Shifted Models Cannot be Used for Predicting Responses of Biodiversity to Global Change: The African Elephant as an Example

Human activities have affected animals’ behaviour, distribution and population structure and this effect is predicted to increase in the future. Considerable effort is therefore being focussed on understanding and predicting such future changes in response to anthropogenic pressures, this to better conserve and restore populations and species. However, there is a risk that scientists and practitioners fail to recognise the extent of past human effects on biodiversity and use the situation they are familiar with as a baseline and a measure against which to assess current or future biodiversity changes. This failure to recognise past changes is known as the shifting baseline syndrome (Pauly 1995). We propose that when models (descriptive, quantitative or conceptual) describing niche features, such as distribution, habitat use, dietary resources or behaviour, are developed using data collected in a system (communities, species or populations) that has already undergone shifted baselines, such models should be referred to as ‘shifted models’. Here we define such shifted models and demonstrate associated shortcomings, focusing on the particular example of shifted environmental niche models in a recently published study on African elephants Loxodonta africana (Dejene et al. 2021).