Halftime seminar: “The role of biological control and host diversity for managing resistance evolution”

Hadis’s thesis focuses on the evolution of pest resistance to biocontrol in several contexts. A major component of the half-time seminar will be a meta-analysis conducted as one chapter of the thesis (see below). Another part of the thesis focuses on AI-based sex classification of adult and pre-adult Caenorhabditis remanei model nematodes using convolutional neural networks. This work will also be briefly presented during the seminar. Opponent at the seminar is Kristina Karlsson Green, SLU Uppsala, and examiner Staffan Andersson, Department of Biological and Environmental Sciences. Hadis’ main supervisor is Luc Bussière (BioEnv & University of Agder, Norge) with Mats Olsson (BioEnv), Matthew Tinsley (Biological and Environmental Sciences, University of Stirling, UK) and  Rosemary Mangan (Biological and Environmental Sciences, University of Stirling, UK ) as co-supervisors.

Short summary

A key presumed advantage of using microbial biopesticides, which contain living organisms, is their low risk of inducing resistance in hosts. Nevertheless, microbial biopesticides can also lead to the evolution of resistance, but the magnitude of this risk, and the nature of moderators that affect it remain poorly quantified. The risk of resistance evolution should theoretically be linked to the biochemical complexity of interactions between pests and pathogens, which might make distinct classes of biopesticide more or less evolutionarily sustainable. We conducted a meta-analysis of studies published between 1990 and 2025 (Web of Science) that quantified insect susceptibility to biopesticides across multiple generations. Across studies, we detected a negative evolutionary response on average, indicating a general tendency toward reduced host susceptibility over time, but with substantial heterogeneity among systems. Across analyses, the number of generations under selection emerged as the most consistent predictor of resistance evolution, with evolutionary responses strengthening as selection persisted. Designs including a component of relaxed selection before measurement were consistently associated with weaker estimates of evolutionary responses, highlighting the importance of accounting for transgenerational plasticity and environmental carryover effects. In addition, selection pressure and biocontrol agent type showed modest main effects, with evidence for context-dependent interactions that were sensitive to influential studies and constrained by the structured nature of the literature. Together, these results demonstrate that resistance evolution to biocontrol agents is a general consequence of sustained selection, while emphasizing that its tempo and trajectory are shaped primarily by selection duration, and experimental context rather than by biocontrol agent class alone.