Let's DAG in: how directed acyclic graphs can help behavioural ecology be more transparent

Directed acyclic graphs (DAGs) are powerful tools for visualizing assumptions/hypotheses and causal inference. Although their use is becoming more widespread across various disciplines, they remain underutilised in behavioural ecology and evolution. Here, we point out why DAGs can serve as highly valuable tools in this field, particularly in the context of observational and field studies, which can feature many variables with complex relationships. Using concrete examples, we show that including DAGs in empirical studies helps clarify and summarize the key underlying assumptions, which are often implicit. With that, DAGs can be used to make researchers aware of bad controls and help them to explicitly think through the relationship between variables and their inclusion in statistical models. In addition, providing DAGs makes the work of reviewers and meta-analysis researchers easier, more rigorous and reliable. Overall, DAGs enhance understanding and transparency, ultimately improving study reproducibility and contributing to greater reliability and replicability across the field. With this paper, we hope to encourage all behavioural ecologists to include DAGs in their papers.