This research was conducted from November 2024 to April 2025 at the Plant Pest Laboratory, Department of Plant Pests and Diseases, Faculty of Agriculture, Universitas Brawijaya (UB). The experimental design was used a completely randomized design (CRD) with twelve treatments (maize varieties) and four replication. The maize varieties tested were: Bisi 2, BTS 5, Bisma, Lamuru, Perkasa, Nusa 01, Nusa 03, Pertiwi 3, Sukmaraga, Bima 20 URI, Bima 14 Batara, and Srikandi Kuning.

Insect rearing. The coffee bean weevil, A. fasciculatus, specimens were obtained from the established insect collection at the Plant Pest Laboratory, UB. The insects were reared using a modified method based on (Novo & Baptista, 1998) and (Atikah et al., 2022). Five hundred grams (500 g) of dried cassava were placed into a rearing box (16 cm × 16 cm × 10 cm). Two hundred (200) unsexed A. fasciculatus adults were introduced into the box for seven days for oviposition. After this period, the adults were removed, and the infested diet was maintained in the box until the F1 progeny emerged. The adult A. fasciculatus used for the preference test were aged 1–2 weeks old (Novo & Baptista, 1998).

Diet sterilization. All diets used in the study, dried cassava for rearing and the twelve maize varieties for the treatments, were heat-sterilized to prevent contamination (Pratiwi & Ananda, 2020). The diets were placed in trays (29 cm × 21 cm × 4.5 cm) and heated in an oven at 40 °C for four hours. After sterilization, the diets were removed and allowed to cool at room temperature for at least 24 hours prior to use in the experiment.

Maize grain analysis. To characterize the experimental diets, both physical and chemical analyses were conducted on twelve maize varieties. Physical characterization, specifically grain hardness, was measured at the Laboratory of Agricultural Food Technology, Faculty of Agricultural Technology, Universitas Gadjah Mada (UGM), Yogyakarta. Chemical profiling included the determination of total phenolic content and proximate composition. The phenolic analysis was performed at the Plant Pest Laboratory, Department of Plant Pests and Diseases, Universitas Brawijaya, Malang, while proximate analysis was carried out at the Chemistry Laboratory, Balai Pengujian dan Sertifikasi Mutu Barang (BPSMB), Malang, to determine moisture, ash, crude protein, lipid, and carbohydrate contents.

The results demonstrate that these maize varieties possess distinct physical and chemical profiles (Table 1). Grain hardness varied substantially, ranging from 379.95 N to 866.10 N. Regarding chemical composition, phenol contents fluctuated between 6.09 ppm and 9.46 ppm, and protein levels ranged from 4.59% to 8.02%. Furthermore, carbohydrate levels remained consistently high across all varieties (73.94%–79.82%), indicating that the selected varieties differ markedly in both their structural integrity and specific nutrient concentrations.

The host preference of A. fasciculatus was evaluated using the free choice test method (FCTM) within custom-made, transparent glass preference cages (± 3 mm thick) equipped with internal glass partitions to create separate chambers. The study employed two types of cages: one with twelve chambers (Figure 1A) and two separate units with six chambers each (Figure 1B). The experiments using the twelve-chamber and six-chamber configurations were conducted to verify the consistency of adult presence across different variety compositions. The twelve maize varieties were randomly assigned to these three test compositions: a single 12-variety test (all twelve varieties tested simultaneously), a first six-variety test (comprising Bisi 2, Bisma, Perkasa, Nusa 01, Sukmaraga, and Bima 20 URI), and a second six-variety test (comprising BTS 5, Lamuru, Nusa 03, Pertiwi 3, Bima 14 Batara, and Srikandi Kuning).

For each test, thirty grams (30 g) of the respective maize variety were placed into each chamber of the preference cage (Chijindu & Boateng, 2008). Infestation was initiated by introducing sixty pairs (120 individuals) of 7- to 14-day-old A. fasciculatus adults into the center of the cage (Chijindu, 2002). The cages were then sealed with a modified lid and covered with cloth. The number of adults present in each variety chamber was recorded daily for a total of seven days. The final count at 7 days after infestation (DAI) was recorded both as the total number of individuals and as a percentage to analyze trends across the different test compositions. On the seventh day, the adults were removed, and a final count was conducted, distinguishing between males and females for sex ratio determination. Following adult removal, the maize seeds were immediately inspected under a microscope to determine the number of eggs laid in each maize variety, thereby assessing the oviposition preference.

Total adult presence. The composition of maize varieties significantly influenced the total presence of A. fasciculatus adults across all preference tests conducted over the seven-day infestation period (12-variety test: F = 8.93, P < 0.001; first six varieties: F = 33.88, P < 0.001; second six varieties: F = 134.60, P < 0.001). In the 12-variety preference test, the Pertiwi 3 variety (21.86 individuals) was the most preferred by A. fasciculatus adults, followed by Bisi 2 (14.22 individuals) and Perkasa (13.54 individuals) (Table 2). When tested in combinations of six, the adults showed an even stronger preference for certain varieties. In the first six-variety test, adults preferred Perkasa (30.29 individuals) and Pertiwi 3 (49.79 individuals) significantly more than the other four varieties (Table 3 and Table 4). These combined results demonstrate that A. fasciculatus adults consistently favored Pertiwi 3 and Perkasa when presented with different combinations of maize varieties. The distribution pattern of the adults across the chambers was determined to be clumped in all three preference tests, as indicated by the index of dispersion (ID) values (12 varieties: ID = 2.47; first six varieties: ID = 2.12; second six varieties: ID = 9.62). Furthermore, a generalized linear model (GLM) analysis revealed a positive correlation between the moisture content of the maize and the total adult presence (Estimate = 1.475; P = 0.001) (Table 5).

Presence by sex and sex ratio. An analysis of variance confirmed that the maize variety composition significantly affected the presence of both males and females at 7 DAI across all tests. In the 12-variety test, both male and female A. fasciculatus preferred the Pertiwi 3 variety (males: 10.75 individuals; females: 10.50 individuals), followed by the Perkasa variety (males: 6.00 individuals; females: 7.25 individuals) (Table 2). Similarly, in the first six-variety test, both sexes showed a higher preference for Perkasa (males: 14.50; females: 13.75 individuals) and Pertiwi 3 (males: 24.00; females: 25.50 individuals) over the other varieties (Table 3 and Table 4). Regarding distribution, the index of dispersion showed a clumped distribution pattern for both males and females in the 12-variety and second six-variety tests. The only exception was the females in the first six-variety test, which showed a uniform distribution pattern (ID = 0.75). GLM analysis further revealed a positive correlation between moisture content and the presence of both males (Estimate = 0.577; P = 0.012) and females (Estimate = 0.507; P = 0.039) (Table 5).

The sex ratio of A. fasciculatus adults varied across the preference tests. In the preference test of the twelve varieties of maize, the sex ratio of A. fasciculatus ranged from 1:0.60 to 1:1.36 (Table 2). The preference test on the first six varieties of maize showed that the sex ratio of A. fasciculatus ranged from 1:0.95 to 1:1.31 (Table 3). In the preference test of the second six varieties of maize, the sex ratio ranged from 1:0.63 to 1:1.06 (Table 4).

The effect of maize variety composition on the number of eggs laid was significant in the 12-variety test (F = 7.98, P < 0.001) but not significant in the first sixvariety test (F = 1.03, P < 0.001). However, the effect was significant again in the second six-variety test (F = 15.06, P < 0.001). Consistent with the adult presence results, the Pertiwi 3 variety received the highest number of eggs in both the 12-variety test (22.75 eggs) and the second six-variety test (45.00 eggs) (Table 2 and Table 4). These findings confirm that females consistently chose Pertiwi 3 as the preferred variety for egg-laying across the tests where significant differences were detected. The distribution of eggs was clumped in the first six-variety test (ID = 2.66) and the second six-variety test (ID = 17.37). However, the 12-variety test showed a uniform distribution pattern for oviposition. Finally, GLM analysis identified a negative correlation between the phenol content of the maize grain and the number of eggs laid (Estimate = -2.908; P = 0.048) (Table 5).