The lead author is C. Fred T. Andrus, a postdoctoral associate at UGA. Co-authors include Douglas E. Crowe, Elizabeth J. Reitz and Christopher S. Romanek of UGA and Daniel H. Sandweiss of UMaine. Romanek also works at the Savannah River Ecology Laboratory in Aiken, S. C. The paper presents the results of chemical analyses of fish bones known as otoliths from a species of sea catfish (Galeichthys peruvianus) living along the Peruvian coast.
"Our data strengthen the argument that El Niño, as we know it, began relatively recently - since 5000 years ago," said Andrus. (El Niño is, in large part, warming of the ocean waters off Peru at irregular intervals.) "This is more evidence that climate change is the norm, and climate stability is the exception in the earth's history, even in relatively recent times. Given the enormous global impact of El Niño, it's important to understand that climate is a naturally variable system, and that just six thousand years ago El Niño was less frequent."
Andrus said that otoliths from sea catfish are a good candidate for such analysis because the fish do not migrate. They tend to stay near shore in relatively small areas away from river mouths where fresh water can complicate the oxygen results.
"Otoliths act like miniature temperature recording devices," said Crowe. "Throughout the life of a fish, they grow concentrically larger, and the ratio of oxygen isotopes in each individual growth band allows us to determine the temperature of the water at that time. By looking at the entire otolith, we can reconstruct the water temperature history throughout the life of the fish, from season to season and year to year.
In 1996, Sandweiss, Reitz and three other scientists co-authored a paper in Science presenting new evidence that conditions indicating the onset of El Niño about 5,000 years ago. Their conclusions were based on the identification of mollusk and fish remains from archeological sites.
Some critics of that paper suggested that the mollusks might have lived in warm water embayments that would not accurately reflect temperatures in the open ocean.
"The evidence in this paper, based on fish that live off shore, strongly supports the conclusions that we reached in 1996," says Sandweiss.
The current paper in Science is allied to two other papers by Andrus and Crowe, now in press at the Journal of Archaeological Sciences and the journal Paleoceanography.
The JAS paper showed that that cooking and subsequent burial for 5,000 to 8,000 years didn't alter the oxygen isotope signatures. The research to be published soon in Paleoceanography establishes the use of otoliths as proxies for temperature-using modern fish to show the technique works. The new Science paper in a sense brings the research together.
Andrus has participated in archeological activities led by Sandweiss in Peru. He performed analyses of oxygen isotopes on otoliths from modern fish as well as seven sea catfish otoliths collected from the Ostra site in 1991 and five from the Siches site in 1995. Siches is located in far northern Peru near the border with Ecuador, whereas Ostra is located about midway between Siches and Lima, the Peruvian capital.
The results at Ostra showed that, although winter water temperatures were similar to those of today, summer temperatures prior to 5,000 years ago were much warmer than they are today. Thus, the average annual temperature was about three degrees Celsius warmer than present. At Siches, although there was little seasonal variation, temperatures were also about three to four degrees warmer than they are today.
"A change in El Niño frequency and the related increase in upwelling about 5000 years ago may be related to changes in fishing resources and increased cultural complexity," said Reitz.
These results are also consistent with a theory posed by Sandweiss and his colleagues in a paper in Geology in 2001. That paper suggested that cultural changes such as the construction of cities and large monuments may have been related to the onset of El Niño 5000 years ago and later to changes in its frequency and intensity.
"Our research shows that the current El Niño cycle is significantly different from what it was five to eight thousand years ago," said Andrus. "Our hope is that our data can be used to build accurate predictive models of future El Niño events.
The research was funded by the National Science Foundation, the Geological Society of America and the Explorer's Club International.
WRITERS: Nick Houtman, Univ. Maine, 207-581-3777, Houtman@maine.edu
Phil Williams, 706/542-8501, phil@franklin.uga.edu
CONTACT: C. Fred T. Andrus, 706/542-0127, andrus@gly.uga.edu
Daniel H. Sandweiss, 207/581-1889, dan.sandweiss@umit.maine.edu.
Douglas Crowe, 706, 542-2382, crowe@gly.uga.edu
Journal
Science