4.1 Influence of sampling time and position on community composition 
Our results indicated that differences in community composition were caused by seasonality but not sampling position, which matched our expectations despite positional compositional differences occurring in other studies examining fish and macroinvertebrates (e.g. Berger et al., 2020; Macher & Leese, 2017; Thalinger et al., 2021). High similarity among sampling positions likely occurred due to the complete mixing of eDNA in the water body, with smaller compositional differences occurring by replication, the three water samples taken within the sampling site, and not by systematic differences between positions (Macher et al., 2021). The stream sampled in our study had a water depth <1 m and sampling positions were less then 10 m apart with a mostly turbulent flow, all of which can reduce community heterogeneity (see Shogren et al., 2017; Fremier et al., 2019 for further discussions). As we sampled only one location in a single temperate stream, it is not possible to derive general predictions from our results regarding eDNA detection probability of macroinvertebrates between different sampling positions in other streams and regions. As a next step, it would be of interest to systematically compare sampling position differences among divergent stream types.
Temporal beta-diversity shifted in a seasonally cyclical and gradual pattern, which was primarily driven by turnover, with turnover and nestedness both returning to the same level after one year. The potential of eDNA in detecting seasonal shifts in community composition has been already shown for marine environments (Jensen et al., 2022; Liu et al., 2022) but not for freshwater.