Cognitive and behavioural plasticity in the intertidal Cocos Frillgoby (Bathygobius cocosensis)

The age-old dichotomy of nature versus nurture continues to spark debate in the context of plastic behaviour, and not just when it comes to humans. Every year there is further evidence illustrating how individual experience shapes personality, behaviour and cognitive ability, all of which are fundamental aspects of an individual’s phenotype. Ultimately, experience is intricately linked to an individual’s environment, and numerous studies have focused on population differences in behaviour. Key explanatory variables include variation in predation, competition and habitat stability/physical complexity. Despite their evolutionary history being embedded in ancient lineages, and the associated popularist idea that they are ‘simple and primitive,’ fishes demonstrate highly flexible behaviour. Moreover, they have become one of the leading vertebrate taxa in behavioural plasticity studies using comparative approaches, not least because of the wide range of aquatic habitats 20they occupy. In marine coastal environments, the intertidal zone is a highly dynamic habitat and home to one of the largest and most successful groups of fishes (>2000 spp), the family Gobiidae. This family is well-adapted to the intertidal zone morphologically and behaviourally, both aspects of which differ widely in species occupying different micro-habitats. Their sheer diversity offers exciting opportunities for comparative studies which attempt to untangle the relative influence of genes versus experience in shaping behaviour. Despite the diversity of intertidal gobies, niche overlap is common and competition for resources plays a vital role in behaviours such as foraging, suggesting that behavioural plasticity would be beneficial when securing resources. Furthermore, food resources in the intertidal zone vary on temporal and spatial scales, so that individuals with flexible behaviour can adjust to these changes and thus reap the fitness benefits. Chapter 2 of this thesis focuses on the niche overlap of two sympatric goby species, one rockpool specialist and one sand specialist, and how this overlap changes seasonally and ontogenetically. In addition, the trophic niche of a third allopatric species was investigated, to determine how diet changes seasonally and ontogenetically in the absence of interspecific competition. In the sympatric species, we found a seasonal switch in diet complexity, where the sand species consumed a variety of prey taxa in winter but not summer, and vice versa for the rockpool specialist. In contrast, the allopatric species showed similar diet complexity across seasons but shifted toward a specialised diet later in ontogeny. Seasonal change has been tied to variation in cognition, whereby changes in cognitive function are linked to reproductive demands and can differ dramatically between sexes, depending on life-history strategies. Although this has been demonstrated in several mammalian species, few have investigated sexually dimorphic cognitive ability in fishes, and none in the context of reproductive strategy. Rockpool gobies demonstrate exceptional cognitive function by way of spatial learning ability so Chapter 3 focused on the male nest-guarding mating system in the intertidal goby Bathygobius cocosensis and how it influenced male and female cognitive abilities in a spatial learning task in each season. Males and females performed similarly in all seasons except spring, which marks the breeding season in this species. Males showed a substantial decrease in cognitive ability while females did not. I suggest that the decreased cognitive ability observed in males during the breeding season is linked to their reproductive strategy; males are site-attached whilst they guard their nests and forgo foraging excursions. This study highlights the importance of cognitive plasticity and how individuals manage the trade-off between costs and benefits associated with enhanced cognition over relatively short temporal scales. The ecological cognition hypothesis suggests that an individual’s brain and behaviour are greatly influenced by environmental characteristics such as stability and predictability. However, whether the plasticity of these aspects is finite or otherwise constrained by inherited genetic mechanisms shaped by evolutionary pressures over multiple generations, remains unexplored in gobies. To that end, Chapters 4, 5 and 6 examined environmental drivers of spatial learning ability, anti-predator behaviour and laterality in gobies, using a wild-captive comparative framework. Previous studies have shown that rockpool gobies possess superior spatial learning abilities compared to sand specialist gobies, associated with the selective environmental pressures of living in a structurally complex habitat. However, whether this ability is fixed and innate, or flexible and shaped by experience, remains unknown. In Chapter 4, I reared juvenile gobies in 4 different habitats that varied in the degree of physical complexity and trained them to solve a simple spatial learning task. I found that gobies reared in structurally complex habitats solved the task faster than those reared in the simpler regimes, suggesting that, although spatial learning ability may have an innate component, life experience shaped by environmental heterogeneity continues to alter learning ability in later ontogenetic stages. In addition to controlled manipulation of the physical environment, captivity also allows researchers to regulate the social environment, such as predation pressure. Wild gobies were captured as adults with experience in assessing visual and olfactory cues to ascertain predation risk and altering their behaviour accordingly, while captive gobies were captured as juveniles and reared in the absence of predation risk. In Chapter 5, wild and captive-reared gobies were exposed to a series of cues from a sympatric predatory crab species and their anti-predator behaviours observed. In addition, I paid close attention to correlations between behaviours which may indicate population-level behavioural syndromes which, in other taxa, are most often manifested in high-predation contexts. Captive-reared gobies showed little differences in behaviour, regardless of cue treatment, although larger individuals generally spent less time moving in the presence of the predator. In the wild population, large individuals spent less time moving than smaller individuals, and gobies exposed to olfactory cues were less active than those in visual and control treatments. The relationship between activity and size emphasises the importance of body size in risk-related behaviour and the influence of captive- rearing on animal behaviour more broadly. As predicted, behavioural syndromes were only observed in wild fish when exposed to olfactory cues (olfactory cues alone or in combination with visual cues) emanating from predators, which aligns well with the existing literature. Previous studies have shown that behavioural differences in gobies from contrasting environments are mirrored in brain morphology, exemplified by larger telencephala in rockpool specialists compared to sand specialists. The question remains whether laterality, the preferred use of one brain hemisphere over another when assessing information, is similarly influenced by habitat complexity. Existing literature suggests that laterality is influenced by both habitat complexity and predation pressure. Chapter 6 focused on whether gobies exhibit population-level laterality and if differences exist between captive and wild populations. Trials were conducted using a mirror test, where the body position of gobies was observed, and eye use preference recorded. I found no evidence of population-level laterality in either group, although there was a tendency toward stronger lateralisation with increasing size, indicating laterality remains plastic throughout ontogeny. Moreover, this data supports the existing literature which suggests that population-level laterality occurs most often in highly social species. In summary, the research outlined in this thesis emphasises the plasticity of behaviour in a species that occupies a complex and dynamic habitat, and how the extent of this plasticity can be altered with controlled manipulation of environment in early ontogeny. It also highlights the strengths of a comparative framework, particularly when captive experience is improved with environmental enrichment to encourage natural behaviours in fishes.