Our Recent Paper in Four Tweets

As promised, this is the tl;dr version of my previous post, where I have tried to reduce our open access paper into four tweet-length snippets per sub-heading. Here goes:

The History: A particular plankton shell used to reconstruct climate is purported to have 2 morphs that live in different depths of the ocean.

The Importance: If true, previous studies that attempt to quantify past oceanic climates from non-selective morphs of that plankton species are biased.

The Study: We analyzed pairs of extreme & intermediate morphs, & with a model, found that all morphs live in the top (<30 m) part of the ocean. 

The Implications: We conclude that morph-based uncertainty in this species when used for studying ancient (Holocene) climates is little-to-none.

Foraminiferal Morphotypes: Birds of a Feather?

We have a new open access paper (yes, anyone, including you, can access it!) out in Scientific Reports titled Globigerinoides ruber morphotypes in the Gulf of Mexico: A test of null hypothesis. Here is a breakdown of the paper:

The History

  • Globingerinoides ruber (G. ruber) is a rather famous planktic foraminfer (or foram for short), whose shell chemistry has been widely (and successfully) used to reconstruct ancient surface ocean parameters such as temperature and salinity. This foram lives in the upper ocean and creates a shell for its protection; the shell later sinks to the seafloor after its death.
  • G. ruber shells were first identified and reported by French naturalist Alcide d’Orbigny in 1839. Since then, several morphotypes of the species have been reported. These morphotypes have seemingly minor variations in their shell characteristics (e.g. smaller aperture hole, more arched chambers etc.)
  • In 2000, a core-top (or near-modern) study by Chinese paleoceanographer, Luejiang Wang (who tragically passed away drilling corals in the South China Sea), analyzed stable isotopes in the two principal morphotypes of G. ruber‘s white variety: sensu alto – sl & sensu stricto – ss (as he christened them).
  • The study seemed to indicate differences in the stable isotopic signatures of these morphotypes: ss seemed to have a warmer signature while sl was cooler.
  • Wang suggested that sl might live deeper than ss and is hence, cold-biased (the deeper you go in the ocean, the colder it gets!)
  • More recent studies seemed to find equivocal/ambiguous results for similar analyses i.e. some found significant differences but others didn’t. However, nobody sought out to perform a comprehensive, controlled experiment specifically for G. ruber morphotypes.

G. ruber morphotypes (1) a and b: sensu lato; (2) c and d: sensu stricto. There are numerous intermediate transitional forms between these.

G. ruber morphotypes (1) a and b: sensu lato; (2) c and d: sensu stricto. There are numerous intermediate transitional forms between these.

The Importance

  • A lot of our knowledge about past climate change in the oceans comes from studies analyzing G. ruber shells.
  • If these studies did not selectively discriminate between the two morphotypes prior to analyses, the Wang, 2000 study and others suggest that these reconstructions could be biased as we would be averaging signals from two different depths. Thus, our quantitative understanding of climate change itself may be biased!
  • Furthermore, all our calibrations and verification exercises on G. ruber have been done on non-selective mixtures of these morphotypes.
  • It is logistically very difficult to observe these critters in the wild. Here is a nice video that details the challenging process of culturing forams.
  • It is NON-TRIVIAL to differentiate between these two morphotypes as there are numerous transitional shell forms between ss and sl. It is HIGHLY subjective! (one man’s sensu stricto is another’s sensu lato)
  • Genetic work shows that it is NON-TRIVIAL to select different genotypes based on the shell morphology alone.
  • As a birder, here is an analogy with birds: two birds that look very, very similar may, in reality, be different species and have completely different habitats and/or eating habits etc. If we are looking to gain information from the physiological chemistry of these birds (say, their feathers) to infer something about the environment they live in – it would be prudent NOT to mix samples of both the birds, correct?
  • But… it is impractical to perform pilot genetic studies on living forams in tandem with paleoceanographic reconstructions using foram shells.
  • So, how much would it matter if we did not perform genetic analyses accompanying paleoclimate reconstructions in the curious case of these two G. ruber morphotypes? Do they really live at different depths? How much does it matter if they did?

The Study

  • To shed some light on (some) of these important questions, we turned to the abundant resources that are available to us in the northern Gulf of Mexico. These include:
    1. A Sediment Trap: A device that collects foram shells before they hit the seafloor.
    2. Core-tops: The topmost portion of the seafloor, where recently dead foram shells accumulate.
    3. Downcore material: Cores spanning the last 4,000 years containing ancient foram shells.
  • We sat down and decided to chalk out a strategy to be consistent in how we selected the stereotypical ss and sl morphotype sample.
  • We decided to perform a geochemical test of “null hypothesis“, where, along with the stereotypical ss and sl morphotypes, we analyzed samples of ‘intermediate’ morphotypes that had transitional shell characteristics to these extreme morphotypes:
    • If the geochemical variability between the sets of ‘intermediate’ morphotypes was consistently different from the ss-sl sets, then the shape of the shell dictates its stable isotope signature, and hence provides evidence for cold/warm biases.
    • On the contrary, if the ‘intermediate’ sets showed comparable variability to the ss-sl sets, then we cannot reject the null hypothesis that morphotypical variability has no effect on the stable isotope signature.

The Results

  • We found that the  ss-sl isotopic signatures for 37 sets from was statistically indistinguishable.
  • The ‘intermediate’ sets showed variability very similar to the offsets in the ss-sl pairs.
  • The sediment trap results indicated that the ss, sl, and intermediate morphotypes are good indicators of sea-surface conditions (and not deeper).
  • They also revealed no seasonal differences between these morphotypes (i.e. all morphotypes grow throughout the year)
  • Using a forward model and our observations, we found that both ss and sl morphotypes live and calcify in the upper ~35 m of the water column in the Gulf of Mexico.

The Implications

  • In the Gulf of Mexico, the uncertainty due to morphotypes in Holocene-based reconstructions is little-to-none.
  • G. ruber (at least in this part of the world) appears to calcify in the topmost portion of the surface ocean.
  • Not all previous reconstructions, calibrations, verification experiments that didn’t discriminate between G. ruber morphotypes are wrong.

Propagation of a Taxonomic Error

Taxonomic identification lies at the crux of foraminiferal paleoceanography. I have written before about the importance of properly identifying and reporting the species of foraminifera used for geochemical analysis in a study. This has implications not only for placing the extracted geochemical signals in a physical context with appropriate uncertainty bounds, but also for communicating the methodologies employed in the study for future replication and reproducibility purposes, a tenet of science.

Recently, Yi Ge Zhang and colleagues had an article in Science magazine about their work on a 12-million-year-old temperature reconstruction in the tropical Pacific Ocean. The study produces new, long and detailed tropical sea-surface temperature records using the TEX86 paleothermometer. Their results are remarkable as they show a sustained zonal tropical Pacific temperature gradient through the Pliocene, going against the paradigm of “permanent El Niño”. The Pliocene is a time period where the tropical Pacific gradient was thought to have collapsed, similar to what happens during a brief El Niño event. This inference comes from paleoceanographic records produced from the magnesium-to-calcium (Mg/Ca) ratios of planktic foraminifera. The authors make the case that this technique has limitations compared to TEX86 reconstructions on these time scales.

Discussing these previous studies that investigated Pliocene climate using planktic foraminifera, the authors write:

Published temperature records based on magnesium-to-calcium ratios (Mg/Ca) of the planktonic foraminifera Globoritalia sacculifer, from Ocean Drilling Program (ODP) site 806 (0°N, 159°E) (Fig. 1) (8), suggest that warm pool temperatures remained relatively constant as Earth cooled over the past 5 million years.

I was immediately struck by ‘Globoritalia’ – I had never heard of the genus, only Globorotalia. Furthermore, the only sacculifer species I was familiar with was Globigerinoides sacculifer. Being relatively inexperienced in non-Quaternary foraminifera, I assumed Globoritalia sacculifer was some kind of Pliocene planktic species. Skimming through foraminiferal identification books, I could not find any mention of the species. Google Scholar too yielded only five results for ‘Globoritalia sacculifer (including Zhang et al.) with all of them discussing the Pliocene mean state hypothesis. I was increasingly beginning to suspect that the species did not exist. The final nail came when I questioned my more experienced colleagues including foraminiferal expert Dick Poore, all of whom immediately recognized the error when I pointed it out: that it should be Globigerinoides sacculifer and not Globoritalia sacculifer (and Globorotalia not Globoritalia).

The origin of the taxonomic error seems to stem from another Science article, Permanent El Niño-Like Conditions During the Pliocene Warm Period by Michael Wara and colleagues, published in 2005. They write:

To track changes in the mean thermocline depth at the EEP site, we used ∆δ18O, the difference in δ18O between surface-dwelling Globoritalia sacculifer (without sac) and G. tumida (355 to 425 mm) (Fig. 2A), which occupies the base of the photic zone.

The error also seems to have propagated into a Journal of Climate article by Carl Wunsch (where Globigerinoides sacculifer is written correctly but Globorotalia tumida is written as Globoritalia tumida) and an article by Guodong Ji and others published in Geophysical Research Letters.

Though Globoritalia sacculifer was probably the result of an innocuous spelling oversight that somehow made its way past editing and though this error has no bearing on the scientific results contained therein, I find it very impressive that it has made its way into Science magazine, a highly prestigious journal, not once but twice over the last nine years.

Academia Bizarro: Apocalypse π

There’s a potential Ig-Nobel worthy paper on arXiv that was submitted earlier this month: A Ballistic Monte Carlo Approximation of π. The authors use a Mossberg 500 pump-action shotgun and an aluminium foil to calculate π with a Monte Carlo approximation. What is their motivation, you ask? From the introduction:

The ratio between a circle’s circumference and its diameter, named π, is a mathematical constant of crucial importance to science, yet most scientists rely on pre-computed approximations of π for their research. This is problematic, because scientific progress relies on information that will very likely disappear in case of a cataclysmic event, such as a zombie apocalypse. In such case, scientific progress might even stop entirely. This motivates the need for a robust, yet easily applicable method to estimate π.

How cool is that! Thanks to Gail Gutowski for the tip.

Off to London, guvnor!

IPCC-ENSOJud Partin and I are off to University College London to attend a workshop entitled Palaeovariability: Data-Model Comparisons. The main aim of the workshop is to facilitate interaction between paleoclimate data gatherers (or “proxy people” such as myself who produce paleoclimate records) and paleoclimate modelers (who use state-of-the-art GCMs to model ancient climates) to figure out novel and optimal methods to actually compare model output to proxy data.

Climate models used to forecast future climates can also be used to simulate past climates based on different climate forcings in the past (such as variations in the Earth’s orbit, CO2 etc.) Paleoclimate records of temperature, salinity, rainfall etc. that are generated from proxy data, can be used to test and validate the hindcasts of these climate models. However, there are multiple sources of uncertainty, both from the modeling side and the proxy side, that hinder straightforward comparison between paleoclimate proxy data and paleoclimate model output. For example, the world in models is divided into multiple grids, the size of which defines the spatial resolution of the model. Paleoclimate records obtained from proxy data on the other hand, are generated from one spatial point (eg. a coral, sediment cores, ice cores etc.) What is the best way to compare data and model output? Grid-to-spatial point? Extrapolation? Interpolation? These questions don’t have straightforward answers either. Further, there can be multiple age-related uncertainties in the proxy (analytical radiogenic error, bioturbation etc.) that might make you think you are comparing apples and apples from the same time period, when they are actually apples belonging to totally different eras. Most importantly, how well can different climate models simulate climates in a particular region/time and how well can different proxies actually reconstruct climatic parameters?

Well, I hope to gain some insights into all these questions and more, during my time in London! I will be talking about interpreting individual foraminiferal analyses data and how best to reconcile them with model output. My talk is entitled: Paleoceanographic reconstructions using individual foraminifera in the tropical Pacific: Records of annual cycle or ENSO variability? Jud will be bringing his tropical paleoclimate expertise to the table with a talk entitled Tropical Climate Variability in the Western Pacific on Sub-annual to Orbital Timescales. All in all, I’m looking forward to a great week in London, guvnor!

Academia Bizarro: A Rhyme with Reason?


Here’s a paper entitled Habitat Provision for Meiofauna by Fucus serratus Epifauna with Particular Data on the Flatworm Monocelis lineata by Patrick S. Boaden published in Marine Ecology. Why is it fit for Academia Bizarro? Well, the entire paper is in rhyme. The entire thing! Including materials, methods, results, discussion, and conclusions! The abstract:


Hat tip to AWI graduate student Marieke Feis for sending this paper my way and hats off to the author (and the editor) for publishing this enjoyable read!

Heavy Metal Haikus

Here are some of my favorite metal albums released in 2013, with a haiku to go with each one (click on album art for YouTube links):

Gorguts – Colored Sands

Colored Sands

Non-linear waveforms
Meditated dissonance
Pushing boundaries

Paysage D’Hiver – Das Tor

Das Tor

Wintertime voyage
Into black forest darkness
Cold serenity

Rotting Christ – Κατά τον δαίμονα εαυτού


Ode to the ancients
Channeling folk symphonies
Unique black metal

A Pale Horse Named Death – Lay My Soul To Waste

Lay My Soul To Waste

Downtempo structures
Unabashedly gothic
A catchy affair

Satan – Life Sentence

Life Sentence

Dueling guitars
Swift rhythms and harmonies
Always hail Satan

Power Trip – Manifest Decimation

Manifest Decimation

Texan thrash outfit
Uses hardcore and death riffs
To make your head nod

Inquisition – Obscure Verses for the Multiverse

Obscure Multiverse

Croaking the obverse
Profoundly, divisively
Scaled classically

Castevet – Obsian


Sharp, serrated notes
As a river erodes stones
Shaped by melody

Summoning – Old Mornings Dawn


The clarion calls
A return to Middle-Earth
As the war drum beats

Agrimonia – Rites of Separation

Rites of Sep

Meaty, drawn-out sludge
Hosting vast influences
Complex songwriting

Hateform – Sactuary In Abyss


Brisk melodic thrash
Tempered with technical flair
Progressive delight

In Solitude – Sister


Nostalgic metal
Reverberates melodic
To merciful fates

Windhand – Soma


Dark, menacing chords
Tinted with insanity
Heavy as a thing

VHÖL – Vhöl


An all-star lineup
Listens to d-beat punk rock
And makes black metal

Other notable releases: Carcass – Surgical Steel, Exhumed – Necrocracy, Ruins of Beverast – Blood Vaults, Ulcerate – Vermis, Deafhaven – Sunbather, Intronaut – Habitual Levitations, Church of Misery – Kingdom Scum, SPEkTER – Cypher, Inter Arma – Sky Burial, Tribulation – The Formulas of Death, Clandestine Blaze – Harmony of Struggle.