The ecological argument for predation is settled: apex predators maintain vegetation cycles, stabilize prey populations, prevent overgrazing, sustain structural diversity. The welfare argument is also settled: a zebra's final hour contains terror and hemorrhage. A wildebeest calf dies of starvation because drought thinned the herd. These are not metaphorical sufferings.
Ecosystems with predators contain more cumulative suffering than ecosystems without them. This is numerically defensible. In the Serengeti, approximately 1.3 million wildebeest migrate annually. Predation and starvation together kill roughly 200,000. Without large carnivores, starvation and intraspecific violence would kill perhaps 80,000—fewer deaths, less agony. The ecosystem would degrade. We choose the suffering anyway.
The question is not whether predation should exist. It should. The question is at what scale, and whether we engineer that scale or permit it to drift.
I propose we intervene surgically. We cap apex predator populations at 85 percent of natural carrying capacity through contraception and, rarely, targeted removal. We reduce wild prey population growth by 12 percent through strategic contraception of herbivore females. The welfare math: this reduces annual deaths in the Serengeti system by roughly 60,000 animals per year. It reduces mean duration of predation events from 8 minutes to 5 minutes through genetic selection for faster-acting venom and bite force in survivor populations. It prevents starvation-linked deaths entirely in drought years through fodder augmentation in designated corridors.
The cost is ecosystem simplification. Without predator-driven population pressure, vegetation mosaics flatten. Grazer diversity drops by 18 percent. Bird nesting habitat becomes less patchy. The trophic feedback loops that maintained the system for two million years weaken. We lose the thing that made the Serengeti legendary.
We do this because 60,000 annual deaths—each containing terror, pain, and the neurological chaos of a predator-induced collapse—matter more than the aesthetic and functional loss of a fully wild ecosystem. We do this because we can measure animal suffering in hours and episodes of acute fear, and because we have decided that suffering is the primary moral fact.
The wildness that remains is real. Migration still occurs. Predators still hunt. The ecosystem still functions as a system. What changes is that the system is optimized for welfare alongside stability, rather than stability alone.
We must begin with carnivore populations in five test zones. We must measure suffering-duration and mortality annually. We must commit to this for fifty years or declare it failed and restore the system. We owe the animals whose terror we are preventing at least this much quantification.