Genetics of personalities: no simple answers for complex traits



Staffan Bensch, Fax: +46464716; E-mail:


Identifying the genes that underlie phenotypic variation in natural populations, and assessing the consequences of polymorphisms at these loci for individual fitness are major objectives in evolutionary biology. Yet, with the exception of a few success stories, little progress has been made, and our understanding of the link between genotype and phenotype is still in its infancy. For example, although body length in humans is largely genetically determined, with heritability estimates greater than 0.8, massive genome-wide association studies (GWAS) have only been able to account for a very small proportion of this variation (Gudbjartsson et al. 2008). If it is so difficult to explain the genetics behind relatively ‘simple’ traits, can we envision that it will at all be possible to find genes underlying complex behavioural traits in wild non-model organisms? Some notable examples suggest that this can indeed be a worthwhile endeavour. Recently, the circadian rhythm gene Clock has been associated with timing of breeding in a wild blue tit population (Johnsen et al. 2007; Liedvogel et al. 2009) and the Pgi gene to variation in dispersal and flight endurance in Glanville fritillary butterflies (Niitepold et al. 2009). A promising candidate gene for influencing complex animal personality traits, also known as behavioural syndromes (Sih et al. 2004), is the dopamine receptor D4 (DRD4) gene. Within the last decade, polymorphisms in this gene have been associated with variation in novelty seeking and exploration behaviour in a range of species, from humans to great tits (Schinka et al. 2002; Fidler et al. 2007). In this issue, Korsten et al. (2010) attempt to replicate this previously observed association in wild-living birds, and test for the generality of the association between DRD4 and personality across a number of European great tit populations.

Great tits (Fig. 1) show consistent individual differences in novelty seeking and exploratory behaviour (i.e. personality) as measured in simple testing arenas (Fig. 2), with estimated heritabilities between 0.22 and 0.54 (Dingemanse et al. 2002; Drent et al. 2003; Quinn et al. 2009). The recent finding of an association between DRD4 polymorphisms and variation in personality in captive great tits (Fidler et al. 2007) was a major step forward in our understanding of the genetic basis of animal personalities. Yet, as is well known from association studies in humans and animal model species, SNP–phenotype correlations are rarely robust across populations and environments.

Figure 1.

 Bold or cautious? Part of the answer might lie in the dopamine receptor D4 (DRD4) gene. Photo by Henk Dikkers.

Figure 2.

 Assessment of great tit personality in the testing arena. The hops and flights between perches on the artificial trees are used as a proxy for exploration behaviour. Photo by Anne Rutten.

Korsten et al.’s (2010) study is remarkable because the authors did what is rarely done: using an independent sample of free-living birds from the same study population they tried to replicate Fidler et al.’s (2007) initial finding and, in addition, tested for a similar association between DRD4 polymorphisms and variation in novelty seeking behaviour in other European great tit populations. They could replicate the association between one of the DRD4 polymorphisms and explorative behaviour in the original population. Moreover, the amount of variation in avian personality explained by the DRD4 SNP was remarkably similar (4.5–5.8%) compared to the previous study (5.6–6.0%). Interestingly, however, the association was weak to completely absent in the other populations, despite similar heritabilities in these populations (Dingemanse et al. 2002; Quinn et al. 2009).

Whereas the authors cannot rule out that methodological issues, for example the slightly different protocols used to measure personality in the different study populations, contribute to these findings, the differences among populations in the association between DRD4 genotype and personality might be genuine. Indeed, although a meta-analysis of more than 30 studies confirmed that the DRD4 gene is associated with novelty seeking in humans, also here there were large differences between populations (Munafòet al. 2008). Hence, both in humans and great tits we are far from having a simple genetic marker that can predict personalities, or can be used to track changes in gene frequencies in response to selection.

The future of personality genetics

A remaining issue in need of more research is to establish whether the identified DRD4 SNP is the actual site under selection. The SNP is synonymous and although it is possible that such a substitution in the DNA sequence can have phenotypic effects, for example via differential splicing, it is more likely that it is in linkage disequilibrium with other functionally significant polymorphisms. Although Fidler et al. (2007) screened for DNA polymorphisms along an almost 10 kb DNA fragment including four DRD4 exons, covering larger flanking regions might be worthwhile.

The candidate gene approach was clearly a rewarding stepping stone into the genetics of personality. However, there are most likely many genes that are contributing to such complex phenotypes. Identifying more genes governing variation in behavioural traits within and across populations will thus be a logical next step, which would also allow for the study of epistasis and genotype × environment interactions (Dingemanse et al. 2010). Recent technical advances in high-throughput sequencing have now made it technically possible to perform genome-wide association studies in any species (Ellegren & Sheldon 2008), and the amount of behavioural data and pedigree depth available for many great tit populations might make this a particularly fruitful approach to identify more loci contributing to variation in complex personality traits.

However, the basis of personality differences might not (only) lay in the DNA sequence. Few genetic changes can result in large differences in gene expression, as exemplified by differences between males and females (e.g. Mank 2009), or different casts in social hymenoptera, which are genetically identical but express very different phenotypes (Evans & Wheeler 1999). If personality differences are manifested by substantial differences in gene expression, either through slight genetic differences or governed by epigenetic processes (Groothuis & Carere 2005), microarray analyses or transcriptome sequencing might be the way to go.

With these new tools within reach it is only a matter of time (and sizable grants) until the genetics of personality and other complex behavioural traits will really take off. The study by Korsten et al. (2010) is certainly a first step in the right direction, and although one could interpret the different genotype–phenotype relationships among study populations as a negative result, it should rather be seen as an appetizer to dig deeper into the mechanistic basis of animal behaviour. It is also a reminder that we should probably not expect genotype–phenotype associations to be simple, especially not for highly complex traits such as animal personality.

Barbara Tschirren is an evolutionary ecologist interested in linking genetic polymorphisms to variation in behaviour, life history and parasite resistance in wild-living birds and mammals. Staffan Bensch is Professor in Animal Ecology at Lund University and has a broad interest in molecular ecology with current research projects involving population genetics of migratory birds and avian malaria parasites.