Germplasm and data generation
A panel of 289 diverse maize inbred lines was used for this study (Supplemental Table 1). The panel included both FAW resistant and susceptible lines from various breeding programs, some of which formed association panels published previously (Warburton et al., 2013; Flint-Garcia et al., 2005). These lines were genotyped and phenotyped as inbreds. They were grown at the R.R. Foil Plant Science Research Center, Mississippi State, Mississippi (MS) in 2019, 2020, and 2021. Seeds were planted in 4 m single-row plots with 0.97 m spacing between rows and thinned to 20 plants plot-1. The experiment was planted in a randomized complete block design with two replications in 2019 and 2020 and three replications in 2021. Standard production practices were followed, and supplemental irrigation was applied as needed.
The fall armyworm neonate larvae used in this study were obtained from the laboratory colony at the insect rearing facility of the Corn Host Plant Resistance Research Unit (CHPRRU), USDA-ARS, Mississippi State. The larvae were reared using artificial diet following the procedure described by Davis (1989). Within 24 hours following the egg hatch, first instar larvae (neonates) were mixed with corn cob grits. The mixture was dispensed into the maize whorl utilizing a mechanical applicator (Wiseman et al., 1980). Each individual plant was infested with approximately 40 FAW neonates at the mid-whorl stage (V-7 of maize growth stage). All plants within experimental plots were rated individually to determine the leaf feeding damage caused by FAW. The damage was visually rated at 7 days and 14 days after infestation. The 7-day and 14-day leaf feeding ratings were conducted on a 0 (no visible leaf feeding damage) to 9 scale according to Davis et al. (1992; Supplemental Table 2). The 7-day ratings were conducted by one person and the 14-day ratings were conducted independently by two people.
The mean rating for each entry was calculated as an adjusted mean (Best Linear Unbiased Estimator; BLUE) using mixed model methodology in the PROC GLIMMIX procedure of SAS 9.4 (SAS Institute, Cary, NC) and can be found in Supplemental Table 1. For the 7-day ratings, the individual plants within a plot were treated as subsamples. Plot means were calculated within years treating genotype as a fixed effect and replication and genotype by replication as random effects. To combine the 7-day ratings over years, genotype was treated as a fixed effect and year, replication nested within year, year by genotype interaction, and genotype by replication nested in year were all treated as random effects. For the 14-day ratings two sets of individual plant ratings, one set per rater, were used as subsamples for each plot. Plot means were calculated similarly to the 7-day ratings, but additional model terms were added to account for the variance due to different raters. Within years, genotype was treated as a fixed effect and replication, rater, rater by replication, genotype by replication, and genotype by rater by replication were all treated as random effects. When the data was combined across three years, genotype was treated as a fixed effect and year, replication nested in year, rater, rater by replication nested in year, year by genotype interaction, year by rater interaction, genotype by replication nested in year, and rater by genotype by replication nested in year were all treated as random effects.