Range Edges Are Set by Competitor Noise

01The evidence

• Demographic stochasticity and resilience, rather than mean fitness alone, were shown to matter at tree species climatic edges — and transplant experiments have repeatedly found positive growth beyond observed edges, which niche models cannot accommodate.Tree species climatic edges

02The tension

Correlative niche models predict edges from climate and are used for everything from conservation planning to invasion risk. They routinely misplace edges, and the standard fix is to add more climate variables. The alternative, that the edge is set by biotic interactions, has been hard to test because it requires measuring variance rather than means.

03The theory

At the warm edge a population is small and near its physiological limit, so its growth rate is low and its variance is high. Small populations with high variance go extinct from demographic noise even when the mean growth rate is positive. Competition amplifies this by raising the variance of the growth rate.

The edge is therefore where the probability of a run of bad years exceeds the probability of recovery, not where mean fitness crosses zero. That makes the edge a property of the variance and of the competitor community, and it predicts that removing competitors extends the range without changing the climate.

This explains the transplant results. Plants grow beyond the edge when competitors are absent, and the edge in the field is set by the interaction between a marginal physiology and a fluctuating competitor load. It also predicts that edges should be fuzzy in variable climates and sharp in stable ones, because variance is what drives the extinction.

The practical consequence is that assisted migration and invasion risk are both mis-forecast by climate-only models, and that a variance-aware model would need competitor abundance as an input.

04What this is not

05Predictions

• P1Edge position should correlate with competitor density after climate is controlled.
• P2Experimental removal of competitors should extend the range edge without warming.
• P3Populations at the edge should show higher variance in growth rate than populations at the centre, at matched density.

06What would refute it

Edges that track isotherms regardless of competitor removal would restore the climatic account.

07What it would change

It would change invasion risk forecasts and make competitor data a required input for range projections.