The mere¹ presence of a predator² causes enough stress to kill a dragonfly, even when the predator cannot actually get at its prey³ to eat it, say biologists at the University of Toronto. "How prey respond to the fear of being eaten is an important topic in ecology, and we've learned a great deal about how these responses affect predator and prey interactions," says Professor Locke Rowe, chair of the Department of Ecology and Evolutionary⁴ Biology (EEB) and co-principal investigator⁵ of a study conducted at U of T's Koffler Scientific Reserve.

"As we learn more about how animals respond to stressful conditions -- whether it's the presence of predators⁶ or stresses from other natural or human-caused disruptions -- we increasingly find that stress brings a greater risk of death, presumably（大概） from things such as infections that normally wouldn't kill them," says Rowe.

Shannon McCauley, a post-doctoral fellow, and EEB professors Marie-Josée Fortin and Rowe raised juvenile⁷ dragonfly larvae⁸ (Leucorrhinia intacta) in aquariums⁹（水族馆） or tanks along with their predators. The two groups were separated so that while the dragonflies could see and smell their predators, the predators could not actually eat them.

"What we found was unexpected -- more of the dragonflies died when predators shared their habitat," says Rowe. Larvae exposed to predatory（掠夺的，食肉的） fish or aquatic¹⁰ insects had survival rates 2.5 to 4.3 times less than those not exposed.

In a second experiment, 11 per cent of larvae exposed to fish died as they attempted to metamorphose（变质，变形） into their adult stage, compared to only two per cent of those growing in a fish-free environment. "We allowed the juvenile dragonflies to go through metamorphosis to become adult dragonflies, and found those that had grown up around predators were more likely to fail to complete metamorphosis successfully, more often dying in the process," says Rowe.

The scientists suggest that their findings could apply to all organisms facing any amount of stress, and that the experiment could be used as a model for future studies on the lethal¹¹ effects of stress.

The research is described in a paper titled "The deadly effects of 'nonlethal' predators," published in Ecology and highlighted in Nature this week. It was supported by grants to Fortin and Rowe from the Canada Research Chairs program and the Natural Sciences and Engineering Research Council of Canada, and a post-doctoral fellowship awarded to McCauley.