Abstract
Good experimental design is critical for sound empirical ecology and evolution. However, many contemporary studies fail to replicate at the appropriate biological or organizational level, so causal inference might have less vigorous support than often assumed. Here, I provide a guide for how to identify the appropriate scale of replication for a range of common experimental designs in ecological and evolutionary studies. I discuss the merits of replicating multiple scales of biological organization. I suggest that experimental design be discussed in terms of the scale of replication relative to the scale at which inferences are sought when designing, discussing and reviewing experiments in ecology and evolution. I also suggest that more conversations about experimental design are needed, and I hope this piece stimulates such conversation.