Water yam (Dioscorea alata L.) development as affected by photoperiod and temperature: Experiment and modeling

Water yam (Dioscorea alata L.) is one of the most Popular tuber crops in the tropics. Although it is well known that photoperiod greatly affects yam development, little information is available on the effect of temperature on different developmental phases, and there has been no attempt to model the effect of both environmental factors. The aim of this study was to assess the combined effect of photoperiod and temperature on the development of two early maturing varieties of yam with similar growth pattern. For this We used a model proposed for potatoes grown under tropical conditions. Experimental information was obtained from 15 field experiments carried out in Guadeloupe (French West Indies) covering a wide range of planting dates. Two yam phases were analyzed: from emergence (EM) to tuber initiation (TI), and from TI to harvest (HA). The EM-HA period varied from 3 to 6 months, with the longer cycles corresponding to early planting dates (e.g. April). On average, EM-TI represented one-third of the EM-HA period, and was mainly affected by photoperiod and to a lesser extent by temperature. Both factors also affected the duration of TI-HA but their effects were less noticeable. The observed mean temperature during TI-HA was near the estimated optimum and its effect was less than that of photoperiod. The variation of the phase duration was higher for EM-TI (CV 45%) than for TI-HA (8%), which was satisfactorily explained by the model, For EM-TI there was a positive interaction between both environmental factors which, together with the greater influence of climatic conditions, resulted in much variability in its duration. For TI-HA, a small influence of environmental factors, coupled with a compensatory effect between environmental factors, resulted in a relatively smaller variation in the duration of this phase. The model estimated satisfactorily the dates and the duration of each phase. The root mean square error (RMSE) was 7.4 d which corresponded to 13% and 7% of the observed mean duration of EM-TI and TI-HA, respectively. Our results showed that small changes in photoperiod and temperature, which are very usual in the tropics, have a big effect on the tested yam varieties. We also showed that the model applied in this study can be a useful tool to predict yam development for management purposes as well as for the modeling of yam growth. (C) 2009 Elsevier B.V. All rights reserved.