Résumé
Taro [
Colocasia esculenta
(L.) Schott] is cultivated for its starchy corm consumed baked or boiled and processed in snacks (chips or French fries) to satisfy growing urban markets. Most consumers prefer non sweet taros. High content of sucrose (non-reducing sugar), glucose and fructose (reducing sugars) represent undesirable characteristics because they cause browning of the snacks. Breeding taro for improved corm quality is complex and phenotypic recurrent selection is impaired by the long growth cycle and the low vegetative propagation ratio. New high-throughput phenotyping tools are needed to select suitable hybrids in early clonal generations. The aim of the present study was to develop an HPTLC protocol for the quantitation of sugars in the fresh corm (FW). The individual sugars values in 60 hybrids released by four different breeding programs were compared with 300 cultivars from six different countries. Sucrose, glucose, fructose, maltose and ribose were quantitated. Mean total sugars varied from 1.83 to 6.28%FW in hybrids and from 1.32 to 7.69%FW in cultivars. The ratio sucrose/reducing sugars varied from 0.06 to 4.34 in hybrids and from 0.04 to 4.82 in cultivars. The protocol developed in this study is rapid, cost efficient, environment-friendly and more accurate than previous techniques based on dry matter because sample preparation is known to affect chemical composition of individual sugars. This technique can be used in taro breeding programmes for the early detection of undesirable hybrids with high levels of reducing sugars.