**Background:** Helicoverpa armigera is a major pest of pigeonpea (Cajanus cajan), causing substantial yield losses, particularly in short-duration cultivars. Despite extensive screening for host-plant resistance, progress has been limited, and many studies rely on detached-plant-part assays that may not reflect real-world insect behavior. This study aimed to evaluate oviposition preference and larval performance on whole plants of three short-duration pigeonpea cultivars: ICPL 87 (susceptible control), ICPL 86012, and ICPL 88039 (both previously reported to have low-to-moderate resistance).
**Methods:** Pigeonpea plants were grown in a controlled-temperature glasshouse (27±4°C day, 25±4°C night) in Toowoomba, Australia. A laboratory colony of H. armigera was established from field collections in southeast Queensland. For oviposition choice experiments, single mated female moths were placed in cages containing one flowering plant of each cultivar (n=17 replicates). For no-choice experiments, single females were placed with one plant of a single cultivar (n=34 moths that laid eggs). Eggs were counted and their locations (plant structure and substructure) recorded. For larval performance, neonate larvae were placed on flowers or leaves (10 per plant, 5 per node) and recovered after 72 h; survival, weight, and developmental stage were recorded. Trichome densities (Types A–D) were quantified on flower calyxes, adaxial and abaxial leaf surfaces under 8× magnification using an ocular grid. Statistical analyses included ANOVA, Kruskal–Wallis tests, and Fisher's LSD post-hoc tests in R.
**Key Results:** In the choice experiment, a higher proportion of eggs was laid on ICPL 88039 (F₂,₄₈=3.35, P=0.044). In the no-choice experiment, total egg numbers did not differ among cultivars (χ²₂=4.52, P=0.10). Most eggs were laid on floral structures; the proportion on floral structures was highest for ICPL 88039 (F₂,₂₆=5.20, P=0.013). Within floral structures, eggs on bud initials and buds were predominantly on calyxes (98% and 71%, respectively), while on flowers and spent flowers eggs were evenly distributed between calyxes and petals. On leaves, 86% of eggs were on the abaxial surface. Larval survival did not differ among cultivars (F₂,₂₀=2.03, P=0.16) but was higher when neonates were placed on flowers (82%) versus leaves (67%) (F₁,₂₀=10.80, P=0.004). Larval weight was affected by cultivar (F₂,₂₀=7.26, P=0.004) and placement (F₁,₂₀=14.86, P<0.001); larvae were heaviest on ICPL 86012. Larval development (proportion reaching second instar) was also affected by cultivar (F₂,₂₀=4.7, P=0.021) and placement (F₁,₂₀=11.58, P=0.003), with faster development on ICPL 86012 and on flowers. Trichome densities varied significantly among cultivars and plant structures. On calyxes, ICPL 87 had the most Type A trichomes (2.09±0.36 per mm²) but fewer Types B, C, and D. ICPL 86012 and ICPL 88039 had higher densities of non-glandular trichomes on calyxes. Trichomes were absent from flower petals.
**Clinical Implications:** This study demonstrates that whole-plant assays yield different conclusions than detached-part bioassays for H. armigera resistance in pigeonpea. Cultivars previously reported as resistant (ICPL 86012 and ICPL 88039) did not show ovipositional non-preference or larval antibiosis when tested on whole plants; in fact, larvae performed best on ICPL 86012. The strong preference of moths and larvae for floral structures, and the superior larval performance on flowers, indicates that resistance screening must account for insect behavior and plant architecture. Future research should use whole-plant assays, consider the role of trichomes on pods (not just flowers and leaves), and examine how plant architecture (determinate vs. indeterminate growth) influences susceptibility. These findings have direct implications for breeding programs aiming to develop H. armigera-resistant pigeonpea varieties.