A rapid assessment method to estimate the distribution of juvenile chinook salmon (Oncorhynchus tshawytscha) in an Interior Alaska river basin

Abstract

Identification and protection of water bodies used by anadromous species in Alaska are critical in light of increasing threats to fish populations, yet challenging given budgetary and logistical limitations. Non-invasive, rapid assessment sampling techniques may reduce costs and effort while increasing species detection efficiencies. I used an intrinsic potential (IP) habitat model to identify high quality Chinook Salmon Oncorhynchus tshawytscha rearing habitats and select sites to sample throughout the Chena River basin for juvenile occupancy using environmental DNA (eDNA) and distribution within tributaries using snorkel surveys. Water samples were collected from 75 tributary sites in 2014 and 2015. The presence of Chinook Salmon DNA in water samples was assessed using a quantitative polymerase chain reaction (qPCR) assay targeting that species. Snorkel surveys were conducted and physical habitat was measured for a subset of tributaries examined with the eDNA approach. Juvenile salmon were counted within 50 m reaches starting at the tributary confluence and continuing upstream until no juvenile salmon were observed. The IP model predicted over 900 stream km in the basin to support high quality (IP ≥ 0.75) rearing habitat. Occupancy estimation based on eDNA samples indicated that 80.2% (± 4.3 SE) of previously unsampled sites classified as high IP and 56.4% of previously unsampled sites classified as low IP were occupied. The probability of detection of Chinook Salmon DNA from three replicate water samples was high (0.76 ± 1.9 SE) but varied with drainage area. A power analysis indicated power to detect proportional changes in occupancy based on parameter values estimated from eDNA occupancy models. Results of snorkel surveys showed that the upper extent of juvenile Chinook Salmon within tributaries was from 200 to 1,350 m upstream of tributary confluences. Occurrence estimates based on eDNA and snorkel surveys generally agreed, but care should be taken to ensure that little temporal gap exists between samples as juvenile salmon use of tributary habitats is likely often intermittent. Overall, the combination of IP habitat modeling, occupancy estimation based on eDNA, and snorkel surveys provided a useful, rapid-assessment method to predict and subsequently quantify the distribution of juvenile salmon in previously unsampled tributary habitats. These methods will provide tools for managers to rapidly and efficiently map critical rearing habitats and prioritize sampling efforts to expand the known distribution of juvenile salmon in interior Alaska streams.