Article contents
Edge Effects and Interactions with Environment in Cocoa
Published online by Cambridge University Press: 03 October 2008
Summary
In trials of seedling cocoa in Ghana, biases in the estimates of yield and tree size arising on the perimeters of experiments and on the interfaces of unguarded plots were strongly correlated, characteristic of progenies and stable across environments, but were inconsistent for black pod losses. No variate showed progeny × site interactions, but those of Caribbean yield data were interpreted by principal components analysis. The results show that 25-tree plots may be used for initial tests of new varieties and, as long as the environmental variation does not exceed that of the Ghanaian trials, at single sites.
- Type
- Research Article
- Information
- Copyright
- Copyright © Cambridge University Press 1979
References
REFERENCES
- 5
- Cited by