- News Home
5 December 2013 11:26 am ,
Vol. 342 ,
- 5 December 2013 11:26 am , Vol. 342 , #6163
- About Us
Plant Experiments Underestimate Climate Change Effects
2 May 2012 1:00 pm
The past mild winter in the United States brought spring flowers before April's showers, fueling concerns about how climate change might upset the synchrony between plants and the organisms they interact with. A new study suggests that the mismatch may be worse than previously thought. Experiments designed to test the effects of warming failed to predict just how much earlier plants will develop.
When plants leaf or bloom—part of a field called phenology—"provides one of the most sensitive indicators of how ecosystems are responding to climate change," says Elsa Cleland, an ecologist at the University of California, San Diego. Typically, ecologists assess plant phenology in one of two ways. In observational studies, they monitor each plant in its natural setting as it starts to grow, taking note year after year of the temperature and other environmental conditions. In warming studies, researchers take matters into their own hands, growing plants outside at specific temperatures by using heaters and then noting how phenology differs between those plants and others growing naturally. Researchers considered both approaches equivalent, but Cleland and her colleagues compared results from the two types of studies to make sure.
While in Cleland's lab, Elizabeth Wolkovich, now an ecologist at the University of British Columbia, Vancouver, and her colleagues combed the literature and gathered data sets, coming up with 50 relevant studies spanning four continents. The records covered 1643 plant species. The researchers determined how much sooner each species flowered and leafed per degree Celsius of temperature rise. For 36 species, they had data from observational studies and experimental warming studies.
Plants leafed four times earlier and flowered eight times earlier in the observational studies than in the warming studies, Wolkovich, Cleland, and their colleagues report online today in Nature. Warming experiments showed advances in flowering or leafing time of less than 1 day to 1.6 days per degree rise. But in the observational studies, the plants advanced 5 to 6 days per degree. "We were surprised at how different the observations and the experiments really were," says co-author Benjamin Cook, a climatologist at the NASA Goddard Institute for Space Studies in New York City.
The disparity could pose difficulties for researchers who use warming data to build climate models that also predict how ecosystems will change, says Johanna Schmitt, a plant ecologist at Brown University who was not involved with the work. There are relatively few observational data sets for modelers to use, so they often depend on results from warming experiments. But the new results could mean that ecosystems will change faster than the models would suggest, Schmitt says.
Researchers can't explain the difference between the warming and the observation studies. It could be because warming experiments are conducted on a small scale and for short periods of time and thus may not reflect the complete set of conditions plants confront over the long term, such as increases in carbon dioxide. And depending on the heat source and the way temperature is measured, the plant may not warm as much as expected. "Given the results, we should be more careful about how we design and implement these experiments," says Wolkovich. Along those lines, more recent warming experiments are measuring the temperature more accurately and trying to make sure other conditions, such as soil moisture, are the same between warmed and unwarmed plots.
The discovery that warming experiments are so far off worries some researchers. "That a trait like plant phenology, which is frequently thought to have a straightforward relationship with temperature, could be so mispredicted by warming experiments is a bit sobering," says Abraham Miller-Rushing, an ecologist with the National Park Service in Winter Harbor, Maine. He wants to be able to predict phenology accurately to make better resource management decisions. For example, the best time to get rid of invasive plants is after they leaf but before they flower, so it's important to know when those two events happen.