ARCHIVED CONTENT: In December 2020, the CZO program was succeeded by the Critical Zone Collaborative Network (CZ Net) ×

Kelson, 2018


Evolutionary Ecology of Partial Migration: A Case Study from a Pacific Salmonid Fish, Oncorhynchus mykiss

Kelson, S.J. (2018)
University of California, Berkeley  


Intraspecific diversity, or trait differences among individuals of the same species, is important for ecological interactions. Through my dissertation research, I explore the linkages between genotypes, phenotypes, and ecology in nature. I highlight partially migratory populations as systems that are ripe for understanding these linkages. Partially migratory populations are ones that are comprised of migratory and resident individuals – a strong form of intraspecific variation. There is growing appreciation that many migratory populations are in fact comprised of both migratory and resident individuals, including culturally and economically valuable populations of ungulates and salmonid fishes.

I focused on a salmonid fish native to the Pacific Rim, Oncorhynchus mykiss. In this system, both migratory and resident forms breed and rear in freshwater. The migratory form (“steelhead trout”) then out-migrates to the ocean to feed and grow, returning to freshwater to breed. In contrast, the resident form (“rainbow trout”) completes its entire life cycle in freshwater. Recent research by Pearse et al. (2014, Proc. Roy. Soc. B) revealed that the genetic basis of migration in O. mykiss is linked to a narrow region of the genome, which opened the door to understanding genotype-phenotype-ecology linkages. My dissertation explores these connections in partially migratory populations of O. mykiss in two streams, Fox and Elder Creek, which are tributaries to the South Fork Eel River in coastal Northern California.

In my first chapter, I explored the spatial distribution of migration-linked genetic variation in these streams. I predicted that small natural barriers would limit the upstream distribution of migratory genotypes, and that the effect of these barriers would be greater in dry years when there is less opportunity for adult steelhead passage. In Elder Creek, the largest barrier is a waterfall located 2 km from the mouth of the stream, and is passable under a narrow range of stream flows. In Fox Creek, the largest barrier is at the mouth. I captured fish from pools distributed longitudinally from the mouth to the upper extent of fish in each stream. I conducted RAD-capture on over 3,000 individuals and then classified these individuals as migratory, heterozygous, or resident genotypes using over 400 single nucleotide polymorphisms (SNPs) located on the migration-linked region of the genome. The partial barrier in Elder Creek reduced the frequency of migratory genotypes (migratory allele frequency of 0.60 below the barrier vs. 0.31 above the barrier). In Fox Creek, the proportion of migratory alleles varied greatly among years, ranging from 0.30-0.68. Years when migratory allele frequency was low were also years when there was not a storm event in February, which is peak breeding season for O. mykiss in the South Fork Eel River, suggesting that the frequency of migratory genotypes is tied to adult steelhead access in this creek. Overall, I found that there was spatial variation in migration-linked genotypes, with migratory genotype-fish being more common below the waterfall in Elder Creek, and rare in some years in Fox Creek. Furthermore, inter-annual variation was associated with water year type (dry or wet) and the timing of rainfall events.

In my second chapter, I determined the correlation between life history genotypes and phenotypes at the individual level. I installed stationary antennas at the mouth of Fox and Elder creeks to detect individuals who expressed migration, and assigned individuals to the resident phenotype using a size threshold. I found that resident-phenotype fish were dominated by resident genotypes (55% resident, 39% heterozygous, and 6% migratory genotypes), but migratory-phenotype fish were comprised of a mix of genotypes (25% resident, 45% heterozygous, and 30% migratory genotypes). Females are more likely to express migration in salmonid systems, given that larger females are more fecund. Therefore, I predicted that including information on sex would improve our ability to explain phenotypic variation. Genetic sex typing confirmed that females were more likely to express migration: migratory-phenotype fish were 62% female while resident-phenotype fish were 79% male. This is the first study to use life history genotypes and sex to predict individual phenotypes in partially migratory O. mykiss.

In my third chapter, I explored the connections between genotype and aspects of ecology, including population ecology (density and size structure of O. mykiss) and community ecology (food chain length and trophic cascades). During fish sampling, I estimated the density and size structure of fish in study pools. I found that stream reaches dominated by migratory genotypes were characterized by double the density of juvenile fish as compared to resident-dominated reaches (0.46 vs 0.26 individuals/m2), presumably reflecting the higher fecundity of migratory females, but half as many older fish (0.05 vs. 0.13 individuals/m2). Furthermore, differences in size structure were linked to differences in trophic structure; stable isotope analyses revealed that larger, old fish, were feeding higher on the food web (6.1± 0.62 ‰ δ15N vs for age-0 fish and 7.8 ± 0.83 ‰ δ15N for older fish). Overall, pools within the migratory-dominated region were characterized by many young fish (simple size structure) and a shorter food chain than pools sampled in regions dominated by resident genotypes.

Finally, I explored how inter-annual variation in precipitation influenced two key aspects of O. mykiss ecology: downstream migration timing and over-summer growth. My research occurred during the multi-year drought in California and included two dry years (2014 and 2015) and two wet years (2016 and 2017). Despite large differences in overall rainfall magnitudes, out-migration timing and over-summer growth differed little among years, which highlights the value of shaded, groundwater fed streams with a water-storing lithology for the conservation of salmonid fishes in warming river systems.

Overall, my dissertation provides empirical support for linkages among genotypes, phenotypes, and ecology, while also highlighting partially migratory O. mykiss populations as a model system for investigating the ecological consequences of intraspecific variation. Partially migratory populations may be common systems where heritable intraspecific variation is associated with ecological change. Migration is on the decline globally, and it is important to understand the ecological consequences of shifting ratios of migratory to resident individuals.


Kelson, S.J. (2018): Evolutionary Ecology of Partial Migration: A Case Study from a Pacific Salmonid Fish, Oncorhynchus mykiss. University of California, Berkeley.

This Paper/Book acknowledges NSF CZO grant support.

Associated Files

(3 MB pdf)