A Six‐Year Multisite Monitoring Network Reveals a Regionally Structured Hierarchy of Incidence Among Cotton Bollworms in Benin

ABSTRACT

Cotton bollworms are a major constraint to cotton production in Benin despite insecticide use, yet their spatio‐temporal distribution remains poorly characterised. Cotton pest management still relies on calendar‐based programmes that do not account for local variability in pest pressure. We characterised the incidence of six bollworm species ( Helicoverpa armigera (Hübner), Earias spp., Pectinophora gossypiella (Saunders), Diparopsis watersi (Rothschild), Spodoptera littoralis (Boisduval), and Spodoptera frugiperda (J.E. Smith)) using weekly larval counts from a six‐year (2018–2023) network of 977 on‐station and farmer‐managed plots across Benin. For the four most abundant species, Bayesian hurdle models with commune × year random effects separated occurrence (whether a species was detected in a plot) from intensity (larval numbers where it was present). Commune‐level multiannual trends were then derived from the random effects. A clear abundance hierarchy emerged: Earias spp. and H. armigera dominated, P. gossypiella and D. watersi were intermediate, and both Spodoptera species were occasional. Geography structured occurrence but not intensity. Earias spp. and H. armigera were near‐ubiquitous. Occurrence of H. armigera and D. watersi increased gradually northwards, whereas occurrence of P. gossypiella dropped abruptly north of 9° N. On experimental stations, the standard calendar programme reduced larval intensity by about 60% in three species without affecting occurrence. The intensive programme also reduced occurrence and provided a credible additional benefit over the standard programme on at least one component for every species. Unexpectedly, farmer‐managed plots, whose owners were advised to follow the standard programme, showed detection rates 2.1–3.0 times higher and 2.3–2.9 times higher larval intensity than standard experimental plots. No national interannual trend was detected, but local trajectories diverged. These findings provide the first national‐scale reference for cotton bollworm dynamics in Benin. They suggest treating occurrence and intensity as distinct management targets, and indicate that improving the conditions under which recommended programmes are applied is a more immediate priority than modulating them by region.