Modelling the Sterile Insect Technique in an agricultural context: examination of biological and technical factors likely to reduce its effectiveness

President of the jury :                       

• Céline Casenave, UMR MISTEA (INRAE) 

Rapporteurs :                     

• Pierre-Alexandre Bliman, Inria-Paris 
• Daniele Bevacqua, PSH (INRAE)                                 

Examiners

• Camille Coron, AgroParisTech (INRAE) 
• Louise van Oudenhove, ISA (INRAE) 

Thesis Directors :

• Ludovic Mailleret, ISA (INRAE) 

Abstract :

The Sterile Insect Technique (SIT) is a biological control method that consists in the massive release of sterilized males in order to reduce pest insect populations. In agriculture, females of certain pest species, such as the Mediterranean fruit fly (Ceratitis capitata) or the spotted-wing drosophila (Drosophila suzukii), lay their larvae inside fruits, causing them to rot and leading to considerable economic losses for farmers. Although promising, the effectiveness of SIT can be weakened by several biological and technical factors. During my PhD, I developed different mathematical models (differential equations, individual-based models, impulsive approaches) to analyze these limitations and to identify conditions that enable the control of pest populations.

Three main research axes structured this work: (i) residual fertility: despite sterilization, some males retain a low level of fertility. I showed, through a model applied to Ceratitis capitata, that even a small rate of residual fertility can compromise the eradication of pest populations; (ii) multiple mating: in many pest species, females mate several times, which complicates the impact of sterilized male releases. I compared a compartmental model and an individual-based model calibrated on Drosophila suzukii, and demonstrated that multiple matings are disadvantageous for long-term control but may be beneficial in the short term. Moreover, the individual-based model allowed me to explore the effect of different sperm use biases (first, last, mixed, preference for fertile or sterile sperm) on SIT efficiency, which ultimately appeared to have little influence on control capacity; (iii) the agricultural context: by integrating economic constraints and crop damage related to sterile oviposition, I studied discrete release strategies under limited budgets. The results highlight a frequency–amplitude trade-off in release strategies, as well as a ``hidden cost'' associated with sterile punctures (oviposition punctures by females mated with sterilized males) that reduce fruit quality, although less severely than fertile punctures.

This thesis highlights the biological (residual fertility, multiple matings) and operational (budget, frequency and intensity of releases) conditions that influence the success of SIT. It relies on different modeling frameworks, including differential equations, individual-based models and impulsive approaches, to analyze these factors. Altogether, this work provides perspectives to improve the effectiveness of SIT in agriculture, by reconciling pest population control with the reduction of crop losses.

Keywords :

Mathematical modeling; SIT; Residual fertility; Multiple mating; Ordinary differential equations; Eradication threshold; Population dynamics

In person or via Zoom: 

https://inrae-fr.zoom.us/j/8869452953?omn=98369150338