ADAGIO is a joint U.S.-German study of the effects of ocean acidification on particle aggregation. The project is funded by the National Science Foundation Chemical Oceanography Program. It is directed by Cindy Lee of the School of Marine and Atmospheric Science (SoMAS) at Stony Brook University in Stony Brook, New York, and Anja Engel of the Helmholtz-Zentrum für Ozeanforschung Kiel (GEOMAR), Germany, and SoMAS. Two experiments make up the bulk of ADAGIO work, a 2010 Chemostat Experiment and an 2011 Mesocosm Experiment.

Large-scale changes in surface ocean chemical equilibrium and elemental cycling are occurring due to ocean acidification. The ocean is the largest global sink for anthropogenic CO₂, and as the CO₂ inventory of the surface ocean increases due to enhanced uptake from the atmosphere, the pH of seawater decreases. This relative acidification of the surface ocean is expected to intensify in the future. By the year 2300 a further decrease in surface ocean pH of 0.7 units is expected, which will exceed anything experienced in the past 300,000 yr of earth history. CO₂-related changes in the carbonate chemistry of the ocean affect biological processes like photosynthesis and calcification.

Calcifying phytoplankton species are thus likely to be particularly sensitive to changes in CO₂ concentration. Biominerals (especially CaCO₃) play an important role in exporting organic matter (OM) to the deep ocean.  Minerals in association with OM act as ballast, thereby increasing OM sinking rate and hence the depth to which OM is exported. Biominerals affect the rate of formation and properties of aggregates and can delay their microbial degradation. Aggregates play a pivotal role in organic matter cycling and export, as they mediate vertical mass fluxes and elemental cycling in the ocean.  Aggregation processes in the ocean range over a wide size spectrum, cascading from the nano scale up to the size of fast settling marine snow, primarily by the process of gel particle formation from organic polymers. In general, the reactivity and rigidity of a gel depends on its molecular composition and on the type of binding and interactions between the individual polymer chains. Gel particle formation is likely to become affected by ocean acidification due to changes in the reactivity and production rate of organic polymers. The ADAGIO project will investigate some of the mechanisms of aggregation of marine particles that specifically relate to organic matter-ballast mineral interactions and their sensitivity to ocean acidification. Our intention is to understand the mechanisms but also consider the properties of aggregation that might most result in alterations of sedimentation and export in the future ocean. Three hypotheses will be tested:

1. Gel particles enhance aggregate formation. The production of gel particles is sensitive to ocean acidification.
2. Biomineral ballast (particularly calcium carbonate) is an important factor in carbon export in the ocean. Lower pH will decrease ballast-particle interactions and consequently export.
3. In addition to carbohydrates, gel particles also include proteins and lipids that will affect rate and mechanisms of organic matter aggregation in the ocean. Changes in the composition of gel particles in seawater are expected at different pH levels.

 

2010 Chemostat Experiment

In January through March, 2010, we grew phytoplankton in controlled chemostats, and then measured aggregate formation in roller tables (Fig. 1).

Participants
PI’s:	Cindy Lee and Anja Engel
Project Graduate Students:	Carolina Cisternas-Novoa, SBU, Tiantian Tang, SBU, Sonja Endres, AWI, Mascha Wurst, AWI
Helped with sampling:	Nicole Händel, AWI, Corinna Borchard, AWI, Tobias Mattfeldt, AWI, Helge-Ansgar Giebel, AWI
Helped with analysis:	Karin Woudsma (nutrient analyses), AWI, Laura Wischnewski (nutrient analyses), AWI, Nathalie Pülmanns (TEP/CSP microscopy), AWI, David Hirschberg (POC, PON), SBU
Student assistants involved with the preparation:	Aneta Mista, AWI, Claudia Buchholz, AWI, Tobias Mattfeldt, AWI

Data

Roller Tank Nutrients |  CHN |  Cell counts  |  TEP counts  |  Alkalinity

Hypothesis 1:  Gel particles enhance aggregate formation.  The production of gel particles increases with ocean acidification, and thus export increases.

Hypothesis 2:  Biomineral ballast (particularly calcium carbonate) is an important factor in carbon export in the ocean.  Lower pH will decrease ballast-particle interactions, and export consequently decreases.

Chemostat experiments with Emiliania huxleyi were conducted at naturally low phytoplankton cell abundances (<105 ml-1, Lampert et al. 2002) to prevent unrealistic effects of biological processes on the seawater carbonate system and to allow for a better control of pH.  To describe changes in the seawater carbonate system due to biogenic calcification, three variables of the carbonate system, i.e., pH, DIC and total alkalinity, were determined during all experiments.

Acidification effects were investigated in the chemostats at 180, 380, and 750 ppm CO₂ partial pressure.  Three chemostats (180ppm, 380ppm, 750ppm) were operated at a temperature of 10°C. The flow rate was set at 0.1 d-1, because we previously observed high rates of calcification and gel particle formation at this low growth rate.  Chemostats were sampled three times within a 3-week period. Samples will be taken for parameters as described in Table 1.  The pH and temperature in the chemostats were continuously monitored during the experiment by means of pH and temperature electrodes coupled with data logger units.  On each sampling occasion, additional samples were taken to conduct roller table experiments.

Samples from the chemostats grown at three different CO₂ levels as described above were subjected to aggregate formation using roller tables (Shanks and Edmondson, 1989).  Samples were divided and transferred to cylindrical Plexiglas tanks and kept in the dark at 10°C. All tanks were placed on roller tables that rotate at 0.66 rpm so that aggregates could form under conditions that allow for continuous particle settling.  To simulate the sinking of aggregates into deeper and hence more dilute waters, the concentration of suspended cells was lowered before the actual experiment began. Samples were taken directly after aggregate formation (day 2-4) and after 1 week for chemical analyses of the aggregates and of the suspended particles in the seawater surrounding the aggregates. To determine the size and abundance of aggregates, as well as the time of their formation, the roller tanks were monitored using a video camera as described in Engel et al. (2008).  Aggregates from the tanks will be collected and chemically analyzed as above.  In addition to the measurement of settling velocity, TEP, CSP, lipophilic gel tracer, and the bulk parameters POC, PN, DOC, DON, we will follow degradation of the particulate fraction by measuring amino acids and pigments as in Engel et al. (2008b).

Expected Results (H1):  If hypothesis 1 is correct, we expect cell numbers and Chl a to be similar in each of the chemostats with different levels of CO₂ since the phytoplankton abundance will be determined by the nutrient levels.  However, TEP, CSP, and LSP should be higher in experiments with higher CO₂ since from past studies we expect gel production to be higher.  The ratio of TEP:CSP (or (C:N) in the gel particles will be higher with higher CO₂ since N will be limiting.  If H1 is correct, in experiments with higher CO₂, POC should increase as more gels are formed, and PIC should decrease as calcification slows down.  Because of higher phytoplankton exudation, DOC should be higher in experiments with higher CO₂, but will decrease once gel formation begins.

Expected Results (H2): If hypothesis 2 is correct, then we expect to find the number of aggregates higher in tanks with higher CO₂.  Aggregates at higher CO₂ should have lower PIC:POC ratios.  They should have lower settling velocities as lower CaCO₃ due to less ballast.  DOC and DON will depend on the balance between microbial uptake and phytoplankton exudation.  Particulate amino acids and carbohydrate compositions and amounts will demonstrate that there is more decomposition at higher CO₂ because the CaCO₃ that protects the organic matter at will be reduced.

Table 1 summarizes the methods and instruments that were or will be used. All analyses will be conducted in replicate.

Parameter	Method	Reference
DIC	Photometry	Stoll et al. 2001
Total Alkalinity	Gran Titration	Gran 1952
pH/T	Electrochemical detection	WTW Operation Protocol
CO2 gas	Infrared Analyzer	Li-Cor Operation Protocol
Inorganic Nutrients	Colorimety	Koroleff and Grasshof 1983
POC, PN	Combustion, CHN analyzer	JGOFS Protocoll 1994
DOC, DON	High Temperature combustion	JGOFS Protocoll 1994
Cell abundance	Flow Cytometry	Vaulot 1989
Bacterial abundance	Flow Cytometry	Vaulot 1989
TEP concentration and size distribution	Microscopy and Colorimetry after staining with Alcian Blue	Passow and Alldredge 1995, Engel 2008
CSP concentration and size distribution	Microscopy after staining with Coomassie Blue	Long and Azam 1996, Engel 2008
Dissolved CHO	High Performance Anion Exchange Chromatography with Pulsed Amperometric Detection (HPAEC-PAD)	Mopper et al  1992, Engel and Händel 2008
Total CHO	HPAEC-PAD	Mopper et al 1992, Engel and Händel 2008
Dissolved AA	HPLC following derivatization	Lee et al 2000
Total AA	HPLC following derivatization	Lee et al 2000
Coccolith morphology	Scanning Electron Microscopy (SEM)	Paerl and Shimp 1973
Aggregate size	Video-imaging and Image analysis	Engel et al 2008
Aggregate settling velocity	Video-imaging and Image analysis	Engel et al 2008
Lipid classes	Iatroscan (Madeleine Goutx – see support letter)	Striby et al. 1999

 

2011 Mesocosm Experiment

In May through September, 2011, we built 12 controlled mesocosms with different CO2 and temperature conditions, and then measured changes in phytoplankton growth, community structure, aggregate formation, etc..

Participants
PI’s:	Cindy Lee, SBU, Anja Engel, GEOMAR
Senior Staff:	Roman de Jesus, SBU
Project Graduate Students:	Carolina Cisternas-Novoa, SBU, Tiantian Tang, SBU, Thea Dammrich, AWI
Analytical help:	David Hirschberg (POC, PON), SBU
Student assistants involved with the preparation:	Sumantro Ray, SBU, Trevor Shea, East Hampton High School, NY, Claire Beaulieu, Garcia Program, SBU, Coralie Beaulieu, Garcia Program, SBU, Jason Kuan, Garcia Program, SBU, Nazac Saleh, Garcia Program, SBU