To include or not to include (the invader in community analyses)? That is the question

An increasing number of studies report impacts from invasive species on community metrics or ecosystem functions. We draw attention to an issue arising whenever impact is measured on a community where the invader is an integrated part: should or shouldn't the attributes of the invader itself be included in the data-analysis? We identify many examples from the published literature showing inconsistency in whether or not data for the invader is included or excluded, and discuss potential implications for ecological interpretations. We also provide a case study to show that the invasive seaweed Undaria pinnatifida can be interpreted to have strong or no impact on seaweed communities, depending on its inclusion or exclusion in the data analysis. We conclude that it is critical for studies to (1) clearly state in the methods section, if the invaders are included or excluded from the data-analysis, (2) acknowledge potential differences in outcomes when comparing results based on different methods, and (3) analyze, if possible, impacts both with and without the invader. Finally, we note that this 'inclusion versus exclusion' conundrum is not only relevant to invasion biology, but to any field where the test-object of interest can be an integrated part of the response, such as when impact of seaweed blooms are analysed on community productivity or community effects are quantified over time from ecological pulse-perturbation experiments.