Evaluating the Quality of Taxonomic Publications: A Simple Alternative to Citations and Effort

Abstract

Despite the fundamental importance of taxonomic work, its impacts go largely unrecorded and unacknowledged because of a reliance on citations in its evaluation (Valdecasas et al. 2000). In published articles, the names of subject species are provided but generally without citing the taxonomic publications that described the species or higher taxonomic levels (Valdecasas et al. 2000, Venu and Sanjappa 2011). The continued use of indices based on citations to evaluate taxonomic publications is therefore unfair and unreasonable. The recently proposed T-Index is an alternative method for evaluating taxonomic publications and is based on the nature and extent of the work or effort involved (Valdecasas 2011). This index considers each taxonomic level separately (e.g., genus, family, order) and assumes that taxonomic effort increases with the number of new taxa described (newly discovered and revised combined) but decreases with the consequent total number of taxa within each level (Valdecasas 2011). More precisely, it takes the square of the number of new taxa within each taxonomic level, divided by the resulting total number of taxa within the level, and summed over relevant taxonomic levels (Valdecasas 2011). Some minor adjustments to this index have been suggested and accepted (PallGergely 2014, Valdecasas 2014). The T-Index may assess the performance of taxonomic work and would obviously get away from citation-based indices, thus avoiding their inherent problems (Krell 2000, 2002, Valdecasas et al. 2000), but it provides little or no assessment of taxonomic impact. In assessing such impact, it would surely be necessary to determine the usefulness of being able to identify and name particular taxa, something the T-Index does not do. Therefore, given the increasing and understandable attempts to consider research impact in evaluating individual researchers, their institutions, grant proposals, and so on (Pyke 2013, 2014), a new approach to evaluating taxonomic publications is still required. The impact of a description of a new taxon seems best assessed by how many subsequent articles mention that taxon (Wagele et al. 2011), which provides a simple basis for an alternative index for evaluating taxonomic publications. With computer technology, mentions of particular taxa could be tallied across publications, presumably counting multiple mentions within a single document just once. Each mention of a taxon could also be taken as a mention of all relevant higher-level taxa, with decreasing weight assigned to successively higher taxa on the basis of the increasing number of included species. A simple way to achieve this would be to take the reciprocal of the total number of included species as the weight for each mention of a particular taxon. Summing weighted mentions for each taxonomic publication across the described taxa could provide an overall index for that publication. I shall refer to this index as the M-Index. The following example illustrates how the M-Index might work in practice. Suppose a taxonomic publication described a new genus and three new species and that a subsequent publication mentions two of these species. In that case, the second publication would count as two mentions of the original publication at the species level and a one-third-weighted mention of the genus, giving a total weighted mention of 2.33 for the original taxonomic publication. This process could be similarly repeated across all publications that mention either one of the three species or the genus, with the weighted mentions summed over all such publications. Descriptions at higher taxonomic levels could be evaluated in the same way. Of course, the M-Index would be based on a changing classification system across all species, but this need not impede its implementation. Each time a taxon is newly described or revised, the overall system would change, but keeping track of such changes should not be difficult with available computer technology. For the evaluation of taxonomic publications, my proposed M-Index therefore provides a simple alternative to measures based on citations or effort. However, it might warrant modification and refinement, and it could be used in combination with other indices.