Agroforestry Systems

, Volume 53, Issue 2, pp 195–203

Research methods for multistrata agroforestry systems with coffee and cacao: recommendations from two decades of research at CATIE


  • E. Somarriba
    • Department of Agroforestry and Watershed Management
  • J. Beer
    • Department of Agroforestry and Watershed Management
  • R. G. Muschler

DOI: 10.1023/A:1013380605176

Cite this article as:
Somarriba, E., Beer, J. & Muschler, R.G. Agroforestry Systems (2001) 53: 195. doi:10.1023/A:1013380605176


This paper reviews the research themes and methodologies used by CATIE in agroforestry research with shade trees over coffee (Coffea arabica) and cacao (Theobroma cacao) during the past 20 years. Initially research focused on characterization and production studies (of crop and timber including border areas) of traditional systems using temporary and permanent sample plots on private farms. The assessment area of traditional shade-coffee (or cacao) systems should be the whole plot, including the border areas, and not some subjectively selected central area which supposedly represents unit area productivity. Uncontrolled crop, tree, and management heterogeneity limited extrapolation of early on-farm research results to other farmers' fields. Replicated case studies of best bet technologies (traditional or experimental) on different farms are often preferable to the use of formal experimental designs. On-station research included the use of systematic spacing designs to test extreme shade tree density treatments of coffee. Most nutrient cycling studies were also carried out on-station, using service and timber shade species over coffee and cacao to evaluate the ability of these agroforestry systems to maintain nutrient reserves and diversify production. Plot size (even 36 × 36 m) was limiting for long term research because of inter-plot interference, both below- and above ground, when using fast growing, tall timber trees as shade. These experiences suggest a minimum plot size of 2,500 m2. Individual tree designs and tree-crop interface studies (e.g. regression analysis of data taken along transects) are promising experimental/sampling approaches that need further development. The principal research thrusts proposed for the next five years are bio-physical process research on coffee responses to shade and competition with trees (growth, carbon allocation, phenology, disease-pest tolerance, yields and coffee quality effects) and socioeconomic analyses of both traditional and new or improved shade – coffee combinations vs. monocultures.

experimental designson-farm researchsampling criteriashade trees

Copyright information

© Kluwer Academic Publishers 2001