[Defended PhD Thesis] Julien Lamour: Analysis of spatialised data from production to improve agronomic diagnosis in banana plantations – Consideration of crop asynchronism

PhD Thesis Defence: Friday 12 July, 14 h 00, Montpellier SupAgro, 2 Place Viala, amphi 206

Julien Lamour is one of the first #DigitAg labeled PhDs

Analysis of spatialised data from production to improve agronomic diagnosis in banana plantations – Consideration of crop asynchronism

  • Start Date: June 2016
  • Defence: 12 July 2019
  • University: MUSE Montpellier University of Excellence / Montpellier SupAgro
  • PhD School:  GAIA, APAB, Montpellier (France)
  • Field(s): Agronomy / Plant Physiology – Precision Agriculture/ GIS – Statistics
  • Doctoral Thesis Advisor:  Bruno Tisseyre (Montpellier SupAgro)
  • Co-supervisors :  Olivier Naud (Irstea, UMR ITAP), Mathieu Léchaudel (Cirad, UMR QualiSud), Alain Normand (Compagnie Fruitière)
  • Funding: Cifre Agreement Irstea-Compagnie Fruitière
  • #DigitAg: Labeled PhD – Challenge 8 (ICT and agricultural development in Southern countries (Africa))

Keywords: Precision agriculture, Variogram, geostatistics, Cosineogram, Observational study, Banana


Exported bananas are produced on industrial plantations covering large areas and relying most of the times on clones of the Cavendish cultivars. This crop is input and labour intensive and is subject to environmental and societal constraints that impose an improvement in agricultural practices. Precision Agriculture (PA) is a methodological approach that has emerged on arable crops to optimize their yield, quality and reduce their environmental impact. This approach uses geo-referenced producer’s data and studies their spatial variability to increase the farming performance by adjusting the management by area. Its objective is to highlight the variable growth conditions and identify those that are manageable in order to optimize production processes according to the specific potential of the sites on the farm.

Unlike arable crops, bananas exhibit unique characteristics that must be taken into account when studying production and yields. In particular, banana plants are asynchronous, their development cycle is not seasonal and is not synchronized by the cropping system. As a result, the fields are made up of banana plants at different phenological stages. Thus, unlike arable crops which are synchronous, the observations that can be made on a banana plantation at a given time depend not only on growing conditions but also on the variable phenological stages of the plants.

The objective of this doctoral thesis was to propose new methods for using data produced in banana plantations in order to facilitate agronomic diagnosis in a PA approach. The data used are mainly those recorded to manage the harvests. These are observations classically recorded in banana plantations. We also studied remote sensing data and proposed analytical methods to study spatial variability by reducing the bias due to the asynchronism of banana plants.

The first part of the thesis work consisted in proposing methods to characterize the asynchronism of the fields. For this purpose, we defined several indicators: the average duration of the banana plant development cycle; the heterogeneity of growing conditions; and finally the average phenological stage of a field and the within field variability of the stages. These methods were applied to flowering data from an industrial plantation in Cameroon. An effect of the environment and producers’ practices on asynchronism was highlighted.

A second part of the thesis consisted in proposing a model that identifies variability related to the environment and producers’ choices using production data. The purpose of this method was to generate maps that could be interpreted agronomically, without the asynchronism bias. The model we proposed was applied to a particular property: the time between flowering and commercial maturity. It is assumed that this method is general enough to be applied to other agronomic properties such as the weight of bunches at harvest.

Finally, the last subject we studied was the relevancy of remote sensing in banana plantations despite the diversity of phenological stages. We assessed the importance of different sources of variability on the chlorophyll content of banana leaves measured from a pedestrian sensor and constructed an index to predict this content by drone. Maps of within field variability of chlorophyll content have been produced by this method, but agronomic interpretation must be done with caution because the phenological stages are not known and their effects cannot be corrected. The conditions for interpreting these maps are discussed.

Contact: julien.lamour [AT] fruitiere.fr – Tél : (+33) 6 79 67 06 22


Communications /Publications

Lamour, J., Le Moguedec, G., Naud, O., Léchaudel, M., Taylor, J., et al. (2020) Evaluating the drivers of banana flowering cycle duration using a stochastic model and on farm production data, Precision Agriculture

Lamour, J., Leroux, C., Naud, O., Lechaudel, M., Le Moguédec G. and Tisseyre, B. (Accepted 2019). Disentangling the sources of chlorophyll-content variability in banana crops using an optical chlorophyll meter. 12th European  Conference  on  Precision  Agriculture  (ECPA  2019),  SupAgro, Montpellier,  France,  July  8–11  2019

Rabatel, G., Lamour, J., Moura, D. and Naud, O. (Accepted 2019). A multispectral processing chain for chlorophyll content assessment in banana fields by UAV imagery. 12th European  Conference  on  Precision  Agriculture  (ECPA  2019),  SupAgro, Montpellier,  France,  July  8–11  2019

Leroux, C., Jones, H., Pichon, L., Guillaume, S., Lamour, J., Taylor, J., … and Tisseyre, B. (2018). GeoFIS: An Open Source, Decision-Support Tool for Precision Agriculture Data. Agriculture, 8(6), 73. DOI : doi: https://doi.org/10.3390/agriculture8060073

J. Lamour, O. Naud, M. Lechaudel  & B. Tisseyre (2017), Mapping properties of an asynchronous crop: the example of time interval between flowering and maturity of banana. Paper presented at the 11th European Conference on Precision Agriculture (ECPA 2017), John McIntyre Centre, Edinburgh, UK, July 16–20 2017, Advances in Animal Biosciences , 8(2), 481-486. DOI: https://doi.org/10.1017/S2040470017000449