These Caterpillars Go Bananas for Fruity Smells—and So Do Their Offspring

Bicyclus anynana caterpillars feed on maize leaves in the lab of Antónia Monteiro, Ph.D., at the National University of Singapore. Researchers there have found that B. anynana caterpillars learn to prefer the smell of banana- and mango- coated leaves and pass that preference on to their offspring, a rare example of inheritance of a learned trait. (Photo by Antónia Monteiro, Ph.D., and William Piel)
By Crystal Ho
Remember the time you ate something bad and had an upset stomach for days? Perhaps you shudder when you smell the same food now.
Similarly, mice that receive electric shocks while smelling fruits jump when they smell fruits, even without being shocked. But what’s amazing is their offspring exhibit similar behaviour. Now, imagine your kids avoiding the same food even without you ever telling them about your experience.
Roundworms, meanwhile, gravitate toward the same food smells over 40 generations.
And now, yet another species has been found to have similarly inherited a learned trait. As it turns out, Lamarck may not have been entirely wrong.
A team of researchers from National University of Singapore (NUS) and Yale-NUS College has found that the butterfly species Bicyclus anynana passes down the preference for leaves that smell like banana to its next generation, according to a study published in October 2019 in the journal Evolution.
Lamarck May Not Have Been Entirely Wrong
The study initially began as a pilot project by Florence Monteiro Piel, daughter of Antónia Monteiro, Ph.D., associate professor at NUS and Yale-NUS and corresponding author of the study.
“This study has shown how exposure to certain environments can impact how offspring behave toward that environment,” says Monteiro, “showing support for inheritance of acquired traits, the phrase promoted by Lamarck, which is not really accepted by many people.”
Although inheritance of learned behaviours has been shown in mice and roundworms, exploring this possibility in other animal models can shed some light on why and how it happens and the adaptive benefits of the inheritability of such traits.
Not a Picky Eater
The butterfly species Bicyclus anynana prefers certain food plants but is not particularly choosy about where it lays its eggs if its preferred host plants are unavailable, making it a good choice to study food preferences for a plant-eating insect.
The flexible food preferences show the species’ potential to evolve novel food preferences, perhaps due to exposure to new food plants or via learning new food odors.
While B. anynana was shown to recognize novel pheromone odors and pass down these abilities to the next generation, research on inheritance of learned food preference has never been done.

Researchers in the lab of Antónia Monteiro, Ph.D., at the National University of Singapore use Bicyclus anynana and other Bicyclus butterfly species as model organisms to study various elements of evolutionary developmental biology. (Photo by Antónia Monteiro, Ph.D., and William Piel)
Scented Food Choice Assays
The researchers tested whether caterpillars preferred certain smells by exposing them to new, artificial odors—banana, mango, almond, and coffee—on their regular lab food, maize leaves. Ethanol-coated leaves were used as a control.
Despite initially avoiding banana-coated leaves, the caterpillars quickly learned to prefer banana- and mango-coated leaves when feeding on them, but they did not learn a preference for coffee- or almond-coated leaves.
Only caterpillars that fed on banana- and mango-coated leaves were used to test whether their new smell preferences, as larvae, affected egg-laying choice on banana- and mango-coated leaves.
Testing for Male Germline Effect on Heritability of Learned Traits
These caterpillars were allowed to develop fully into butterflies. Within each group of the mango-preferring and banana-preferring caterpillars, the butterflies were mated with each other and some male butterflies were mated to female butterflies exposed to no odor.
The researchers hoped to see the effect of the male germline on the next generation: “We mated odor-exposed males to wild-type females. The offspring showed a preference towards banana-scented leaves,” says V. Gowri, the study’s first author and a Ph.D. student in the Department of Biological Sciences at NUS.
“In comparison to the egg, the sperm is much smaller—yet it seems that the sperm may hold some information that transmits odor preference,” Gowri says. “The original belief is that the environment cannot affect the germline, but more experiments are proving that the environment can affect the germline as well.”
The diet of the parents, the diet of the father alone, and the odor of the leaf where the eggs were laid all affected the odor preference of the young.
Regardless of egg location or whether just the father or both parents ate banana-coated leaves, offspring of banana odor-preferring parents chose to feed on banana-coated leaves.
Similarly, regardless of whether just the father or both parents ate mango-coated leaves, offspring of mango odor-preferring parents chose to feed on mango-coated leaves more often than offspring of parent who never fed on these mango-coated odors.
“The most exciting part is that these results can be seen from one generation alone,” Gowri says. “I am now exploring whether these preferences fade after a few generations of larvae returning to their normal diet and whether they may persist longer-term if I expose multiple generations to these odors.”
Interestingly, the caterpillars’ food preferences did not affect egg-laying choice when grown, which contradicts the natal habitat preference induction hypothesis—adults lay eggs on the plants they consume when young.
Why Do B. anynana Caterpillars Prefer Fruity Smells?
While B. anynana caterpillars do not feed on fruit, B. anynana adults do. The odor receptor genes detecting fruit odors may be expressed before the adult stage—an innate genetic sensitivity that may have caused caterpillars to learn to prefer the fruity odors as opposed to coffee and almond odors.
Different starting concentrations of coffee and almond odors due to manufacturing conditions by the supplier may also have caused these results—while standard 2 percent odor solutions were used for all treatments, the exact molecular composition and concentration of chemicals may have differed between the coffee, almond, and fruit odors.
“We don’t know the exact mechanism yet—the results could be due to epigenetic marks or other factors in the cytoplasm,” Monteiro says.
More Bananas in the Future for B. anynana
Gowri will be repeating the study with the banana odor itself, as this study used commercially available banana essence, which may have additional flavoring and other chemicals including ethylene glycol. Gowri is also exploring the effect of pear and strawberry odors.
The separation of the odor from the food will help the researchers understand whether the results observed in this study was due to associative learning—the caterpillars linking the new odors to the food reward and searching for the odor to get food, or odor imprinting—whether the unrewarded odor exposure leads to preference of the odor.
“We are trying to test this over five generations to see if the learned food preferences can indeed be passed down over more than one generation,” Monteiro says.
Read More
Evolution
Crystal Ho is a recent science communication graduate from the Australian National University and National University of Singapore. She has previously written for Eco-Business. Twitter: @crystalhzf. Email: crystalho.com@gmail.com.