Biologists at the Montana Department of Fish, Wildlife and Parks may never know how a controlled poisoning of Cherry Creek got out of hand and accidentally killed hundreds of trout last week, an official overseeing the project said Friday.

Still, FWP Region 3 fisheries manager Travis Horton said the agency intends to continue treating the upper portion of the creek with rotenone, the chemical that killed upwards of 1,000 fish last Friday.

And, Horton said, the affected stretch of Cherry Creek should be recovered by next year as trout migrate up the stream from the Madison River to spawn.

Since 2004, FWP has been putting rotenone in the upper 60 miles of Cherry Creek in an effort to remove non-native species from the waters. After stretches of water are treated with rotenone, they are stocked with westslope cutthroat trout, the only trout species native to the area, which have been losing territory to introduced species like rainbow and brook trout.

Last week's rotenone treatment was the second to last planned in the project. But it overran its mark, killing fish within a few miles of where the creek flows into the Madison River, a stretch of water that was supposed to be left unscathed by the controversial project.

Officials investigated whether a bottle of rotenone was mislabeled, causing workers to put too much of the chemical in the water, or whether the chemical got into some groundwater and re-emerged in lower portions of the creek.

Tests have shown the bottles were correctly labeled.

Horton said he will "never know" exactly how the rotenone traveled farther than intended.

"The main point is that the rotenone survived over 13 miles and over 13 hours as the stream doubled in size. It just doesn't make sense that it could do that," he said.

But while the accident remains mysterious, FWP plans to proceed with the final rotenone treatment, Horton said. The agency plans to take extra precautions, and will be treating a different stretch of the river, which he expects will ward off a repeat incident. Rotenone is commonly used across the west, and Horton has said it is extremely rare for a mishap like the one that happened last week.

"We're 100 percent confident we can stop the rotenone as we've done hundreds of other times," he said.

Since its inception, the project has drawn strong support and opposition. Adding fuel to the flames was the fact that much of the project took place on Ted Turner's land and that Turner helped fund the $500,000 effort.

Following news of the accidental poisoning, many anglers renewed their critique that it was an unnecessary and dangerous plan.

"It is beyond sickening to me," was how John Trayser, of Bozeman, described it earlier this week.

But others said the effects will be short-term and expressed relief that the poison doesn't seem to have reached either the Madison River or the most popular portion of Cherry Creek.

"I don't think there are any villains in the deal," said Kris Kumlien, general manager of Montana Troutfitters in Bozeman.

Tests done this week show that fish are alive - albeit in heavily depleted numbers - four and a half miles downstream of where the poison was supposed to stop, Horton said.

Tests of the water a half-mile from where Cherry Creek feeds into the Madison show a normal fish population there, suggesting the rotenone did not get that far down the creek.

The final rotenone treatment is planned for early September, Horton said.

Daniel Person can be reached at dperson@dailychronicle.com or 582-2665.