Parental effects on animal personality

Abstract

Parental effects are any effect parents may have on the phenotype of their offspring over and above direct genetic transmission. By adaptively adjusting the phenotypes of their offspring to suit future environmental conditions, parents may increase their own fitness. A likely target of this parental programming is behavior, and the resulting variation in individual behavior can lead to diverse animal personalities. Here, I argue that parental effects may be an important source of variation in behavior and that by synthesizing the fields of parental effects and animal personality, we can gain novel insights into the ecological and evolutionary causes and consequences of individual variation.

INTRODUCTION

The study of animal personality and parental effects are 2 major research areas that have recently grown to prominence within behavioral ecology (for touchstone reviews, see Mousseau and Fox 1998; Sih et al. 2004, respectively). An increasing number of investigations of animal personality (also known as temperaments, coping styles, or behavioral syndromes) have shown that there is a tendency for individuals from the same population to differ consistently in their behavioral response to a given context and to show generality across multiple contexts (Réale et al. 2007; Dingemanse and Wolf 2009; Sih et al. 2010).

Over the same period, a different group of researchers has been investigating the ways parents can shape their offspring's phenotypes over and above genetic inheritance. Parental effects occur when the phenotype or experienced environment of one or both parents influences the phenotype of their offspring apart from the effects of direct genetic transmission (Marshall and Uller 2007). Parental effects can be a result of one or both parents influencing offspring phenotypes via nest site selection, nutrient provisioning, hormone transfer, providing particular social experiences or serving as a mentor or model from which to socially learn, and epigenetic modification of gene expression may be an important proximate mechanism (Bagot and Meaney 2010; Crews 2010; Curley et al. 2011). Behavior can act as both a generator and a target of parental effects as parental behaviors may affect nonbehavioral traits in the offspring (for example, parental provisioning behavior may affect offspring size) and nonbehavioral traits in the parents may affect offspring behavior (for example, the deposition of androgens in eggs may increase the aggressiveness of the offspring) (for recent reviews, see Marshall and Uller 2007; Badyaev and Uller 2009; Maestripieri and Mateo 2009; Wolf and Wade 2009). The developmental conditions provided by the parents that affect the phenotype of the offspring may be shaped by heritable traits in the parents, in which case parental effects can be thought of as indirect genetic effects (Wolf et al. 1998).

The parental effects and animal personality research traditions have progressed for the most part independently of one another. However, both share the common goal of uncovering the underlying causes and consequences of individual phenotypic variation. In this forum article, I argue that parental effects are a mechanism that can generate the behavioral variation that is typified by animal personalities. Parental effects can program offspring personalities in ways that maximize parental fitness, and as such parental effects may be an important, but previously underappreciated, force in shaping animal personality. I argue that parental effects may be an evolved system that allows parents to modulate the personality characteristics of their offspring in order to maximize parental fitness.

Variation is the raw material for natural selection, and it is important that we understand the developmental, physiological, ecological, and evolutionary processes that generate and maintain phenotypic variability (Sih et al. 2010). To date, scientific effort has focused on the proximate mechanisms and evolutionary implications of animal personality (Dingemanse and Wolf 2009; Sih et al. 2010). Meanwhile, the developmental underpinnings of variation in animal personalities have received less attention (Stamps 2003; Stamps and Groothuis 2010a, 2010b). This lack of emphasis on development is a general phenomenon throughout behavioral ecology (Stamps 2003) but has been particularly acute in the subfield of animal personality. Developmental processes should influence the observed structure of personality and need to be carefully considered when attempting to unravel the evolutionary and ecological importance of animal personality (Stamps and Groothuis 2010a, 2010b). In particular, parental effects on development may be important for understanding individual variation in personality.

CONTEXT-DEPENDENT PARENTAL EFFECTS

There are 2 general categories of parental effects: 1) developmentally entrenched effects, which are an obligate aspect of the developmental process and 2) context-dependent effects that have variable outcomes depending on the environmental context in which they are expressed (Badyaev and Uller 2009). Developmentally entrenched effects are fundamental to the successful development of the organism, and failures in their implementation are likely to result in poor fitness outcomes (Badyaev and Uller 2009). For example, provisioning the developing offspring can be a developmentally entrenched parental effect because without a minimum level of nutrients, the offspring will not survive or develop normally. Developmentally entrenched parental effects occur regardless of the environmental or social context that the parents experience and will generally reduce phenotypic variation within a generation by standardizing the developmental process. By contrast, context-dependent parental effects are expressed differentially depending on the environmental or social conditions experienced by the parents. Context-dependent effects may act to increase phenotypic variation within a generation because different parental experiences lead to different programming of the offspring generation (Crews and Groothuis 2005). For example, parents in certain contexts may provide additional provisioning to the propagule or offspring over and above the minimum required level thereby increasing growth rate or adult body size which may have important effects on behavior (Bernardo 1996a). Increased provisioning may permanently alter phonotypic characteristics of the offspring, or may have transitory effects that dissipate by adulthood, but alter the offspring's developmental trajectory during which time other characteristics may be modified (Bernardo 1996a).

Context-dependent parental effects can be further subdivided into 2 distinct categories (Marshall and Uller 2007): anticipatory and diversifying effects, each of which can generate phenotypic variation and variation in personality. Anticipatory parental effects (sometimes called directional parental effects) occur when the parent's current environment or some cue the parents perceive of the likely future environment has a predictable unidirectional influence on offspring phenotype (Marshall and Uller 2007). Anticipatory parental effects occur when particular offspring phenotypes are more fit in particular environmental conditions, and parents can act to bias their offspring toward the phenotype that is best suited to the environment those offspring are likely to face. Anticipatory parental effects may give rise to individual variation in personality if there is sufficient spatiotemporal variation in the environment over scales relevant to the individual. Because different personality characteristics may be more successful in certain environmental or social contexts, parents should produce offspring with different personality types that will perform well within their individual local environment. These local contexts may be subtle and so when researchers sample from a population of animals in the wild, they may draw on individuals that developed under different conditions and received different parental programming. As a result, anticipatory parental effects can contribute to measured behavioral variation at the population level.

A clear example of such possible anticipatory effects comes from the Norway rat. Rat mothers differ in the degree to which they lick and groom (LG) their pups during their first week postparturition, and the amount of LG a pup receives has lifelong consequences on its physiological and behavioral responses to stress (Champagne et al. 2003; Cameron et al. 2005). Rat pups that receive a high level of LG have a reduced response to acute stress compared with pups that received low LG. Low LG pups are shyer, less exploratory, less social, less aggressive, and less dominant than high LG pups throughout their lifespan (Meaney 2001; Cameron et al. 2005). Cross fostering studies have demonstrated that the difference in neuroendocrine stress response and behavior is a maternal effect with pups showing behavior and stress profiles consistent with their rearing environment and not with their genetic mother (Francis et al. 1999). Chronically stressed mothers show low LG behavior and thus produce highly stress reactive offspring (Cameron et al. 2005). Being highly stress reactive may be beneficial in challenging environments as stress responses can improve defensive reactions and may increase survival under high predation threat (Zhang et al. 2004). Thus, mothers may transfer information about current environmental conditions to their offspring via differences in maternal behavior, and as a result, parents may increase their offspring's survival or reproductive fitness (Champagne et al. 2003).

The second type of context-dependent parental effect is known as a diversifying parental effect or a bet-hedging strategy. Diversifying parental effects occur when parents increase the variability of offspring phenotypes in response to environmental uncertainty (Marshall and Uller 2007; Crean and Marshall 2009). When the environment is particularly unstable, or the parents have poor information as to what kind of environment their offspring will face, the best strategy may be to diversify offspring phenotypes to ensure that at least some of them will be a good phenotypic match for their environment (Crean and Marshall 2009; Russell and Lummaa 2009). As argued above, different personality types may be more successful in different environments (Dingemanse and Wolf 2009; Sih et al. 2010), and by diversifying the personalities of their offspring, parents may ensure that at least some of their offspring have the ideal personality type for the social and environmental conditions they will face. For example, in some bird species (e.g., black-headed gulls, Larus ridibundus; zebra finches, Taeniopygia guttata; pied flycatchers, Ficedula hypoleuca), mothers adjust level of androgens deposited in their eggs generating variation between eggs both within and between clutches (Schwabl 1993; Gil et al. 1999; Groothuis et al. 2005). Embryonic androgen exposure has lifelong consequences for the phenotype of the developing chick (Forstmeier et al. 2004; Eising et al. 2006; Groothuis et al. 2008; Ruuskanen and Laaksonen 2010). Chicks exposed to higher androgen levels during development grow into bolder, more aggressive, and exploratory adults than individuals exposed to lower androgen levels (reviewed in Groothuis et al. 2005). Mothers may increase the androgen levels deposited in later laid eggs in order to increase the competitive ability of later hatched chicks thereby reducing the effects of hatching asynchrony (Groothuis et al. 2005). However, avian mothers may also vary androgen levels within a clutch to diversify the personality of their chicks to ensure at least some of them will be successful in the current environmental context (Eising et al. 2006).

WHY SHOULD THERE BE PARENTAL EFFECTS ON PERSONALITY?

Personality has widespread effects on a diversity of behaviors across a variety of contexts (Sih et al. 2010). Amongst the possible mechanisms modulating personality, parental effects are particularly well suited for designing offspring that will behave in a way that enhances their survival and reproduction in a given environment. Parental effects typically occur during early development when an organism's neural and bodily systems are being constructed (Meaney 2001; Zhang et al. 2004) and may be altered in a way that would be costly or impossible later on. As a result, the period of early development is a critical time for the integration of environmental information into the behavioral phenotype of the organism (Stamps and Groothuis 2010a, 2010b). During this period, which often coincides with the time of parental dependence, the phenotype of the organism may be fine-tuned to improve its match to the current environmental conditions (Stamps 2003). As such, the parents of a developing organism are in a unique position to guide its development and alter the offspring's personality to better match the environment it will face (Crews and Groothuis 2005). Parental programming of offspring personality may be adaptive as a response to the inherent imbalance in information about the environment between the parents and the offspring during early development (Meaney 2001). Because parents have survived to reproduce in their current environment, they will have gathered information about the social and ecological conditions that exist in that environment. A lack of experience and sensory or mobility limitations may limit an offspring's ability to directly assess their environment during early development (Bernardo 1996b; Mousseau and Fox 1998). If the parental environment is a good predictor of offspring environment, then it may pay parents to alter the personality of their offspring to best cope with that environment (Marshall and Uller 2007). Cues of impending environmental change, for example, shortening day lengths or increasing food supply, may also affect parental behavior resulting in adjustment of the offspring personality to be adapted to conditions that are likely to occur during their lifetime (Bernardo 1996b; Mousseau and Fox 1998). When parents have good information about the environment the offspring are likely to face, we should expect anticipatory parental effects to predominate. On the other hand, if parents have poor information about the environment their offspring are likely to face, it may pay parents to hedge their bets by producing offspring with a variety of personality types and diversifying effects may be more prevalent (Crean and Marshall 2009). In either case, parental effects may be an important force in the generation and maintenance of individual variation and should be carefully considered in studies investigating animal personality.

CONCLUSIONS

Parental effects are an important concept in biology and have wide-reaching ecological and evolutionary implications. Parental effects may be particularly important in behavioral development and may be an underappreciated source of individual variation in animal personality. Parents may adjust the personality of their offspring to improve the phenotype–environment correspondence. Alternatively, when the environment is highly unpredictable, parents may hedge their bets and actively diversify the personality of their offspring thereby ensuring at least some of their offspring will be a good fit. In either scenario, parental effects may be an important mechanism to generate and maintain behavioral variation and therefore may have substantial evolutionary implications. Researchers investigating personality in animals should carefully consider the early developmental conditions of their subjects and pay particular attention to the importance of parent–offspring interactions in shaping personality characteristics. In the future, careful experimentation will be required to fully understand the relationship between parental experience and offspring personality. Successful modeling of animal personality may have to account for parental effects, and the explicit development of personality theory with parental effects in mind would be a fruitful area for future work.

FUNDING

A.R.R. is supported by a Natural Sciences and Engineering Research Council of Canada Alexandra Graham Bell Canada Graduate Scholarship (Doctoral).

I thank Sigal Balshine, Reuven Dukas, David Shore, Jim Quinn, Connie O’Connor, Cody Dey, Tara Farrell, Lauren Guillette, and Cristian Gutiérrez-Ibáñez for their feedback on earlier versions of this manuscript.

References