidate tree species and potentially relevant criteria, a flexible input database with species performance scores upon these criteria, and a ready-to-use multi-criteria decision support tool are presented. Two examples of species selection under different scenarios have been worked out in detail, with highest scores obtained for Cordia africana and Dodonaea angustifolia, as well as Eucalyptus spp., Acacia abyssinica, Acacia saligna, Olea europaea and Faidherbia albida. Sensitivity to criteria weights, and reliability and lack of knowledge on particular species attributes remain constraints towards applicability, particularly when many species are jointly evaluated. Nonetheless, the amount and diversity of the knowledge pooled in the presented database is high, covering 91 species and 45 attributes.


Afforestation Dryland restoration Local ecological knowledge Multi-purpose tree Rural appraisal Species selection tool