When microbes behave like prey: what bacterial 'herds' and amphibian diet shifts tell us about a warming planet
Two new studies out of Queen Mary University of London suggest the smallest organisms are already reorganising their lives around a hotter world, with limits.
On 15 July 2026, researchers at Queen Mary University of London published a finding that upended an assumption baked into decades of microbial ecology: when predatory microbes attack, some photosynthetic bacteria don't flee, they cluster. The behaviour, described in a Queen Mary release and reported by Phys.org on 15 July 2026, resembles the way wildebeest bunch on the Serengeti or anchovies tighten into bait balls. The cells form protective "herds" that sacrifice some individuals to keep the colony alive.
The work matters beyond microbiology. These bacteria sit at the base of freshwater food webs and pump a meaningful share of fixed carbon into inland waters. If predation pressure shapes how they aggregate, it also shapes how much carbon stays in the water column and how much settles to the sediment. A behavioural trick at the micrometre scale becomes a knob on a planetary dial.
The discovery in plain terms
Queen Mary microbiologists found that the green-pigmented bacterium Chlorella forms dense aggregates within minutes when exposed to a predator, the ciliate Tetrahymena. Individual cells on the outside of a cluster are far more likely to be consumed; those packed inside survive. The strategy trades some of the colony for the rest, an arrangement the researchers describe in language borrowed from animal ecology. In controlled experiments reported by Phys.org, aggregated populations persisted longer than dispersed ones under sustained predation.
The mechanism is mechanical as much as chemical. Dense clusters create local viscosity that slows the ciliate's strike rate, and the outer cell layer acts as a physical buffer. Predators appear to lose interest when the cost of capture rises.
The amphibian parallel
A day earlier, on 14 July 2026, a separate team at Queen Mary published a parallel finding. Amphibians, they reported via Phys.org, can adjust their diet under heat stress, shifting toward prey that delivers more water or different nutrients. Tadpoles and newts raised under simulated warming ate differently from their siblings in cooler tanks. The flexibility is real. So is its ceiling.
When the temperature gap grew too wide, the amphibians stopped eating altogether. Diet plasticity, the researchers concluded, is a stopgap, not a long-term answer. Beyond a threshold, an animal cannot swap its way out of a thermal regime it did not evolve for.
Two stories, one structural point
Read together, the two papers describe the same logic at different scales. When the environment changes faster than a lineage can evolve, behaviour fills the gap. Bacteria clump rather than mutate. Amphibians retune their diet rather than relocate. The question that runs through both studies is the same: how much rearranging can a population do before the rearrangement stops working?
This is the part of climate biology that rarely makes headlines. Genome-scale adaptation takes generations; for many freshwater microbes, that window is weeks. Behavioural plasticity is faster, but it draws on a finite toolkit. The Queen Mary work on Chlorella shows what that toolkit looks like at its most elegant. The amphibian work shows what it looks like at its edge.
What remains uncertain
Neither study, on the public record, claims to have solved the carbon-cycle implication. The bacterial-herd result is a controlled-tank finding; whether the same dynamics hold in real lakes, with their turbulence, mixed grazer communities and seasonal swings, is open. The Phys.org coverage notes the team plans field follow-up, but it does not report any estimate of how much the clumping behaviour shifts carbon export in a natural system.
The amphibian paper carries a similar caveat. Diet shifts were observed in the lab. Translating that into survival rates in the wild, where food webs are already disrupted by chytrid fungus, habitat loss and introduced predators, is a different question. The researchers are explicit, per the Phys.org write-up, that behavioural flexibility has hard limits under sustained warming. Where those limits sit for any given species remains a research question, not a number.
Why the wiring matters
The two papers also point at a structural point that goes unsaid. When the inputs to a system change faster than its components can evolve, the components start behaving differently first. Aggregation, dietary switching, altered breeding windows: these are behavioural signals of an environment that is moving out from under its inhabitants.
For policymakers, the message is not comforting. Conservation frameworks built around species ranges and habitat maps tend to assume populations hold still long enough to be counted. The Queen Mary work, taken together, suggests the behaviour is the leading edge of the change, and the morphology, the genetics and eventually the extinction risk are following it. Understanding what microbial herds do in a warming lake, and what an amphibian will or will not eat at 28 degrees Celsius, is not a footnote to climate science. It is part of how the science has to be done from here on.
The next test is whether the bacterial behaviour survives contact with a real lake, and whether the amphibian flexibility collapses at the temperatures projected for the second half of this century. Both are answerable. Neither will be answered this year.
Monexus framed these two findings as one story, not two. The wires treated them as separate items; the structural connection only emerges when the reading is done side by side.