A Brown Pelican waits to be cleaned after the Deepwater Horizon spill in 2010. Photo: Kim Hubbard


BP's Sleight of Hand

A BP-authored report claiming that the Gulf has recovered is inaccurate and insulting—here’s why.

Originally published on the National Audubon Society's website

Nearly five years after the largest accidental marine oil spill in U.S. history, BP is doing its best to convince the public that the 4.9[1] million barrels of oil that leaked into the Gulf of Mexico have done no lasting damage. That’s the message found in The Gulf of Mexico Environmental Recovery and Restoration report the company released three weeks ago.

Based on what the report does present, it’s obvious that BP chose to ignore recent research that shows the ongoing impact of the oil spill, particularly on birds (pp 15-17 of report). That BP ignores peer-reviewed science is nothing new—they have been denying culpability and the validity of peer-reviewed, published research since the spill itself. Meanwhile, their own conclusions lack peer-review and they have not released their methodology. Because of this lack of transparency, BP has made claims that are impossible to refute or verify.

One of the government agencies responsible for monitoring the Gulf health following the oil spill—the NOAA Damage Assessment, Remediation, and Restoration Program—is expected to hold itself to a high standard of independent scientific peer-review as it completes its Natural Resource Damage Assessment (NRDA). While we await the NRDA studies to be completed and released, BP has flaunted such professional standards in its allusions to NRDA findings that cannot be verified at this stage. The trustees of NRDA themselves called BP’s recent report “inappropriate as well as premature.”

Along with others in the scientific community, the National Audubon Society rejects BP’s claims. While their report leaves much to criticize, here are its five main flaws:

  • Their conclusions are not credible given lack of peer-review, questionable methodology and premature conclusions. 

​They claim “search teams likely found 97 percent of large birds and 78 percent of small birds.” These numbers are inconceivable to any researcher or birder who has looked for birds on a beach. It stands to reason that a detection rate of 97 percent as claimed by BP is unlikely, even for living and mobile birds; their estimate seems grossly optimistic compared to detection rates published widely in the scientific literature. In addition, BP compared their estimate of searcher efficiency for carcasses placed on beaches to the searcher efficiency across sandy beach, rocky beach, and marsh habitats. Detection rates are known to vary widely based on factors such as carcass coloration and habitat, and the actual searches in the Deepwater Horizon disaster occurred across marsh, mangrove, rocky beach, and sandy beach habitats.  Attacking a single parameter from a complex study, and doing so using an apples-to-oranges comparison, is disingenuous and misleading. 

  • The Gulf of Mexico ecosystem has not returned to a pre-oil spill “baseline condition”. 

First of all, the oil is not gone. Significant deposits (up to 230,000 barrels) of oil have been discovered in the sediment on the Gulf floor. This oil regularly washes ashore during storms, re-oiling marshes and beaches. In October 2014, researchers discovered a 1,250-square-mile “bathtub ring” of oil on the Gulf floor–equivalent to the size of Rhode Island. Indeed, as the BP report was released a 22,000-pound tar mat was being removed from East Grand Terre Island in Barataria Bay, LA.

BP’s conclusions are incredibly premature. With lagged effects, trophic cascades, food web effects, and repeated re-oiling of the coast, we could be seeing surprises and environmental effects for years to come on birds and on the entire ecosystem.

Erosion is forever. Land and crucial habitats have disappeared. It’s impossible to return to a “baseline” when the land that composed that baseline and is crucial to coastal bird populations has disappeared. Studies show that marsh that was heavily oiled was undercut and eroded back, creating ragged marsh edges that were unoiled or lightly oiled. The scientists have then seen a wave of erosion of these unoiled or lightly oiled marsh, as waves reshape the ragged marsh edge to create a new, smoother edge. Not only has much of the heavily oiled marsh eroded, but marsh that was less damaged has followed.

  • Damage to birds from the BP oil spill is evident and has been well documented, contrary to what BP reports.

In the 95 days following the oil spill, conservative estimates are that hundreds of thousands to more than a million birds died.

This acute damage does not take into account the molecular, cellular, reproductive, and developmental damage that may accumulate in birds for years to come. That is significant, it is long-term, and it matters.

Birds exposed to oil during their early development may accumulate damage that causes failed reproduction. Additionally, many long-lived species like pelicans, terns, and egrets do not breed until their second, third, or fourth year of life, meaning the young exposed to oil early in their lives hadn’t started breeding in 2011, making BP’s conclusions based on 2011 information premature.

  • More studies are needed to understand the ongoing and long-term effects on bird populations.

In other species, lagged and trophic level effects have resulted in increasing effects from oil being detected over time since the disaster. Very few studies on bird productivity have been published that looked beyond the 2010-2011 nesting season. More studies are needed to draw broad conclusions.

Scientifically, the most reliable way to study productivity is to compare productivity of birds exposed to oil and not exposed to oil within the same season. The only study of which we are aware that has used this reliable methodology is a study of the Seaside Sparrow, which showed reduced nest success for sparrows on oiled sites in 2012 and 2013.

  • BP’s use of Audubon’s Christmas Bird Count (CBC) data is flawed and their summaries of other data are incomplete and misleading. 

BP misused CBC data to obscure potential impacts. It’s likely they used data from well outside the impacted areas. Particularly, Sandwich Terns and Northern Gannets are not counted by CBC participants 10-100 times more than Piping Plovers in the impacted areas. The numbers they used for Red Knots appear to include the entire rufa subspecies, which mainly winters in the Tierra del Fuego region in southern Argentina and the Maranhão region of northeastern Brazil. It is difficult to tell how CBC data were used, but it is clear from the relative abundance of some species that data from outside the Gulf were included in analyses. Using broader data from other regions could mask any local and regional changes in populations due to the oil spill. Without describing their methods, it is apparent that BP’s report represents a misuse of Audubon’s publicly available data.

BP reported oiling rates from long after the oil well was capped, resulting in oiling rates that appear to be very low: In its report, BP cites data saying “researchers saw no visible oil in 99 percent of the roughly 500,000 live bird observations from May 2010 to March 2011.” A study following the Deepwater Horizon spill showed that 8.6 percent of captured shorebirds showed evidence of oiling, indicating that BP’s observational evidence may have underestimated exposure by as much as an order of magnitude.

BP selected a study that appeared to show no effects on birds to highlight a success that is likely just a lack of use of the oiled region:  In BP’s reference to the Northern Gannet publication, researchers in this paper used light-level geolocators to determine wintering locations of Northern Gannets. Digging into the details of the paper, it is clear that only a small sample of the researcher’s birds wintered in the Gulf of Mexico (8 birds in their study) and that the Gulf birds foraged in areas primarily >100 miles away from the spill site, i.e., well outside of area of the most devastating impacts. Thus it may be no surprise that these few birds did not show elevated polycyclic aromatic hydrocarbon (harmful chemicals from oil) levels as they were not in the spill zone. The U.S. Fish and Wildlife Service’s published account of bird deaths following the spill paints a different picture—Northern Gannets were one of the most frequently encountered dead bird after the spill.

BP incorrectly concluded from a published study that researchers did not find evidence that variations in age classes of Brown Pelicans on Louisiana barrier islands were related to the oil spill. However, the paper specifically states “…additional research is required to evaluate potential long-term population trends.” Also, “Care should be taken in interpretation of our result as a lack of effect [of oil] on demography because there are combinations of spill-related reductions in fecundity and age-specific mortality that could result in no changes in age structure.” One scenario that’s easy to imagine is that all age classes were reduced similarly. In short, like the authors state, there are many possible scenarios that could influence their findings.

Despite BP’s claims, we may not know the full impact of the oil spill for decades to come. Even more, we will not achieve lasting, meaningful restoration in the Gulf Coast until BP quits stalling, quits the legal and publicity campaign and actually funds full restoration.

[1] BP did collect or burn off some of this oil before or soon after it entered the environment, convincing the Federal District Court in New Orleans to rule on Clean Water Act penalties based on 3.2 million barrels of oil. But scientific analysis estimates total spillage was 4.9 million barrels.

How you can help, right now