Predators, Pools, and the Art of Risk-Taking: Lessons from Mosquito Mothers

Professor Manvi Sharma’s recent study focuses on how animals respond to danger and behave differently, ranging from risk-aversion to boldly facing predators. The study shows that while animal decision-making may appear to be “inconsistent” or “nonsensical”, in reality, their choices can reflect logical and sensible responses to the unpredictable and messy world around them.

Professor Manvi Sharma from Ashoka University and Professor Kavita Isvaran from the Indian Institute of Science worked together to find answers to the question: Why do animals that face danger behave so differently? Why are some animals highly risk-averse, while others face predators boldly?

Professor Sharma and Professor Isvaran explored this puzzle by studying a simple yet revealing system: rock pools on a sheet of rock where the mosquito Aedes vexans lays eggs, and dragonfly nymphs (Bradinopyga geminata) eat the mosquito larvae. By measuring predator numbers across space and time, the researchers explored if predation risk is predictable, i.e., you can tell where danger will be, or unpredictable, i.e., it is a roll of the dice, and how female mosquitoes respond to the nature of this risk when choosing egg-laying sites.

The professors studied the rock-pool ecosystem on the Rishi Valley School campus in Andhra Pradesh, India, mapping 89 rock pools and sampling them monthly for dragonfly nymphs during the breeding season to quantify predation risk associated with pools of different sizes. What they found was surprising! While unpacking these findings, Professor Sharma offers an interesting stock market analogy. She says, “You can think of pools as fund investment portfolios: small-sized pools that rarely have predators, large ones that almost always do – these are the reliable funds, but the medium-sized pools are a draw of luck – these are the extremely unpredictable funds.”

Next, the researchers designed carefully controlled experiments to test how mosquito mothers are investing in egg-laying when faced with these pools.

For field experiments, 31 pools were used: researchers added or removed dragonfly nymphs in a rotating design, then collected eggs laid on cloth strips (ovistrips) placed along pool edges every 24 hours. The team conducted these manipulative trials for two years and used statistical models to tease apart predictable versus unpredictable variation and female responses.

Mosquito mothers consistently avoided large pools (reliable danger), especially when predators were present. What surprised the researchers was that mosquito behaviour changed with the season: when background risk was low, females preferred predator-free pools, but when predators were common everywhere, females stopped discriminating – likely because searching for a rare, safe spot became too costly and time-consuming. There was only weak evidence that females behaved more unpredictably when pools themselves were unpredictable, though medium pools did show more variable egg-laying. Overall, females appear to respond to predictable signals strongly and treat unpredictable signals more flexibly – much like a wise mutual fund manager!

This study shows why it may seem that animal decision-making is “inconsistent” or “nonsensical”, but in fact, their choices can reflect a sensible response to a messy world. By measuring both the landscape of predation risk and prey behaviour in the wild, the paper reveals that predictable danger (big, predator-filled pools) leads to consistent avoidance. In contrast, unpredictable danger can favour flexible or bet-hedging strategies (spreading eggs across multiple environments). These insights matter not only for understanding how traits persist under variable selection but also for practical applications, such as vector-control management, where understanding mosquito oviposition rules could inform control strategies.

Professor Sharma and Professor Isvaran’s study also provides a neat field template for studying how predictable vs unpredictable selection shapes behaviour in other systems.