Abstract
Rapid-cycling recurrent selection has been implemented to produce cassava varieties that deliver provitamin A carotenoids to human populations that are affected by vitamin A deficiency. Rapid-cycling
selection relies on quantification in just one plant per genotype. This study was conducted to assess the
reliability of single-plant evaluations and the effect of age of the plant on carotenoid content in cassava
roots. Variation in aliquot quantifications from the same root was negligible indicating a reliable
experimental procedure. A large source of variation for carotenoids was due to differences among
genotypes. Root-to-root variation from the same plant in some cases was surprisingly high and
accounted for an average of 25% of the total variation. Plant-to-plant variation was not as high and
accounted for 20% of the total variance. Carotenoid content was shown to vary depending on the age of
the plant. Single-plant evaluations for carotenoid content in cassava, which is a requirement for rapidcycling recurrent selection is acceptable considering that it reduces in half the time required for
evaluation and selection. However, it is suggested that 2 to 3 roots per plant be combined together in a
sample to represent each genotype at a standard plant age.
Key words: Plant age, environmental influence, rapid cycling, genetic gains, -carotene, repeatability
