On the problems of not accounting for measurement- and sampling uncertainty in ecological and environmental models

In many applied cases of ecological and environmental modelling, there is a sizeable variation among the measured variables due to measurement- and sampling error. Such measurement- and sampling error among the independent variables may lead to regression dilution and biased prediction intervals in traditional empirical modelling. It is possible to avoid these shortcomings by subsampling and hierarchical modelling, but this is still not a common practice. Here, it is recommended to model the measurement- and sampling errors by integrating the stochastic modelling of the errors into hierarchical ecological and environmental models.