Patterns and drivers of juvenile steelhead survival throughout a stream network

Background/Question/Methods When managing and conserving migratory populations, it is important to consider population parameters at each location or life history stage. For anadromous fish, the freshwater rearing phase can be important to population regulation and persistence. Both natural and anthropogenic factors can drive survival of rearing juvenile fish, thus impacting the timing of outmigration, life history expression and ultimately population fecundity in terms of the number and size of spawning adults. Over a four year period we quantified demographic and habitat variables of a population of federally threatened juvenile steelhead, Oncorhynchus mykiss, in an anthropogenically altered headwater system. During the base flow season when we expected anthropogenically reduced flows to reduce habitat quality and create population bottlenecks, we repeatedly sampled sites located throughout the stream network (4 sites the first two years and 14 sites the second two years, covering four tributaries). We used mark-recapture analyses and information theoretic model selection to quantify spatial and temporal patterns of apparent survival and suggest drivers of these patterns. Specifically, we considered hierarchical spatial groupings (i.e. site, stream, watershed), temporal trends over the base flow period, density, growth, elevation, water temperature and presence of water withdrawal (2 impacted streams vs. 2 reference streams) as potential correlates/drivers of survival. Additionally, we considered individual covariates of body length and condition factor.

Results/Conclusions We found that in some years survival varied spatially, whereas in others temporally, and that the locations or time periods of relatively high or low survival were not necessarily similar among years. Such variability highlights the dynamic nature of upper trophic level populations in stream ecosystems and the need to consider spatially and temporally detailed measures. Sometimes there was a clear driver of survival, as when high density reduced survival. At the highest densities individual covariates were especially useful at predicting survival, where larger fish more often perished. Density effects were likely exacerbated by low prey availability and high temperatures associated with base flow conditions, particularly for larger fish, which must consume more relative to smaller fish. In lower density years, temporal variability or spatial variability unrelated to density tended to prevail. In some cases it was still possible to suggest drivers of survival, such as differences in habitat across streams, by considering the detailed survival patterns that our models produced. Such information may be utilized to direct management activities and improve freshwater rearing conditions for this and other imperiled salmonid populations.