Natural selection, evolution, and demography of salmonine populations experiencing intrusion from non-local stock

Many salmonine taxa experience intrusion into their wild, free-living populations from non-local conspecific and heterospecific individuals. Such intrusion arises most commonly as a result of releases from captive breeding programmes for conservation, or to provide a demographic excess that can be exploited commercially or recreationally. Furthermore, the relocation of conspecific individuals from one population to another, the deliberate stocking of sexually compatible foreign taxa into areas outside of their natural range, and domesticated individuals escaping from fish farms provide additional pathways for intrusion. The relative fitness of non-local to local fish, as well as the effects intrusion has on wild populations, is highly dependent on the ecological context that both types of fish experience and the level of adaptation displayed by the introduced fish. In this thesis, I examine how natural selection and, thus, evolution affect the performance of free-living salmonine populations that have experienced intrusion from non-local (captive-bred, translocated, domesticated) stock. Exploring this interplay can help to identify what conditions allow for, and the extent to which, non-local fish successfully breed in a given wild setting, as well as to determine the effects their breeding has on the demography of the recipient wild population. A better understanding of the roles natural selection and evolution have on demography and population viability is crucial for designing better captive breeding programmes and mitigating against the negative effects sexually compatible foreign taxa and fish farm escapees can have when they spawn in the wild. In Chapter 2, I use a molecular pedigree to estimate the lifetime reproductive success of individual Atlantic salmon and demonstrate that captive-bred fish are 64% less fit than their wild-bred conspecifics when both spawn together in the wild. Furthermore, I found evidence of a transgenerational carry-over effect from the hatchery where the wild-spawned offspring of two captive-bred salmon experienced lower survival to adulthood than the offspring of two wild-bred fish. Finally, I used 43 years of population census data to determine that in years where the proportion of spawners that were captive-bred was larger, the productivity of the whole population was reduced. In Chapter 3, I used the molecular pedigree to explore the evolutionary dynamics of female body size at spawning in Atlantic salmon by applying the variance decomposition methods of quantitative genetics. Female (but not male) salmon experienced positive directional selection for the trait but displayed no phenotypic or evolutionary response despite the trait being genetically heritable and, via the Breeder’s Equation, being predicted to evolve towards larger sizes. By utilising a Bayesian regression technique for decomposing the selection gradient into genetic and environmental components, I determined that the female univariate selection gradient used to predict evolution was upwardly biased by one or more unmeasured, but genetically correlated, traits. This highlights the need to measure more than just the focal trait when examining the evolutionary trajectories of populations as management decisions for intruded populations based on biased predictions could result in unforeseen, negative consequences. Chapter 4 is a review of 91 genetically-explicit eco-evolutionary models in fisheries science, with a particular focus on the genetic architecture employed in the models. With relevance to my thesis, only 15 studies (16.5%) examined captive-wild interactions. 14 of these studies were parameterised for or motivated by the effects of captive-bred releases into wild salmonine populations. Five of these modelled trait inheritance through quantitative genetic architectures, nine used explicit Mendelian models of inheritance, and one study allowed for the independent inheritance of both a quantitative trait and a biallelic locus. Together, the results from Chapters 3 and 4 informed the design of my own genetically-explicit eco-evolutionary model in Chapter 5. In Chapter 5, the aforementioned genetically-explicit eco-evolutionary model was used to explore how soft and hard selection interacted to affect evolution and demography in a salmonine population that had experienced intrusion from non-local stock. Non-local alleles were purged faster from the population when soft selection was stronger. Soft selection also indirectly influenced the strength of hard selection. By limiting the number of maladapted individuals that could breed, soft selection reduced introgression from maladapted alleles into the wild population. This caused a reduction in the number of fish displaying maladaptive phenotypes that would be selected against by hard selection, thus, weakening the strength of hard selection. Furthermore, the weakening of hard selection buffered against the demographic declines associated with this form of natural selection. The results of this chapter demonstrate how unexpected evolutionary dynamics can emerge in populations and how variation in the ecological context (in this case, the form and strength of both soft and hard selection) affects evolution and demography. This thesis emphasises how variation in natural selection and the extent of intrusion/introgression impacts the evolutionary dynamics observed in salmonine populations that experience intrusion/introgression from non-local fish. Chapter 2 adds to the ever-growing list of studies describing reduced fitness for captive- relative to wild-bred fish while also providing a known example of how increasing numbers of captive-bred fish lead to a concomitant reduction in population productivity. The genetically-explicit eco-evolutionary model resulting from work done for Chapters 3, 4, and 5 stresses how the form and strength of natural selection can vary depending not just on prevailing ecological conditions but also on the genotypic/phenotypic composition of both the wild population and introduced individuals. A greater understanding of how evolutionary dynamics emerge and what conditions lead to a particular set of dynamics is critical in mitigating against the negative impacts arising from the deliberate or accidental release of non-local individuals into wild populations