The use of quantitative indices to quantify the importance of a plant species to human societies is widespread. While quantifi- cation may yield support for standardized methodologies and facilitate generalizations, it is important to examine the potential limitations of these indices. Moreover, because these indices are calculated at the species level, failure to control for phylogenetic relatedness in predictive models may yield misleading conclusions. We test if commonly used cultural importance indices predict species cultural keystone status among the Shipibo-Konibo community of Paoyhan in the Peruvian Amazon. Eleven of the 12 indices were correlated with each other indicating most cultural importance indices are redundant. Most indices did not predict species cultural keystone status. Phylogenetic control improved our models indicating a significant part of the predictive power of even the best index was explained by species shared evolutionary history. Our findings highlight the need for the cautious use of cultural importance indices to infer species cultural keystone status. Newly developed indices should be tested for correlation with existing indices to avoid redundancy.

Data was collected via ethnobotanical methods decribed in detail in our manuscript entitled "Most cultural importance indices do not predict species cultural keystone status."

All predictor variables were estimated via cultural importance indices cited in our manuscript entitled "Most cultural importance indices do not predict species cultural keystone status." The response variable was estimated via PCA using a correlation  matrix.  Details are provided in the data analysis section of our manuscript. All analysis were conducted in R.