A metric of species' charisma

Link To Share This Poster: https://cdmcd.co/PJ9xQZ
Live Discussion Link: https://cdmcd.co/Dwm4DG

Hayden Freedman and Andre van der Hoek, University of California, Irvine, Bill Tomlinson, Informatics, University of California, Irvine, Irvine, CA, Andrew Torrance, University of Kansas

Background/Question/Methods
Species are often targeted for conservation based on their "charisma"; that is, the level of interest that they elicit from humans (Martín-López et al. 2009). However, charisma can confound conservation law. For example, while the Endangered Species Act of 1973 was written to apply equally to species from all non-microbial branches of the tree of life, in practice charismatic species are disproportionately listed as endangered (Weitzman et al. 1996). A comprehensive, quantitative metric for species charisma could improve conservation efforts and help ensure that even low-charisma species are protected. We seek to identify such a metric.
We compared three Wikipedia-based metrics to determine which of them correlated most closely with prior work on species' charisma (Albert et al. 2018). Specifically, we evaluated PageRank score (the algorithm originally used by Google search), total number of links to and from each article (an alternative connectedness metric), and article length for 253 species and subspecies drawn from the IUCN red list of critically endangered mammals. We used PageRank data from (Thalhammer and Hees 2020), and calculated link count and article length via custom software.
Results/Conclusions
A point-biserial correlation revealed highly significant correlations between all three metrics and presence on Albert et al.'s list of most charismatic animals, with article length as the most significant of the three (rpb = 0.57; p < .00001**). Sixteen of the twenty species with the longest articles were present on Albert et al.'s list. By contrast, none of the twenty species with the shortest articles was included on that list. A two-tailed, two-proportion Z-test revealed a highly significant difference between these two groups (z = 5.164; p < .00001**).
We recognize that metrics derived from Wikipedia cannot perfectly capture species charisma; nevertheless, in the absence of any other metric that may be easily and rapidly applied across thousands of species, we propose that these metrics may provide a suitable approximation of a species' charisma.
We see two main applications for these metrics. First, they could be useful to identify relatively uncharismatic species objectively deserving of federal legal protection. Second, they could aid in conservation by identifying charismatic flagship/“umbrella” species whose own protection via public or private efforts might conveniently protect other species that live in the same habitat. We view this work as an initial effort to develop broad and easily-applied metrics for species’ charisma.