Spatial estimation of yield for a perennial crop in West Africa: the case of mango case in Senegal

Hello, my name is Julien Sarron. I come from a family of farmers and have always been curious and interested in the natural sciences, particularly the environment and agriculture. I graduated from Agrocampus Ouest with a specialization in Plant Production Science (Agrosystems Engineering option).  During this period I did 2 6-month internships, including one in a horticultural research center in Vietnam. My end-of-study internship focused on the agronomic diagnosis of stagnating sunflower yields in France, and I was welcomed at INRA's UMR AGIR in Toulouse.
I chose to continue my thesis at CIRAD for several reasons. Firstly, to deepen my scientific knowledge and participate in the research effort, while remaining focused on concrete, medium-term applications. Secondly, because CIRAD is involved in finding solutions to global problems, such as climate change. Finally, I want to compare the tools already available in the world of research, particularly the use of technologies in agriculture and digital agriculture, with the reality in the field and its constraints and limits.

• Starting date: October 2016
• University: University of Montpellier / Institut Agro
• PhD school: GAIA , Montpellier
• Scientific field: Agronomic sciences
• Thesis management: Eric Malézieux (Cirad, HortSys)
• Thesis management: Emile Faye (Cirad, HortSys)
• Funding: Cirad
•  #DigitAg : Labeled PhD – Axe 6 – Challenges 1 et 8

Keywords: Fruit tree, mango tree, agroforestry, cropping system, remote sensing, digital agriculture, Sahel, West Africa, agroecology

Abstract: In Africa, crop yield estimation is a critical and strategic challenge to face development issues and reduce population vulnerability to global changes. However, tools for harvest estimation are still vague for most of annual crops. In the case of perennial crops (e.g. fruit trees), it exists an important knowledge gap and methods are still missing because of the lowest scientific interest for this kind of crops; although specific questions arise for these crops (e.g. featuring the variability in plant phenology across and between trees, and orchards?). Thus, the aim of this Ph.D. is to develop an operative tool for a multi-scale spatial estimation of mango yield from the tree to the regional scale (the region is the Niayes area in West Senegal, the main horticultural production area of the country). This work will rely on various image analysis methodologies brought together: at regional scale, satellite images will be used to classify mango orchards following a typology of mango-based cropping systems; at orchard and tree scales, yield heterogeneity at the intra and inter orchard levels will be quantified within a monitoring network of 30 orchards representing the cropping system variability of the study area. For this purpose, the number of tree organs (sprouts, inflorescences, green and mature fruits) will be counted by analysis of RGB images at the tree scale, and RGB and multispectral orthomosaics will be acquired by U.A.V. and analysed for agronomic indices at the orchard scale. Agronomic factors of the tree (variety, crown size, height and trunk diameter), agroecological parameters (crop management, cultivated biodiversity, orchards structure) and data describing the environment (landscape structure, climate, pedology) will be taken into account at each scale. In fine, mango production estimated at tree and orchard scales by image analysis will be extrapolated at regional scale by integrating these factors. Finally, this thesis could bring opportunities to develop an innovative tool for multi-scale spatial estimation of mango yields.

Contact : Eric.Malezieux [AT] cirad.fr

Social networks: ResearchGate – LinkedIn

Download the thsis manuscript: Spatial estimation of yield for a perennial crop in West Africa: the case of mango case in Senegal: https://agritrop.cirad.fr/596637/

Communications and Papers

• Sarron, J.; Malézieux, É.; Sané, C.A.B.; Faye, É. (2018). Mango Yield Mapping at the Orchard Scale Based on Tree Structure and Land Cover Assessed by UAV. Remote Sensing, 10(2), 1900. https://doi.org/10.3390/rs10121900
• Sarron J., Sané C.A.B., Diatta P., Malézieux É, Faye É, Plant diversity affects the productivity of Senegalese mango orchards: evidences from UAV photogrammetry, 4th World Congress of Agroforestry (2019) – https://agritrop.cirad.fr/592664/
• P. Borianne, J. Sarron, F. Borne and E. Faye, Deep mangoes: from fruit detection to cultivar identification in color images of mango trees, DISP’19 – International Conference on Digital image and Signal Processing (2019) – https://hal.archives-ouvertes.fr/hal-02295256

PixYield – L’analyse d’images photographiques comme outil pour l’estimation de rendement : le cas du manguier au Sénégal (poster)

Read more : Estimer les rendements de manguiers en Afrique de l’Ouest : des outils issus de la thèse de Julien Sarron