Phenotypic expressions of Serranus scriba at individual level: behavioural traits, oxygen consumption and morphology

Abstract

[eng] The pace-of-life syndrome (POLS) theory proposes that several animal features are correlated each other in order to maximize the cumulated fitness of a given individual across its entire life span. It is expected that the animals from a given population can largely differ each other in several ways but, for example, fast growing animals would display accelerated metabolism and reach sexual maturity earlier in life. The original POLS considered only conventional life history traits such a growth rate or age-at-maturity, but other traits related with behaviour, physiology and morphology has been recently considered too. Concerning behaviour, a twofold pattern is expected. At one hand, behaviour should be partially plastic (i.e., a behavioural response could change depending on the balance between external resource availability and internal energy demand). On the other hand, the POLS theory predicts that behaviour should also have a consistent and repeatable component, in the sense that the same individual should tend to respond similarly at different moments and at different environmental contexts. For example, according POLS, a bold animal should be successfully discriminated form a shy animal, independently that they were living at different environments. In this work we describe the extend of between individual variation in (1) three behavioural variables (Activity, Exposure and Feeding), (2) one physiological variable (oxygen consumption at resting) and (3) body shape, using a coastal marine fish (Serranus scriba) as study case. According POLS, the working hypothesis under testing is that all those variables should be linearly correlated. For example, fish taking more risk would move more actively, would remain more time exposed, would respond quickly to a feed item, would display an accelerated metabolism (e.g., it would show an increased oxygen consumption) and would display morphological differences related with, for example, a quicker escaping response to a potential predator, which could imply larger muscular mass around the tail. The experimental settings for assessing behaviour was based in measuring (through video analysis) the three behavioural metrics described above from fish keep in tanks with four levels of environmental complexity and during 10 days. After completing these behavioural experiments, oxygen consumption of the same fish was measured continuously over night with a submerged respirometer. Finally, body shape was quantified using geometric morphometry from standardized fish images. Between-individual variability in all the five variables considered was noticeable. For example, up to five-fold between-individual differences in oxygen consumption is reported here, which is remarkable because such large individual variability has been rarely described for fishes. As expected, behaviour response depended on the environmental complexity of the tank (for example, fish in average tend to move more actively at more complex tanks) but it was more relevant in the context of the current Thesis that each fish has been successfully scored in terms of behaviour. These corrected individual-specific scores are independent of the complexity of the tank and of any other potentially confounding effect (e.g., fish size), thus allowing proper between-fish comparisons. Finally, concerning body shape, the anatomical regions experiencing more between-individual variability are those located on the tail peduncle, the anal fin and near the opercula. Nevertheless, almost all the correlational pattern between the five variables considered were non-significant, thus the results reported here no or weakly support POLS predictions. Concerning behaviour, Exposure and Activity are strongly (but negatively) correlated and Feeding was independent from the two former behavioural variables. Variability in oxygen consumption was not explained by none of the behavioural variables, but it should be noted that the effects of Exposure was marginally significant (Prob. = 0.065), thus this specific relationship deserves further attention. Finally, body shape is not explained by neither the three behavioural variables nor by oxygen consumption. A plausible explanation for such a failure may be that cause-effect relationship between the variables considered may be largely non-linear. Assuming linear relationships could be an oversimplification and more complex, mechanistic models relating the variables considered (specifically, between behaviour and energy balance) should be explored in the future. Specifically, the recent coupling between a bioenergetics model and a movement model strongly suggests that between-individual differences in exploration movement, body length and reproductive output will be sharp at low-resource environments but smooth or even inexistent when resource is at saturating levels. The results presented here seems to point in the same direction: The conceptual framework of the POLS hypothesis is a good starting point but sound hypothesis could only be properly tested after clarifying the specific mechanistic links between behaviour and metabolism.