The impact of coinfection on population stability and chaos

By combining experiments with flour beetles and a modified stage-structured model, this study demonstrates that while transmission mode is a key determinant of population stability, facilitation between co-infecting parasites is a critical mechanism that rapidly drives host populations into oscillations and chaos.

The Gist

Imagine a bustling city of flour beetles living in a jar of flour. In this city, life is a delicate balancing act. The beetles lay eggs, which hatch into larvae, grow into pupae, and finally become adults. But there's a catch: the adults and larvae are cannibals. They eat the eggs and pupae of their neighbors.

In the world of these beetles, this cannibalism acts like a traffic light.

• If the traffic light is green (low cannibalism), the population grows smoothly.
• If it's red (high cannibalism), the population crashes.
• If it's blinking yellow (just the right amount), the population might start bouncing up and down in a predictable rhythm, or it might go completely crazy and chaotic, with numbers spiking and crashing unpredictably.

For decades, scientists have studied this "traffic light" system. But in the real world, beetles rarely face just one problem. They often get hit by two different parasites at the same time. This paper asks a big question: What happens to the city's stability when the beetles are fighting two different infections instead of one?

The Two Invaders

The researchers studied two types of "invaders" (parasites) that behave very differently:

1. The "Suicide Bomber" (Obligate Killer): Think of this parasite as a terrorist who needs the host to die to spread. It's like a virus that turns the beetle into a walking bomb. The beetle must die and explode (release spores) for the next generation of parasites to infect others.
2. The "Sneaky Squatter" (Direct Transmitter): This parasite is more like a squatter living in an apartment. It doesn't want to kill the host; it just wants to live there, eat some food, and leave eggs behind while the host is still alive.

The Experiment: A Jar of Chaos

The scientists set up little "cities" (microcosms) in jars. Some had no parasites, some had one type, and some had both. They watched to see how the beetles fared.

What they found in the jars:
When the beetles had both parasites, they didn't necessarily get sicker or die faster in a simple way. However, the two parasites seemed to have a weird relationship. Sometimes they fought each other (antagonism), and sometimes they helped each other (facilitation). In the jars, the beetles with both parasites seemed to have slightly higher loads of the "Sneaky Squatter," suggesting the two parasites might be weakly helping each other out, though the effect wasn't huge.

The Computer Simulation: The Crystal Ball

Since you can't watch a beetle city for 300 years to see if it goes extinct, the scientists built a computer model. This model is like a video game simulator that predicts the future of the beetle population based on the rules of cannibalism and infection.

They ran the simulation with different settings to see how the "traffic light" (cannibalism) changed when parasites were added.

The Big Discoveries:

1. One Bad Apple vs. Two:

   • Having one parasite actually made the population more stable in some situations. It was like adding a second brake to a car; it helped keep the population from swinging too wildly.
   • Having two parasites, however, made the system much more fragile. It was like having two people fighting over the steering wheel.

2. The "Helping Hand" Effect (Facilitation):
   This is the most critical finding. The model showed that if the two parasites helped each other (facilitation)—for example, if one parasite weakened the beetle's immune system, making it easier for the second parasite to spread—the population went completely chaotic.

   • Analogy: Imagine a choir where everyone is singing in tune. If one person starts singing off-key, it's annoying. But if two people start helping each other sing off-key in a way that disrupts the whole rhythm, the entire song collapses into noise.
   • When the parasites helped each other, the beetle population didn't just fluctuate; it went into a state of chaos, where you couldn't predict the population size at all.

3. The "Suicide Bomber" is the Real Trouble:
   The chaos was most intense when the "Suicide Bomber" parasite was involved. If the "Sneaky Squatter" helped the "Suicide Bomber" spread its deadly spores, the population dynamics became wild and unpredictable.

The Takeaway

The main lesson of this paper is that nature is messy, and two problems are often worse than the sum of their parts.

If you are trying to manage a population (like a pest you want to control, or a rare species you want to save), you can't just look at one disease at a time. You have to ask: Do these diseases fight each other, or do they help each other?

• If they fight (antagonism), the population might actually be safer.
• If they help (facilitation), especially if one of them is a killer, the whole system could spiral into chaos, leading to unpredictable boom-and-bust cycles that could wipe out the population or cause it to explode uncontrollably.

In short, the stability of a population isn't just about how many beetles there are or how many eat each other; it's about the hidden, complex relationships between the invisible enemies living inside them.

Technical Summary

1. Problem Statement

Parasites are known to regulate host population dynamics, often driving populations from stability to oscillations or chaos. However, most natural hosts face coinfection (simultaneous infection by multiple parasite species) rather than single infections. Within-host interactions between parasites can be antagonistic (competition for resources, cross-immunity) or facilitative (immunosuppression, resource sharing).

While theoretical models exist for single infections, it remains unclear whether coinfection simply reinforces stability shifts seen in single infections or creates a fundamentally different stability landscape. Specifically, the study addresses the gap in understanding how transmission modes (obligate killer vs. directly transmitted) and within-host interaction mechanisms (facilitation vs. antagonism) interact to alter the stability of stage-structured host populations.

2. Methodology

The study employed a hybrid approach combining empirical mesocosm experiments with advanced mathematical modeling.

A. Empirical Experiments:

• System: Tribolium castaneum (flour beetles).
• Parasites:
  • Farinocystis tribolii (Neogregarine): An obligate killer requiring host death for transmission.
  • Eugregarines: Directly transmitted protozoa that reside in the gut without requiring host death.
• Design: Larvae were reared in four treatment groups: Uninfected, Single infection (Farinocystis), Single infection (Eugregarines), and Co-infection.
• Metrics: Prevalence (presence/absence) and intensity (parasite load) were quantified via dissection and microscopy. Statistical analysis (Chi-square, Kruskal-Wallis) was used to determine if coinfection altered prevalence or intensity compared to single infections.

B. Mathematical Modeling:

• Framework: A modified LPA (Larva-Pupa-Adult) model, a classic discrete-time, stage-structured model known for exhibiting chaos driven by cannibalism.
• Modifications:
  • Introduced explicit larval infection compartments and environmental spore reservoirs.
  • Differentiated transmission modes:
    • Obligate Killers (OB): Spores released upon host death; mortality rate ($\alpha$) drives transmission.
    • Direct Transmission (D): Spores released by living hosts; survival is required for transmission.
  • Interaction Coefficient ($\phi$): Added a parameter to the spore production equations to model within-host interactions.
    • $\phi < 1$: Antagonism.
    • $\phi = 1$: Neutral.
    • $\phi > 1$: Facilitation.
• Analysis Techniques:
  • Bifurcation Analysis: Swept the adult-on-pupa cannibalism rate ($c_{pa}$) and disease-induced mortality ($\alpha$) to map transitions from equilibrium to chaos.
  • Lyapunov Exponents (LE): Calculated the maximum LE to quantify stability (negative LE = stable; positive LE = chaotic).
  • Global Sensitivity Analysis: Used Sobol total-effect indices to identify which parameters (transmission, mortality, interaction) most strongly influenced stability.

3. Key Contributions

• Integration of Within-Host and Population Dynamics: The study bridges the gap between micro-scale parasite interactions (within-host facilitation/antagonism) and macro-scale population stability.
• Transmission Mode Differentiation: It explicitly models the distinct stability impacts of obligate killers (which require host death) versus directly transmitted parasites, showing they create different dynamical regimes.
• Quantification of Interaction Effects: The model isolates the specific role of the interaction coefficient ($\phi$), demonstrating that facilitation is a primary driver of chaos, whereas antagonism promotes stability.

4. Key Results

A. Empirical Findings:

• Prevalence: No significant difference was found in the prevalence of either parasite between single and co-infection groups.
• Coinfection Rate: Observed coinfection prevalence (17.6%) was slightly higher than the expected independent probability (11%), suggesting a weak, non-significant facilitative interaction.
• Intensity: Co-infected larvae showed a trend toward higher median eugregarine intensity, though not statistically significant.

B. Modeling & Stability Analysis:

• Single vs. Co-infection:
  • Single Infections: Generally increased stability at low and high cannibalism rates but exacerbated instability at intermediate rates.
  • Co-infection: Produced a unique stability profile. At intermediate cannibalism rates, co-infection (specifically involving obligate killers) created a wider range of regular periodic cycles compared to single infections. However, at high cannibalism rates, co-infection led to significantly less stability (more chaos) than single infections.
• Transmission Mode Impact:
  • Obligate Killers: Population stability was highly sensitive to disease-induced mortality ($\alpha$). Small changes in mortality caused abrupt shifts between stability, quasiperiodicity, and chaos.
  • Direct Transmission: Stability was less sensitive to mortality changes and more dependent on cannibalism rates.
  • Mixed Transmission: The presence of an obligate killer in a mixed system dictated the stability profile, overriding the stabilizing effects of the directly transmitted parasite.
• The Role of Facilitation ($\phi$):
  • Antagonism ($\phi \leq 0.5$): Strongly stabilized populations, resulting in negative Lyapunov exponents (stable equilibrium).
  • Neutrality ($\phi \approx 1$): Led to quasiperiodicity and mild chaos.
  • Facilitation ($\phi > 1$): Rapidly drove the system into chaos.
  • Thresholds: Systems with two obligate killers required very little facilitation to become chaotic. Mixed-transmission systems required higher levels of facilitation to trigger chaos.

5. Significance

• Mechanism for Variability: The study provides a mechanistic explanation for why host populations with similar demographics exhibit vastly different stability patterns. It suggests that facilitative interactions between co-infecting parasites are a critical, often overlooked driver of population chaos.
• Predictive Power: The findings highlight that ignoring within-host interactions (specifically facilitation) can lead to inaccurate predictions in conservation and biological control. For instance, introducing a parasite to control a pest might inadvertently facilitate a second pathogen, destabilizing the host population rather than suppressing it.
• Generalizability: While focused on Tribolium, the framework applies to any stage-structured system (e.g., insects, amphibians) where multiple pathogens interact. It emphasizes that the transmission mode of the dominant pathogen is a key determinant of whether a population will remain stable or collapse into chaos under coinfection pressure.

In conclusion, the paper demonstrates that coinfection is not merely an additive stressor but a dynamical perturbation that can qualitatively alter the stability landscape of host populations, with facilitation acting as a potent catalyst for chaotic dynamics.