1. Stream ecologists have been puzzled by the apparent paradox that invertebrate populations persist in headwater streams despite the high frequency with which individuals drift downstream. To resolve this ‘drift paradox’, directions and distances of both larval and adult movement must be identified. Using over 50 interception traps in combination with results from several mark–capture experiments using 15N as a label, we tested the assumption that interception traps accurately represent the ultimate direction of adult insect flight.
2. In several streams in the Hubbard Brook Experimental Forest, 76% of 15N-labelled stoneflies (Leuctra ferruginea) had flown upstream from where they emerged to where they were captured. In contrast, over 60% of stoneflies were flying downstream when captured, i.e. on the upstream side of an interception trap.
3. The instantaneous direction, as indicated by the side of the interception trap on which they were captured, indicated the ultimate flight direction for fewer than 1/3 of the individuals captured. Thus, such traps did not accurately reflect the ultimate flight patterns of individuals, as indicated by mark–capture data.
4. Conclusions drawn from interception trap counts regarding the direction of movement and the distribution and persistence of populations may need to be re-evaluated. We suggest that better tracking methods, including mass mark–capture studies using stable isotopes, be used to evaluate the potentially complex patterns of adult insect movement and the consequences of that movement for individuals and populations.