Discussion
The relative brain volume, i.e. after controlling for fish FL, was larger in lake habitats compared to streams. In addition, volumes of telencephalon and optic tectum were, in relation to other brain regions, disproportionately larger in lakes compared to streams, while volumes of other brain regions changed in close correlation to the overall brain volume. The telencephalon of stream fish was even smaller than telencephalon of hatchery individuals. Similarly to the study of Ahmed et al. (2017) on wild populations of stickleback, our findings do not support the hypothesis that in brown trout volume of brain and regions important for navigation and decision-making increases in physically complex stream compared to simpler lake and hatchery habitat.
We posit that the discrepancy between theoretical predictions and our findings can be explained by two main mutually non-exclusive factors. First is the non-linear association between physical habitat complexity and selection for larger brains, particularly for larger telencephala and optic tecta (Boogert et al. , 2018). This explanation assumes that habitat complexity beyond a certain threshold may favor simple behavioural strategies to operate effectively in those complex environments, because their success probability is comparable to more complicated and cognitively demanding behaviours requiring costly investment in brain development (Morand-Ferron et al. 2015). Therefore, the high complexity of the rapidly changing stream habitat may favour simpler behavioural strategies than the less complex lake habitat. Stationary behaviour and a sit-and-wait foraging strategy of stream dwelling trout (Jonsson & Jonsson 2011) is an example of such simple behavioural adaptations to the complex stream habitat. In contrast, alternation in foraging between pelagic and littoral zone and more common piscivory of lake-dwelling brown trout (Sánchez-Hernández 2020) may induce selection for large telencephala and optic tecta due to the need for visual cues processing and relatively complex navigation and decision-making skills that they require (Edmunds et al. , 2016).
The second possible explanation for differences in brain size between the habitats is differences in available diet quality, which limit the nutrient supply for brain development. Lake-dwelling trout feeding on zooplankton or other fish acquire more n-3 LC-PUFA than stream-dwelling trout, which rely on a mix of macro-zoobenthos and n-3 LC-PUFA-poor terrestrial insect (Heissenberger et al. , 2010; Sánchez-Hernández & Cobo 2016). Previous laboratory studies that have shown that availability of dietary n-3 LC-PUFA has positive effect on overall brain size (Lund et al. , 2012) and on size of optic tecta (Ishizakiet al. , 2001). Therefore, the higher dietary intake of these nutrients may facilitate brain size development in lake-dwelling brown trout, compared to their conspecifics from stream habitats. An extremely n-3 LC-PUFA-rich diet is also typical for hatchery-reared fish (Heissenberger et al. , 2010). This high-quality diet may loosen the selection trade-off between the cost of brain development and benefits of high cognitive capacity that typically shapes brain morphology in wild animals (Morand-Ferron et al. 2015; Boogert et al. , 2018). Thus, a high-quality diet, which enables rapid brain growth in hatchery fish, can explain the findings of this and other studies (Näslund et al. , 2012; Kotrschal et al., 2012) that hatchery salmonids can have similar or larger brain and telencephalon volume than stream-dwelling salmonids, despite the low physical complexity of hatchery habitat. The relative brain size (i.e. , encephalization) of hatchery trout was in our study difficult to compare with the wild fish due to the difference of allometric relationship between the FL and brain in hatchery and wild individuals, but it appears that hatchery fish brain was of intermediate size between fish of stream and lake origin. The lack of positive correlation between the brain size and FL in hatchery individuals is unusual for wild fishes where body size is a strong predictor of brain volume (e.g. , Triki et al. 2021). Previous studies on evolution of encephalization in mammals have suggested that relative brain size depends on selection pressure on size of the brain as well as on the overall body size (Smaers et al. 2012). Therefore, high-quality diet and selection for fast body growth in hatchery fish could uncouple the developmental link between the brain and the rest of the body (Kotrschal et al.,2012; Smaers et al. 2012).
Other factors that have been shown to influence brain size and morphology in fishes and were not explicitly considered in our comparative study are predation pressure (Kotrschal et al. , 2017), sex (Kolm et al. , 2009; Näslud 2018), and ontogeny (Abrahao et al. 2021). Potential predators, that is eel and large trout, were present in all sampled wild populations (see method section), and thus were unlikely to explain differences between the lake and stream habitat. Proportion of males and females in brown trout populations in generally even (e.g. , Baglinière et al.1989) and thus should not differ between the groups compared in this study, but some studies suggest that female brown trout are more common in pelagic lake habitat (e.g. , Jonsson 1989). There were clearly ontogenetic differences between the compared groups, as hatchery trout were young-of-year, while the sizes of all wild trout corresponded to adults. In addition, lake trout were larger than stream trout, and thus some of lake individuals could have been older than the rest of the sample, but these differences could also stem from differences in growth rates in lake and stream trout (Jonsson & Jonsson 2011; Sánchez-Hernández 2020). The differences in growth rates among the environments compared in this study makes it impossible to collect individuals of the same body size and age, and thus confounding of these two factors is an inherent shortcoming of any such comparative study.
In conclusion, our study provides an example of among population variability of brain size and morphology, which is a topic still widely understudied in the wild. We suggest that besides the cognitive demands of the environment (e.g., habitat complexity) future studies should also consider the availability of dietary essential fatty acids as a possible key driver of brain evolution and development in wild fishes.