Experimental species removals impact the architecture of pollination networks

Most network simulation studies exploring the consequences of species losses assume that network structure stays fixed after species losses. In these replicated field experiments we show that network structure is altered, in ways that can be predicted from basic ecological theory.