Date of Award


Level of Access Assigned by Author

Campus-Only Dissertation

Degree Name

Doctor of Philosophy (PhD)




Michael T. Kinnsion

Second Committee Member

Julian J. Dodson

Third Committee Member

John F. Kocik


In this dissertation I consider how Atlantic salmon {Salmo salar L.) restoration may influence, or be influenced by, phenotypic trait variation in juvenile salmon stocked into the wild First, I assessed individual growth, survival and movement histories in a single stream in Maine undergoing restoration stocking of juvenile salmon. I studied six life stages over 18 months for four cohorts of presmolt Atlantic salmon. My findings show that DDE on growth can be detected at multiple stages well past young-of-the-year in juvenile salmon. I also noted an apparent tendency for mean fish sizes to become more variable in low-density reaches. Shifting scale of analysis from individuals to group means spanning larger spatial scales had pronounced effects on inferences of density dependence. Second, I showed that normal hatchery practices significantly advance the development of endangered Atlantic salmon fry. I in-turn investigated patterns of ontogenetic selection acting on hatchery produced salmon fry by experimentally manipulating fry development stage at stocking. I found evidence for strong stabilizing selection on the ontogeny of unfed hatchery fry (coefficients = -0.051 and -0.091), with weaker evidence for positive directional selection on fed fry. These selection patterns support a seasonally-independent tradeoff between selection favoring advanced development and physiological constraints of starvation. I heuristically show how such selection on ontogeny may impede restoration by limiting fry productivity and reducing effective population sizes (by as much as 81%). Finally, I quantified geographic variation in selection on fry size during the six-week period following stocking in the wild. I found substantial evidence for size-selection in five out of six selection trials. Interestingly, selection varied extensively among sites (from -0.25 to 0.89 standard selection gradients). Model selection suggested that this variation in phenotypic selection is associated with the presence of large woody debris and the slope of the stream channel. The strong selection coefficients I observed should be a concern for endangered salmon restoration, whether they reflect natural processes and an opportunity to maintain adaptation, or the potentially deleterious phenotypic consequences of hatchery practices.