Bias is inherent to any kind of data. And so is biodiversity data. There are many biases in the distribution of species records, as well as spatial biases in the location of points, and in the environmental variables used for modelling. So it comes as no surprise that accouting for such bias is one of the key problems that need to be solved when modelling biodiversity trends. In a paper in npj Biodiversity, a large group of experts in different types of biodiversity modelling led by Duccio Rocchini discuss different ways of accouting for several sources of bias to improve the reliability and accuracy of any kind of biodiversity model.
You can read the article at https://jhortal.com/project/rocchini-et-al-npj-biodiv-2023-a-quixotic-view-of-spatial-bias-in-biodiversity-modelling/