Final Report Summary - PARA-PEST (The role of host personality and stress responses in parasite dynamics)
Personality - that is, consistent differences in behaviour between individuals - has been demonstrated in a wide range of animal taxa. Currently, both the mechanisms that cause this variation between individuals, and the effects of the variation in terms of evolutionary costs and benefits, are largely unknown. Thus, there is limited scientific evidence explaining how and why animal personality exists and is maintained in natural populations. This fellowship focused on the links between personality and stress physiology, one potential underlying mechanism that causes personality variation. The costs and benefits of having different personality traits were investigated using parasite infection. Parasitism and related disease impose important costs to animals, and previous research suggests that both host personality and physiology may affect host-parasite dynamics, and hence the growth and spread of parasite populations.
This fellowship therefore used a multi-disciplinary approach, drawing techniques from behavioural ecology, stress physiology and parasitology to determine relationships between stress hormone release and personality under a range of contexts, including during host development. It then explored how these factors affect an individual host’s susceptibility to infection and the likelihood that parasite infections will be transmitted between individuals.
A fish host-parasite model system was used: the guppy (Poecilia reticulata), a well-known fish from the aquarium trade, and one of its parasitic worms (the monogenean Gyrodactylus turnbulli). These parasites live on the outside of the fish, where they can reproduce rapidly, and transmit directly to conspecifics. This model system is amenable for lab studies, but can also be studied in the wild in Trinidadian rivers, allowing the importance of links between physiology, behaviour and parasitology to be placed in a natural context. In addition, there are many hundreds of gyrodactylid species, some of which are major fish parasites, causing significant economic losses, both within the EU and worldwide. This is true of both tropical fish species used in the aquarium trade, such as guppies, and species used to provide wild and cultured fish stocks for human consumption, such as salmonids infected with G. salaris, which has had a devastating effect on the salmon fishing industry. In addition, identifying links between animals’ physiological and behavioural responses to stressful events, and consequences for the spread of parasite infections has implications for captive animal welfare. In agriculture, aquaculture and captive breeding for conservation, animals are exposed to human-induced stressors on multiple levels and timescales, but at present we have little understanding of how this will shape future responses to stress, behaviour and disease dynamics.
Methodological techniques used
1. Personality traits can be identified by demonstrating that individuals behave in a repeatable manner when tested at different times or in different contexts. We assayed exploration and activity personality traits using multiple tests in a novel arena.
2. Stress hormones were collected and extracted from water in which fish had been isolated. This non-invasive technique is based on the fact that fish excrete hormones across their gills in proportion to their circulating levels. It utilizes an assay developed for human cortisol, that we validated for guppies. Cortisol samples were taken before and after fish had experienced stressful events, to quantify both baseline (unstressed) and elevated (stressed) hormone levels. Stressful events consisted of being chased by a net in a standardized manner, to mimic an unsuccessful predation attempt. They were imposed both as one off events to test short-term hormone and behavioural responses, and repeatedly to test long-term behavioural and physiological adaptation to stress.
3. Parasite susceptibility studies involved tracking the outcome of an experimental infection on individually housed fish. Parasite transmission studies involved infecting one fish in a pair or small group, and then following the increase in number of parasites and spread of infection between individuals.
Stress hormone release was a predictor of personality with, as expected, fish that released more cortisol being more cautious and less explorative. However, neither cortisol release nor personality were a predictor of which fish died or survived an experimental parasite infection, or the variables that described infection progression such as maximum number of parasites sustained during the infection. Instead, susceptibility to infection was determined by the size and body condition of the fish. This suggests that internal energy resources, rather than stress physiology, is the primary factor determining how infections progressed, and whether or not guppies could survive infection.
Despite this, it is possible that personality (and underlying stress physiology) can affect host-parasite dynamics via differential effects on transmission rates. A field study using guppies taken from the wild found no relationship between their personality and their naturally acquired gyrodactylid loads. However, in order to tease apart whether personality could affect infection via transmission, or indeed whether infection may change personalities, experimental manipulations were necessary. Two experiments found no evidence that infection caused changes in personality, at least at levels of infection seen naturally in the wild. (In the lab, infections rates are frequently seen that are not recorded in natural populations, and these very high parasite burdens are known to cause ‘sickness behaviours’.) These results suggest that gyrodactylids do not manipulate the behaviour of their guppy hosts. Rather, transmission is affected by innate host factors, such as personality, and external factors, such as stressful events in their environment that affect how and when individuals are in contact.
Fish responded to short-term stressful events by releasing elevated levels of stress hormones and by changing their behaviour. As expected when quantifying a personality trait, non-stressed fish exhibited repeatable behaviour across pairs of behavioural tests. However, experience of a short-term stressor prior to the second test disrupted this relationship, both by changing the mean rate of exploration and activity behaviours and by rendering behaviours no longer repeatable. In group tests, application of a behavioural stressor caused fish to school more, but there was no evidence that this increased transmission, perhaps due to the relatively short duration of the trials and/or the short-term effect of the stressor on schooling.
In a study investigating the effects of personality on transmission, fish personality did not differ overall between tests when they were alone versus those with a randomly chosen conspecific. However, the more exploratory fish within pairs did become less exploratory, and the less exploratory fish became more exploratory. Furthermore, both personality and body size within the pairs affected how long the fish spent in proximity to one another, and increased time in proximity decreased the time taken for an experimental infection to transmit from one member of a pair to the other. These data clearly show that personality traits in individuals, although stable, can be affected by the personalities of others within a group, and that personality and body size can, via their affects on group behaviour and hence how long individuals spend in proximity, affect parasite transmission rates.
Finally to investigate the long-term effects of stressful events during development on adult stress responses, personality and host-parasite dynamics, juvenile guppies were reared under non-stressful or stressful conditions. At adulthood, there were, surprisingly, no effects of long-term repeated stress on the amount of stress hormone released, personality or susceptibility to infection. Given that we have demonstrated the effects of short-term stress on hormone release and behaviour, and that rates of hormone release and behaviour did not differ from those in other experiments, this suggests that fish can adjust during development to the presence of a stressor, and can therefore downgrade their physiological and behavioural stress response to adapt to living in more stressful conditions. It remains to determined, however, at what levels of stress fish become unable to do this, thus imposing costs of stress that cannot be easily mitigated.
Fish personality co-varied with stress physiology. Although this variation did not affect susceptibility to parasite infection, it may have import implications for host-parasite dynamics via effects on transmission rates. Transmission was affected by the personality and body size of individuals within groups, because these factors affected how long individuals spent in proximity, allowing direct transmission to occur.
Fish responded to short-term stressful events by increasing their output of stress hormones and changing their behaviour, in ways that disrupted their personality traits. However, long-term exposure to stressful events did not cause a change in these measures, suggesting physiological and behavioural adaptation during development.
By integrating techniques from different scientific disciplines, a key component of the EU 7th Research Framework, we have therefore increased knowledge of both the mechanisms that determine personality and the effect that this host variation has on host-parasite dynamics, with consequent implications for our understanding of how disease spreads in both captive and wild animal populations.