The state's $23 billion plan to restore native fish in the Delta would continue killing some of those fish at the same rate in drought years, according to a new pile of documents released Wednesday on the project.

The documents are part of the Bay Delta Conservation Plan, a project led by the California Department of Water Resources. It aims to restore endangered fish and improve reliability of freshwater supplies diverted from the Sacramento-San Joaquin Delta.

The project would divert a portion of the Sacramento River into two tunnels, 35 miles long, built beneath the delta. Diversions would occur at three massive new intakes proposed near Courtland in Sacramento County.

The tunnels would route water to existing state and federal water diversion canals in the south Delta near Tracy. The canals serve 25 million Californians and 3 million acres of farmland from San Jose to San Diego.

The plan has been in the works for seven years to resolve decades of conflict over Delta water supplies. Gov. Jerry Brown last year made it a centerpiece of his own political agenda.

The plan's goal, in part, is to reduce the number of fish killed now by Delta water diversions. Those diversions reverse the natural currents in the estuary, confusing fish such as Delta smelt and juvenile winter-run salmon, both endangered species.

Officials involved in the conservation plan assert that the new diversions would better protect fish because they would feature modern fish screens and would pump only when least harmful to fish.

Among many other findings, the new documents reveal that south Delta pumps near Tracy would still be used to divert water in dry years. Under those conditions, Delta smelt, a threatened native fish, would still be killed at a rate similar to today.

This is just one effect the project would have, and not necessarily the most significant.

Officials said readers must consider the entire range of effects. For instance, fewer smelt would be killed under other conditions when water diversions occur at the new northern intakes. And thousands of acres of restored habitat are expected to help breed more smelt and more food for them, reducing the significance of each smelt death.

"We don't expect to have improvement in every time period, every season, every life-history period for each species," said Mark Cowin, director of the state Department of Water Resources. "But we do intend to have a net effect that provides for meeting the recovery standard."

Fishery advocates said there is uncertainty about whether the new habitat would produce sufficient benefits.

"Proponents of the peripheral tunnels are asking us to just trust them when it comes to keeping our salmon fishery alive," said John McManus, executive director of the Golden Gate Salmon Association. "We're skeptical, to say the least."

The new plumbing is estimated to cost $14 billion, to be paid by farm and urban water ratepayers who rely on Delta water, via rate increases to repay bonds issued by DWR.

The plan also calls for converting many Delta islands into 145,000 acres of restored wildlife habitat. This is estimated to cost $4 billion, which the planners expect all California taxpayers to fund.

The balance of the estimated $23 billion cost consists of long-term operations and maintenance expenses, which recipients of the water would also pay.

The most crucial chapter released Wednesday is called "Effects Analysis," which describes how the diversion tunnels would alter the environment. The chapter is 656 pages long – not counting 18 appendices. Just one appendix, dealing with effects on fish migration, is 406 pages long.

Other unexpected effects are found deep in the details.

An appendix that deals with upstream environmental effects (656 pages) reveals potential negative effects on salmon and steelhead spawning in the American River, which flows through the Sacramento metropolitan area upstream from the proposed tunnels.

The Bay Delta plan would cause small flow changes in the American River, altering the water available to these fish and the temperatures they can withstand. The changes are small, but potentially worrisome, especially in the fall months of drought years when the river can become too warm to sustain fish under present conditions.

"We're not seeing a pattern of any negative adverse change, although there could be some sporadic changes in some fall years," said consultant Jennifer Pierre of ICF International, which conducted much of the analysis under contract with the state. "We think that's something we need to continue to investigate."

Tom Gohring, executive director of the Sacramento Water Forum, questioned this analysis. After skimming the appendix, he noted the analysis is based on a computer model that relies on temperature and flow data on a monthly scale. Which means, he said, that even if the documents show no ill effects, there could be days or weeks within a month that are deadly to fish which never show up in the results.

The Water Forum, a coalition of water agencies and environmental groups, is preparing its own study of American River flows. It relies on a computer model from the U.S. Army Corps of Engineers that reveals detail on a daily or shorter time scale.

Gohring noted that steelhead, in particular, are affected by temperature cumulatively. If they get too warm on a hot day, their bodies don't reset overnight when the river cools down. If the next day is hot, the harmful effects keep adding up. "The fish don't experience temperature effects on a monthly-average basis," he said. "They experience it on an hour-by-hour, minute-by-minute basis."

Contact The Bee's Matt Weiser at (916) 321-1264. Follow him on Twitter @matt_weiser.

© Copyright The Sacramento Bee. All rights reserved.

• Read more articles by Matt Weiser

Order Reprint