The role of personality research in contemporary behavioral ecology: a comment on Beekman and Jordan

In their review, Beekman and Jordan (2017) ask whether animal personality research has provided any new insights for behavioral ecology. This is an important and timely question as much research in behavioral ecology is currently dedicated to this topic, while its contribution to behavioral ecology is subject to debate. Beekman and Jordan describe how they perceive the field’s objectives, approaches, and achievements. Beekman and Jordan’s thesis usefully exemplifies the confusion among behavioral ecologists on what students of animal personality research actually try to achieve. This confusion has arisen because personality is used as a flagship term in 2 distinct fields of research. First, ethologists and psychologists study the structure of individual behavior, often purposefully seeking comparisons with (approaches and paradigms used in) human personality research (Carter et al. 2013). Beekman and Jordan criticize their lack of rigorous testing of adaptive hypotheses and their phenomenological approach. Their criticism appears justified but should rather have been phrased as a critic to the field of evolutionary psychology. Second, behavioral ecologists study personality because they seek to understand the ecological conditions under which selection favors among-individual variation in repeatedly expressed behaviors. This line of research does not simply replicate previous theoretical work (e.g., in optimal foraging) as it focuses on predicting among-individual (co)variation in repeatedly expressed traits rather than just predicting phenotypic (co)variation in behavior (the norm in behavioral ecology), and thus represents a clear advance in our understanding of hierarchically structured phenotypic traits, such as behaviors. Personality research has consequently not only developed a body of adaptive theory, but has, in fact, also put forward a set of testable assumptions and predictions that have been put to the test (e.g., Nicolaus et al. 2012), exactly in the way as proposed by Beekman and Jordan. Beekman and Jordan do largely recognize the usefulness of personality research in introducing variance partitioning approaches in behavioral ecology. This integration of concepts developed in quantitative genetics, such as those related to reaction norms, character states, and indirect genetic effects, should, however, not merely be viewed as a simple application of existing methods from other fields, but rather as an important way to help facilitate the long-overdue incorporation of quantitative genetics theory into evolutionary behavioral ecology research (Westneat and Fox 2010). The focus of personality researchers on the occurrence of variation, and covariation, at multiple hierarchical levels, has, moreover, contributed importantly to an increasing awareness that many of our biological hypotheses concern patterns at specific hierarchical levels of biological organization (e.g., Nicolaus et al. 2016). For example, the role of trade-offs in maintaining individual variation in life-history strategies implies the existence of patterns of covariance between life-history traits at the among-individual level specifically (Araya-Ajoy et al. 2016), while the norm in behavioral ecology research would be to inappropriately test such hypotheses using simple (unpartitioned) phenotypic data. Personality researchers have, notably, actively worked on getting such messages across within the behavioral ecology community, for example by producing educational tools for behavioral ecologists interested in learning, or teaching, analyses, and interpretation of multilevel data (Allegue et al. 2016). The statistical angle to the personality field has also resolved issues that are repeatedly highlighted by Beekman and Jordan: is the labeling of among-individual variances as “personality” and among-individual covariances as “behavioral syndromes” useful? There is probably little disagreement here; many personality researchers have already adopted the variance partitioning terminology used in evolutionary ecology, while others will likely follow suit. The challenge for behavioral ecology therefore is not whether or not to pursue personality research, or whether among-individual variance in behavior should or should not be called “personality.” Those questions have been resolved. The question at hand should rather be why behavioral ecologists have not yet broadly acknowledged the fundamental importance of animal behavior varying across hierarchical levels, and why the interesting biology underpinning such complex forms of variation are still not studied more broadly. Achieving such a shift would finally move the field beyond the tyranny of the gold mean and better connect behavioral ecology to other fields of evolutionary biology.