Empty waters, harder hunts: rethinking the "halo" around the world's seabird cities

On 15 July 2026, a paper in Physical Review Letters put a fresh twist on one of ecology's most cited ideas. The authors argue that for large seabird colonies the surrounding "halo" of depleted waters is shaped less by how much prey exists and more by whether the birds can reach it. The framing matters because the colonies in question are not curiosities: they are some of the largest animal aggregations on Earth, and the foraging radius around them governs how much energy the population can burn, and on what.

The headline finding, stripped of the maths, is this. A penguin standing on the rim of a big colony does not behave the way a model that only counts prey density would predict. Density is high. Yet the birds fly, swim and dive further than the prey map alone seems to require. The paper attributes the gap to accessibility: currents, sea-ice cover and the vertical position of prey in the water column all intervene. Where prey is locked under ice, or pushed into deeper, faster layers, the colony effectively sees a smaller ocean.

What the halo actually is

The phenomenon has a name older than most of the scientists arguing about it. In 1963, the ornithologist David Ashmole documented that seabirds on Ascension Island travelled further to forage as the local breeding population grew, even when prey in the wider ocean did not appear scarce. The pattern took the shorthand Ashmole's halo. The intuitive reading is that the colony itself eats out its neighbourhood, leaving birds to commute longer for each meal.

The paper published this week accepts that intuition but insists it is incomplete. The authors build a reaction-diffusion model in which prey are born, drift, and are consumed, and then ask what the optimal foraging radius looks like once the geometry of access is fixed. The result is a halo whose shape depends on ocean structure as much as on biomass. Northern krill clusters that sit under fast-moving surface water are as good as absent to a diving bird, even when biomass numbers look generous.

Where the prey actually lives

The point is concrete. Antarctic krill, the dietary mainstay of penguin colonies on the Antarctic Peninsula and around the Scotia Arc, has shown population swings of an order of magnitude over the past four decades. Yet foraging radii at some of the largest emperor and gentoo colonies have not collapsed in step with that biomass. The mismatch has nagged at field biologists for years. If prey had simply thinned, one would expect a gradual contraction of foraging trips as prey near the colony becomes lucrative again. In some seasons that does not happen.

The accessibility hypothesis offers an explanation. Sea-ice extent around the Antarctic Peninsula has been highly variable year to year since the late 2000s, and ice concentrates krill near the underside of the floes. When ice retreats early, krill disperse into the open water column at depths and light levels that penguins reach less efficiently. The biomass is there. The birds cannot get to it.

What the model changes, and what it does not

There is a counter-read. Some ecologists will argue that a reaction-diffusion model, however elegant, smooths over heterogeneity that matters in the field. Real colonies forage along corridors, not radially. Local upwellings, polynyas and currents near the South Shetlands and the Weddell Sea break the assumption of an isotropic ocean. If the modelled halo does not capture those corridors, the conclusion about accessibility could be partly an artefact of how the grid is drawn.

The authors' response is partial. Their model does not, and is not meant to, reproduce specific colony geometries. It is a theoretical scaffold for an empirical pattern. The strongest claim is the conditional one: that predator behaviour in the halo is jointly set by prey density and prey accessibility, and that theoretical work ignoring the second variable under-fits the field data. That conditional claim is more useful than a sweeping one.

The practical stakes are not small. Conservation planners who set marine protected areas around penguin colonies tend to draw circles. The radius is calibrated to the colony's foraging range, with the implicit assumption that any water inside the circle is roughly equivalent prey habitat. If the halo is asymmetric, and access is the binding constraint, then a circular reserve can leave the actually productive water outside the boundary and the actually empty water inside it. The paper does not settle that question, but it sharpens it.

What is next on the water

Two things to watch over the coming southern-hemisphere summer. First, whether the model's predictions hold up against the kind of high-resolution biologging data now being collected at colonies such as those at Cape Crozier and on the Antarctic Peninsula's Danco Coast. If accessibility really is doing the heavy lifting, the correlation between foraging radius and sea-ice conditions should be tighter than the correlation with krill-density estimates alone. Second, whether the framework travels beyond penguins. Ashmole's halo was first described for tropical seabirds, and the argument that geometry of access matters as much as biomass has obvious echoes in the fishing-policy debates around forage fish, where quota systems tend to be set on biomass surveys that say nothing about where, vertically or horizontally, the fish actually sit.

The unresolved question, on this reading, is whether the reaction-diffusion world the paper builds will survive contact with the noisier one the field biologists live in. The hypothesis is plausible. The data to confirm it at the relevant scale is still being collected, and the model itself does not yet say which observation would falsify it cleanly. Until that test is built in, the accessibility argument sits alongside the abundance argument, not in its place.

This piece was framed by Monexus against a single, recent publication in Physical Review Letters; field biologists' wider critique is drawn from the long-running debate around Ashmole's halo, not from a second peer-reviewed source within the present thread.