Unveiling invasive insect threats to plant biodiversity: Leveraging eDNA metabarcoding and saturated salt trap solutions for biosurveillance

difficult and costly. In the 2022 collection, there were 25 individual detections with 11 of those rat- ified by both molecular and morphological approaches; one detection was morphological only and the remaining 13 were molecular only. The case of morphological detection only was repre- sented by a single specimen, which might have been trapped towards the last moment of collec- tion and may have not shed enough genetic material into the trap fluid before it was removed for morphological identification. Of the 13 detections that were molecular detection only, six of them represented cases where the species was validated in the same collection site but from another trap in that same site. If matching traditional morphological identification to molecular detection is considered the gold standard for validating eDNA results, the current success rate combining both scenarios is 68% (17 out of 25). However, as discussed above, there are well-known scenarios where molecular and morphological detections would not allow for 100% agreement, emphasiz- ing the need to carefully evaluate molecular tools in regular biomonitoring and biosurveillance programs on a sample-by-sample basis. At the bioinformatic level, MetaWorks [19] resulted in more coincidences between molecular and morphological identifications than mBRAVE [20], although the number of IAS detected when using both bioinformatic pipelines was comparable. Altogether, it demonstrates the impact of the data analysis strategy on the final output and the

need for using more than one bioinformatic pipeline for maximizing detection probability. Trap height There were significant differences in IAS detectability, with ground-level traps detecting all five species in the 2020 collection, while canopy level-traps detected only one species (Lyman- tria dispar). On the contrary, most IAS detected in 2022 were found at the canopy level com- pared to the ground-level sampling. These differences can be partially explained through species behaviours. However, similarly to current findings, studies often find mixed results concerning trap height and should be considered on a species-by-species basis [53, 54]. Previ- ous behavioural studies show that Japanese beetles (Popillia japonica) are more attracted to traps near ground level [55, 56]; this is where males often fly in search of females [44]. Some studies show Emerald ash borer (Agrilus planipennis) (Coleoptera, Buprestidae) prefers traps placed in mid to high canopy in contrast to those placed near the ground [57, 58]; this corre- sponds with previous findings [59] that observed most adult flights of emerald ash borer occurring in the upper and mid-canopy region. Studies on other Buprestid and Cerambycid species have mixed results, with both taxa either preferring the canopy, ground, or with no preference depending on the species [53, 54, 60–62]. The specific flight behaviour of T. cam- pestris (Coleoptera: Cerambycidae) in relation to forest canopy height is currently unknown, although the standard trap placement used is 1.5 m [63]. Grape phylloxera (Daktulosphaira

vitifoliae) is commonly found near ground level since they emerge from the soil. In Ontario, grape phylloxera is suggested to prefer a trap height of 1.3 m in contrast to 0.6 m [64], while another study found no significant difference between traps at 1.5, 2.1, and 2.7 m [65]. Spongy moth (Lymantria dispar) males are strongly attracted to vertical objects and commonly fly up and down tree trunks, where females would most likely be encountered [66]. Various studies note greater success for ground-level traps generally below 4 m [67–72]. Collins and Potts (1932) found that traps placed at 21.3 m height did not capture any males [73] while Richerson et al. (1976a) showed that males primarily search for females at the height of the highest phero- mone source [74], which tends to be below 4 meters [75]. In summary, our study showed greater success for ground-level sampling at 2020 and at the canopy-level in the 2022 collec- tion, which partially corresponds to specific insect behaviour in each case. However, other studies of similar taxa show mixed results, therefore we suggest sampling at both the ground and canopy in future studies for better species representation until more data is gathered. PLOS ONE Unveiling invasive insect threats to plant biodiversity PLOS ONE | https://doi.org/10.1371/journal.pone.0290036 August 11, 2023 19 / 24