In the standard model of evolution, traits are "use it or lose it." Maintaining a complex defense—whether it is a protein that pumps out antibiotics or a behavioral response to a predator—requires energy and resources. If the threat disappears, individuals that stop wasting energy on those defenses should outcompete those that don't. Over time, the population should revert to a "susceptible" or "defenseless" state.
However, a growing body of research across bacteria, insects, and aquatic invertebrates shows that this reversion often fails to happen. Instead, adaptive traits frequently persist for decades after the selective pressure has vanished. This phenomenon, sometimes called the "ghosts of selection past," suggests that the biological cost of resistance is much lower than previously assumed.
The Myth of the Heavy Metabolic Burden
The primary reason we expect traits to disappear is the "fitness cost." If a bacterium carries a plasmid (a small loop of extra DNA) encoding resistance, it must spend energy replicating that DNA and synthesizing the associated proteins. In a world without antibiotics, this bacterium should grow slower than its peers.
Recent genomic studies challenge this. Guoxiu Xiang and colleagues, studying Morganella morganii in China, found that high-risk lineages (Clade 1) can accumulate a massive "toolkit" of mobile genetic elements without sacrificing growth rates Clinical Molecular and Genomic Epidemiology of Morganella morganii in China. These bacteria have evolved to minimize the metabolic burden of their extra genetic load, allowing them to remain "armed" even in non-clinical environments.
Similarly, Z. Y. Li and Guibo Song demonstrated that certain small, non-conjugative plasmids are so stable and "lightweight" that they persist indefinitely across different bacterial lineages Characterization of OXA232-Producing Carbapenem-Resistant Klebsiella pneumoniae. When the cost of a trait approaches zero, there is no evolutionary incentive to get rid of it.
Persistence Across Decades
This stability isn't limited to the rapid generation times of bacteria. It is visible in complex multicellular organisms as well.
- Insecticides: In Europe, the western corn rootworm remains resistant to the insecticide aldrin, even though the chemical has been banned for decades. Marc Ciosi and Blair D. Siegfried argue that this resistance was "pre-imported" from North America and has persisted because it carries no significant fitness cost European populations of Diabrotica virgifera virgifera are resistant to aldrin. A similar pattern was found in Aedes aegypti mosquitoes in Cabo Verde, which maintain dieldrin resistance mutations long after the chemical's removal A next generation targeted amplicon sequencing method to screen for insecticide resistance mutations.
- Antimalarials: In the Democratic Republic of Congo, malaria parasites (Plasmodium falciparum) maintained high levels of chloroquine resistance for over a decade after the drug was retired as a first-line treatment. Molly Deutsch-Feldman and Jeffrey A. Bailey suggest that the lack of a biological penalty prevents the re-emergence of drug-sensitive strains The changing landscape of Plasmodium falciparum drug resistance in the Democratic Republic of Congo.
Ghosts of Selection Past
Why do these traits stay? Beyond low costs, two mechanisms stand out: gene flow and sensory machinery maintenance.
In aquatic ecology, Mirosław Ślusarczyk and Agnieszka Ochocka studied Daphnia magna (water fleas) that produce dormant eggs when they "smell" fish. They found that even Daphnia from habitats that haven't had fish for generations still retain this response Prevalence of kairomone-induced diapause in Daphnia magna. They suggest that either the cost of keeping the "sensory machinery" is negligible, or that high gene flow from fish-heavy habitats keeps the trait circulating in the population.
The "One Health" Reservoir
The persistence of these traits creates a "hidden" reservoir of resistance in the environment. Hongwei Zhou and Ning Dong illustrated this by finding a single fly on a sheep farm carrying four different species of carbapenem-resistant bacteria Isolation of four carbapenem-resistant gram-negative species from a single fly.
If resistance genes do not fade away in the wild, then every insect, animal, and water source becomes a permanent storage locker for "superbugs." This makes the "One Health" approach—tracking resistance across humans, animals, and the environment—not just a recommendation, but a necessity. We cannot simply wait for resistance to "die out" after we stop using a drug; we must account for the fact that once a population learns a survival trick, it rarely forgets it.