Taxonomic bias in animal behaviour publications

Evidence suggests that certain taxonomic groups are more thoroughly studied than others across a wide range of biological disciplines. Such taxonomic biases have the potential to define our understanding of theory, and limit the generality of our insights. To assess the distribution of taxonomic representation in current and historical animal behaviour research, we constructed a data set containing article metrics and taxonomic information for all research articles published in the journal Animal Behaviour between 1953 and 2015. We found significant taxonomic bias, with chordate papers making up 70% of all publications in the past 15 years, despite accounting for less than 7% of all animal species. Within chordates, Animal Behaviour content is biased towards endotherms, with birds and mammals comprising more than 50% of all publications. In sum, six animal orders account for more than half of all publications, with the most commonly published order, Passeriformes, representing one in five articles. Our findings confirm that a relatively narrow group of ‘model’ taxa represent the vast majority of articles, and may have a disproportionate influence on our understanding of behavioural patterns and processes. Furthermore, we find evidence of a citation bias, with chordate studies receiving on average four citations more per paper than arthropod studies. While historical trends suggest that the publication gap between arthropods and chordates has been shrinking for the past 45 years, our findings show that a considerable bias still remains. These biases may originate from human preferences for certain animal types, but we argue that they are likely maintained by a mixture of taxonomic prejudices, cultural aspects of behavioural ecology as a field, and of academia in general. We suggest that the patterns are clear and their implications serious, and that it is time that both researchers and journals give serious consideration to addressing them.