Date of Award

Spring 2021

Project Type


Program or Major

Biological Sciences

Degree Name

Master of Science

First Advisor

Nathan Furey

Second Advisor

Nathan Furey

Third Advisor

John Kocik


Atlantic salmon Salmo salar are an endangered migratory fish with recovery projects in their last remaining habitats in Maine focused on juvenile stocking and habitat restoration efforts. Population productivity of Atlantic salmon has been linked to size of juveniles, suggesting that conditions that facilitate growth in stream habitats are important to recovery. Broadly, the objectives of my thesis were to examine how different habitat and biological variables affect the size of juvenile parr throughout Maine and to assess long-term trends (1969 – 2017) among four drainages (East Machias, Narraguagus, Sheepscot, and Penobscot). Extensive electrofishing surveys captured ~250,000 individual Atlantic salmon across the four drainages within 61 years (1956 – 2017). I fit generalized additive mixed models to determine relationships between parr size (fork length) and environmental and biological variables as well as through time. Site- and drainage-specific variables evaluated included temperature (mean summer air temperature in Chapter 2, and mean summer stream temperature in Chapter 3), elevation, mean channel width, juvenile salmon density, age-0 parr stocked, estimated fry stocked (a metric to combine egg and fry stocking efforts into one metric), and year. The averaged models for each of the drainages allowed for comparisons within and across drainages. The largest increase in parr size throughout the time period was ~5mm, with smolting occurring early at between 1 and 4 years of age. Consistently among the drainages, sites with low-elevation, wide channels, and warmer temperatures were associated with longer parr. The Sheepscot drainage demonstrated the most positive trend of size over time with a ~5mm increase in FL in both Chapters. Density dependence was evident in every model that included a density related variable. Stronger FL relationships were found with air temperature rather than stream temperature, which implied that temporal variations in temperature were more important to parr size than the difference between stream and air. Although my models found important linkages between parr size and environmental and biological factors, the percent deviance explained was only moderate (12% - 40% among all) and the inclusion of further variables (such as habitat complexity, distance from dams, and stream discharge) would improve further work. These studies conclude that rather than increasing stocking intensities, managers should prioritize stream restoration to improve habitat quality and access to better benefit parr growth.