“…Theories of range shifts, population declines and extinctions are now being tested directly using historical and palaeo-reconstructions (Fordham et al, 2021(Fordham et al, , 2022, permitting inferences of how biodiversity is likely to respond to future environmental change (Fordham et al, 2020). However, reconstructing past demographic changes at landscape scales poses unique modelling challenges, including the reliance on indirect proxies to draw inferences about range collapses and the timing and location of extinction (Dietl et al, 2015); uncertainty in reconstructions of past climates (Rutherford et al, 2005) and human-driven environmental threats (Ellis et al, 2021;Pilowsky, Manica, et al, 2022); and a lack of information on the ecological lifestyles and traits of many species (Fordham et al, 2016). Some of these issues can be addressed, at least in part, using process-explicit models, particularly if they are combined with pattern-oriented modelling (POM) techniques (Box 1).…”