Ragen rebukes scientist

07/26/2012

A 20-page letter sent by Marine Mammal Commission head Tim Ragen blasting a report by Marshall scientist Corey Goodman was released in the lead-up to the National Research Council meeting on Drake’s Bay Oyster Company in Irvine.

Dr. Goodman’s report critiqued the 2011 Marine Mammal Commission (MMC) report evaluating the oyster company’s effects on the movements of harbor seals in Drake’s Estero, and data on those effects gathered by scientists at Point Reyes National Seashore. In his letter, Dr. Ragen pointed to “fundamental errors” in some of Dr. Goodman’s interpretations of data, which he said rendered them “incorrect,” “invalid” and “unreliable.” He also argued that Dr. Goodman did not apply the same level of scrutiny to the oyster company’s records as he did to those of the National Park Service (NPS).

He also disagreed with Dr. Goodman’s characterization of the NPS report on seal disturbances, co-authored by Ben Becker, Sarah Allen and David Press, calling it “incomplete, inaccurate and misleading.” He said Dr. Goodman inappropriately cast nuanced scientific issues in black and white, and removed statements by NPS scientists from their context.

“My view of this case has not changed,” Dr. Ragen concluded in the letter. “I continue to believe the Commission’s report summarized the situation accurately. The park service has provided ‘some support for the conclusion that harbor seal habitat-use patterns and mariculture activities in Drake’s Estero are at lease correlated.’ The evidence is not overwhelming, but also cannot be dismissed.”

Dr. Goodman, a member of the National Academy of Sciences who has been volunteering his analysis of scientific data on the oyster company since Marin County Supervisor Steve Kinsey asked him to do so in 2007, has been the most vocal critic of NPS scientists in this debate. He prepared his critique of the MMC report between August 2011 and January 2012 with the help of David Lewis, director and watershed management advisor for the University of California Cooperative Extension.

Among other critiques, Dr. Goodman suggested the exclusion of data from 2003—a year seal numbers in the estero spiked and an elephant seal lowered seal counts at Double Point—as an outlier.

Dr. Goodman and Dr. Lewis sought the advice of statistics faculty at the University of California, Davis and Stanford University before submitting the critique, Dr. Goodman said.

“In particular, we consulted with the Stanford faculty member about our response, and he concurred with our analysis,” Dr. Goodman said. “He completely concurred with the appropriateness of the outlier analysis, and told me so in writing.”

Dr. Goodman continued: “We were not arguing that we could explain all of the movements of the harbor seals, as Ragen’s letter puts forward as a straw man argument. But rather we were arguing that NPS data are too thin and too highly leveraged by a [random] event in 2003 to support the NPS correlation between harbor seals and oyster activity. That was our main thesis, and remains to this day.

“Furthermore, Dr. Ragen accepted disturbance studies that are not credible and contain flawed data, even disturbance records that the [National Academy of Sciences] panel did not accept.”

He said that although he and Dr. Ragen exchanged emails about the possibility of meeting to discuss the critique, the latter never substantively responded until June 17, when he sent the letter.

“No meeting took place. There was never any scientific discussion,” Dr. Goodman said. “I was accused of being disruptive and was told that NPS, [National Parks Conservation Association] and Gordon Bennett told the MMC a meeting to discuss the science would be of little value and they would not participate if I was in the room.”

David Weiman, a lobbyist and consultant who has advised oyster farm owner Kevin Lunny, said that in a June 4 phone conversation, Mr. Ragen indicated that the letter he was planning to send would be private and sent only to Dr. Goodman “to keep it contained.”

But that was not to be.

Though Dr. Ragen confirmed that the communication was intended to be private, he also said he forwarded it to seashore superintendent Cicely Muldoon at her request, after he asked NPS scientists to run analyses in preparation for his response.

The letter, undated, unsigned and not written on letterhead, was emailed to Dr. Goodman two days before the National Research Council’s meeting in Irvine, which Dr. Goodman had already decided not to attend. National Parks Conservation Association Pacific Region Associate Director Neal Desai acquired it through a Freedom of Information Act request ten days later.

On Wednesday, July 11 the letter appeared on the seashore’s website, along with related emails and other documents. Amy Trainer, executive director of the Environmental Action Committee of West Marin, then blasted it to members of the media, including the Light.

The MMC report included counts of harbor seals present in different locations throughout the region in 1982 and 1983, and between 1997 and 2010. In 2003, the number of seals in Drake’s Estero jumped from 968 to 1,158, a leap NPS scientists argue could have been caused by a drop in oyster harvest as the Johnson family’s operations wound down.

Dr. Ragen wrote that the counts, taken during the pupping season, were not a “perfect measure,” but could still be used as a relatively stable index of harbor seal hauling patterns.

“Those are probably the best [index] we have,” he said.

The NPS report’s hypothesis was that when oyster operations decreased in the estero in 2003, the seal population went up—suggesting that increased actions of the farm could be preventing the seals from hauling out there.

“The big question is, do we see this [as related to] hauling patterns, and is this just a co-occurrence, or is there something more to it,” Dr. Ragen said. “The data isn’t sufficient to make a confident conclusion. To understand it better you would have to investigate the cause and effect, and do some adaptive management experiments.”

The presence of an aggressive elephant seal at Double Point, on the southern end of the Point Reyes peninsula, in 2003 may have killed as many as 40 harbor seals, causing many more to flee to safety in the estero—resulting in a higher count that year unrelated to the oyster farm. Dr. Goodman has argued that the elephant seal, and not the oyster farm, was responsible for the population shift.

“I believe everyone agrees that the elephant seal at Double Point had a significant effect on harbor seals at that site in 2003, with recovery occurring over the next few years,” Dr. Ragen admits in his letter, “the Double Point effect on harbor seal hauling patterns in Drake’s Estero is not so clear.”

While the NPS and MMC reports concluded that seal behavior and mariculture in the estero were connected, the data did not show evidence of “harm,”—just displacement. However, groups keen to oust the farm were quick to accept the suggestion as evidence enough.

But conclusive evidence proving harm to the flora or fauna of Drakes Estero has increasingly appeared tenuous.

Still, efforts to discover what negative environmental impact, if any, the oyster farm has on the estero have increased in the run-up to this November, when a 40-year lease on the on-shore facilities expires and the Department of the Interior may choose to renew or terminate the lease. Since Mr. Lunny and his family, neighboring ranchers, took over the lease from the Johnson family in 2005, a vicious debate has raged over whether or not they should be allowed to continue operations.

While some have argued that the intention was always for Drakes Estero, a “potential wilderness” area, to be given full wilderness status—and that this status necessitated the removal of the oyster farm—others have insisted this was not the case. John Burton, the primary author of the Point Reyes Wilderness Act, and others have argued that the intention was always for agricultural operations—and the oyster farm—to continue.

Uncertainty over the legality of terminating the oyster farm’s lease has made it all the more important for its detractors to find scientific evidence of environmental harm. So far, Dr. Goodman says that even what tenuous proof they have—that activity from the oyster farm may decrease the number of seals hauling out in parts of the estero, causing them to go elsewhere—is unfounded, and sought to prove the assertion using NPS data. But Mr. Ragen said he was mistaken.

“To be clear,” Mr. Ragen wrote, “such errors are not trivial in any regard. They are fundamental flaws in your application of the multiple regression model and they invalidate your results.”

Le-Minh Ho, a Ph.D. student in the Yale Graduate School Statistics Department, took a look at the analysis, and expressed similar concerns.

“I agree with Ragen’s criticism,” she told the Light in an email. “It is reasonable to think that the attacks drove the seals away from [Double Point] to [Drake’s Estero]—and this may indeed have happened—but even if it did happen, [Dr. Goodman] does not have enough evidence to conclude that it was the main reason behind the 2003 counts…[He] uses off-the-shelf statistical methods to study seal hauling patterns. As [Mr. Ragen] points out, though, he makes elementary mistakes in his modeling assumptions. The mistakes mean that Dr. Goodman’s conclusions are unreliable.”

Meanwhile, Mr. Lunny has been insisting that the indicator chosen to represent an increase in oyster activity—annual production—is fundamentally flawed, and does not reflect actual activity that could cause the seals to change their hauling patterns.

“The park service didn’t make any effort to ask us,” he said. “We would have been very happy to share how the farm works, but none of the authors of the Becker report have even been with us in a boat, or even talked with us about how we do what we do. It’s all speculation.”

Of the more than 40 harvest areas, Mr. Lunny said only three or four were within a half-mile of where seals haul out. Further, he said that since the NPS began to make claims about the farm’s activities, the Lunnys purchased and installed GPS devices to track the movements of all of their boats.

“Those data are all available,” Mr. Lunny said. “We know how many trips we take, and where our boats go.”

He went on to say that the environment and not the farm was in control of the production level, and that the methods for measuring oyster production—the gallons of shucked oysters that were harvested—did not accurately indicate the farm’s impact or activity level in the estero.

“We set out hundreds of millions of oyster larvae per year,” Mr. Lunny said. Even a one percent change in the survival  of that seed could indicate an increase in production, just because there were fewer mortalities, he explained.

“We’re getting lost in the weeds of statistics when there is something really easy to understand here,” Mr. Lunny said. “The high-low harvest does not have a linear relationship with our efforts in the bay that could translate into additional concerns for seals. There is nothing true in that high-low means more boat trips, or more disturbances or more anything. It’s something they guessed might be true, and they guessed wrong.”

Dr. Ragen also stressed that a more site-specific, collaborate approach was necessary in order to obtain more usable data.

“The more productive thing would be to set up a system where you could investigate these things better,” he said. “To understand the relationship [between seal hauling patterns and the oyster farm], you’d need some adaptive management approach, and to adapt our management [of the estero] as you learn from those experiments.”