The movement of predator larvae were not significantly impeded by the surface of C. sativa leaves

Specifically, the behavior, including mobility of the larvae of C. carnea was unaffected by leaf surface . In addition, the larvae of A. aphidimyza spent significantly more time crawling on C. sativa leaves compared to sweet-pepper leaves. This might be explained by the different strategies and morphologies of the larvae of the two predators and also by the morphology of the C. sativa leaves that are covered with more trichomes compared to sweet pepper and thus, may annoy the apodous larvae of A. aphidimyza, that in turn, might move as a response to avoid those trichomes. To our knowledge, no previous studies have examined the locomotion of A. aphidimyza larvae and interaction with prey as affected by plant non-glandular trichomes. Madahi et al.,  evaluated the oviposition and life-history characteristics of A. aphidimyza on six plant species and with different types of leaf trichomes  infested with the A. gossypii. Although females oviposited on all host plants,marijuana grow system the lowest fecundity and longest larva developmental time were observed on tomato leaves that had a higher density of unicellular and multicellular glandular trichomes.

Thus, A. aphidimyza larvae seem to have difficulties in moving and attacking prey on the leaves covered with glandular trichomes. However, this is not the case in C. sativa because the leaves are covered mainly by non-glandular trichomes. Both predators were able to engage in interactions with the aphid A. gossypii on C. sativa and sweet-pepper leaf disks. For the larvae of C. carnea, no significant differences between plant species for any of the behavioral parameters examined were found, confirming the capacity of this predator to move efficiently on the C. sativa leaf surface. For instance, larvae mobility was not affected by non-glandular trichomes on cotton . On the other hand, non-glandular trichomes in Dutchman’s pipevine  negatively affected walking speed, search efficiency, and predation success of C. carnea larvae . Although walking speed was not measured in our study, C. carnea larvae moved and engaged in interactions with prey on C. sativa leaves dominated by non-glandular trichomes. We suggest that the morphology of non-glandular trichomes, e.g., height, may influence greatly the mobility of C. carnea larvae. For the larvae of A. aphidimyza, more encounters with prey were registered on sweet pepper. This can be due to the smoother surface  in combination with the morphology of the larvae that are in direct contact with the plant surface .

It is important to highlight the low attack and success rates registered for both predators. These results might be explained by the presence of larger aphid instars in the experimental arenas that defended themselves more efficiently when predatory larvae were nearby or immediately after the encounter . Ideally, we should have performed these experiments using a homogeneous mix of aphid instars. Similarly, the proportion of the larvae showing no interest might be due to the artificial experimental conditions. Finally, our results regarding predator interaction with prey on C. sativa and sweet-pepper were obtained with A. gossypii because the cannabis aphid does not infest sweet pepper. Given that both predator species are able to prey upon several aphid pest species,cannabis vertical farming it is highly likely that the predatory behavior would be similar against P. cannabis on C. sativa leaves, but not on inflorescences where the mobility of A. aphidimyza is severely reduced . Overall, the mobility of both predators was severely reduced on C. sativa inflorescences. In fact, the larvae of A. aphidimyza spent most of the time inactive. However, C. carnea larvae were more active than the A. aphidimyza larvae. Apparently, the capitate-stalked glandular trichomes that prevail in the inflorescences, pose serious difficulties to the movement of the apodous A. aphidimyza larvae that are in direct contact with the plant surface.

The legged C. carnea larvae, on the other hand, were able to surpass the height of the capitate-stalked glandular trichomes and moved more efficiently on the C. sativa inflorescences, although with more difficulty compared to the movement on the C. sativa leaves. Similarly to our results, Gassmann and Hare, , showed that the mobility and predation success of Chrysoperla spp. larvae was hampered by glandular trichomes of jimsonweed . All in all, our results suggest that the C. carnea larvae are better suited for biological control programs targeting the cannabis aphids on inflorescences compared to A. aphidimyza. It is important to consider that although P. cannabis can disperse on the whole plant, they are mostly located on leaves and stems  where the larvae of both predators can move and interact with their prey. A very low number of interactions with prey was registered on C. sativa inflorescences for the larvae of both predators. This was expected for A. aphidimyza, given that larvae spent most of their time inactive on the inflorescences.