Vol. 2, Issue 1 (2017)
Statistical models for coconut (Cocos nucifera L.) pollen fertility prediction according to its age and biochemical composition
Author(s): Yao Saraka Didier Martial, Konan Brou Roger, Deffan Ange Benedicte Zranseu, Koffi Eric Blanchard Zadjehi, Yoboue Koffi, Diarrassouba Nafan, Konan Konan Jean Louis, Sie Raoul Sylvère
Abstract: This investigation was conducted to propose predicting models of stored pollen fertility through simple linear regression equations to improve the yields of coconut (Cocos nucifera L.) seeds produced by controlled pollination method. Therefore, pollen samples extracted from both coconut populations Comoro Moheli Tall (CMT) and Malayan Yellow Dwarf (MYD) planted in field coconut genebank at Port-Bouët, Côte d’Ivoire, were used. Therefore, 2 pollen fertility components, fruit-set rate and in vitro pollen germination rate and 5 biochemical traits that are moisture, lipid, protein, total sugars and reducing sugars contents in pollen extracted initially (0 month) then conditioned vacuum and stored in deep-freezer at -15°C to reach the pollen ages that are 1, 2, 3 and 4 months were studied. The results showed that in vitro pollen germination rate assessed at the laboratory was the best indicator having explained 64% fluctuations of fruit-set rate at the field in Tall coconut ecotype. Rates of in vitro pollen germination were predicted in 56.25% of the cases by its total sugars content and in 79.21% by pollen age respectively in Tall and Dwarf coconut palms studied. Mathematical models relatively of the fruit-set and in vitro pollen germination rates revealed in this study could be helpful to an efficient management of pollen bank for increase seed production by hand pollinations in coconut populations.