A Dynamic Action Threshold for Spodoptera frugiperda in Tomato Integrating Crop Phenology and Natural-Enemy Pressure

Abstract

Background: Static action thresholds for Spodoptera frugiperda in tomato are agronomically flawed because they ignore the crop's changing susceptibility and the suppressive capacity of natural enemies. This study addresses that gap by developing a dynamic action threshold that explicitly integrates crop phenology (vegetative, flowering, fruiting) with observed predator‑to‑pest pressure to produce a field‑ready, auditable decision rule. Methods: Stage‑specific Economic Injury Levels (EILs) were established at 3.2, 1.8, and 2.1 larvae per 10 plants for vegetative, flowering, and fruiting stages, respectively. A transparent piecewise adjustment rule was then applied, adding 0.0, 0.7, 1.7, or 2.7 larvae per 10 plants based on four predator‑to‑pest ratio bands (<0.2, 0.2–0.5, 0.5–1.0, >1.0). This framework was field‑validated against a static threshold (2 larvae/10 plants) and a seven‑day calendar spray programme, with all strategies assessed for application frequency, marketable yield, and direct control cost. Results: The dynamic strategy required only 5.2 insecticide applications, representing a 35.0% reduction compared to the calendar programme (8.0 applications) and a 22.4% reduction versus the static threshold (6.7 applications). Direct control cost fell significantly from PKR 70,000 ha⁻¹ (calendar) to PKR 45,000 ha⁻¹ (dynamic), while reported marketable yield remained statistically similar across all treatments (range: 31.8–33.2 t ha⁻¹). The lowest EIL during flowering (1.8) ensured protection of the most economically sensitive stage, while high natural‑enemy pressure (>1.0 ratio) raised the threshold to 4.5–5.9 larvae, safely deferring sprays when biological control was active. Implications: The framework demonstrates that a transparent, phenology‑ and enemy‑aware decision rule can substantially reduce pesticide use and expenditure without compromising harvest outcomes. However, because the damage‑function data and taxon‑specific predator weights remain unverified, the authors explicitly caution that this is a provisional tool, not a universal recommendation. Full deployment requires rigorous equivalence testing, multi‑season validation, and recalibration of economic parameters to local market conditions, ensuring that the threshold remains both ecologically sound and economically defensible.