Microbial carriage can keep a population alive long enough for adaptation to catch up
Rapid environmental change can drive a host population toward extinction before beneficial host alleles have time to spread. This preprint treats microbe-assisted rescue as a population-genetics problem rather than a metaphor, asking whether an associated microbe changes the odds that a declining host population survives long enough to recover. The model replaces one locus in a classical two-locus rescue framework with microbial carriage and follows the coupled dynamics of host allele frequency and population abundance under weak selection. It is a compact mathematical move, but it captures two very different biological effects that can operate at the same time: direct improvement in host performance and altered selection on the rescue allele itself. One effect buys time. The other changes the speed of evolutionary repair. Those are not interchangeable outcomes, and the model keeps them separate from the start.
Under the weak-selection approximation, microbial variables settle into a quasi-microbial equilibrium conditional on the host allele. That reduction matters because it strips the system down to the variables that actually control rescue: how many hosts remain, and how quickly the rescue allele rises. The framework identifies two microbial contributions. First, additive benefits of microbial carriage raise host fitness immediately, lifting population growth without changing the evolutionary trajectory of the host allele. Second, positive interactions between the rescue allele and the microbe strengthen selection on that allele, which speeds adaptation but does not cushion the initial demographic crash in the same way. The two mechanisms can look similar at the bedside of the population, yet their signatures are distinct in the model. One buffers decline. The other accelerates genetic response.
A microbe can therefore function as demographic insurance, an evolutionary catalyst, or both.
Direct fitness gains and altered selection pull rescue in different directions
The model separates the direct and interaction effects cleanly enough to show how each reshapes the classic rescue U, the relationship between environmental stress and the rescue genotype’s ability to prevent population collapse. Additive microbial benefits raise the floor of that U. In practical terms, hosts carrying the microbe start from a higher fitness baseline, so the population does not fall as far or as fast while the rescue allele is still rare. This is the cushioning effect most people would expect from a helpful symbiont: better growth, less demographic drag, more time for evolution to work.
The interaction term behaves differently. When the microbe and the rescue allele reinforce one another, selection on the allele becomes stronger, so the allele rises faster through the host population. Yet that same strengthening narrows the rescue U rather than buffering the demographic trough. Faster adaptation does not automatically mean a safer demographic trajectory. If the allele is pushed upward mainly because its association with the microbe makes it more advantageous, the population can still pass through a severe bottleneck before recovery arrives. The model’s point is sharp: microbial influence on evolution is not the same as microbial support for population size.
Imperfect transmission creates a failure mode during recovery
The most clinically interesting result in this theoretical framework is the failure mode it names selection-amplified dilution. When selection acts through fecundity and carriers transmit the microbe to only some offspring, adaptation itself can erode the beneficial microbe. Successful hosts produce more descendants, but some of those descendants do not inherit the microbe, so the very process of adaptation generates a growing pool of non-carriers faster than transmission can replenish it. If that dilution is strong enough, recovery fails even though the allele, the microbe, and their interaction are each beneficial on their own.
This is a subtle but important outcome. The population does not collapse because the microbe is useless, nor because the host allele lacks benefit, nor because the interaction is negative. It collapses because the epidemiology of transmission lags behind the genetics of selection. The model therefore links two literatures that often run on parallel tracks: evolutionary rescue theory and evolutionary epidemiology. In a real host population, that means a helpful symbiont may need more than favorable biology; it may also need faithful inheritance to stay relevant during a crisis. Weak transmission can turn a temporary aid into a fading advantage right when the population needs it most.
Why the model matters for host-microbe rescue theory
The framework clarifies three qualitatively different routes to survival. A microbe can raise host growth directly and give the population time. A microbe can strengthen selection on a rescue allele and speed evolutionary recovery. Or both effects can be necessary, with neither sufficient alone. These are not just semantic distinctions. They define which biological measurements matter most when trying to judge whether a host-associated microbe is likely to prevent extinction under environmental stress. If the key effect is demographic buffering, then host growth rate and immediate fitness consequences dominate. If the key effect is evolutionary acceleration, then the strength of the host allele–microbe interaction becomes central. If transmission is imperfect, then the persistence of carriage across generations becomes a limiting step rather than a background detail.
The model also explains why beneficial microbes can still fail under the wrong transmission regime. A population can be helped by a microbe in the short term and undermined by the same microbe’s instability in inheritance over the longer term. That sequence is biologically plausible in both animal and plant hosts, where microbial associations can be partially transmitted rather than strictly inherited. The preprint does not claim those outcomes are common in nature, only that the logic exists and can be derived formally. In that sense, the value of the model lies in separating mechanisms that are often blended together in verbal accounts of symbiosis.
A framework for identifying when symbionts rescue hosts and when they do not
The authors position the model as a general account of how associated microbes can enhance or compromise evolutionary rescue. The central contribution is conceptual discipline: it distinguishes direct microbial support from interaction-driven acceleration and then adds imperfect inheritance as a route to failure. That makes the framework useful for interpreting future empirical systems in which host decline, microbial carriage, and adaptive recovery happen at the same time. It also sets up a clear warning for any rescue story built on symbiosis alone: a beneficial microbe is not automatically a stable one, and selection can work against carriage even while it works for the host allele.
Because this is a preprint and not peer reviewed, the model should be treated as a theoretical advance rather than a settled description of any one natural system. Its contribution is the structure it gives to a question that has been hard to pin down: when does a microbe merely accompany survival, and when does it become part of the mechanism that makes survival possible? The answer depends on direct fitness effects, allele-specific interactions, and the reliability of transmission across generations. Miss one of those pieces, and the rescue story can change direction entirely.