Brown marmorated stink bug (Halyomorpha halys) feeding damage determines early drop in olive crops

The brown marmorated stink bug (BMSB), Halyomorpha halys Stål, is an invasive species and a polyphagous pest. BMSB feeding activity was suspected to be responsible for olive damage. To evaluate the effect of feeding damage from adults and nymphs of BMSB, 30 rearing sleeves were positioned in an olive grove, at an early stage of drupe development. The individuals were kept in the rearing sleeves for 48 h, and the number of olives in each sleeve was checked weekly, visually assessing signs of damage and measuring their volume. After the 48‐hr exposure, the number of early dropped olives was significantly higher for rearing sleeves containing BMSB adults and nymphs compared with control, with visible signs of damage. The volume of olives still attached was significantly lower for rearing sleeves with adults. These results provide key evidence on BMSB damage in developing olives. If the numbers of BMSB keep increasing in Mediterranean regions (where most of the olive production occurs), actions should be taken to prevent economic losses.

the severity of the damage with the time and if it varied between adults and nymphs.

| Experimental assessment of olive damage
In an IPM olive grove of the 'Casaliva' cultivar located near Cavedine, Italy (46.0142677 N, 10.9622535 E), rearing sleeves (BugDorm 40 × 20 cm, nylon netting 104 × 94 mesh) were used to evaluate BMSB feeding ( Figure S1, Supplementary Material). We considered three treatments: (1) Exposure of olives to adults for 48 h; (2) to nymphs for 48 h and (3) as a control, rearing sleeves with no individuals. Thirty rearing sleeves were distributed on ten trees, 3 per plant, placed on singles branches randomly selected (with at least 10 fruits), to have a sleeve per treatment on each tree (10 per treatment).
Brown marmorated stink bug individuals were collected at the beginning of July 2021 using attractive traps placed along a hedgerow at the Fondazione Edmund Mach campus, San Michele, Italy (46.1862381 N, 11.1339682 E). They were reared on green beans (Phaseolus vulgaris L.), tomato (Solanum lycopersicum L.), carrots (Dacus carota L.) and peanuts (Arachis hypogaea L.), in plastic cages (BugDorm-4M3030). They were kept in controlled conditions (25 ± 3°C and 60% ± 10% RH) with 16:8 h L:D photoperiod, before they were placed in the rearing sleeves.
To distinguish olive phenological growth stages, the international

| Data analysis
Data analysis and plotting were performed with R (R Core Team, 2018). Kruskal-Wallis test was used to compare the number and the volume of olives for each treatment at each session, followed by Dunn's post hoc test with Bonferroni correction.
To model the observed olive early drop, a beta-binomial generalized mixed model (GLMM) with a logit link was applied (Brooks et al., 2017). Rearing sleeve ID was used as a random factor, while session number and treatment were used as explanatory variables.
Model assumptions were verified following Zuur and Ieno (2016) and with package DHARMa (Hartig, 2021). Akaike information criterion (AIC) was used for model selection. Predictor effect plots (Fox & Weisberg, 2018) were used to visualize the fitted coefficients.

| RE SULTS
The percentage of dropped olives in the rearing sleeves at the last session was on average (±SD) 66.83% ± 30.85% for adults, 56.06 ± 26.97 for nymphs and 4.19 ± 8.58 for the control. The  1, Figure 2). The olives exposed to BMSB adults and nymphs presented external and internal feeding injury, with evident punctures and stylet sheaths (Figure 3). At 80 BBCH, the volume of the olives still attached to the trees was significantly lower (Kruskal-Wallis: H = 12.82, df = 2, p < 0.005) for the rearing sleeves that contained adult BMSB compared to those exposed to nymphs and control, while there was not significant difference between nymphs and control ( Figure 4).

| DISCUSS ION
As the list of crops attacked by BMSB still increases, new symptoms appear related to its feeding activity. Since several factors can be responsible for fruit drop (e.g., physiology, biotic and abiotic factors), this study offers key evidence of the potential role of BMSB as olive pest. Our work aimed at establishing that olives are subjected to BMSB feeding damage, which is expressed in terms of both early olive drop and reduced volume, as previously suggested (D'Ascenzo et al., 2020). Even if confining insects represents an artificial condition, it is an effective approach to correctly identify and characterize feeding damage under field conditions, a useful information to detect BMSB damage to olives in time.
To our knowledge, this is the first study that has experimentally documented symptoms associated with BMSB feeding damage on olive, when the fruits are at an intermediate stage of development.