Relationship between societal and scientific interest across the eukaryotic Tree of Life.

A) Relationship between number of views in Wikipedia (popular interest) and number of papers in the Web of Science (scientific interest) for each species. Both axes are log-scaled to ease visualization. Density functions are provided for both scientific (above scatter plot) and societal interests (right of scatter plot) to illustrate the distribution of values. Color coding refers to the three realms of Animalia, Fungi and Plantae. The regression line is obtained by fitting a Gaussian generalized additive model through the data (F1,3017 = 2497.5; p< 0.001). The farther away a dot is from the fitted line, the more the attention is unbalanced towards either scientific (negative residuals) or societal interest (positive residuals). B) Distribution of negative and positive residuals (from the regression line in A) across the species sampled for each Phylum/Division. Phyla/Divisions with only one sampled species are represented with dots.

Influence of species-level traits (blue) and cultural factors (red) on the scientific and societal interest across the eukaryotic Tree of Life.

A) Forest plots summarize the estimated parameters based on Negative binomial generalized linear mixed models (Eq. 1). Baseline levels for multilevel factor variables are: Domain [Multiple] and IUCN [Not Evaluated]. Error bars mark 95% confidence intervals. Variance explained is reported as marginal R2, i.e. those explained by fixed factors. Asterisks (*) mark significant effects (α = 0.01). Exact estimated regression parameters and p-values are in Tables S1. B) Outcomes of the variance partitioning analysis, whereby we partitioned out the relative contribution of species-level traits (blue) and culture factors (red). Joint explained variance (Species + Culture) is highlighted in purple. Unexplained variance is the amount of unexplained variance after considering the contribution of random factors related to species’ taxonomy and biogeographic origin (as obtained via conditional R2).

Influence of species-level traits (blue) and cultural factors (red) on scientific (A) and societal (B) interest for Arthopoda, Chordata, and Tracheophyta.

Forest plots summarize the estimated parameters based on Negative binomial generalized linear mixed models (Eq. 2). Baseline levels for multilevel factor variables are: Domain [Multiple] and IUCN [Not Evaluated]. Error bars mark 95% confidence intervals. Variance explained is reported as marginal R2, i.e. those explained by fixed factors. Asterisks (*) mark significant effects (α = 0.01). Estimated regression parameters and p-values are in Tables S2 (A) and Table S3 (B).