Predictor variables considered in construction of regression models which modelled the nocturnality of each independent detection event of a given species as a function of a suite of predictor variables.

Prior to modeling, predictor variables were tested for excessive multicollinearity. Due to anticipated excessive multicollinearity (Pearson’s r > 0.7) of all direct human presence variables, only one variable from this category was included in each species’ model. Direct human presence variable selection was performed by regressing nocturnality against each direct human presence predictor independently, and comparing Bayes Factors of these models against a null (intercept-only) model. The variable from the model with the highest Bayes Factor was selected, so long as Bayes Factor was > 1. If no Bayes Factors were > 1, a direct human presence variable was not included in the species’ final model. a CT: camera trap, b GIS: Geographic Information Systems (Acquisition through ArcGIS Pro), c BC Vegetation Resources Inventory (2019 data; (https://www2.gov.bc.ca/gov/content/industry/forestry/managing-our-forest-resources/forest-inventory/data-management-and-access), accessed May 29, 2020.