Tuesday, February 28, 2017

How NOAA and Bad Modeling Invented an “Ocean Acidification” Icon: Part 1 - Sea Butterflies

If you google “ocean acidification,” the first 3 websites presented according to “Google’s truth rankings” are: 1) Wikipedia, 2) NOAA’s PMEL site featuring the graphic cartoon shown below with a dissolving pteropod shell (a sea butterfly) as the icon of ocean acidification, and 3) the Smithsonian’s Ocean Portal site similarly featuring a dissolving sea butterfly shell. However NOAA’s illustration incorrectly implies shells are dissolving near the surface due to invading anthropogenic atmospheric CO2. As will be shown, the depiction would be far more accurate if it was turned upside down, so that the downward arrows point upwards to illustrate shell dissolution happens when old carbon stored at depth is upwelled to the surface. Furthermore the horizontal depiction of extreme dissolution illustrated by their intact (green) sea butterfly shell dissolving into an extremely shriveled shell (red), rarely if ever happens in the ocean’s upper layers. Surface waters are supersaturated regards calcium carbonates. Although upwelling causes some near surface dissolution, dead sea-butterfly shells only experience such extreme dissolution when they sink to depths containing ancient corrosive waters.

NOAA's Upside Down Cartoon of Sea Butterfly SHell Dissolution

As for most organisms, pteropod populations fluctuate over the short term. But research finds no significant long-term trends in pteropod abundance. Nonetheless NOAA’s Nina Bednarsek has been preparing a preliminary report arguing sea butterflies should be listed as endangered and NOAA’s cartoon appears to be an attempt to gain support for her claims. To warrant endangered status, Bednarsek presents a hypothesis that increasing CO2 has reduced critical pteropod habitat by raising the depth of calcium carbonate saturation horizons.  The threshold above which high concentrations of carbonate ions (CO32-) promote abiotic calcium carbonate precipitation but below which favors dissolution, is referred to as the saturation horizon. The horizon is quantified as 1, and higher numbers characterize supersaturated water that favor calcification. As seen below in Fig 10 from Jiang 2016, most of the globe’s surface oceans are supersaturated.

Global calicum carbonate supersaturation from Jiang 2016

Bednarsek assumes anthropogenic carbon is mostly accumulating near the surface based on modeling results. However as detailed in Part 2, all ocean acidification models are deeply flawed based on an incorrect assumption that CO2 enters the ocean and is then transported like an inert tracer. But CO2 is not inert! When CO2 first invades sunlit surface waters, it indeed dissolves into 3 forms of inorganic carbon (DIC) and lowers pH (DIC is discussed in How Gaia and Coral Reefs Regulate Ocean pH). But in contrast to those models, DIC is rapidly assimilated into particulate organic carbon via photosynthesis, which raises pH.  Particulate organic carbon (alive or dead) is heavy, and if not consumed and recycled, it sinks. For millions of years, this process created and maintained a DIC/pH gradient with high pH/low DIC near the surface and low pH/higher DIC at depth.

Gravity drives the biological pump and removes a significant proportion of organic carbon (assimilated from both natural and anthropogenic carbon). That carbon is transported to depths where it can be harmlessly sequestered for hundreds to thousands of years. However NOAA’s models fail to account for the biological pump, based on the narrow belief that carbon storage is strictly “a chemical and physical response to rising atmospheric CO2” (Sabine 2010). In contrast to Bednarsek’s anthropogenic hypothesis, an increase in the assimilation of CO2 and an efficient biological pump can prevent a decrease in surface pH and calcium carbonate saturation. In fact experiments show CO2 is often a limiting nutrient. Mesocosm experiments found that when atmospheric CO2 was increased, primary production by plankton community consumed 39% more DIC. When primary production increases, more carbon is shuttled to depth.

Meet the Sea Butterflies



Sea Butterfly  Limacina helicina


Sea butterflies are pteropods, a kind of snail exclusively living in the open ocean. A cubic meter of seawater may contain 50 to several thousand individuals. Unlike their terrestrial relatives that plod along on a slimy “foot”, pteropods transformed their foot into a pair of wings to “fly” through ocean waters.

Pteropods are divided into two main groups: sea butterflies with extremely thin, coiled or cone-shaped shells, and “naked” sea angels that evolved a way to shut off their shell-making genes completely when larvae. Sea butterflies feed by suspending themselves in the water column and extruding a web of mucus that passively catches sinking plankton and other organic particles. In contrast sea angels specialized to aggressively prey on sea butterflies. Abandoning their shell suggests whatever benefits a shell may have provided, those benefits were not critical, but losing the shell increased their maneuverability for the hunt. When encountering a sea butterfly, the bizarre sea angel shoots out tentacles from its head. The tentacles dig into the butterflies’ shells and if properly grasped, the tentacles give the angel leverage to extract the butterfly from its shell opening. Below is a 2-minute video below of a sea angel attacking a sea butterfly (not in English). Fish and whales also feed on sea butterflies, gulping mouthfuls at a time. So overall the butterfly’s shell offers precious little protection from their main predators.

While Bednarsek fears pteropods might not adapt quickly enough to rising atmospheric CO2, pteropod behavior argues they are already well adapted. Sea butterflies prefer to graze in highly productive regions generated by nutrient rich but corrosive upwelled waters. Accordingly upwelled regions are typically key reproductive habitat supporting an abundance of juveniles. Sea butterflies are most abundant in the upper 50 meters of the ocean, grazing on abundant phytoplankton. However depending on the species, the population, and location, most sea butterflies migrate daily to depths of 100 meters or more (sometimes below 500 meters) where pH can drop to around 7.6 and waters become corrosive. Similarly they will migrate to deeper more “acidic” depths to over-winter. And although tropical waters are the most supersaturated and the most unlikely to promote shell dissolution, pteropods are least abundant in those tropical waters. In contrast they are very abundant in marginally supersaturated waters around Antarctica.

All calcifying organisms have a protective organic layer that minimizes sensitivity to any changes in seawater pH and all isolate their calcifying chambers from ambient water conditions. Mollusks like clams, oysters and snails have a protective outer layer of organic tissue called the periostracum. The mollusk periostracum has allowed them to colonize the acidic depths of ocean floors, colonize freshwater lakes and streams where pH falls to truly acidic levels below 6.0, and to colonize the flanks of submerged volcanoes where escaping CO2 naturally lowers the pH between 7.3 and 5.39.

Single cell foraminifera and coccolithophorids have some of the thinnest organic layers that effectively prevent dissolution, and the petite sea butterfly has one of the thinnest mollusk periostraca. How well it protects the sea butterfly has created a debate between Bednarsek and other pteropod researchers. Bednarsek argues the sea butterfly’s thin periostracum, especially in juveniles, offers very little to no protection from low pH water suggesting they are very susceptible to life threatening shell dissolution. In contrast other researchers argue shell dissolution occurs when the periostracum is damaged. During the sea butterflies’ short life, which can be less than a year, they are under constant attacks from predators like sea angels that can damage their periostracum. There also exists a whole range of shell-inhabiting/shell-digesting organisms from bacteria to sponges and worms (aka the epibiont) that drill through a mollusk’s periostracum. Thus researchers argue when the periostracum remains intact, “the shell appears pristine with no sign of dissolution”, even when exposed to undersaturated waters. Only the damaged shells showed dissolution when exposed to undersaturated water. Based on observations, they concluded sea butterflies “are perhaps not as vulnerable to ocean acidification as previously claimed, at least not from direct shell dissolution.”

Dissolution of Sea Butterfly Shell from Bednarsek 2014

Furthermore to counteract shell dissolution in damaged areas, sea butterflies rapidly repair their shells by adding more calcium carbonate to the inside of the shell. Biogenic calcification happens at much greater rates than dissolution, and such rapid repair mechanisms would be expected for an animal seeking low pH upwelled waters to graze.

Ironically Bednarsek’s electron microscope images of corroded sea butterfly shells, provide evidence that supports her detractors. The shell (Figure 2 above) from Bednarsek’s 2014 paper, shows severe dissolution (labeled “b”) on the innermost whorls, similar to her other images (not shown here) showing widespread dissolution on juvenile shells. However despite dissolution during its juvenile stage, the snail clearly survived and continued to grow. The subsequent growth shows very little dissolution (labeled “a”). That suggests two possible scenarios that are not mutually exclusive. The snail’s exposure to corrosive waters was limited to short-term episodic upwelling during its earliest years, and was an insignificant cost of grazing in highly productive waters. Or if corrosive conditions continued, then the more developed periostracum protected the shell from further dissolution. The small area of severe dissolution within the section of unharmed shell (the patch to the right of region b) further supports the argument that dissolution only happens where there is damage. Otherwise, if the periostracum provided little protection, then the whole region would have suffered dissolution not just the isolated patch. Thus life-threatening dissolution is just conjecture, and as Bednarsek later admits, “dissolution-driven mortality in pteropods has not been directly confirmed.”

Catastrophic Dissolution?

Long before the politics of climate wars emerged, Mark Twain quipped, “There is something fascinating about science. One gets such wholesale returns of conjecture out of such a trifling investment of fact.”  Indeed the oft-cited Orr 2005 “ocean acidification” paper is an iconic example illustrating Twain’s observation. Orr et al examined just 14 living sea butterflies (Clio pyramidata, see photograph below), captured in the subarctic Pacific. Orr exposed them for 48 hours to experimentally induced under-saturated water that he predicted would occur in the future around Antarctica due to rising atmospheric CO2. All specimens survived. However scanning electron microscopy revealed etching on their shells.

Although it is highly unlikely such minor etching would be of any consequence, Orr 2005 jumped to hypothesizing that pteropod species “will not be able to adapt quickly enough to live in the undersaturated conditions”. Orr 2005 lamented rising atmospheric CO2 would reduce pteropod habitat and survival, leading to ecosystem collapse in polar regions. He speculated arctic pteropods could be forced southward to warmer waters that were more saturated. Due to the butterfly’s imagined extirpation via dissolution, he predicted sea butterfly predators such as sea angels, fish and whales would all suffer. Such was their wholesale catastrophic conjecture, published in a top journal, based on etching in 14 specimens. 

Observed Dissolution due to Ancient CO2 Enriched Waters

Nina Bednarsek followed with a 2012 paper in which she too attributed shell dissolution of a sea butterfly (Limacina helicina) to increasing anthropogenic CO2. Media outlets promoted her conclusions in articles like Science Daily’s “First Evidence Of Ocean Acidification Affecting Live Marine Creatures In The Southern Ocean. Unfortunately the term “ocean acidification” is indiscriminately used to describe any reduction of pH even if naturally induced by upwelling.

On her cruise through the Scotia Sea (just north of the Antarctic peninsula), Bednarsek collected snails at depths of 200 meters from 6 different stations. Snails at five stations showed no evidence of shell dissolution. All stations exhibited supersaturated surface waters. Her shell dissolution was limited to just one station.

All stations experienced upwelling and Bednarsek acknowledged, “these upwelled waters are approximately 1000 years old”. Thousand-year old water means it had not contacted the surface for 1000 years, and those deep waters had not absorbed any anthropogenic CO2. The water’s corrosiveness was due to a millennium of decomposing sunken organic matter that increasingly released CO2 and lowered pH. Still she conjectured the observed dissolution was due to anthropogenic CO2.

Because winds will mix supersaturated surface waters with corrosive upwelled waters in the upper 200 meters, mixing can neutralize any corrosive effects of upwelled waters. To attribute her observed dissolution to anthropogenic CO2, Bednarsek argued recent invasions of anthropogenic CO2 into the surface water had lowered its surface pH to such an extent, mixing no longer counteracted the low pH of upwelled water. It was a reasonable hypothesis, however there was no evidence to support it.

At the one and only station where snails had experienced dissolution, surface waters were far more supersaturated than at any other station. In contrast to Bednarsek’s narrative, that station’s supersaturated waters should have had the most neutralizing effect. Her anthropogenic attribution was simply not consistent with observations. More parsimoniously either that one station experienced greater upwelling, or there was less wind mixing to deepen the neutralizing effect. It is puzzling why peer-review or the editors at Nature allowed her unsupported anthropogenic conclusion to be published.

In 2014, NOAA News promoted another paper by Bednarsek about dissolving sea butterfly shells with the headlines, “NOAA-Led Researchers Discover Ocean Acidity Is Dissolving Shells Of Tiny Snails Off The U.S. West Coast.”  NOAA’s press release explicitly stated the term “ocean acidification” described the process of ocean water becoming corrosive as a result of absorbing nearly a third of the carbon dioxide released into the atmosphere from human sources.” In contrast, her observed shell dissolution only happened where upwelling was the greatest, along Oregon and northern California.  Along southern California where upwelling was minimal, Bednarsek found no dissolution. If acidification was due to atmospheric CO2, we would expect a more uniform pattern of dissolution. But again Bednarsek set forth a scenario to blame anthropogenic CO2, arguing upwelled waters had been directly “acidified” by anthropogenic CO2.

She speculated the upwelled waters had been near the surface 50 years ago, during which time it equilibrated with the 1960s atmosphere. Those waters then sank to depths of 80 to 200 meters, and were now upwelled to the surface. However the problem with this scenario is the source of upwelled waters along the California Oregon coast can be traced back to the California Undercurrent. The California Undercurrent originates in equatorial regions primarily at depths 100 to 200 meters and flows poleward beneath the equatorward flowing California Current. The undercurrent is supplied with low pH, low oxygen, and high inorganic carbon waters from the eastern tropical Pacific. Due to the accumulation of ancient carbon, the eastern tropical Pacific contains some of the oldest waters on earth and more than any other region on earth ventilates tremendous amounts of CO2 from the ocean into the air.

Studies of the changing characteristics of the California Undercurrent conclude all its “changes are consistent with an increasing influence of Pacific equatorial waters” over the past decades. During a negative Pacific Decadal Oscillation (PDO) or a La Nina, research shows the California Undercurrent acquires increased amounts of those ancient waters and upwelling is stronger. Although those studies did not consider possible anthropogenic contributions, the trends in lower oxygen and lower pH were explained by natural increased mixing of older water masses. That mixing trend could represent decadal variability or the centuries long increasing upwelling trend documented under the Peru Current (detailed in Part 2).

The same corrosive upwelling associated with Bednarsek’s sea butterfly dissolution was also responsible for the corrosive waters that oyster fisherman unwittingly pumped into larval-rearing tanks in 2008-9. (To make matters worse, oyster fishermen pumped water in the early morning when nighttime respiration further acidified the water). And consistent with multidecadal variability, during the previous negative PDO from 1940 to the late 70s, corrosive upwelled waters had similarly reduced survival of larval oysters in those bays. Clearly natural oscillations episodically upwell more nutrient-rich, oxygen-poor corrosive waters. And because oceans contain the greatest concentration of inorganic carbon by far, the question remains has anthropogenic CO2 significantly exacerbated the corrosiveness of natural upwelling? In contrast, NOAA’s Richard Feely believes anthropogenic CO2 is the primary factor. Thus he answers that question with an upside down perspective, stating upwelling “exacerbates [anthropogenic] ocean acidification”. And as detailed in part 2, NOAA's models have an upside down representation of carbon distribution.

The sea butterfly joins the parade of icons like polar bears, penguins, pika, mangroves and Parmesan’s butterflies where the effects of natural climate variability or direct human interference are obscured and falsely promoted as catastrophic climate change.