Using environmental DNA to assess the response of steelhead/Rainbow Trout and Coastrange Sculpin populations to postfire debris flows in coastal streams of Big Sur, California
Objective Debris flows are among the most extreme disturbances to streams and are predicted to become more frequent under climate change. We assessed the response of steelhead Oncorhynchus mykiss (anadromous Rainbow Trout)/Rainbow Trout (hereafter, collectively referred to as O. mykiss ) and Coastrange Sculpin Cottus aleuticus populations to major postfire debris flows in two small coastal basins of California using noninvasive environmental DNA (eDNA) sampling. Methods We analyzed water samples from disturbed reaches for eDNA in the first two summers after debris flows. In Big Creek, all Coastrange Sculpin habitat was disturbed by debris flows, but a major tributary occupied by O. mykiss was not affected. In Mill Creek, all available habitat for both species was disturbed. We also sampled two unburned basins as undisturbed reference streams. Result In Big Creek, O. mykiss eDNA was detected in all water samples during both years, and concentrations in some samples approached the lowest concentrations in the reference streams. Coastrange Sculpin eDNA was detected only in a single water sample in the first year and was not detected in the second year. In Mill Creek, neither species was detected in the first year, but during the second year, O. mykiss eDNA was detected at very low concentrations in 40% of samples and Coastrange Sculpin eDNA was detected in a single sample. Conclusion Our results indicate that O. mykiss and Coastrange Sculpins either survived in or quickly reoccupied the disturbed reaches from within the basins. However, the detection rates for cases in which all habitat for a species in a basin was affected by debris flows indicate that abundances were very low, suggesting that persistence may be uncertain unless there is successful reproduction or immigration. Finally, eDNA appears to be effective for monitoring sensitive fish populations after a major disturbance; however, detection rates in individual samples may be low and require appropriate sampling designs to achieve the desired detection probability.