Penn scientists tracked fruit flies in outdoor cages for four months; seasonal shifts were linked to genetic changes across more than 60% of the genome

Imagine a group of fruit flies living in tent-like shelters in Philadelphia with biologists keen to test whether evolution, normally thought of as a slow, gradual process, could actually happen fast enough to track the changing seasons. According to a 2022 Penn Today report on the University of Pennsylvania experiment, researchers observed the flies for four months, and over 60% of the genome changed as a result. The study’s findings were published in Science in 2022 under the title “Direct observation of adaptive tracking on ecological time scales in Drosophila,” and raise a significant question: how quickly can an organism adapt to changes in its environment?

An orchard with walls

The researchers created an experimental orchard within Pennovation Works not far from the centre of the Penn campus. As reported by Penn Today, the fruit flies were put in a shelter within the orchard where they were exposed to actual weather conditions, including cold, heat, and rain. The shelter contained no external fruit flies, meaning no new flies would enter or exit. Previous studies had been undertaken at other orchards in the Philadelphia area. However, this study managed to avoid any new flies entering the population, as previous studies had identified differences between the early season and late fall populations but couldn’t rule out the possibility of new fruit flies entering the population.

In July 2014, 1000 Drosophila melanogaster fruit flies were introduced into 10 different shelters. During this time, the populations grew to around 100,000. Every month, 2,500 eggs and fruit flies were taken from each shelter and bred individually. The researchers looked at six physical characteristics, including reproductive success and cold resistance. From each shelter, 100 fruit flies were chosen randomly and their genomes sequenced. The flies were tracked until the first hard frost on 7 November 2014, about four months after the cages were founded. Because fly generations overlapped rather than occurring in strict, evenly spaced cycles, the researchers estimated this period corresponded to roughly one to four generations between each sampling point, or about 10 to 12 generations in total: an approximation based on generation-time estimates rather than a literal fixed cadence.

Over 60% Evolved

Whilst large numbers make great headlines, it should be noted that whilst over 60% of the genome did evolve directly or indirectly, this does not necessarily mean that more than half was subject to selection. Some areas of DNA were simply caught up in the changes to sections close by. This is known as the genetic draft.

Throughout the study, adaptation changed direction several times. It was surprising that a trait was evolving for a few weeks, and then the next month it was going the other way, according to co-lead author of the Science paper Seth Rudman. It was seen that there was a frequent change in the direction and reason for adaptation, which is consistent with intense and rapidly changing selection. Previous studies looking at just two time points may have underestimated the speed of adaptation because of this back-and-forth pattern.

Not the first clue, but a closer look

Previous research had indicated this. A 2014 study titled “Genomic Evidence of Rapid and Stable Adaptive Oscillations over Seasonal Time Scales in Drosophila,” was undertaken by Alan Bergland et al., and published in the journal PLOS Genetics. It looked at fruit flies collected over 3 years, from 2009 to 2011. Fruit flies were collected in spring and fall within an orchard in Pennsylvania. Sequencing revealed that the model was consistent with rapid climatic changes over generations, helping to maintain genetic diversity within the population for fitness-related traits. Paul Schmidt, Senior Author of the 2022 paper, was one of the authors for this earlier paper.

Why should this matter

Whilst this study was looking at fruit flies, the researchers do state that the findings are relevant to longer-lived species such as humans. Schmidt argues that although these processes may be common, they can be harder to measure. He mentions possible selective pressures such as climate change, agriculture, and using milk as a food source. The direct observation of rapid, genome-wide adaptive tracking in this experiment is established, peer-reviewed data, not a hypothesis. What remains a hypothesis is how far this pattern extends to longer-lived, slower-reproducing species like humans, where such shifts would be far harder to measure.