Reply to Bortolus: what's in a name?

r guides for authors should implement more stringen requirements for citing taxonomic materials or processe used to arrive at species names for publication in ecologica journals [1]. Although we agree with the general sentimen that greater diligence is necessary when attributing spe cies names to taxonomic units, the suggestions he has pu forward are unrealistic, and could diminish, rather than improve, the overall quality of published studies in ecology and evolution. Before throwing out studies with incomplete (or suspect taxonomy, one must first ask why it is important to know the ‘true’ taxonomic designations of entities in those stud ies, and then ask where the use of taxonomy might best be applied and improved. Good taxonomy is critical for studie of evolution. It ensures continuity among multiple studie on the same taxon, and it prevents amplification of error through time and misleading results in phylogenetically controlled meta-analyses [2]. However, studies with incom plete taxonomy can still have valid and important finding when focusing on functionally, morphologically, or geneti cally similar species, as is likely to be the case when the risk of making taxonomic errors is highest [3]. This i particularly true where the topic of interest is a broad response to experimental manipulation rather than an evolutionary response, or comparison of responses within a single taxon among multiple studies. For example, researchers working in experimental com munity ecology frequently make use of highly diverse taxa that are easy to sample and/or manipulate [4,5]. Rathe than focusing on precisely determining and recording a formally recognized taxon of each individual, they work with diversity indices or measurements of species richnes that only require the researchers to distinguish reliably among unique taxa. Experiments of this type typically require researchers to work with a large number of taxa that have received insufficient attention from experts and therefore, with ‘morphospecies’ or ‘operational taxonomi units’ rather than with Linnaean species. Recent work shows that, when examining species richness, the majority of information on among-site patterns can be retained even with large numbers of species removed from the data set and statistical power can be maximized when increasing sample size at the expense of difficult-to-identify specie [6]. Other methods used to address these topics involve