Do Farm Characteristics Explain Differences in Tree Species Diversity among Western Kenyan Farms?

• Published in: Agroforestry systems 2004-12, Vol.63 (1), p.63-74
• Main Authors: Kindt, R., Simons, A. J., Van Damme, P.
• Format: Article
• Language: English
• Subjects: Agricultural land; Biodiversity; Diversity indices; Farmers; Farms; Plant diversity; Plant species; Species composition; Species diversity; Species richness
• ISSN: 0167-4366; 1572-9680
• DOI: 10.1023/B:AGFO.0000049434.54654.97

With the objective of planning diversification of on-farm tree species composition, a survey was conducted in western Kenya involving a complete tree census (trees were defined as all woody or ligneous plants, including shrubs and lianas) and collection of ethnobotanical information on 201 farms. Differences between farms in diversity of the 12 most frequent use-groups were analysed by species richness, Shannon, Simpson and Berger-Parker diversity indices, and Shannon evenness and equitability. A large range of values was detected among farms and use-groups. Multiple linear regression of diversity statistics on household characteristics indicated significant relationships. However, these relationships generally explained low percentages of variation (ranging 2-44%). The connection between household characteristics and use-group diversity allows targeting specific household types with lower diversity. Farm size had a positive relationship with diversity of most use-groups. However, accumulation curves revealed that the same area carried a larger abundance and diversity when it was composed of a greater number of smaller farms. If the pattern of further subdivision of farmland in the survey area continues and the same differences between smaller and larger farms prevail, then larger diversity per unit area can be expected. Because smaller farms contain smaller diversity, however, diversification with the aim of enhancing or stabilising productivity of individual farms may become an important priority in the survey area. The results presented allow for the identification of individual farms, use-groups, and household types for which diversification is more relevant, and at the same time allow for impact monitoring.[PUBLICATION ABSTRACT]