Research overview: sampling bias and correction in presence-only felid models

Wildlife occurrence data for cats are often opportunistically collected, biased toward roads, reserves, and accessible terrain, which can distort distribution models if uncorrected. This overview surveys how sampling bias arises and the methods used to address it, including target-group background sampling, spatial filtering of records, and bias files in MaxEnt. It reviews how uncorrected bias can make models reflect survey effort rather than true habitat, with serious consequences for climate projections. The synthesis frames bias correction as a critical but underappreciated step in felid distribution modeling, explaining why two studies of the same species can disagree and why careful handling of where data came from is as important as the climate variables themselves.