Can Removing One Predator Change a River? The Claim Is Contested

The idea is that a change at the top of a food chain propagates downwards in alternating directions. Remove the predator and its prey increase, which means the plants those prey eat decrease. Return the predator and the sequence reverses. The alternation is the characteristic signature, since the effect on each level is opposite to the effect on the level above it, and in a longer chain the alternation continues. The effect can work through numbers, with the predator actually killing enough prey to matter, or through behaviour, with the prey avoiding places where it is dangerous to feed even though few are killed.

The strongest evidence comes from simpler systems that can be manipulated:

• •Sea otters, urchins and kelp forests along the Pacific coast
• •Removing starfish from rocky shores, which collapses the species mix
• •Fish added to or removed from small lakes, which changes the algae
• •Islands cut off by a reservoir, which lose predators and change entirely
• •Enclosure experiments where predators are excluded from a plot
• •Insect systems on single plants, where whole chains fit in a cage

The story that wolves returned to Yellowstone in 1995, reduced elk numbers and moved the survivors away from valley bottoms, allowing willow and aspen to regrow, stabilising banks and changing the course of rivers, has been told in a widely shared film and repeated everywhere. Ecologists working there have challenged most of the chain. Elk numbers fell for several reasons including hunting, drought and bears, not wolves alone. Evidence that elk avoid valleys because of wolves is weak. Willow recovery is patchy and appears driven mainly by water availability and by beaver activity. Changes to channels are hard to attribute. The current view is that the system changed considerably and that assigning it to one cause was premature.

The mechanism that works through behaviour rather than through killing is worth separating, because it is both the more interesting claim and the harder one to demonstrate. The argument is that prey avoid places where a predator is likely, feed less there, and stay alert instead of eating, so the plants in those places recover even though the number of prey has barely changed. If true it means a small number of predators can have a large effect, which matters for conservation. Testing it requires showing that prey behaviour actually changed, that it changed because of the predator, and that plants responded, and studies that measure all three carefully find effects that are real but generally smaller than early claims.

A story of this shape is attractive for reasons worth recognising, since the same pattern recurs across popular science. It is a clean causal chain with a single lever, which is far easier to tell and to remember than a list of interacting factors. It carries a satisfying moral about restoring what was removed. It gives conservation a powerful argument, which makes advocates reluctant to complicate it. And the original researchers did publish evidence, so the claim was not invented, it was overstated and then simplified further in retelling. The underlying point survives the correction, since top predators genuinely do shape ecosystems, and the correction is about how much and how directly.

A change at the top of a food chain propagates downward with alternating effects, working through numbers killed or through where prey dare to feed. Otters and urchins, starfish on rocky shores and fish in small lakes provide clear evidence. The Yellowstone wolf story has been challenged by ecologists working there, who find elk declined for several reasons and willow recovery tracks water rather than wolves.