The harvest of animals from the wild is a pervasive selective force, especially in fisheries, where harvesting often targets individuals with specific traits. Early research on trait-based selection was particularly focussed on traits related to age and size at maturation, but there has been a more recent shift toward understanding how traits beyond growth, such as behavioural and physiological traits, may also be targets of selection. Indeed, there is now emerging evidence from commercial and recreational fisheries that individuals with physiological traits related to metabolism and swimming performance are more likely to be captured, as are fish with behavioural traits related to exploration, risk-taking, and group cohesion. However, the direction of selection has been found to be largely dependent on the type of gear used, with more active gears such as trawls targeting traits relating to swimming capacity, and more passive gears such as traps targeting traits related to increased risk-taking and exploratory behaviour. Many of the traits under selection are also influenced by environmental factors such as temperature and oxygen availability, but how environmental factors modulate capture vulnerability is currently not well understood. Another important gap in fisheries selection literature is that some fishery sectors have been completely overlooked including smaller-scale artisanal fisheries such as the ornamental trade, a global industry with important repercussions for sustainability, biodiversity, and the livelihoods of local communities. In this thesis, for the first time, I aimed to uncover if there was a potential for trait-based selection in the ornamental fishing industry, specifically in the Amazon, which is home to a large proportion of wild-caught ornamental species. Using a combination of lab-based studies and field observations, I investigated capture vulnerability and compared phenotypes of key ornamental species caught using different gears. I also examined the impact of environmental factors on behavioural and physiological traits to determine if the environment can modulate relationships between individual phenotype and capture vulnerability. Chapter 1 introduced the main research themes and highlighted the key knowledge gaps that were addressed in this thesis. In chapter 2, I used a scaled-down fisheries simulation that has been used in previous studies focussed on commercial and recreational fisheries. Unlike previous studies, which have used surrogate species, I used an actual target species of the trade, the cardinal tetra (Paracheirodon axelrodi), which is a highly popular ornamental species. I repeatedly caught individuals to determine their vulnerability to capture and tested whether vulnerability to capture can be predicted by phenotype. I revealed that capture vulnerability is predicted by a suite of traits, namely size, swimming endurance, exploration, and risk-taking. In chapter 3, I caught wild populations of cardinal tetra using active (net) and passive (trap) gear types and examined differences in behavioural and physiological traits in the lab. I used a range of assays including constant acceleration tests, maze tests, open field, and novel object tests. Trap-caught fish were larger and had a lower swimming performance compared to net-caught fish. In contrast, net-caught fish were more cohesive and active, but specific results varied across assays. Social and environmental factors also influenced findings, with differences in risk-taking revealed between individuals and groups and even small variations in ambient temperature during trials significantly influencing behaviour. In chapter 4, I used a completely field-based approach to investigate the ornamental capture process in-situ. Using underwater recording of traps in the Amazon, I observed behaviours for two species, the spotted tetra (Copella nattereri) and Hemigrammus Spp., including pre-capture behaviours such as passes and inspections, which have not been observed in the majority of fisheries selection studies. The observations revealed that the majority of fish that inspected traps did not enter them, and while the likelihood of capture was similar for both species, once a given trap had caught one species, it would not catch the other. There were also differences between species in the frequency and timing of behaviours, such as passes and inspections. There was also a relationship between environmental factors and behaviours, but these differed between species, highlighting the importance of integrating multiple species in studies. In chapter 5, I captured spotted tetra using active (net) and passive (trap) gears and examined behavioural traits of groups using open field and novel object tests after a short acclimation time. I also investigated the impact of an acute temperature increase on behavioural traits. I found that trap-caught fish were more exploratory and showed greater group cohesion, while net-caught fish were more active. Interestingly, there was limited evidence that exposure to an elevated temperature influenced any behavioural traits apart from cohesion. However, group identity accounted for a large amount of behavioural variation, as individual responses were masked at the group level, highlighting an important dynamic to consider when interpreting the potential for selection. Finally, in chapter 6, I summarised the key findings and overarching themes of my thesis. I discussed the potential for selection in the ornamental fishing industry and highlighted how my findings contribute to the existing body of literature. I then provided an overview of the challenges and limitations of my thesis and provided some future directions for the field. Taken together, my thesis provides an important foundation towards understanding the potential for selection in ornamental fisheries. Using lab simulations and field observations across different environmental contexts, this thesis shows that capture is not random, but trait based. This thesis also highlights the potential for gear-based selectivity in the ornamental fishing trade, which can have important repercussions for biodiversity and conservation. This thesis also raises questions about how selection operates in different environmental conditions or how group dynamics can mask individual responses. In summary, this thesis contributes a novel perspective on fisheries-induced selection by focusing on a previously overlooked system and broadens our understanding of how artisanal fisheries can shape wild populations.
Mar Deniz Yerli Pineda (Thu,) studied this question.