Background: The use of environmental DNA analysis has revolutionized biodiversity monitoring. Initially, eDNA monitoring surveys in aquatic environments utilized a targeted approach, but there has been a steady shift toward whole community assessments (eDNA metabarcoding). Both approaches can increase the detection sensitivity for rare and elusive species, compared to more conventional methods. However, it is important to understand the benefits and limitations of targeted and whole community eDNA monitoring to tailor surveys to research questions and management objectives. Aims: Here, we aimed to test the relative merits of targeted eDNA analysis versus eDNA metabarcoding in an intermittent river system. Methods: First, samples collected during different seasons were used to assess the influence of seasonality on the detection probabilities of both methods. Second, detection probabilities from the two monitoring approaches for one focal species were compared to evaluate the sensitivity of both methods. Finally, the data from an eDNA metabarcoding survey conducted across the outer distribution limits of an invasive species were used to evaluate whether species interactions can be inferred by this method. Results: Analyses showed that sampling intermittent river systems during low flow events increases the performance of the targeted eDNA surveys, while sampling season does not influence the performance of eDNA metabarcoding surveys. Environmental DNA metabarcoding was found to be less sensitive than a targeted monitoring approach, thus making the latter more suitable for generating detailed distribution data. Nevertheless, eDNA metabarcoding survey data can be interpreted in a semiquantitative manner and can provide insights into biological interactions.