Farmer participatory evaluation of agroforestry trees in eastern Zambia Elias Kuntashula * , Paramu L. Mafongoya Zambia-ICRAF Agroforestry Project, P.O. Box 510089, Chipata, Zambia Received 17 July 2003; revised 14 April 2004; accepted for publication 1 June 2004 Abstract The use of participatory research methods in evaluating tree legumes has received little at- tention among researchers because of the widely-held assumption that data generated in this way are qualitative and not amenable to statistical testing. Using a participatory tool called a bao game, 112 farmers in eastern Zambia rated 11 agroforestry trees for soil fertility improve- ment, source of fuel wood, light construction materials, poles and fodder. Leucaena collinsii (provenance 45/85) was rated highest for provision of all the above benefits. Gliricidia sepium, Acacia angustissima and Calliandra calothyrsus (provenance ex Embu) had high scores for soil fertility improvement, while Senna siamea, Leucaena esculenta (provenance 52/87) and Leuc- aena pallida were rated highly for fuel wood and pole production. These later species, with the exception of Senna siamea, were also rated highly for light construction materials. A. an- gustissima and C. calothyrsus were rated highly for fodder production. L. diversifolia (prove- nance 35/88), L. diversifolia (provenance 53/88), S. spectabilis and L. esculenta (provenance ex Machakos) were rated low for all the above benefits. This work shows that the data generated using the bao game can be quantitatively analyzed in a statistically rigorous manner. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Participatory research; Tree legumes; bao game; Qualitative; Quantitative 0308-521X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.agsy.2004.06.009 * Corresponding author. Tel./fax: 260-06-221404. E-mail address: kelias@zamnet.zm (E. Kuntashula). www.elsevier.com/locate/agsy Agricultural Systems 84 (2005) 39–53 AGRICULTURAL SYSTEMS