Abstract
Urbanization represents an unintentional global experiment that can provide insights into how species will respond and interact under future global change scenarios. Cities produce many conditions that are predicted to occur widely in the future, such as warmer temperatures, higher carbon dioxide (CO(2)) concentrations and exacerbated droughts. In using cities as surrogates for global change, it is challenging to disentangle climate variables-such as temperature-from co-occurring or confounding urban variables-such as impervious surface-and then to understand the interactive effects of multiple climate variables on both individual species and species interactions. However, such interactions are also difficult to replicate experimentally, and thus the challenges of cities are also their unique advantage. Here, we review insights gained from cities, with a focus on plants and arthropods, and how urban findings agree or disagree with experimental predictions and historical data. We discuss the types of hypotheses that can be best tested in cities, caveats to urban research and how to further validate cities as surrogates for global change. Lastly, we summarize how to achieve the goal of using urban species responses to predict broader regional- and ecosystem-level patterns in the future.