S16 - Session O1 - Variability of mango tree production (cv. 'Kent') in Northern Ivory Coast

Authors: Jean Mathias Koffi *, Julien Sarron, Dognimeton Soro, Frederic Normand, Charles Konan Kouakou, Lassina Fondio, Emile Faye

Understanding the factors that determine the variability of intra- and inter-orchard production is key to better guide agronomic research and improve the level of production of fruit trees. Few studies have focused on identifying these factors in mango orchards in Cote d'Ivoire, yet one of the main producing countries in West Africa. This work consists in quantifying the production of the mango tree and analysing the effects of the tree structure and the year on the production levels. In 23 orchards, 230 mango trees (cv. 'Kent') were selected to be representative of the tree observed in the Poro region in Northern Cote d'Ivoire. Different variables of tree structure were measured: tree height, trunk height, trunk circumference and crown dimensions. Individual tree production (in kg of fruit per tree) was estimated in 2019 and 2020 by image analysis. Each tree was photographed before harvesting on two opposite sides with a RGB camera. Images were analysed with a machine learning algorithm combined with mixed linear model. Similarly to other countries in West Africa, a high variability was observed between the average production of the trees monitored orchards in 2019 (CV=84.61%) and in 2020 (CV= 130.62%). The average production per tree was significantly higher in 2019 (124.1 +- 105.0 kg of fruits) compared to 2020 (34.2 +- 44.7 kg of fruits). In 2019, the tree production was positively correlated with structure variables, unlike in 2020 where these correlations were weak and negative. Individual tree production in 2020 was significantly but weakly and negatively correlated (r= - 0.34) with that of 2019. Mango tree production in Northern Cote d'Ivoire showed a high variability between trees and between years that could not be explained by tree structure only.