Theory 19/Paleontology & Mass Extinction /Macroecology/3 min
Extinction Runs Bottom-Up
The big animals disappear from the record. They may be the last to go, not the first.
Mass extinctions propagate from the base of food webs upward rather than from the loss of top predators, so the order of extinction should be predictable from trophic position and body size, with primary producers and small-bodied consumers going first.
01The evidence
- Responses of complex food webs to realistic extinction sequences have been modelled, and the order of removal changes the outcome, which is the property a bottom-up account requires.Food web extinction sequences
- Work on recovery from mass extinctions has emphasised the role of ecological interactions in stabilising the system after the event, which is the other half of a bottom-up story.Ecological interactions in recovery
02The tension
Extinction is read from the fossil record, which is biased toward large, skeletonised animals. The apparent order is therefore top-down by default, and the possibility that the base collapsed first is hard to see because the base is poorly preserved.
03The theory
A food web is sustained by primary production. If the base is disrupted, for instance by a productivity collapse or by a change in the plankton, the effect propagates upward as starvation and as loss of habitat structure.
On this account the large animals disappear last, after their prey and their habitat have gone, but their disappearance is better recorded because they are easier to fossilise. The record therefore inverts the causal order.
The prediction is that extinction selectivity should depend on trophic position once preservation bias is modelled, and that the base should show a decline before the top in well-sampled sections. It also predicts that recovery should begin at the base.
The consequence is that extinction drivers should be sought in productivity and in primary producers rather than in the predators that dominate the narrative.
04What this is not
Top-down accounts read the order from the record, which over-represents large animals. This predicts the opposite order and treats the apparent top-down pattern as a preservation artefact.
05Predictions
- P1Well-sampled sections should show base decline before top decline.
- P2Extinction selectivity should depend on trophic position after modelling preservation.
- P3Recovery should begin at the base of the web.
06What would refute it
A base that persists while the top collapses, in well-sampled sections, would restore the top-down order.
07What it would change
It would redirect the search for extinction drivers toward productivity.
08Evidence base
- Response of complex food webs to realistic extinction sequenceshttps://www.semanticscholar.org/paper/Response-of-complex-food-webs-to-realistic-Srinivasan-Dunne/a06ee4b481453a13e892bbec074212954cf3eb65
- Stabilising the unstable: ecological interactions in recovery from mass extinctions — GSAhttps://gsa.confex.com/gsa/2013AM/webprogram/Handout/Paper228915/Roopnarine_upload.pdf