Mangel Pflugeisen compared the amounts two key elements, strontium and calcium, at the core of each otolith -- the part of the bone that grew just after the fish hatched.
Bedrock beneath the Sandusky contains more strontium than bedrock beneath the Maumee. Yet both sites contain roughly equal amounts of calcium. So, she reasoned, fish hatched in the Sandusky should have absorbed much more strontium from the water during their early life, and stored much higher concentrations of strontium in their otoliths from that time.
Once she isolated a unique chemical signature for the two rivers, she used a statistical technique known as Bayesian hierarchical mixture modeling to analyze the data. The task was difficult, because the model had to account for the ratio of elements in the otoliths and in the water of both rivers at the same time.
The analysis showed that about 92 percent of the fish that were caught at the Maumee had also hatched in the Maumee, with a very small percentage having originated at the Sandusky.
At the Sandusky, however, only about 66 percent of the fish that were caught were returners -- that is, had been hatched in the Sandusky -- and about 30 percent originated at the Maumee.
The results confirmed what Marschall already suspected: the Maumee fish were straying to the Sandusky to spawn, but not vice versa.
"Dr. Marschall already had strong reason to believe that's what was happening, so I was not surprised by the results," Mangel Pflugeisen said. "But it was really neat to be able to back up her strong, ecologically-based sense of what was going on with a statistical analysis that yielded the same general trend, while also giving numerical
|Contact: Catherine Calder|
Ohio State University