Abstract
Understanding insect behaviour and its underlying drivers is vital for interpreting changes in local biodiversity and predicting future trends. Conventional insect traps are typically limited to assess the composition of local insect communities over longer time periods and provide only limited insights into the effects of abiotic factors, such as light on species activity. Achieving finer temporal resolution is labour-intensive or only possible under laboratory conditions. Here, we demonstrate that time-controlled insect sampling using an automated Malaise trap in combination with metabarcoding allows for the observation and documentation of taxon-specific activity patterns. Furthermore, these recorded activity patterns can provide valuable insights into the underlying ecological processes. Insect activity curves, derived from predicted detection numbers using generalised linear latent variable models, reveal distinct differences in activity patterns at higher and lower taxonomic level. While our findings align with existing literature, they also reveal that the activity patterns of some species are more complex than previously known. Additionally, a comparison of the assessed activity patterns across taxa suggest potential, previously undescribed parasitoid-host relationships. Within taxonomic groups, we observe variations in both the timing and duration of activity patterns, which can be linked to differences in mating strategies among closely related species. By capturing circadian rhythms of insect activity through time-controlled bulk sampling, we can expand our knowledge on species behaviour, ecology and temporal interactions. This contributes significantly to the advancement of chronoecology, allowing for further exploration of the roles of species and benefits in natural and anthropogenic ecosystems, alongside their potentially significant threat.