When the same pathogen gets two different invasion thresholds
A pathogen can be poised to invade in one stable version of a population and unable to gain a foothold in another—even when the species and environmental parameters are the same. That is the striking result of a mathematical comparison by Carli Peterson, Erika Gallo and Christopher Kribs: in their tick-host model, the invasion threshold changes depending on whether ticks and hosts settle into steady numbers or repeat a two-year cycle. In a second model, a host-parasitoid system, the same kind of demographic bistability leaves the host-driven threshold unchanged (Peterson et al., 2026).
The contrast matters because public-health and ecological models often compress invasion potential into one number, the basic reproductive number, . If the population that a pathogen encounters can settle into more than one stable pattern, “the” threshold may not be singular. The paper’s answer is not that cycles always make disease more dangerous. It is more precise: the outcome depends on how infection moves through the populations involved.
This is a theoretical study, not evidence that Lyme disease behaves differently in a particular county or that any measured outbreak is being missed. Its value is conceptual. It shows why an invasion calculation must be matched to the demographic state and transmission route it is meant to describe.
The Science
The basic reproductive number asks, in simplified terms, how many new infections one infected individual would generate in a population where infection is initially rare. When is below one, each infection tends to replace itself with fewer than one further infection, and a disease-free state is locally stable. Above one, infection can grow from a small introduction. This is a threshold for invasion near a particular disease-free state—not a guarantee that an epidemic will occur or a forecast of its eventual size.
Many textbook calculations assume the uninfected population settles to one stable equilibrium: a fixed point, where the population sizes stop changing from one time step to the next. But nonlinear populations can have more than one stable destination under the same conditions. A stable cycle is one alternative: numbers rise and fall in a repeating sequence. The word “bistability” describes the coexistence of two stable outcomes; which one the system reaches can depend on its starting conditions.
Peterson and colleagues compare two discrete-time host-parasite models. Discrete-time models update populations in steps, making them natural for seasonal or generational life cycles. In the first, ticks and several vertebrate host groups are linked in a model motivated by transmission of Borrelia burgdorferi, the bacterium associated with Lyme disease in the northeastern United States. The model distinguishes tick life stages—larvae, nymphs and adults—and susceptible and infected hosts. Its demographic dynamics can support both a positive fixed point and a stable two-cycle.
The tick cycle reflects a particularly consequential detail: the blacklegged tick’s stages feed in sequence across seasons, and the complete life cycle spans more than one year. In the model’s synchronous cycle, the two annual states differ sharply: one is larval-only, while the other contains nymphs and adults. A separate stable, strictly positive two-cycle also emerges over part of the parameter range. Figure 1 summarizes how tick population outcomes change as the demographic reproductive number—the model’s measure of tick reproduction and survival—increases.
The researchers then ask what happens when infection is introduced in small numbers around each stable disease-free state. They use the next-generation operator method, which tracks the expected production of new infections while accounting for transitions among infected classes. At a fixed point, the calculation uses a constant demographic background. On a cycle, it must account for transmission across an entire repeating period, not just one annual step.
The second model is a version of the Nicholson–Bailey host-parasitoid framework, modified to include intraspecific competition and infection. Hosts and parasitoids can acquire infection through horizontal transmission, and the model also includes imperfect vertical transmission—transfer from parent to offspring that does not occur every time. The paper uses it as a contrasting transmission architecture: infection can circulate between hosts and parasitoids, but it does not follow the same full tick–host–tick route as the vector-borne model.
What They Found
The tick-host result is attractor-dependent. The basic reproductive number calculated around the stable demographic cycle differs from the one calculated around the positive coexistence fixed point. Consequently, there are parameter regions in which infection can invade from rarity on one attractor but not on the other (Peterson et al., 2026). Put plainly: knowing the model parameters is not enough to name a single invasion threshold unless one also knows which disease-free population pattern is being considered.
The route of infection helps explain why. In the Lyme-motivated model, transmission involves a loop: infected hosts can infect feeding ticks, and infected ticks can later infect hosts. The abundance and host-contact pattern of the tick stages change over the cycle. Those changes affect successive links in the transmission chain. A cycle can therefore alter the combined potential of the host–vector–host pathway, rather than merely shifting the timing of a single transmission event.
The authors’ fixed-point calculation makes this structure visible. The infection subsystem can be reduced to the relevant larval and host components because, under the model’s assumptions, adult ticks feed only on deer, which are treated as incompetent reservoirs. The resulting expression combines contributions from competent host groups, tick survival and feeding success, transmission probabilities, and host survival. On a cycle, the corresponding calculation must carry infection through both points of the repeating pattern. Those are not interchangeable demographic backdrops.
The host-parasitoid comparison supplies the important counterexample. Here, the host-driven invasion threshold is necessarily the same at the positive coexistence fixed point and on the host-only periodic cycles. In other words, multiple stable demographic attractors do not automatically generate different values of the threshold that determines whether infection grows through the host pathway. The model’s transmission structure constrains what demographic variation can do.
That contrast is the paper’s central finding, and Figure 2 places the two tick-host reproduction numbers beside one another across demographic conditions. The thin and thick curves distinguish values calculated at the stable fixed point and stable two-cycle, respectively. The figure is useful as a qualitative map of where the measures diverge; the supplied paper material does not include the underlying plotted values, so this digest does not reproduce them as a numerical chart.
The familiar boundary of one remains important, but its interpretation needs care. For each attractor, the relevant question is whether the reproduction number calculated around that state is above or below one. If those calculations land on opposite sides of the boundary, a reader cannot responsibly summarize the system with one unqualified answer to “can the infection invade?” The answer is conditional: it depends on which stable disease-free state the population occupies.
Why This Changes Things
This is a refinement of a widely used tool, not a rejection of it. remains a powerful way to assess whether a pathogen can increase when rare. The complication is that in a multistable setting, the number belongs to a particular state and transmission system. A threshold derived at one steady state may not describe invasion around a cycle that is equally stable under the same parameter values.
That matters beyond this tick model. Seasonal disease transmission is often represented with average population sizes or a single equilibrium. Such simplifications can be useful, but they may hide the fact that populations cycle, or that multiple demographic regimes are possible. If the pathogen passes through several host or vector stages, changes in seasonal abundance can affect the whole sequence of transmission opportunities. The paper demonstrates a mathematical mechanism for that effect; it does not establish its size in nature.
The host-parasitoid result guards against an overly simple takeaway. It would be easy to conclude that whenever ecology is complicated, infection thresholds must be complicated in the same way. The models say otherwise. What matters is not just that population numbers fluctuate, but where those numbers enter the infection process. A complete vector-mediated loop can couple the demographic states of hosts and vectors in a way that changes the invasion condition; another transmission arrangement may leave the threshold invariant.
For researchers, the practical implication is to state the demographic context of an invasion calculation. If a model admits a stable equilibrium and a stable cycle, analysts should evaluate the infection dynamics on both rather than silently selecting one. In periodic systems, the calculation should cover the full period. This avoids treating a value that is valid for one attractor as a universal property of the pathogen.
For applied disease ecology, that recommendation is a prompt for better model questions, not an immediate management prescription. Field observations would need to establish whether the relevant populations display the modeled alternative attractors, which one occurs under real conditions, and whether the assumed feeding, survival and transmission pathways are accurate. The tick model includes many host groups, but it also makes simplifying assumptions—for example, infection is transmitted horizontally, persists over an individual’s lifetime, and does not increase mortality. The threshold results depend on such assumptions.
The work also underscores why transmission structure deserves equal attention to population counts. Two places might have comparable average host abundance yet differ in seasonal timing, vector-stage overlap or which hosts are competent reservoirs. A model that keeps only a mean abundance could miss the interactions that determine whether one infected host starts a self-sustaining chain. The paper does not quantify those effects in the field, but it identifies the questions that empirical studies and more detailed models should test.
What’s Next
The provided paper text ends partway through the tick-cycle calculation and does not include the complete host-parasitoid analysis, parameter tables, or numerical data behind the plotted curves. That limits what can be responsibly said about the size of threshold differences, the exact parameter ranges, or which biological mechanisms dominate in particular scenarios. The conclusions summarized here are the explicit qualitative findings stated in the supplied abstract and text, not a reconstruction of missing results.
A natural next step is to connect the mathematical comparison to measured population dynamics. For the tick system, researchers would need time-series data capable of distinguishing stable annual averages from repeating cohorts, together with estimates of host-specific feeding and infection transmission. They could then ask whether the distinct theoretical thresholds correspond to meaningful differences in invasion risk under observed conditions. Similar work could test whether the assumptions about reservoir competence and infection persistence change the contrast.
The broader theoretical task is to identify which transmission architectures make invasion thresholds sensitive to the demographic attractor. The study compares two illustrative cases; it does not supply a universal rule for all multi-host or vector-borne diseases. Extending the analysis to more complex cycles, irregular fluctuations, and additional transmission routes could show when attractor-specific measures are essential and when one threshold is sufficient.
The useful lesson is both modest and consequential: a pathogen’s invasion potential is not always a number detached from its ecological setting. When populations can settle into more than one stable rhythm, the right question is not only “What is ?” It is also “ on which population state, and through which route of transmission?” Answering that more complete question can make disease models more faithful to the living systems they are meant to help us understand.