Â鶹ÒùÔº

Populations live in rapidly changing environments—droughts come and go, food sources change, human activities reshape habitats. For scientists, this raises a fundamental puzzle: How do populations maintain the genetic diversity needed to survive future challenges when natural selection should eliminate variants that aren't useful for long periods?

Researchers at Stanford have addressed this puzzle by tracking the evolution of fruit fly populations in an outdoor orchard where they controlled pesticide exposure over time, and paired experiments with mathematical modeling. Their new paper, Sept. 15 in Nature Ecology & Evolution, offers the first direct evidence to support the theory of "dominance reversal" in a changing environment over time.

Classically, we think of genetic variants (alleles) as strictly dominant or recessive—dominant alleles dominate expression of traits and recessive alleles are only outwardly expressed if there isn't a dominant allele around. In the case of dominance reversal, the same genetic variant is dominant when helpful (providing the flies with resistance in pesticide-rich environments) but recessive when harmful (reducing fitness in pesticide-free environments).

"Let's say you haven't used pesticides for 20 years. The moment you add pesticides again, they'll rapidly respond and resist them," said senior author Dmitri Petrov, professor of biology in the School of Humanities and Sciences (H&S). "It's like the flies have a hidden shield. When they don't need it, it's not in their way. But it's ready as soon as they are threatened."

The researchers suggest this mechanism may be widespread in nature, helping maintain genetic diversity for different environmental challenges that change over time. "What we're seeing could be a general mechanism for populations to hold on to genetic variants they might need for future environmental shifts," said Marianthi Karageorgi, who is the lead author and a research scientist in the Petrov Lab.

"For example, synthetic insecticides are often analogs of plant chemical defenses," Karageorgi added. "So, this mechanism could have been operating in nature for millions of years—helping insects maintain resistance to chemical defenses that vary seasonally with host plant availability."

Evolution in an experimental orchard

Since the 1950s, population geneticists have proposed that dominance reversal could help maintain genetic variation in changing environments over time, but until now there was no way to test if it really happens in nature. The researchers' findings relied on a combination of surveys, lab experiments, field experiments, and mathematical models.

Before experiments began, the researchers analyzed genetic surveys of flies across the world in different environments, including organic farms. Then, using flies bred by collaborator and co-senior author Paul Schmidt at the University of Pennsylvania, they ran lab experiments to assess how the different genetic variations affect the fitness of the flies with and without pesticide exposure.

This work confirmed that pesticide-resistance alleles persist at intermediate frequencies over space and time. It also showed that, when the pesticide-resistant allele is dominant without the presence of pesticides, it negatively affects survival and reproduction. All of this supported the possibility of dominance reversal.

To test this hypothesis outside the lab, the team used experimental evolution in an outdoor orchard developed by Schmidt. In this setup, large fruit fly populations evolved from early summer to late fall under near-natural conditions, in a large outdoor enclosure with a single peach tree to provide shade.

One set of cages was exposed to a pesticide pulse mimicking seasonal insecticide use, while another set remained untreated. Every two generations, the researchers sampled flies from each cage, tracking both pesticide resistance and genome-wide frequencies of gene variants in real time.

In treated cages, populations rapidly responded to the pesticide pulse: Resistance and the associated resistance allele rose sharply with pesticide use and then gradually declined once exposure stopped. But the untreated cages revealed a surprising result.

"When we got results about the untreated cages, we saw that over an extended period of time, both resistant and non-resistant genetic variants were maintained, which was puzzling," said Karageorgi. "If there is a cost associated with resistance, why doesn't resistance drop over time, and why don't the resistance alleles drop?"

Mathematical modeling of allele frequencies in the treated and untreated cages confirmed that dominance reversal was at work. Resistance alleles acted dominant when beneficial in the presence of pesticides but recessive when costly in their absence, allowing the alleles to persist even without pesticide pressure.

In other words, the alleles are not permanently dominant or recessive in terms of their effects on fitness of the individuals but can function as either, depending on the environment. This flexibility allows the pesticide-resistant alleles to quickly provide high levels of resistance when needed, yet hide from natural selection when their presence would be harmful.

Like an earthquake

The researchers also looked beyond the exact locations on the chromosome where the pesticide-resistance alleles reside. It's known that oftentimes evolutionary changes in one place on a chromosome can cause something of a ripple effect, known as a selective sweep, because alleles at different loci on the same chromosome are physically linked.

"When we applied pesticides, we didn't just change allele frequencies at the resistance locus—we affected loci all across the chromosome, which danced to the pesticide pulse, increasing and then decreasing in frequency," said Karageorgi.

"The effects of this reversal are global and very short-lived," said Petrov. "It's a little bit like an earthquake—as if buildings fell in Chicago and we feel the shaking here, and then suddenly in both places it's calm again."

While these non-local effects are expected from theory, the idea of dominance reversal facilitating these effects raises foundational questions about how strong selective natural and anthropogenic pressures impact genomic diversity in changing environments over time. It suggests they might, in some cases, even set the levels of genetic diversity in natural populations.

"This field is trying to understand what forces are involved in evolution, how you measure them, and how much of an effect they have. But often these forces are hidden from us," said Petrov. "So, the big question for us continues to be: How do we wrestle that knowledge from recalcitrant nature?"