There is little information on A. gemmatalis larval movement on soybean plants. In the current study we assessed for the first time the movement of this species on soybean under field conditions and observed that in all the on-plant movement scenarios, larvae had the ability to move constantly over time. For 3rd instar larvae/V7 plant stage, we observed more constancy in the distribution on the plant over the evaluation. However, for 5th instar larvae/V7,R1 plant stage, for example, even in the first evaluations we did not detect larval presence in some plant sections and leaf positions. Larval size and plant growth stage affects the behavior and acceptance of a suitable host, and in turn movement patterns, stimulating larvae gradually to move off the plant. This highly active behavior was already observed for other lepidopteran species. Measuring movement of Ostrinia nubilalis Hübner (Lepidoptera: Crambidae), 50% of the larvae had moved off the plant within 6 hours36.
Larval behavior at a specific instar can contribute to understanding the spatial distribution of a species on a specific host plant. For example, in maize it was reported that small Spodoptera frugiperda Smith (Lepidoptera: Noctuidae) larvae presented an aggregate behavior, while medium and large S. frugiperda larvae exhibited a random pattern37. The causes of this behavior are not always clear and many of these details can influence IPM and IRM strategies. Even though we did not assess feeding behavior, it is known that timing of movement relative to feeding and the age of the larvae are important38. In general, it is relevant to know whether movement consists of discrete steps with each interval of feeding followed by movement or whether movement is mostly one long, continuous step from one to another section in the same plant or the plant of origin to the final host with no feeding between points. We suggest as a future research to characterize the neonate’s dispersion on the plant, once the larvae will start to move and find a suitable feeding site23.
Successful dispersal will depend on the host plant and its architecture, egg load, consequent first instar establishment, and costs of migration; these costs will change with larval age and may have impact on later instars and adult development39. Also, larvae may be affected by environmental conditions. A noticeable characteristic during the evaluation was that we observed higher larval movement to the middle and lower plant section and to the abaxial leaf position during the warm hours of the day. Similar behavior was also reported for Striacosta albicosta (Smith) (Lepidoptera: Noctuidae) in dry beans40. At 14 days after hatch (DAH), most 4th instars larvae were found on the soil surface. No larvae were found on the plant or soil surface after 14 DAH. Observations by time of day clearly showed that the majority of 5th instars spent the day on or in the ground, moving up the plant in the dusk and returning to the ground by sunrise. Even in larger crops, such as corn, when S. frugiperda infestation was done above the ear zone, larvae moved downward to the ear zone over time16.
This on-plant information also helps to understand the dispersion dynamic throughout the day and can improve insecticide spraying efficiency. Based in our observed larval movement, results reinforce the theory to apply insecticides and release entomopathogenic enemies early in the morning or at the dusk, when the insects will be exposed to the products. Cannibalism and predation are also causes that could impact population dynamics, and this behavior was already reported for this species and when interacting with other species41. Another important mortality factor in the agroecosystem is the presence of natural enemies, which was visually detected in our study and might also explain the number of missing larvae. The larvae position and high movement on and between plants can also in turn expose the insect to predators and parasitoids26.
Regarding the movement among plants, no significant, but notable differences in the dispersion were observed. Our results showed that long distances were covered by A. gemmatalis regardless of the position and orientation chosen, indicating the absence of a homogenous movement. This type of movement is a ubiquitous life‐history trait of organisms, with profound consequences on the spatial and temporal dynamics of populations 42. Thus, movement of this species appears to be nondirectional and independent28; however, especially under field conditions, it could be influenced by biotic or abiotic factors. In a similar study with Spodoptera eridania Stoll (Lepidoptera: Noctuidae) larvae, it was also observed that the species move in all directions independent of the season or cultivar14. Pannuti et al. 2016 assessing S. frugiperda and S. albicosta movement observed a dispersal governed by nondirectional sensory information of both species; however, S. frugiperda remained nearer the release point, presented a more aggregated and symmetrical distribution.
The mean distances found here were approximately 40.00 cm in the infested row, and up to 50.00 cm for the other orientations. The mean of maximum distance varied between 38.00 to 142.52 cm, reaching 202.62 cm to the NE. Similarly, in a greenhouse and in open field trials, larvae of Diatraea saccharalis Fabricius (Lepidoptera: Crambidae) were observed to move from initially infested plants to at least four plants away, with the majority of larvae staying within three plants away from the release plant34. Spodoptera frugiperda larvae exhibited greater movement to the North within the infested row. However, no significant differences were observed when considering general N and S orientations, and this northern movement may have been influenced by wind or another environmental factor12.
Despite that in our studies we characterized the movement of A. gemmatalis in non-Bt soybean, we can use this behavior information to apply to a Bt field. Previous research supported the inference that resistance evolves faster at high larval movement rates because of the increased likelihood of larvae moving from non-Bt plants to Bt plants and the increased survival rate of heterozygous larvae relative to homozygous susceptible larvae43. Thus, in practical terms, the distance traveled by the insects may affect the evolution of resistance to Bt crops, in association with configuration and size of the refuge areas18,44. Initial high-dose/refuge strategies were initially designed to delay resistance to Bt crops. Among them, the use of seed mixtures (refuge in the bag ‘RIB’) has been suggested to manage resistance to Bt toxins43. Although RIB can be an option in IRM, insects that exhibit high larval movement across plants generally favor the evolution of resistance in a seed mixture refuge, since such movement exposes insects to sublethal doses of the toxins, especially for single-gene Bt events35,44.
Seed mixtures can increase survival of heterozygous larvae relative to homozygous susceptible larvae when individual larvae feed on both Bt and non-Bt plants45. Consequently, when there is a high larval movement rate, it is expected that plant mixtures increase the effective genetic dominance of resistance to Bt crops, resulting in increased selection for resistance46. There are studies with other pest species and crops that demonstrate RIB is not always appropriate. In a field study conducted with seed mixes of non-Bt and Bt pyramid maize, D. saccharalis larvae were able to move from infested plants, as well as to adjacent rows, so larvae could feed on non-Bt maize until they reached a size that allowed their survival on Bt maize34. This strategy may not be ideal for resistance management of A. gemmatalis to the Cry1Ac toxin, and a more appropriate strategy, such the use of strip or block refuge, may be required. However, possible movement across the border between Bt and non-Bt crops must first be examined, as it may also have an important role in resistance evolution44.
In our open field study, we recovered a lower number of larvae with high variability among plots; however, this is representative on most real conditions. We found a higher frequency of larvae recovered in the different quadrants compared to the infested or across rows, similar to observed for S. albicosta larvae, where about 32% of were recovered within the infested row and 75.3% within a radius of 1.7 m of the infested plant12. Mortality in the early larval stages is commonly high in Lepidoptera 24. The authors reported that almost 50.00% mortality occurs by first instar parasitism. Fanela et al. (2020) observed a higher mean number of recovered larvae at each distance interval for S. eridania. However, the study was conducted in cages, with a more controlled environmental. For instance, the crop environment will strongly influence how predation and desiccation, or the impact of rainfall, will affect survival during larval movement from plant-to-plant or to and from hiding/feeding places38. In a similar study, larval recovery from central non-Bt plants in seed blends was 27.50% less than that from the pure non-Bt plantings47. However, even with larvae moving less in non-Bt plants, a detailed study of H. armigera feeding behavior in Bt and non-Bt treated diet choice bioassays demonstrated larvae could not detect Bt and only avoided it post ingestion48. Recovery and aversion post Bt ingestion is less likely as protein toxicity increases, such as with toxins that approach a high dose. This approach is important since this behavior can impact resistance evolution.
Overall, our results serve as baseline for the support of A. gemmatalis probabilistic models and to improve their predictive ability. The present results corroborate other studies, where the findings suggest that each species must be considered independently, and one should not expect a one-size-fits-all IRM plan to be suitable for all pest species49. Additional A. gemmatalis biological and ecological studies should be conducted in association with fitness costs aiming to elucidate the relationship between larval movement and the evolution of insect resistance. In temperate areas, such as in much of the United States, selected populations of A. gemmatalis would not overwinter, but the selection pressure can be high in the southern United States and in tropical areas such as Brazil. As new transgenic events and/or pyramids become available worldwide, the importance of understanding A. gemmatalis and other target pest species’ behaviors will increase.