Variability in upstream extent of fish distribution in headwater streams in southwest Washington, USA

ABSTRACT Objective Forest practices are regulated in Washington State, USA, with different prescriptions for harvesting timber adjacent to fish-bearing (type-F) and non-fish-bearing (type-N) streams. Prior to implementing forest operations, managers must determine the upstream extent of fish use in affected stream networks. Increased understanding of spatial and temporal variability in fish distributions supports the need for more accurate determinations of accessible and suitable fish habitat when demarcating type-F and type-N stream segments. Further, elucidating the environmental characteristics associated with the locations of fish in a stream network supports broader understanding of aquatic ecological communities and the factors influencing spatial changes in freshwater biodiversity. We aimed to evaluate variability in the location and taxa of the uppermost detected fish (UDF) and to characterize associated stream habitat attributes at those locations across years and seasons in southwest Washington. Methods Between 2015 and 2020, we resurveyed for the location of the UDF and measured associated habitat metrics up to five times on 201 streams. Based on the initial survey, streams were classified as “terminal” (where the UDF occurs along a single stream channel or immediately below the confluence of multiple non-fish stream segments) or “lateral” (a non-fish stream that laterally intersects a fish-bearing main stem). Change in UDF location was assessed relative to the initial survey. We summarized the distance of change by resurvey period and used descriptive statistics and a generalized linear mixed model to examine factors associated with upstream change in UDF location. Results Among 94 terminal sites, we observed a change in the location of the UDF relative to the initial survey during any resurvey at 84 (89%) sites, and the distance of change ranged from 292.9 m downstream to 253.0 m upstream. However, 95% of the sites had a maximum distance of change no greater than 152.4 m (500 ft) over all resurveys. Among 107 lateral sites, 11% contained fish during any resurvey. Results from the generalized linear mixed model indicated that the probability of an upstream change in UDF location was negatively associated with a smaller bankfull width and steeper channel gradient. Conclusions We observed that variability in the UDF location typically occurred over short distances and that the frequency of upstream change in UDF location was higher in streams where the UDF point identified during the initial survey was classified as terminal than in streams where it was classified as lateral. Additionally, an upstream change in the UDF location was associated with local stream habitat characteristics and was less likely in smaller, steeper streams than in wider, low-gradient streams. Advancing the understanding of this variability can help resource managers and regulators to better identify habitat likely to be used by fish based on the results of a single fish survey. This information also expands our understanding of how malleable this ecologically important location is within headwater stream networks and the potential impacts to biodiversity assemblages in these systems.