Weeds are plants of human-modified environments, not plants out of place

The question “What makes a plant a weed?” has persisted in both the agricultural sciences and plant ecology and evolution for well over a hundred years. Various authors—agriculturists focused on crop yield, ecologists examining community assembly, and plant biologists interested in adaptation—have each defined the term weed. However, these definitions, developed from unique vantage points, have led to siloed studies and thus constraints on our understanding of weedy plant evolution. Here we briefly review how weeds have been defined historically and argue that weeds are best defined by the environment they occupy: plants that are found, uncultivated and reproducing, in human-modified environments. We then ask whether weeds are human commensals, populations that depend on human-provided resources and have diverged from non-commensal ancestors, treating human dependence as a gradient to be measured. We highlight critical gaps in our understanding of weedy plant occurrence across agricultural, suburban, and urban environments, and propose a database approach leading to testable hypotheses for studying weed presence and persistence. Important outcomes of our reframing include a bridge between agricultural and (sub)urban research disciplines and a shared, data-driven basis for comparing species across the human-modified environments they occupy, illuminating the forces that underlie the weedy lifestyle.