- I examined the relative roles of biotic and abiotic factors in structuring redband trout
Oncorhynchus mykiss distributions in the South Fork John Day. I first examined the
relationship between the biological traits of the fish assemblage and riparian-geomorphic features in context of prevailing theories of stream ecology stemming from the river continuum concept (Chapter 2). I found that fish traits and habitat characteristics were related to one another in a different manner according to the spatial scale of observation (watershed, tributary, and reach scales) and that mode of species coexistence (niche overlap, niche partitioning, and non-associative) was correlated more strongly with the distribution of traits than with environmental features, suggesting that biotic interactions play an important role in structuring the fish assemblage. Next, I described the 'behaviorscape' of redband trout foraging and aggression in context of prevailing foraging theory (Chapter 3, Part I). I found that redband trout behaviors could be described in terms of expected energetic gains and losses and potentially predation risk, with 'risky' behaviors yielding higher expected energetic gains in habitats with abundant structural refugia. These findings raised the questions of how fish growth and foraging behavior would be affected by increased food resources, questions which I addressed via a supplemental feeding experiment in two streams (Chapter 3, Part II). I observed nearly an order of magnitude increase in instantaneous growth rates of redband trout O. mykiss and a potential competitor, juvenile Chinook salmon O. tshawytscha, upon feeding, demonstrating that food is a limiting factor for salmonid growth in this system. Also upon feeding, redband trout foraging and aggression increased in a stream with no heterospecifics, lower discharge, and colder water temperature, but not in a stream with the opposite characteristics. These findings suggested that O. mykiss and O. tshawytscha
may coexist in a state of interactive segregation, contrary to the common wisdom that the two species are selectively segregated and have therefore evolved fully-distinct habitat preferences. To acquire further evidence for this hypothesis and to elucidate the role of competitor densities on habitat selection at local scales (which would in turn affect fish distribution patterns at larger scales), I employed a multi-species habitat selection model
(Chapter 4). Redband trout habitat selection of enriched patches was negatively affected by conspecific densities as expected, but positively affected by juvenile Chinook densities, a finding that I attribute to habitat effects or heterospecific attraction. Redband trout foraging and aggression decreased in the presence of both conspecifics and heterospecifics, and aggression appeared to decrease with supplemental feeding. These
contrary results between the former and latter experiments testing for aggression as a response to feeding are explained via the manner in which food was delivered to the stream, pulsed at dawn and dusk versus delivered evenly over time, respectively. In the concluding chapter (Chapter 5), I make the case that restoration practices acknowledging a broad range of potential drivers for species distributions—including behaviorally-mediated biotic interactions among the fish assemblage—will better serve a broader range of restoration goals.