Long-term movements of self-marked caddisfly larvae (Trichoptera: Sericostomatidae) in a California coastal mountain stream

1. Over 140 000 larvae of the case‐building caddisfly Gumaga nigricula were self‐marked as they incorporated glitter into small portions of their cases while reared in streamside troughs. These marked individuals were released into stream pools and their movements monitored in the dry season, when base flow was low and no spates occurred, and in the wet season when base flow was high and several spates occurred. 2. Of the 9,000–10 000 larvae released in each of two stream pools in the dry season, 4–20% (i.e. 377–1817 marked individuals) were observed on three sampling dates (4, 11 and 24 d after release). Most larvae (87–93%) remained within 4 m up‐ or downstream of the release line after 24 d. No larvae were found outside of the release pools, even after 37 d. 3. Of the > 120 000 larvae released in one stream pool near the beginning of the wet season, 408 larvae were recaptured 130–167 d later, a period that included 30 days of high flow associated with six spates. Estimated survivorship over this period was 0.7–6.2%; there was no relationship between survival and larval size at release. Most (75%) recaptured larvae were found in the pool where they were originally released. The remaining larvae were found downstream of the release pool. Larvae had generally dispersed only a short distance downstream of the release pool (median = 18 m, maximum = 222 m). In addition, four marked pupae were later found 436 m downstream of the release pool. 4. These results illustrate the sedentary nature of larval G. nigricula as well as the important role that high flow events play in larval mortality and dispersal. These case‐building larvae move very little during low flow periods, even when food resources appear limiting. In contrast, the frequency and distance of larval dispersal are much greater during periods with high flow. 5. Our observations for G. nigricula support previously published inferences that larval dispersal within a stream can be limited for some aquatic insects. However, our observations also suggest that, even for a relatively sedentary species like G. nigricula, larval dispersal during periods with high flow may contribute significantly to gene flow within a stream reach.