Parallel and nonparallel behavioural evolution in response to parasitism and predation in Trinidadian guppies

Natural enemies such as predators and parasites are known to shape intraspecific variability of behaviour and personality in natural populations, yet several key questions remain: (i) What is the relative importance of predation vs. parasitism in shaping intraspecific variation of behaviour across generations? (ii) What are the contributions of genetic and plastic effects to this behavioural divergence? (iii) And to what extent are responses to predation and parasitism repeatable across independent evolutionary lineages? We addressed these questions using Trinidadian guppies (Poecilia reticulata) (i) varying in their exposure to dangerous fish predators and Gyrodactylus ectoparasites for (ii) both wild‐caught F0 and laboratory‐reared F2 individuals and coming from (iii) multiple independent evolutionary lineages (i.e. independent drainages). Several key findings emerged. First, a population's history of predation and parasitism influenced behavioural profiles, but to different extent depending on the behaviour considered (activity, shoaling or boldness). Second, we had evidence for some genetic effects of predation regime on behaviour, with differences in activity of F2 laboratory‐reared individuals, but not for parasitism, which had only plastic effects on the boldness of wild‐caught F0 individuals. Third, the two lineages showed a mixture of parallel and nonparallel responses to predation/parasitism, with parallel responses being stronger for predation than for parasitism and for activity and boldness than for shoaling. These findings suggest that different sets of behaviours provide different pay‐offs in alternative predation/parasitism environments and that parasitism has more transient effects in shaping intraspecific variation of behaviour than does predation.     

Interpopulation Differences in Shoaling Behaviour in Guppies (Poecilia reticulata): Roles of Social Environment and Population Origin

In Trinidad, guppies (Poecilia reticulata) in high‐predation localities show more cohesive shoaling behaviour than those living with less dangerous predators in low‐predation sites. We evaluated the relative contributions of population origin (i.e. genetic and/or maternal effects) and social environment on the expression of shoaling by assessing the behaviour of juveniles reared in a range of social conditions. Focal individuals, offspring of guppies from populations from high‐ or low‐predation localities, were reared in a multifactorial experiment; we created four different social conditions by manipulating the source and demography of the conspecific residents with whom focal individuals interacted. We found that high‐predation fish displayed a stronger propensity to shoal than low‐predation ones. Our results also suggest a role for interactions between the source of the focal individuals, the demography of the group in which they were reared and the origin of the guppies with whom they were reared. Depending on their origin (high‐ vs. low‐ predation) and rearing density, our focal fish were more likely to shoal if they were reared with high‐predation residents. Learning from high‐predation residents, aggressive interactions with low‐predation residents and/or phenotype matching could have played a role in driving this effect of social environment. This effect of the phenotype of conspecifics on shoaling development would enhance heritable differences in shoaling propensity such that both could contribute to the well‐documented difference in shoaling behaviour of high‐ and low‐predation guppies in natural populations.     

The genetic and environmental basis of adaptive differences in shoaling behaviour among populations of Trinidadian guppies,Poecilia reticulata

The degree of plasticity an individual expresses when moving into a new environment is likely to influence the probability of colonization and potential for subsequent evolution. Yet few empirical examples exist where the ancestral and derived conditions suggest a role for plasticity in adaptive genetic divergence of populations. Here we explore the genetic and plastic components of shoaling behaviour in two pairs of populations of Poecilia reticulata (Trinidadian guppies). We contrast shoaling behaviour of guppies derived from high‐ and low‐predation populations from two separate drainages by measuring the shoaling response of second generation laboratory‐reared individuals in the presence and absence of predator induced alarm pheromones. We find persistent differences in mean shoaling cohesion that suggest a genetic basis; when measured under the same conditions high‐predation guppies form more cohesive shoals than low‐predation guppies. Both high and low‐predation guppies also exhibit plasticity in the response to alarm pheromones, by forming tighter, more cohesive shoals. These patterns suggest a conserved capacity for adaptive behavioural plasticity when moving between variable predation communities that are consistent with models of genetic accommodation.     

Additive and nonadditive genetic variation in avian personality traits

Individuals of all vertebrate species differ consistently in their reactions to mildly stressful challenges. These typical reactions, described as personalities or coping strategies, have a clear genetic basis, but the structure of their inheritance in natural populations is almost unknown. We carried out a quantitative genetic analysis of two personality traits (exploration and boldness) and the combination of these two traits (early exploratory behaviour). This study was carried out on the lines resulting from a two-directional artificial selection experiment on early exploratory behaviour (EEB) of great tits (Parus major) originating from a wild population. In analyses using the original lines, reciprocal F(1) and reciprocal first backcross generations, additive, dominance, maternal effects ands sex-dependent expression of exploration, boldness and EEB were estimated. Both additive and dominant genetic effects were important determinants of phenotypic variation in exploratory behaviour and boldness. However, no sex-dependent expression was observed in either of these personality traits. These results are discussed with respect to the maintenance of genetic variation in personality traits, and the expected genetic structure of other behavioural and life history traits in general.     

Isolation and the development of shoaling in two populations of the guppy

Socially reared guppies Poecilia reticulata derived from two wild populations (Upper and Lower Aripo River, Trinidad) showed a significant relationship between body size and shoaling tendency, measured as the proportion of time spent in proximity to a bottle containing conspecifics. Larger females shoaled significantly more than smaller females. Fish from the high-shoaling population (Lower Aripo) showed significantly less shoaling behaviour when reared in isolation. In contrast, low-shoaling fish (Upper Aripo) demonstrated no significant change in their shoaling behaviour in response to social isolation.     

Polyculture production of juvenile fishes for survival in nature

Anecdotal and empirical evidence exists for substantial (up to 40%) declines in growth among Oreochromis populations domesticated in both large and small‐scale fish farms in Africa. These declines are at least partly attributable to poor genetic management, including inadvertent selection, inbreeding, bottle‐necks and founder effects. Due to restricted cash flow and investment capital, genetic management and selective breeding for the improvement of domesticate populations are difficult for small‐scale farmers, but feasible on larger‐scale farms. In managing domesticated gene pools, feral populations can serve as a broodstock reservoir, making the use of indigenous species advantageous. A development model of large‐scale hatcheries producing selected lines of sex‐reversed, indigenous tilapia for sale to smaller‐scale farmers is proposed as a solution to the problems of poor genetic management in African aquaculture.     

Fish cognition: what can we learn, and what do they need to learn?

Fish provide a wonderful opportunity to explore processes that shape and select cognitive ability. In this presentation, I will illustrate three aspects of work that my colleagues and I have used to investigate fish learning and memory over the last decade. First, I will discuss how comparing different populations sampled from contrasting habitats allows differences in cognitive ability to be related to the evolutionary ecology of the fish. I will use examples that have investigated how differences in learning ability between populations of the same species can arise. Here, the examples will be taken from the ubiquitous three‐spined stickleback, and a Panamanian poecilid, Brachyraphis .
The second approach has used fish cognition as a tool to quantify behaviour to enable assessment of different aspects of fish welfare. For example, the recent work investigating pain perception in trout required the use of a learning task to quantify how fish behaviour was modified after noxious stimulation. Ways in which these, and similar, processes can be used in future studies of fish welfare will be discussed.
The final part of the presentation will consider recent work that addresses the problems of releasing hatchery‐reared fish for restocking purposes. Although a common practice, most of the hatchery‐reared fish die shortly after they are released. Much of the observed mortality apparently stems from the fishes' inexperience with a variable environment. Experiments with juvenile cod and brown trout suggest that both age, and the early rearing environment, have profound effects on fish learning and behaviour. I will discuss how simple modifications to current rearing practices may have large beneficial effects on the post‐release survival of hatchery‐reared fish.     

Implications of multiple mating for offspring relatedness and shoaling behaviour in juvenile guppies

Polyandry (female multiple mating) can confer important benefits to females, but few studies have considered its potential costs. One such cost may arise through differences in the relatedness of offspring born to females with different mating histories; offspring born to monandrous females are always full siblings, while those produced by polyandrous females may be full or half siblings. These differences may have important consequences for social interactions among offspring. We used artificial insemination in the guppy (Poecilia reticulata), a promiscuous live-bearing fish, to evaluate shoaling behaviour in polyandrous and monandrous broods. We combined this information with known parentage data for the polyandrous broods to determine whether sibling relatedness influenced offspring shoaling behaviour. While we detected no effect of mating treatment (polyandry/monandry) on shoaling behaviour, we found that pairs of full siblings spent significantly more time shoaling (and in close proximity) than pairs of half siblings. This latter finding confirms the ability of newborn guppies to distinguish brood mates on the basis of kinship, but also suggests an important and hitherto unrealized potential cost of polyandry: a reduction in within-brood relatedness with potentially important implications for offspring social behaviour.     

Boldness and Reproductive Fitness Correlates in the Eastern Mosquitofish, Gambusia holbrooki

Animals are known to exhibit 'personality'; that is, individual differences in behaviour that are consistent across time and/or situations. One axis of personality of particular importance for behavioural ecology is boldness, which can be defined as the tendency of an individual to take risks. The relationship between individual personality and fitness correlates, particularly those involved in reproduction, remains largely unexplored. In the current study, we used female Eastern mosquitofish ( Gambusia holbrooki ) to ascertain whether certain reproductive traits (e.g. stage of pregnancy, fecundity) are correlated with individual personality in two wild populations in New South Wales, Australia. To quantitatively assess this relationship, we tested individual fish for their level of boldness, as measured by their latency to exit a refuge and tendency to shoal in a novel environment. We also quantified individual differences in general activity and tendency to swim near the water surface and substratum. For both populations taken together, bolder individuals tended to be smaller, relatively less fecund (when taking body size into account), and spent more time near the water surface than near the substratum compared with timid individuals. Individual boldness was not correlated with either general activity or stage of pregnancy. To our knowledge, our study characterizes for the first time a relationship between an individual personality trait (boldness) and a reproductive fitness correlate (fecundity) in fishes.     

Behavioural variation in Eurasian perch populations with respect to relative catchability

Animal personalities, i.e. consistent individual differences in behaviour, are currently of high interest among behavioural and evolutionary biologists. The topic has received increasing attention also in fisheries science because selective harvesting of certain behavioural types might impose fishing-induced selection on personality. Here, we ice-fished wild Eurasian perch (Perca fluviatilis) from three native populations and investigated whether differences in relative catchability would explain behavioural differences observed in experimental conditions. We inferred relative catchability differences indirectly by fishing each location first with generally inefficient artificial bait and then by more efficient natural bait. The captured, individually tagged fish were tested in groups for their exploration tendency, activity and boldness under authentic predation risk in semi-natural stream channels. Fish that were easily captured first with artificial bait expressed fast exploration and acute activity, whereas the fish captured with natural bait showed less active and explorative behaviour. Differences in relative catchability did not explain variation in boldness or body size. In conclusion, we found that (1) Eurasian perch differing in relative catchability differ in certain behavioural traits, (2) fast explorers are more common among fish that get easily caught compared to fish that are more difficult to catch, (3) relative catchability explains more behavioural variation in a novel environment than in a familiar one and (4) selectivity of recreational angling on fish behaviour may depend on applied angling method and the effort spent on each location.     

The effect of sex and shoal size on shoaling behaviour in Danio rerio

Male and female zebra fish Danio rerio were given choices of shoals that differed in sex and size. Male zebra fish preferred to associate with female shoals over male shoals, but had no preference when given a choice between a mixed‐sex shoal and either a male or female shoal. Female zebra fish showed no significant preference when given a choice between male and females shoals, nor between mixed‐sex shoals and either male or female shoals. When given choices between shoals of differing size, females preferred to associate with the larger shoal, whether or not they were composed of males or females. Males, however, had no preference for larger shoals over smaller shoals, whether or not they were composed of males or females. These results showed that male zebra fish were capable of distinguishing between males and females solely on the basis of visual cues. Furthermore, these results demonstrated a significant difference between the shoaling choices of male and female zebra fish, which may indicate a difference in the function of shoaling for the two sexes.     

Interdemic variation in haematocrit and lactate dehydrogenase in the African cyprinid Barbus neumayeri

This study evaluated whether the African cyprinid Barbus neumayeri from Rwembaita Swamp (low‐oxygen) and Njuguta River (high‐oxygen) in the Kibale National Park, Uganda differed in traits related to aerobic and anaerobic metabolic potential. Haematocrit was measured as an index of blood oxygen‐carrying capacity, and tissue activities and isozyme composition of lactate dehydrogenase (LDH) were measured as indices of tissue anaerobic capacity. To address whether site‐dependent differences were acute responses v . longer‐term adjustments to environmental conditions, these variables were measured in fish sampled shortly after collection and after laboratory maintenance under well‐aerated conditions. In fish sampled in the field, those from the low‐oxygen site had significantly higher haematocrit, but this difference disappeared after long‐term laboratory maintenance. In contrast, fish from the low‐oxygen site had higher liver LDH activities than fish from the high‐oxygen site, and this difference persisted during laboratory maintenance. Polymorphism was detected at both the LDH‐A and LDH‐B loci, and genotype frequencies for LDH‐B differed significantly between collection sites. These results demonstrate physiological, biochemical and genetic differences in B. neumayeri from habitats differing in dissolved oxygen availability and suggest both acute and long‐term responses to local environmental conditions.     

Dispersal patterns and status change in a co‐operatively breeding cichlid Neolamprologus pulcher: evidence from microsatellite analyses and behavioural observations

Genetic techniques and long‐term behavioural observations were combined to investigate dispersal patterns and changes in social position in Neolamprologus pulcher , a co‐operatively breeding cichlid from Lake Tanganyika. Comparisons of genetic variance ( F _ST) across sub‐populations demonstrated that fish were genetically more similar to individuals from proximate sub‐populations compared to individuals from distant sub‐populations. Microsatellite analyses revealed year‐long philopatry for some individuals and that other individuals dispersed to new territories and sub‐populations. Individuals appeared to disperse farther (across many territories in a sub‐population or to new sub‐populations) to achieve breeding status. Non‐breeding group members (or helpers) were observed to inherit breeding positions and male breeders were replaced faster than female breeders. These results demonstrate that important and difficult to obtain life‐history information can be obtained from genetic sampling.     

Behavioral syndromes and the evolution of correlated behavior in zebrafish

Studies of "behavioral syndromes" in different populations andspecies of animals can be used to shed light on the underlyingmechanisms of evolution. For example, some personality syndromessuggest the existence of an underlying hormonal link, whereasother relationships between boldness and aggression appear tobe the result of similar selective pressures. Here, we used1 wild-derived and 2 laboratory strains of zebrafish (Daniorerio) to examine relationships among 5 behavioral measures:shoaling, activity level, predator approaches, latency to feedafter a disturbance, and biting to a mirror stimulus. We foundevidence of an activity syndrome, as if underlying metaboliccosts influence variation in multiple forms of behavior. Evidencefor a relationship between boldness and aggression was alsoapparent but depended both on strain and which specific behaviorpatterns were identified as measures of "boldness." Althoughsome comparisons of laboratory and wild-derived strains wereconsistent with a domestication syndrome, others were not. Mostobserved relationships were relatively weak and occasionallyinconsistent, arguing against strong underlying genetic linkagesor pleiotropic effects relating any of the behavioral measures.Instead, it may be more important to study the details of selectivecontext or the long-term impact of linkages between some, butnot all, of a large set of genes influencing complex behavioraltraits. We found profound differences among strains in mostbehavior patterns, but few sex differences. One strain (TM1)was consistently different from the others (SH and Nadia) beingmore social, more likely to approach predators, and taking lesstime to recover from disturbance than were the other 2 strains.     

Sex biases in kin shoaling and dispersal in a cichlid fish

Animal dispersal is associated with diverse costs and benefits that vary among individuals based on phenotype and ecological conditions. For example, females may disperse when males benefit more from defending territories in familiar environments. Similarly, size differences in dispersal propensity may occur when dispersal costs are size-dependent. When individuals do disperse, they may adopt behavioral strategies that minimize dispersal costs. Dispersing fish, for example, may travel within shoals to reduce predation risks. Further, kin shoaling may augment inclusive fitness by reducing predation of relatives. However, studies are lacking on the role of kin shoaling in dispersal. We explored how sex and size influence dispersal and kin shoaling in the cichlid Neolamprologus caudopunctatus. We microsatellite genotyped over 900 individuals from two populations separated by a potential dispersal barrier, and documented patterns of population structure, migration and within-shoal relatedness. Genetic differentiation across the barrier was greater for smaller than larger fish, suggesting larger fish had dispersed longer distances. Females exhibited weaker genetic differentiation and 11 times higher migration rates than males, indicating longer-distance female-biased dispersal. Small females frequently shoaled with siblings, possibly offsetting dispersal costs associated with higher predation risks. In contrast, small males appeared to avoid kin shoaling, possibly to avoid local resource competition. In summary, long-distance dispersal in N. caudopunctatus appears to be female-biased, and kin-based shoaling by small females may represent a behavioral adaptation that reduces dispersal costs. Our study appears to be the first to provide evidence that sex differences in dispersal influence sex differences in kin shoaling.     

Behavioral Plasticity in Response to Environmental Manipulation among Zebrafish (Danio rerio) Populations

Plastic responses can have adaptive significance for organisms occurring in unpredictable environments, migratory species and organisms occupying novel environments. Zebrafish (Danio rerio) occur in a wide range of habitats and environments that fluctuate frequently across seasons and habitats. We expect wild populations of fish to be behaviorally more flexible than fish reared in conventional laboratory and hatchery environments. We measured three behavioral traits among 2 wild (U and PN) and 1 laboratory bred (SH) zebrafish populations in four environments differing in water flow and vegetation regimes. We found that the degree of plasticity varied with the type of behavior and also among populations. In general, vegetation increased aggression and water flow decreased latency to feed after a disturbance, but the patterns were population dependent. For example, while wild U fish fed more readily after a disturbance in vegetated and/or flowing habitats, fish from the wild PN population and lab-reared SH strain showed little variation in foraging across different environmental conditions. Zebrafish from all the three populations were more aggressive when tested in an arena with vegetation. In contrast, while there was an inter- population difference in shoaling distances, variation in shoaling distance across environmental conditions within populations was not significant. These results suggest that both foraging and aggression in zebrafish are more plastic and influenced by immediate context than is shoaling distance, which may have a stronger genetic basis. Our findings point to different underlying mechanisms influencing the expression of these traits and warrants further investigations.     

Reasons for the invasive success of a guppy (Poecilia reticulata) population in Trinidad

The introduction of non-native species into new habitats poses a major threat to native populations. Of particular interest, though often overlooked, are introductions of populations that are not fully reproductively isolated from native individuals and can hybridize with them. To address this important topic we used different approaches in a multi-pronged study, combining the effects of mate choice, shoaling behaviour and genetics. Here we present evidence that behavioural traits such as shoaling and mate choice can promote population mixing if individuals do not distinguish between native and foreign conspecifics. We examined this in the context of two guppy (Poecilia reticulata) populations that have been subject to an introduction and subsequent population mixing event in Trinidad. The introduction of Guanapo River guppies into the Turure River more than 50 years ago led to a marked reduction of the original genotype. In our experiments, female guppies did not distinguish between shoaling partners when given the choice between native and foreign individuals. Introduced fish are therefore likely to benefit from the protection of a shoal and will improve their survival chances as a result. The additional finding that male guppies do not discriminate between females on the basis of origin will further increase the process of population mixing, especially if males encounter mixed shoals. In a mesocosm experiment, in which the native and foreign populations were allowed to mate freely, we found, as expected on the basis of these behavioural interactions, that the distribution of offspring genotypes could be predicted from the proportions of the two types of founding fish. This result suggests that stochastic and environmental processes have reinforced the biological ones to bring about the genetic dominance of the invading population in the Turure River. Re-sampling the Turure for genetic analysis using SNP markers confirmed the population mixing process and showed that it is an on-going process in this river and has led to the nearly complete disappearance of the original genotype.     

The relationship between dominance behaviour and growth as a function of habitat stability and community complexity: a test using salmon and trout

Predation shortly after release is the main source of mortality among hatchery‐reared fish used to restore or enhance endangered salmonid populations. We found, that hatchery‐reared salmonid young originating from endangered stocks have weak innate responses to their natural fish predators. The ability to avoid predation in fish can be improved through social learning from experienced to naïve individuals. Huge benefits would be achieved, if social learning processes could be successfully applied on a large scale to enhance viability of hatchery fish prior to release into the wild. By using model predators together with chemical cues from real predators we tested if social learning could be used to train hatchery‐reared salmonid young to avoid fish predators. As there are clear differences in social behaviour among the salmonid species, we first examined whether these differences affect the probability and efficiency of learning anti‐predator skills from trained demonstrators. We compared anti‐predator responses of observers (fish trained by using experienced fish as demonstrators) with those of control fish, which had been 'trained' by untrained naïve conspecifics. We also examined how the efficiency of social learning depends on the ratio of experienced to naïve fish involved in social transmission trials. The results of these experiments will give guidelines how social learning could be utilized in developing hatchery scale training protocols.     

Combining capture–recapture data and pedigree information to assess heritability of demographic parameters in the wild

Quantitative genetic analyses have been increasingly used to estimate the genetic basis of life‐history traits in natural populations. Imperfect detection of individuals is inherent to studies that monitor populations in the wild, yet it is seldom accounted for by quantitative genetic studies, perhaps leading to flawed inference. To facilitate the inclusion of imperfect detection of individuals in such studies, we develop a method to estimate additive genetic variance and assess heritability for binary traits such as survival, using capture–recapture (CR) data. Our approach combines mixed‐effects CR models with a threshold model to incorporate discrete data in a standard ‘animal model’ approach. We employ Markov chain Monte Carlo sampling in a Bayesian framework to estimate model parameters. We illustrate our approach using data from a wild population of blue tits (Cyanistes caeruleus) and present the first estimate of heritability of adult survival in the wild. In agreement with the prediction that selection should deplete additive genetic variance in fitness, we found that survival had low heritability. Because the detection process is incorporated, capture–recapture animal models (CRAM) provide unbiased quantitative genetics analyses of longitudinal data collected in the wild.     

Variation in aggressive behaviour in the poeciliid fish Brachyrhaphis episcopi: Population and sex differences

Aggression is often positively correlated with other behavioural traits such as boldness and activity levels. Comparisons across populations can help to determine factors that promote the evolution of such traits. We quantified these behaviours by testing the responses of wild-caught poeciliid fish, Brachyrhaphis episcopi, to mirror image stimuli. This species occurs in populations that experience either high or low levels of predation pressure. Previous studies have shown that B. episcopi from low predation environments are less bold than those that occur with many predators. We therefore predicted that fish from high predation populations would be more aggressive and more active than fish from low predation populations. However, we found the opposite - low predation fish approached a mirror and a novel object more frequently than high predation fish suggesting that ‘boldness’ and aggression were higher in low predation populations, and that population-level boldness measures may vary depending on context. When tested individually, low predation fish inspected their mirror image more frequently. Females, but not males, from low predation sites were also more aggressive towards their mirror image. Variation in female aggression may be driven by a trade-off between food availability and predation risk. This suggests that the relationship between aggression and boldness has been shaped by adaptation to environmental conditions, and not genetic constraints.