Evolutionary rescue in a consumer-resource system depends on the affected ecological traits and the population's resident life-history

This paper demonstrates that evolutionary rescue in consumer-resource systems depends critically on the specific ecological traits affected by environmental change, the population's resident life-history, and the genetic architecture of evolving traits, revealing that intrinsic growth rates alone are insufficient to predict persistence probabilities.

The Gist

Imagine a bustling city (the consumer population) that relies entirely on a massive, renewable food supply (the resource) to survive. Suddenly, a disaster strikes: the food supply chain is disrupted, or the city's ability to process food is crippled. The city's population begins to shrink rapidly, heading toward total collapse.

This is the scenario of Evolutionary Rescue. It's the race between the city dying out and its citizens evolving new skills fast enough to survive the new, harsh reality.

This paper asks a crucial question: Does it matter how the disaster hits the city? And does the city's original lifestyle matter?

Here is the breakdown of the findings using simple analogies:

1. The Two Types of Cities (Life Histories)

The researchers studied two very different types of cities:

• The "Slow & Steady" City: Citizens live long lives, have few children, and are very careful with resources. (Think of elephants or humans).
• The "Fast & Frantic" City: Citizens have very short lives, reproduce wildly, and consume resources rapidly. (Think of bacteria or insects).

2. The Three Ways the Disaster Strikes

The environmental change could hurt the city in three specific ways:

1. The "Hungry" Strike (Acquisition Rate): The citizens can't find or grab the food as well as before.
2. The "Wasteful" Strike (Conversion Efficiency): The citizens can grab the food, but they can't turn it into energy or babies efficiently. They waste it.
3. The "Fatal" Strike (Survivorship): The citizens are dying faster, regardless of how much food they eat.

3. The Big Discovery: One Size Does Not Fit All

The paper's main finding is that you cannot just look at how fast the population is shrinking to predict if they will survive. You have to know why they are shrinking and what kind of population they are.

Scenario A: The "Slow & Steady" City

If this city is hit by a disaster, they are most likely to be saved if the problem is dying faster (The Fatal Strike).

• Why? Because these citizens live long lives, they have many opportunities to reproduce over time. If the problem is that they are dying young, a mutation that helps them live longer gives them a huge advantage. They have time to pass on that "long life" gene.
• Analogy: Imagine a marathon runner. If the problem is they are getting tired too fast, a mutation that gives them better stamina is a lifesaver.

Scenario B: The "Fast & Frantic" City

If this city is hit by a disaster, they are most likely to be saved if the problem is not being able to find food (The Hungry Strike).

• Why? These citizens live fast and die young. If they can't find food, they starve before they can reproduce. However, if they can evolve to be better at finding food, they can immediately start churning out the massive number of babies needed to save the species.
• Analogy: Imagine a high-speed race car. If the problem is the engine is inefficient (wasteful), it doesn't matter much because the car is already fast. But if the car can't find gas (acquisition), it stops immediately. Fixing the gas intake is the only way to win the race.

4. The "Mutation Lottery"

The paper also explains that survival depends on luck and volume.

• The Lottery Ticket: To survive, the city needs a "winning ticket" (a beneficial mutation) to appear.
• Buying More Tickets: The more babies born, the more lottery tickets are bought.
• The Twist:
  • In the Fast City, if the disaster stops them from having babies (low birth rate), they stop buying tickets. But if the disaster just makes them die faster, they might still have enough time to buy a few tickets before the lights go out.
  • In the Slow City, if the disaster stops them from having babies, they stop buying tickets entirely. But if the disaster just makes them die faster, their long lives mean they keep buying tickets for a long time, increasing their chances of hitting the jackpot.

5. The Takeaway for Real Life

The authors argue that scientists and conservationists often just look at the intrinsic growth rate (a single number that says "this population is shrinking by X%"). They say this is like looking at a car's speedometer and saying "it's slowing down" without checking if the engine is broken, the tires are flat, or the driver is asleep.

To predict if a species will survive climate change or pollution, we need to know:

1. Is the stressor killing them faster?
2. Is it stopping them from eating?
3. Is it stopping them from turning food into babies?
4. Do they live long and slow, or short and fast?

In short: Evolutionary rescue isn't just about how bad the situation is; it's about the specific mechanics of the disaster and the personality of the population trying to survive it.

Technical Summary

1. Problem Statement

Evolutionary rescue (ER) occurs when a declining population adapts to a deteriorating environment, avoiding extinction. While previous theoretical models have explored ER, they often rely on simplifying assumptions that obscure the interplay between ecology and evolution. Specifically, most models assume:

• Populations exist in isolation (ignoring consumer-resource interactions).
• Environmental changes affect only intrinsic growth rates ($r$) or carrying capacity ($K$) in isolation.
• $r$ and $K$ are independent parameters.

In reality, ecological traits (e.g., resource acquisition, conversion efficiency, survivorship) co-vary and differentially impact $r$ and $K$ through density-dependent mechanisms. There is a significant gap in understanding how the specific ecological trait affected by an environmental stressor (e.g., reduced birth rate vs. increased death rate) interacts with the resident life-history strategy (slow-growing vs. fast-growing) to determine rescue probabilities.

2. Methodology

The authors developed a mechanistic consumer-resource model with overlapping generations to investigate ER dynamics.

• Ecological Framework:
  • Consumer Dynamics: Modeled using a discrete-time recurrence relation where consumer abundance ($N_t$) depends on resource acquisition rate ($a$), conversion efficiency ($C$), resource abundance ($R_t$), and survival rate ($S$).
  • Resource Dynamics: Modeled using logistic growth minus consumption.
  • Density Dependence: Arises explicitly through resource depletion; consumer growth is limited by the equilibrium between consumption and resource replenishment.
• Genetic Architectures:
  • One-Locus (De Novo Mutation): Simulated a haploid, asexual population where rescue depends on the fixation of a single beneficial mutation restoring a trait to its pre-stress value.
  • Polygenic (Standing Variation): Simulated a sexual, diploid population with 100 loci underlying a phenotypic trait. The phenotype maps to an ecological trait via a Gaussian performance function. Rescue involves shifting the population mean toward a new phenotypic optimum.
• Analytical Approach:
  • Derived approximate analytical predictions for fixation probabilities ($\psi$), mutational supply ($U$), and time to extinction ($t_e$).
  • Assumptions included weak selection, low individual turnover, and separation of time-scales (resource dynamics are faster than consumer dynamics).
• Simulation Approach:
  • Used SLiM v4.01 for individual-based forward-time simulations to test analytical predictions and explore parameter spaces where assumptions (e.g., weak selection) were violated (e.g., fast-growing populations).
  • Compared "slow-growing" (low fecundity, high survival) and "fast-growing" (high fecundity, low survival) resident strategies under abrupt environmental changes affecting $a$, $C$, or $S$.

3. Key Contributions

• Decoupling $r$ and $K$: The study demonstrates that intrinsic growth rates alone are insufficient to predict rescue probabilities. The mechanism of decline (which trait is affected) dictates the outcome.
• Trait-Specific Rescue Dynamics: It establishes that the probability of rescue depends on the interaction between the affected trait and the resident life-history strategy.
• Analytical Derivation: The paper provides novel analytical formulas for fixation probabilities in a consumer-resource context, accounting for density-dependent resource availability.
• Genetic Architecture Comparison: It contrasts the roles of de novo mutation (driven by mutational supply) versus standing genetic variation (driven by selective advantage) in different ecological contexts.

4. Key Results

A. Dependence on Resident Life-History Strategy

The most favorable trait for rescue changes based on the population's baseline strategy:

• Slow-Growing Populations (High $S$, Low $a/C$):
  • Rescue is most probable when the environmental change reduces survivorship ($S$).
  • Reasoning: Reduced survivorship increases the mutational supply per unit time (more births relative to deaths) and slows the time to extinction, allowing more mutations to arise.
• Fast-Growing Populations (Low $S$, High $a/C$):
  • Rescue is most probable when the environmental change reduces resource acquisition rates ($a$).
  • Reasoning: Reduced acquisition lowers per-capita resource consumption, allowing resource abundance ($R_t$) to recover faster. This boosts the fecundity of rescue mutants early in the decline, increasing their fixation probability. Conversely, reduced survivorship in fast-growers accelerates extinction too rapidly for rescue to occur.

B. Mechanisms Driving Rescue Probabilities

• Mutational Supply vs. Fixation Probability:
  • In one-locus scenarios, rescue is primarily driven by mutational supply. Traits that reduce extinction time (like reduced $S$ in fast-growers) lower the total number of mutations generated, hindering rescue.
  • In polygenic scenarios, where standing variation exists, the advantage of high mutational supply is diminished. Rescue is driven more by the selective advantage of alleles. Here, reduced acquisition rates ($a$) in fast-growers still yield the highest rescue probability due to the rapid recovery of resource abundance benefiting the mutants.
• Resource Feedback:
  • Mutants affecting acquisition rates ($a$) benefit from a "positive feedback" loop: as the consumer population declines, resource abundance increases, disproportionately boosting the fitness of mutants that can utilize the abundant resource.
  • Mutants affecting conversion efficiency ($C$) do not receive this same early boost in resource availability.

C. Analytical vs. Simulation Accuracy

• The analytical approximations work remarkably well for slow-growing populations (weak selection, low turnover).
• For fast-growing populations, analytical models over-predict rescue probabilities because they neglect demographic stochasticity and assume resources reach quasi-equilibrium instantly. In fast-growers, consumer and resource dynamics occur on similar time scales, violating the separation of time-scales assumption.

D. Selection Coefficients

• The selection coefficient ($s$) of a rescue mutant is not constant; it depends on population density and the specific trait evolving.
• At high density, selection favors survival mutants in slow-growers but favors fecundity mutants in fast-growers.
• At low density, fecundity mutants generally have a much larger selective advantage than survival mutants because their benefit is scaled by the carrying capacity of the resource ($K_R$).

5. Significance and Implications

• Beyond $r$ and $K$: The study argues that measuring only the net intrinsic growth rate ($r$) of a declining population is insufficient for conservation planning. Managers must identify which vital rates (birth vs. death) are being suppressed, as this determines the likelihood of evolutionary rescue.
• Conservation Strategy: For fast-growing species facing stressors that reduce resource acquisition (e.g., pollution affecting feeding), rescue is more likely than if the stressor increases mortality. Conversely, for slow-growing species, stressors increasing mortality might paradoxically offer a better chance for rescue via increased mutational supply.
• Empirical Guidance: The authors urge empiricists to distinguish between bacteriostatic (birth-reducing) and bacteriocidal (death-increasing) stressors in antibiotic resistance studies, and to characterize the specific ecological mechanisms of environmental stressors in wildlife.
• Theoretical Advancement: By integrating explicit consumer-resource dynamics with evolutionary rescue theory, the paper bridges the gap between simple population models and complex eco-evolutionary realities, highlighting the critical role of density-dependence and life-history trade-offs.