Species distribution models predict temporal but not spatial variation in forest growth

Area of Science:

• Ecology
• Forestry
• Climate Change Biology

Background:

• Bioclimate envelope models project species distributions under climate change, but their realism and interpretation are debated.
• Understanding the link between predicted climate suitability and actual tree growth is crucial for ecological and forestry applications.

Purpose of the Study:

• To investigate the correlation between climate suitability predictions and tree growth using both inventory plot data and tree-ring records.
• To assess the reliability of species distribution models in predicting tree responses to climate change.

Main Methods:

• Species distribution models were built using the RandomForest classifier and occurrence data for four key European tree species.
• Climate suitability was correlated with tree volume from inventory plots and tree-ring width from chronologies.

Main Results:

• Climate suitability predictions were uncorrelated with spatial tree growth records from inventory plots.
• A moderate correlation (average 0.22) was found between habitat projections and interannual tree-ring width variation.
• Individual site correlations varied widely, from -0.31 to 0.82.

Conclusions:

• Tree responses to climate change are highly site-specific, making local reforestation predictions challenging.
• Projected changes in climatic suitability may indicate average growth trends over larger geographical areas.