Generalized additive models (GAMs) are commonly used in ecological research for their ability to model complex nonlinear relationships. However, assessing predictor importance in the presence of concurvity is difficult due to overlapping variance among predictors.
