Data for: Climate change alone cannot explain boreal caribou range recession in Quebec since 1850

The contraction of species range is one of the most significant symptoms of biodiversity loss worldwide. While anthropogenic activities and habitat alteration are major threats for several species, climate change has not been overlooked either. For species at risk, differentiating the effects of human disturbances and climate change on past and current range transformations is an important step towards improved conservation strategies. We paired historical range maps with global atmospheric reanalyses from different sources (ERA5, CERA-20C, 20CRv3) to assess the potential effects of recent climate change on the observed northward contraction of the distribution range of boreal populations of woodland caribou (Rangifer tarandus caribou) in Quebec (Canada) since 1850. We quantified these effects by highlighting the discrepancies between different southern limits of the caribou’s range (used as references) observed in the past and reconstitutions obtained through the hindcasting of the climate conditions within which caribou are currently found. Hindcasted southern limits moved ~105 km north over time under all reanalysis datasets, a trend drastically different from the ~620 km reported for observed southern limits since 1850. The differences in latitudinal shift through time between the observed and hindcasted southern limits of distribution suggest that caribou range recession should have been only 17% of what has been observed since 1850 if recent climate change had been the only disturbance driver. This limited impact of climate reinforces the scientific consensus stating that caribou range recession in Quebec is mainly caused by anthropogenic drivers (i.e. logging, development of the road network, agriculture, urbanization) that have modified the structure and composition of the forest over the past 160 years, paving the way for habitat-mediated apparent competition and overharvesting. Our results also call for a better consideration of past distribution ranges in models aiming at projecting future distributions, especially for endangered species.

Methods
Adult caribou females were captured and fitted with GPS-collars.

We examined archives and other sources to find reliable information on past boreal caribou distributions.

We extracted the climate data used in this study from three datasets, i.e. ERA5, CERA-20C and 20CRv3. 

We modelled the realized climate niche of boreal caribou by linking current climate conditions and occurrences of caribou and then hindcasted past boreal caribou distributions using the identified climatic niche. Different algorithms were used to do so: Generalized Linear Models (GLM), Generalized Additive Models (GAM), Random Forests (RF) and Boosted Regression Trees (BRT).