Home / News / Newsletters / Issue n°27 - September 2014

Issue n°27 - September 2014

Issue n°27 - September 2014
banner transition NL blank short
In this issue

Lisa Maddison

At the end of June, almost 500 marine reseachers gathered in warm and sunny Bergen in Norway for IMBER’s first Open Science Conference. The theme of the conference was Future Oceans: Reseach for marine sustainability: Multiple stressors, drivers, challenges and solutions. Each morning started with two keynote presentations that considered the gaps, challenges and new insights of a wide range of topics that are an important focus of IMBER. These include ocean acidification, links between marine biogeochemistry and food webs, anthropogenic impacts on continental margins, adaptation of high tropic levels to climate change, sustainability and conservation, and the need to integrate human-ocean interactions in global change research. We also heard about the perspectives of the new research initiative Future Earth. These talks were followed by parallel (see http://www.imber.info/index.php/Meetings/IMBER-OSC-2014/Programme for details) and poster sessions, and some great social events at the medieval Håkonshallen and the conference dinner at the top of Mount Fløyen where we watched the sun set close to midnight!

By all accounts, it was a great success and provided a excellent opportunity for scientists to showcase the IMBER-related work that has been undertaken since the project’s inception in 2005, and to put forward ideas for the direction that IMBER science should take in the future - keeping in mind the need to balance fundamental science with solution-driven science. 

In this edition of IMBER Update, we highlight some of presentations from the conference, and report on some outcomes of the workshops that were held in conjunction with it. We are particularly pleased to include articles by some of the young researchers and students whose presentations earned them "Best Student/Young Researcher Presentation" awards. These include: Toby Boatman, Bernadette Pree, Antonella Rivera and Sarah Peterson. 

banner transition NL

Science Highlights from the IMBER Future Oceans OSC

Predicting Fish from Physics: Strengths, weaknesses and ways forward

Eric Galbraith1, James Watson2 and Myron Peck3

1Earth & Planetary Sciences, McGill University, Montreal, Canada; eric.galbraith@mcgill.ca 
2Stockholm Resilience Centre, Sweden and Princeton University, NJ USA; jrwatson@princeton.edu, james.watson@.su.se
3Institute of Hydrobiology and Fisheries Science, University of Hamburg, Hamburg, Germany; myron.peck@uni-hamburg.de



At present, three-dimensional general circulation models can simulate ocean currents, light, temperature and nutrients with reasonable fidelity. These environmental variables can then be used to drive the dynamics of lower-trophic level organisms – phytoplankton and zooplankton (Stock et al. 2011). Extending these models to predict fish biomass directly from the physics of marine environments is an exciting goal that would allow projections of food-security in the future, but it is difficult (Fulton et al. 2010). Improving models of upper trophic dynamics requires deeper understanding of the feedbacks between upper and lower trophic levels, as well as ocean biogeochemistry (Link et al. 2012).

The goal of the ‘Predicting Fish from Physics’ workshop held at the IMBER Future Oceans conference was to synthesize knowledge of connecting physical-biogeochemical models to upper trophic levels. A number of coupled biogeochemistry-fish models have been developed in recent years. Many of the developers of these models were among the 50 or so workshop participants who gathered beneath the sunny Bergen sky, to discuss five key questions and define the state of the field:

What is the best way to evaluate fish model results and judge model skill?

A model can be conceptualized as consisting of two parts: the structure (i.e. equations), and parameters (i.e. the numerical values of constants used in the equations). The success of both parts – equations and parameters - in simulating reality can only be judged by comparing model results with empirical data. Unfortunately, workshop participants recognized that many models are typically ‘trained’ using all available data. Model evaluation thus requires independent data, not included in the training dataset. This could include variables, or regions or time periods that were not included in the training dataset. Furthermore, the observed variables should match those predicted by the model, rather than a secondary product (e.g. comparing modeled catch data with actual catch data). In addition, relatively simple global models can be compared with complex regional models, such as ECOPATH or ATLANTIS case studies, that are well-constrained by high quality data.

The discussion highlighted the importance of the number of free parameters in the model. The more free parameters, the more ways there are to ‘tune’ the model to fit the data, which may hide mechanistic shortfalls that will affect model skill under different environmental conditions (for example, when projecting beyond the data to future scenarios). A large, complex model can also include ‘hidden’ processes that are fundamentally unrealistic. Thus, goodness of fit to data cannot be taken as the sole criterion for model evaluation - mechanistic integrity and transparency are equally, if not more important.

Participants identified three readily available global data sources for model comparison: the Sea Around Us Project (catch, value, effort; http://www.seaaroundus.org/), the RAM Legacy database (stock, recruitment; http://ramlegacy.marinebiodiversity.ca/), and the ICES database (http://www.ices.dk/marine-data/). Some regional, highly detailed datasets, such as CalCOFI (http://www.calcofi.org/), are also available. However, an important issue that came up was the large amount of highly resolved regional data that remains unavailable for research due to government access restrictions, such as the tuna catch data for the Pacific Ocean. Improving the availability of data will require international organization of common fish and fisheries data. The workshop discussion echoes broader calls for improved data centers, in order to monitor changes in ocean status (e.g., see the recent opinion piece by Banse, 2014).

What are the most critical uncertainties we need to address in the models, how does error propagate?

Three sources of error in ecosystem models can be defined. The first is error in the ecosystem model structure (equations). This can be addressed, to some degree, by using ensembles of models with a variety of structures. Performance metrics, such as population biomass, may help by assigning greater confidence to models that do well when evaluated against data, though this is difficult to achieve objectively. A second source of error is parameter estimates. These are relatively easy to gauge for models with only a few parameters, by testing the sensitivity to the uncertainty range of parameter values using a Monte Carlo approach (Kearney et al. 2012). A third source of error lies within the external forcing due to unresolved features in general circulation models or errors in the lower trophic level models. Errors are interactive within the numerous non-linear functions used in ecosystem models, hence their combined effect can be difficult to evaluate.

Is movement important? If so, how do we model it?

The short answer is: yes, but not always. In many cases, the ‘Field of Dreams’ assumption - if a suitable habitat exists, then the fish will come - is valid. However, movement must be simulated if the ability of fish to move between suitable habitats, whether in larval, juvenile or adult stages, can be limited by their swimming speed. In this case, suitable habitat is not necessarily enough. In addition, given that the habitat relationship with the environment is non-stationary, movement could be an important consideration for simulations of future fish production.


Workshop participants discussed how movement can be modeled using either Lagrangian (individual-based) or Eulerian (concentration-based) approaches. To some extent, the distinction between the two is artificial, and both are equally valid. Movement involves both passive drift by ocean currents, as well as active behaviour, such as swimming after prey, which together can be modeled as an advection/diffusion process. However, active swimming behaviour is an open research topic - for example, does movement respond to gradients of habitat quality or prey density? Numerical tools for integrating movement are also a work in progress, and are not trivial to implement. Given the uncertainties, determining whether or not movement needs to be included depends on the life stage and/or species being modeled, whether it is larvae (Watson et al. 2011) or tuna (Maury 2010).

Is Type II feeding enough? How do we choose the right level of food-web complexity?

Feeding relationships are of key importance. It is well known that simple linear relationships between predator and prey populations can give unstable results, such as population cycles, and prevent the coexistence of multiple species. As a result, many modern marine ecosystem models employ non-linear feeding functions, such as Holling Type-II, Type-III and the foraging-arena model. Including diet preferences and prey-switching can also increase model stability. However, all these sophisticated options have higher parameterization needs, and are often used for the sake of stability, rather than for ecologically-sound reasons.

Individual-based models can explicitly resolve feeding relationships, but it is not clear how one can scale up from individuals to populations and communities while remaining mechanistically consistent. For example, when considering schooling behaviour, feeding functions that include Type-III exponents are commonly used, as they are thought to describe spatial correlation in predators and prey. Yet these are purely phenomenological in nature, and a mechanistic derivation of the effect of schooling on population level feeding rates is still lacking. This is an area of active research for theoretical ecologists studying collective behaviour (Torney et al. 2010).

Another approach to modeling ecosystems is to resolve the flow of energy through the size-spectrum of the community as a whole (Jennings & Mackinson 2013). This greatly simplifies the problem, but excludes interactions between individual species. A third approach is to increase the complexity in only a specific part of an ecosystem (the “rhomboid” approach), resolving the species of greatest interest in more detail and decreasing ecosystem resolution as one moves away from the species or group of interest.

How can we generate future fish harvest scenarios?

Given the tremendous impact that fish harvest has on fish biomass, any attempt to predict future fish biomass must predict fish harvest as well. The IMBER CLIOTOP regional programme is moving towards developing future harvest scenarios that can be related to the IPCC Shared Social-economic Pathways (SSPs). The SSPs represent a range of possible futures, and helps conceptualize how societal actions could influence the collective future. Another strategy, developed by the PBL Netherlands Environmental Assessment Agency, is to assume that the world is sustainable in 2050, and develop possible pathways to arrive at that point. Given the growing importance of aquaculture, that sector must also be considered in combination with future changes in wild fish capture. The very important point was made that IMBER scientists have the capacity to alter the choices of decision-makers. This may require a clear idea of what a ‘sustainable’ 2050 would look like. Workshop participants agreed that carrying out sensitivity tests, to illustrate the impacts that different future fishing scenarios would have on harvest and biomass, would be a very useful next step.

FishMIP: a way forward

It was clear, in light of the discussion, that comparing different models provides great potential for moving forward in terms of model development. In addition, there is significant interest from the climate impacts community to incorporate projections of fish biomass and fish catch into upcoming assessments of global food security under climate change. To address both these targets, the Fish Model Intercomparison Project (FishMIP) will conduct a comparison of regional and global models, under both historical and future conditions. FishMIP is in the process of spinning up, and participation is open to any interested parties with applicable models. For more information or to participate in FishMIP, please contact one of the following: Heike Lotze (co-ordinator), Tyler Eddy (regional models), William Cheung or Eric Galbraith (global models).



  • Banse, K.M. (2014). Assessing ocean changes without data centers? Frontiers in Marine Science, 1 (29): 1-2.
  • Fulton, E.A. (2010). Approaches to end-to-end ecosystem models. Journal of Marine Systems, 81: 171-183.
  • Jennings, S., & Mackinson, S. (2003). Abundance-body mass relationships in size-structured food webs. Ecology Letters, 6(11): 971–974. DOI: 10.1046/j.1461-0248.2003.00529.x
  • Kearney, K.A., Stock, C., Aydin, K., & Sarmiento, J.L. (2012). Coupling planktonic ecosystem and fisheries food web models for a pelagic ecosystem: Description and validation for the subarctic Pacific. Ecological Modelling, 237-238, 43–62. DOI: 10.1016/j.ecolmodel.2012.04.006
  • Link, J. S., Ihde, T., Harvey, C., Gaichas, S. K., Field, J., Brodziak, J., Townsend, H., et al. (2012). Dealing with uncertainty in ecosystem models: the paradox of use for living marine resource management. Progress In Oceanography, 102: 102-114.
  • Maury, O. (2010). An overview of APECOSM, a spatialized mass balanced “Apex Predators ECOSystem Model” to study physiologically structured tuna population dynamics in their ecosystem. Progress In Oceanography, 84(1-2): 113–117. DOI: 10.1016/j.pocean.2009.09.013
  • Stock, C.A., et al. (2011). On the use of IPCC-class models to assess the impact of climate on Living Marine Resources. Progress In Oceanography, 88(1-4): 1–27. DOI: 10.1016/j.pocean.2010.09.001
  • Torney, C.J., Levin, S.A. & Couzin, I. D. (2010). Specialization and evolutionary branching within migratory populations. Proceedings of the National Academy of Sciences of the United States of America, 2010: 1–6. DOI:10.1073/pnas.1014316107
  • Watson, J.R., Hays, C., Raimondi, P., Mitarai, S., Dong, C., McWilliams, J. C., … Siegel, D.A. (2011). Currents connecting communities: nearshore community similarity and ocean circulation. Ecology, 92(6): 1193–1200. DOI: 10.1890/10-1436.1
banner 2nd transition NL

Mesopelagic fishes in the California Current: ecosystem role, climate change impacts and the need for global observations of marine fish populations

Anthony Koslow1, Peter Davison1,2 and Ana Lara-Lopez1,3

1Scripps Institution of Oceanography, University of California, SD, La Jolla, CA 92093 USA

2Farallon Institute for Advanced Ecosystem Research, Petaluma, CA 94952 USA

3Integrated Marine Observing System, University of Tasmania, Hobart, Tas 7001 Australia


Historically, monitoring of fish populations has largely been devolved to those agencies responsible for managing the world’s commercial fisheries; not surprisingly, their attention has focused on exploited fishes. Ecological monitoring of the global ocean has focused on lower trophic levels – the phytoplankton and zooplankton – and the so-called charismatic megafauna: sharks, turtles, marine mammals and seabirds. This has left several notable gaps, including the mesopelagic fishes, which comprise the most abundant and widespread vertebrates on the planet, being found in all open oceans except the Arctic. With a biomass conservatively estimated at 109 t (and probably close to an order of magnitude greater) (Gjøsaeter and Kawaguchi 1980, Irigoien et al. 2014), they are the only vertebrates whose biomass exceeds that of humans. These fishes, along with the entire deep pelagic realm, are exceptionally poorly studied (Robison 2009) and until recently, virtually nothing was known of their interannual variability or sensitivity to climate variations. A series of studies based on the California Cooperative Oceanic Fisheries Investigations (CalCOFI) is striving to fill some of these gaps.

Acoustic and trawl observations on the quarterly CalCOFI cruises that cover the southern California Current (CC) (Fig. 1A) now indicate that mesopelagic fishes are dominant consumers of zooplankton in the CC. Although mesopelagic fishes are less densely distributed than pelagic schooling fishes, they are virtually ubiquitous (Fig. 1B), such that their biomass across the CC is about an order of magnitude larger than the combined biomass of the pelagic schooling planktivores, Pacific sardine, northern anchovy, Pacific and jack mackerel (Table 1) (Davison et al. in press). However, because the metabolism of mesopelagic fishes is significantly lower than that of epipelagic planktivores (Childress et al. 1980), their weight-specific consumption is also lower. Simple mass-balance calculations indicate that the estimated food consumption of mesopelagic fishes is approximately equal to that of the combined epipelagic planktivores (Table 1). Their considerable prey consumption combines with vertical migration behaviour to make mesopelagic fishes biogeochemically significant. Carbon export due to the feeding of mesopelagic fishes and transport to deepwater is estimated to be 15-17% of total carbon export in the region (Davison et al. 2013). Thus, mesopelagic fishes are vital as conduits of energy and nutrients both trophically and biogeochemically.


Figure 1. (A) The core sampling grid for the CalCOFI sampling program. Zooplankton tows are carried out at each station and all ichthyoplankton are removed, identified and enumerated. (B) A composite of the relative daytime backscattering along the six CalCOFI transects with distance offshore. Except in a narrow coastal zone, the backscattering is dominated by mesopelagic fishes in the deep scattering layer. 


Table 1. The mean biomass (millions of tonnes (MT)) and estimated zooplankton prey consumption (MT/yr) from 1951 – 2011 of Pacific sardine, northern anchovy, Pacific mackerel, jack mackerel and mesopelagic vertical migrators and non-migrators. Biomass was estimated from stock assessments, acoustic surveys, and the CalCOFI ichthyoplankton time series; food consumption from biomass and metabolic and growth requirements.


How can we observe potential change in mesopelagic and other fish communities related to climate and other drivers? Ichthyoplankton data sets potentially provide time series on the abundance of a broad suite of fishes, because most species, including mesopelagics, can be readily sampled with simple plankton nets in the upper 200 m of the water column as eggs and larvae. Because the larvae are sampled predominantly at a very early stage, their abundance serves as a proxy for spawning stock biomass.

Since 1951, CalCOFI has systematically sampled the ichthyoplankton of the CC. The dominant pattern (or principal component (PC) 1), which explained 20.5% of the variance in this data set, was based on synchronous changes in the abundance of some 24 taxa of midwater fishes, which closely followed changes in midwater oxygen concentrations (Koslow et al. 2011) (Fig. 2). Their abundance was 63% lower in the 1950s and 2000s, periods of reduced midwater oxygen concentrations compared with the period from 1970 – 1995. Reduced oxygen concentrations are now observed globally in the deep ocean (Helm et al. 2011), and global climate models predict that midwater oxygen concentrations will continue to decline as global warming and increased precipitation in high latitudes increase stratification and reduce deepwater ventilation (Matear & Hirst 2003, Oschlies et al. 2008). Remarkably, there appear to be no other time series in the Pacific Ocean and, perhaps even globally, to track changes in midwater fish populations (Koslow and Couture 2013).


Figure 2. Time series of principal component 1 (PC 1), which predominantly represents changes in abundance of 24 midwater fish species, and mean oxygen concentrations at 200 – 400 m depth in the southern California Current, 1951 – 2008.

PC 2, which explained 12.4% of the variance, was dominated by an assemblage that has been characterized as having northerly affinities (Moser et al. 1987), including six of the seven most abundant taxa in the ichthyoplankton data set: northern anchovy, Pacific hake, Pacific sardine, rockfishes (Sebastes spp.), and two of the most abundant midwater fishes, which were not highly correlated with PC 1, northern lampfish (Stenobrachius leucopsarus) and California smoothtongue (Leuroglossus stilbius) (Koslow et al. 2013). All except Pacific sardine have declined significantly since the late 1960s, leading to an decline of 72% in mean larval fish abundance across all taxa in the southern CC (Koslow et al. submitted). This decline in the CalCOFI ichthyoplankton time series is mirrored by the 78% decline reported for nearshore fishes off southern California based on sampling the cooling-water intakes of power plants along the coast (Miller and McGowan 2013) (MG). In fact the decline in the two data sets is highly correlated, with a correlation of 0.85 between CalCOFI ichthyoplankton PC 2 and MG PC 1 (Fig. 3). Both patterns are correlated with changes in oceanographic conditions in the CC: increased sea-surface temperatures, reduced upwelling, and declining transport of nutrient-rich water into the southern CC, along with patterns of the ENSO cycle, the Pacific Decadal Oscillation, and North Pacific Gyre Oscillation (Koslow et al. submitted).


Figure 3. Time series for PC 1 for nearshore fishes sampled by power plant intakes from Miller and McGowan (2013) (PC1 PPI) and PC 2 from analysis of CalCOFI data (PC 2 CalCOFI). The red line marks an apparently significant shift in abundance in 1989. 

The dramatic decline of broad assemblages of nearshore, mesopelagic, epipelagic and other fishes of the CC is of profound concern, if the changes over the past several decades are a portent of future trends related to climate change. However, these changes were only observed because the CalCOFI program maintains time series for the broad fish community: > 400 taxa are presently identified and enumerated (McClatchie 2014). However, there are no comparable time series with > 20 years data across the Pacific Ocean (Koslow and Couture 2013, submitted). The global community is poorly prepared to assess change in the world’s marine fish communities, other than in the relatively few species of commercial interest. How has this situation come to pass?

Koslow and Couture (2013) point, first, to the failure of the global ocean observation community to go beyond the low hanging fruit: to undertake consistent, systematic measurements of ecologically-relevant variables in the LMEs of the world ocean. Temperature, salinity and chlorophyll are today well-observed from satellites, Argo floats, and coastal platforms; however, the Global Ocean Observing System (GOOS), which was established in the early 1990s, still has not constituted its Biology and Ecology Panel. There is also generally little collaboration between fisheries programs (run by fisheries agencies) and oceanographic and climate observation programs. CalCOFI is a notable exception.

CalCOFI is often viewed as the Cadillac of ocean observation programs. However, its cost (~US$5 million pa), if extended to the approximately 50 large marine ecosystems of the world ocean, would be about US$250 million. This is about 1.5% of NASA’s annual budget (~$18 billion) for space exploration. There can be no adequate ocean stewardship without adequate observations. Is the priority for conservation of our global ocean really two orders of magnitude less than for space exploration? This ballpark estimate for global ocean observations also neglects the fact that there are already ichthyoplankton and trawl surveys throughout much of the world’s oceans. What is missing is collaboration between fisheries and oceanographic institutions and a modest increment in funding and effort to track all species, not just those of commercial value and to take the relevant physical and chemical measurements to put the ecology into an oceanographic context. Furthermore, while a spatially-expanded grid of stations may be necessary for stock assessment of broadly distributed species, a transect covering key habitats (e.g. shelf, slope and offshore) and occupied monthly to quarterly is likely adequate to monitor regional fish communities based on their larvae (Koslow and Wright, in prep.). 

The fish communities of the CC are undergoing massive change, apparently related to climate. It is unlikely that these changes are unique. However, without systematic observations and time series it is not possible to know their extent. It is time for the global community to take up the challenge of developing a representative global network of consistent, systematic observations of the ocean’s ecology, including its fishes, along with its physical state, its chemistry and biogeochemistry.



  • Checkley, D., J. Alheit, Y. Oozeki, C. Roy (2009). Climate Change and Small Pelagic Fish. Cambridge University Press, Cambridge: 392 pp.
  • Childress, J.J., S.M. Taylor, G.M. Cailliet, M.H. Price (1980). Patterns of growth, energy utilization and reproduction in some meso- and bathypelagic fishes off Southern California. Marine Biology, 61: 27-40.
  • Davison, P., A. Lara-Lopez, J.A. Koslow (In press). Mesopelagic fish biomass in the southern California Current Ecosystem. Deep Sea Res. II.
  • Davison, P.C., D.M. Checkley Jr, J.A. Koslow, J. Barlow (2013). Carbon export mediated by mesopelagic fishes in the northeast Pacific Ocean. Progress in Oceanography, 116: 14–30.
  • Gjfsaeter, J., K. Kawaguchi (1980). A review of the world resources of mesopelagic fish. FAO Fisheries Technical Paper, 193: 1-151.
  • Helm, K.P., N.L. Bindoff, J.A. Church (2011). Observed decreases in oxygen content of the global ocean. Geoph. Res. Let, 38: L23602.
  • Irigoien, X., T.A. Klevjer, A. Rfstad, U. Martinez, G. Boyra, J.L. Acuña, A. Bode, F. Echevarria, J.I. Gonzalez-Gordillo, S. Hernandez-Leon, S. Agusti, D.L. Aksnes, C.M. Duarte, S. Kaartvedt (2014). Large mesopelagic fishes biomass and trophic efficiency in the open ocean. Nat Commun, 5.
  • Koslow, J.A., J. Couture (2013). Follow the fish. Nature, 502: 163-164.
  • Koslow, J.A., J. Couture (Submitted). Pacific Ocean observation programs and our ability to assess the state of the ocean. Marine Policy.
  • Koslow, J.A., R. Goericke, A. Lara-Lopez, W. Watson (2011). Impact of declining intermediate-water oxygen on deepwater fishes in the California Current. Mar. Ecol. Prog. Ser, 436: 207–218.
  • Koslow, J.A., R. Goericke, W. Watson (2013). Fish assemblages in the Southern California Current: relationships with climate, 1951 – 2008. Fisheries Oceanography, 22: 207–219.
  • Koslow, J.A., E.F. Miller, J.A. McGowan (Submitted). Trends in fish abundance in the southern California Current Ecosystem: comparison of two independent time series. Marine Ecology Progress Series.
  • Koslow, J.A., M. Wright (In prep.). Sampling requirements for long-term ecological monitoring of fish communities.
  • Matear, R.J., A.C. Hirst (2003). Long-term changes in dissolved oxygen concentrations in the ocean caused by protracted global warming. Global Biogeochemical Cycles, 17(4): 1125.
  • McClatchie, S. (2014). Regional Fisheries Oceanography of the California Current System: the CalCOFI program. Springer, Dordrecht: 235 pp.
  • Miller, E.F., J.A. McGowan (2013) Faunal shift in southern California’s coastal fishes: A new assemblage and trophic structure takes hold. Estuarine coastal and shelf science, 127: 29-36.
  • Moser, H.G., P.E. Smith, L.E. Eber (1987). Larval fish assemblages in the California Current region, 1954-1960, a period of dynamic environmental change. CALCOFI Reports, 28: 91-127.
  • Oschlies, A., K.G. Schultz, U. Riebesell, A. Schmittner (2008). Simulated 21st century’s increase in oceanic suboxia in CO2-enhanced biotic carbon export. Global Biogeochem. Cycles, 22: GB4008.
  • Robison, B. (2009). Conservation of deep pelagic biodiversity. Conservation Biology, 23: 847–858.
  • Matear, R.J., A.C. Hirst (2003). Long-term changes in dissolved oxygen concentrations in the ocean caused by protracted global warming. Global Biogeochemical Cycles, 17(4): 1125.
  • McClatchie, S. (2014). Regional Fisheries Oceanography of the California Current System: the CalCOFI program. Springer, Dordrecht: 235 pp.
  • Miller, E.F., J.A. McGowan (2013). Faunal shift in southern California’s coastal fishes: A new assemblage and trophic structure takes hold. Estuarine coastal and shelf science, 127: 29-36.
  • Moser, H.G., P.E. Smith, L.E. Eber (1987). Larval fish assemblages in the California Current region, 1954-1960, a period of dynamic environmental change. CALCOFI Reports, 28: 91-127.
  • Oschlies, A., K.G. Schultz, U. Riebesell, A. Schmittner (2008). Simulated 21st century’s increase in oceanic suboxia in CO2-enhanced biotic carbon export. Global Biogeochem. Cycles, 22: GB4008.
  • Robison, B. (2009). Conservation of deep pelagic biodiversity. Conservation Biology, 23: 847–858.
banner 2nd transition NL

From watching to acting: adaptation in marine systems

Alistair Hobday

Co-chair of the IMBER CLIOTOP Regional Programme
CSIRO Oceans and Atmosphere, Hobart, Tasmania, Australia


Much of the focus of the climate change scientific community, including many scientists involved in IMBER programmes, is on increasing our understanding about the effect of increased greenhouse gas emissions on the climate system and the responses of the oceans. These efforts inform global mitigation efforts, and have emphasised the urgency of curbing greenhouse gas emission in order to limit the impacts of global warming and ocean acidification. Without effective mitigation, dangerous climate change, defined elsewhere as >2°C of warming above pre-industrial temperatures, is looking increasingly likely. The changes already seen in oceanic systems have impacts on living marine systems, at all levels of the food chain, and ultimate impacts on the continued delivery of ecosystem goods and services on which human societies depend. Without discounting the importance of science that informs the development of mitigation policy and action, it is becoming apparent given the time available and the remaining equivalent carbon dioxide (CO2e) that can be emitted before we exceed 2°C of warming, that there are considerable challenges for many species and human societies around the world. Many experts are now suggesting that planning for 4°C of warming is wise (Stafford-Smith et al. 2011). 


Adaptation – in a climate change context – refers to the autonomous or directed change in a species or human system in response to global warming. Adaptation science was the poor relation to mitigation for a long time, perhaps because it was feared by many, including climate scientists, as a distraction to international efforts to curb greenhouse gas emissions. However, given recent evidence of the impacts of climate change on species and ecosystems in response to the 0.7°C global warming observed to date, it is apparent that many will not be able to naturally keep pace with the rate of climate change over the coming centuries. Many dire predictions for widespread extinction have been made, including irreversible damage to coral reef and other coastal ecosystems. In offshore marine ecosystems, long considered robust to anthropogenic impacts, climate change is one of the most pervasive stressors. Responses to temperature, oxygen, pH changes have been associated with changes in distribution, abundance, phenology and physiology of phytoplankton, zooplankton, fishes such as tuna and marlin, and even to seabirds and marine mammals. The fishing industries that harvest some of these species have also begun to respond to the impacts of climate change, with further large scale changes in the location of human activities projected under future climate scenarios (Bell et al. 2013).

Given that marine species and ecosystems are vulnerable to climate change, and that impacts are already widely observed, with more rapid change to come, developing adaptation strategies for species and systems is becoming important. These strategies will involve intervention – that some may see as attempting to play God before we understand the system. But, we do not have the luxury of studying these systems for another ten or 20 years before we act – this period of time requires some different intervention strategies to be attempted before the situation becomes dire.

NL27-Hobday- Autonomous adaptation

Some scientists may be sceptical that there is much that can be done for marine species in the face of widespread physical change, but examples of adaptation strategies are emerging for a limited number of coastal species (Koehn et al. 2011). Iconic species, particularly those occupying land-based nesting colonies, such as seabirds, marine turtles and sea lions, have been an initial focus in Australia (Fig. 4). Interventions that reduce mortality have emerged from conservation biology experience, and include translocation to new areas (reduce the exposure to climate change), habitat modification (reduce the sensitivity to climate change), and predator control (increase the adaptive capacity) (Hobday et al. 2014). Unfortunately, interventions that are suitable for multiple species appear rare at this stage, and so prioritization is important – which species and where will our efforts be spent?


Figure 4. Adaptation efforts for marine species have been targeted at iconic, coastal or land-based parts of the life history, such as at this albatross colony, when access and intervention are easiest.

Applying these options to offshore pelagic species, such as tuna, that are distributed over large areas will be challenging, but the thinking in this domain is beginning (Salinger & Hobday 2013). For example, genetic intervention, either indirectly through the introduction of heat or acidification tolerant individuals, or directly through selective breeding and release of adapted individuals, may be possible. Engineering may provide new habitats, such as floating islands, in places previously unsuitable for a particular species. Many options may appear novel now, and maturity is needed to have useful discussions around novel ideas. Today’s out-there ideas may become possible tomorrow.

Central to the development of adaptation strategies must be acceptance of an adaptive and flexible approach – termed adaptive management in some sectors (Fig. 5). This involves learning from and changing efforts in response to initial attempts. Monitoring these efforts will be critical. In fact, under an adaptation lens, monitoring is not just about documenting change, it can also inform new approaches, and ocean observing programs that expand to include more biological variables are crucial. 


Figure 5. It is important to test, learn and refine adaptation approaches that may reduce vulnerability under climate change, and this testing should occur before options are foreclosed. Source; Stein et al 2013.

Adaptation efforts alone will not solve the problems associated with climate change, but a global change research agenda that does not include adaptation will be great at documenting change, but will offer society fewer options to respond to change. IMBER programmes, such as CLIOTOP, are expanding efforts to consider directed adaptation responses for marine species and the dependent humans (Hobday et al. 2013), and these represent complimentary efforts to mitigation. Our cleverest thinkers should be attracted to adaptation science, as without additional efforts to develop and implement adaptation options, our oceans will not deliver the benefits we depend on for future well-being, nor will they contain the abundance of species that delight and provide for us today.



  • Bell, J.D., A. Ganachaud, P.C. Gehrke, S.P. Griffiths, A.J. Hobday, O. Hoegh-Guldberg, J.E. Johnson, R. Le Borgne, P. Lehodey, J.M. Lough, R.J. Matear, T.D. Pickering, M.S. Pratchett, A. Sen Gupta, I. Senina and M. Waycott (2013). Mixed responses of tropical Pacific fisheries and aquaculture to climate change. Nature Climate Change, 3(6): 591-599. DOI: 10.1038/NCLIMATE1838.
  • Hobday, A.J., J.W. Young, O. Abe, D.P. Costa, R.K. Cowen, K. Evans, M.A. Gasalla, R. Kloser, O. Maury and K.C. Weng (2013). Climate Impacts and Oceanic Top Predators: Moving from impacts to adaptation in oceanic systems. Reviews in Fish Biology and Fisheries, 23(4): 537-546. DOI: 10.1007/s11160-013-9311-0.
  • Hobday, A.J., L.E. Chambers, J.P.Y. Arnould, T.A. Patterson, C. Wilcox, G.N. Tuck and R.B. Thomson (2014). Developing adaptation options for seabirds and marine mammals impacted by climate change. Final Report. FRDC-DCCEE Marine National Adaptation Research Project 2010/0533.
  • Koehn, J.D., A.J. Hobday, M.S. Pratchett and B.M. Gillanders (2011). Climate change and Australian marine and freshwater environments, fishes and fisheries: synthesis and options for adaptation. Marine and Freshwater Research, 62(9): 1148-1164.
  • Salinger, M. J. and A. J. Hobday (2013). Safeguarding the future of oceanic fisheries under climate change depends on timely preparation. Climatic Change. DOI: 10.1007/s10584-012-0609-z.
  • Stafford-Smith, M., L. Horrocks, A. Harvey and C. Hamilton (2011). Rethinking adaptation for a 4°C world. Philosophical Transactions of the Royal Society A,  Mathematical, Physical and Engineering Sciences, 369: 196-216.
  • Stein, B.A., A. Staudt, M.S. Cross, N.S. Dubois, C. Enquist, R. Griffis, L.J. Hansen, J.J. Hellmann, J.J. Lawler, E.J. Nelson and A. Pairis (2013). Preparing for and managing change: climate adaptation for biodiversity and ecosystems. Frontiers in Ecology and Environment 11(9): 502–510. DOI:10.1890/120277.
banner 2nd transition NL

Trichodesmium Growth Rates: Modelling the Fundamental Niche

Tobias Boatman, Graham Upton and Richard Geider

School of Biological Sciences, University of Essex, Colchester, CO4 3SQ, UK

E-mail: tboatm@essex.ac.uk

Trichodesmium spp. are diazotrophic cyanobacteria found in tropical regions of the open oceans, and are significant contributors to both the global C and N cycles. In the North Atlantic, they account for between 8-47% of total primary production (Carpenter et al. 1997). Global estimates of annual atmospheric N2 fixation are between 100–200Tg; with Trichodesmium spp. accounting for 20 (non-blooming) to 40 Tg (blooming) (Capone et al. 1997; Carpenter & Capone, 1992; Tyrrell et al. 2003). In the past decade, a growing body of research has focused on how Trichodesmium spp. might respond physiologically to the effects of climate change (i.e. increasing sea surface temperatures (SST’s), CO2, etc.) (Barcelos e Ramos et al. 2007; Breitbarth et al. 2007; Garcia et al. 2011; Hutchins et al. 2007; Levitan et al. 2007). With regard to the effect on growth, the literature is somewhat divided about the direction and degree of the response. This is important as changes to growth could have major implications for global biogeochemical cycles, food web dynamics, and the overall productivity of the open oceans.

Cyanobacteria are often subjected to a variety of co-occurring and potentially co-limiting factors (Mills et al. 2004; Moore et al. 2001). For example, DIC varies with salinity and SST’s; surface nitrate depends upon stratification, upwelling, eddy pumping and wave action; iron is uniquely influenced by Aeolian dust deposition; and irradiance varies with incident light, mixing depth and vertical light attenuation (Post et al. 1985).

The main aim of this research was to redefine the combined effects of CO2, temperature and light intensity on Trichodesmium erythraeum IMS101 growth, and to address some of the dissimilarities found in literature. Response curves were used to develop a growth rate function for Trichodesmium IM101, to model global distributions during three distinct time periods (Last glacial maximum (LGM), Present, and Future (est. year 2100)). Prior to data collection, it was important to ensure that the response curves were of high resolution, particularly at the temperature extremes for the forecasted increases in SST’s during the coming decades. It was also crucial that the carbon chemistry was well defined, and that the cultures were properly acclimated to the conditions and growing at balanced growth.

To achieve high data resolution we opted to grow many replicates at low volumes and low biomass to minimise self-shading and CO2 drawdown. Thus, Trichodesmium IMS101 was semi-continuously cultured (5ml volumes) in gas tight glass test tubes in a custom-made water-jacketed metal temperature block. To measure the effect of temperature on growth, the cultures were incubated between 19 and 32°C in approximately 0.5°C increments (Fig 6). At each temperature, the cultures were exposed to low and high light treatments (40µmol m-2s-1 ± 0.4 and 400µmol m-2 s-1 ± 4, respectively), and three CO2 treatments relating to glacial, present and future concentrations (180, 380 and 720ppm, respectively). This resulted in ~120 cultures. To measure the light response on growth a second temperature block was used concurrently. Here the cultures were incubated at optimum temperature (26°C), and the photon flux density was varied using neural density filters to give ~10 light treatments ranging 20 to 500µmol m-2 s-1 at the same three CO2 concentrations, thus giving a total of ~30 cultures. The post-culturing CO2 concentration was calculated using CO2SYS software (Pierrot et al, 2006), with an initial measurement of alkalinity and a post-culturing measurement of pH (Kranz et al. 2010).


Figure 6. Growth rates were quantified daily by fluorescence measurements (Fo) on dark adapted cultures using a FRRFII Fastact Fluorometer (Chelsea Instruments). The cultures were kept in the exponential growth phase at low densities by periodic dilution with YBCII media (0.5ml culture to 4.5ml media); this prevented nutrient limitation, self-shading, and DIC (< 5%) and CO2 (< 75ppm) drawdown (dashed line is the CO2 at the start of culturing). The cultures were maintained in the exponential growth phase for up to nine months to ensure balanced growth had been achieved across the entire temperature and light gradient. Post data collection, all raw data files were processed through an R code, to objectively analyse the data. Firstly, to remove growth rates associated with lag phases, secondly to fit a linear trend line to the natural log of the growth curve associated with the exponential phase, and thirdly to differentiate between the initial (changing) growth rates related to acclimation, and the end growth rates (constant) related to balanced growth. Having calculated the median growth rate of each treatment under balanced growth, the data was used to formulate a function describing how growth rate varied with temperature, pCO2 and photon flux density.

A multiplicative combined light and temperature growth rate function was used to model the growth for each of the three CO2 treatments. If a function parameter did not significantly differ between CO2 treatments it was made into a constant. After parameter optimisation, a 4-parameter growth rate function was obtained which described Trichodesmium’s growth rate at any given temperature and light exposure at any of the three CO2 concentrations.


Figure 7. Trichodesmium growth was modelled at the mixed layer depth (MLD). For this we used the monthly means (1° long/lat resolution) for February and August (Monterey and Levitus, 1997). Data for SST at the MLD was obtained from the CLIMAP 18K model (LGM) ,World Ocean Atlas (WOA) 09 (Present) (Antonov et al, 2010), and GFDL Coupled Climate Model (2.1), based on the IPCC SRESA1B scenario (Future, 2100) (Knutson et al, 2007). Net surface radiation was obtained from the GISS global climate model (CM2.0) (Hansen et al, 1983), converted into a photon flux density and then into light exposure (Forsythe et al, 1995). The Lambert Beer equation was used to model the light exposure at the MLD (Stavn, 1988), having calculated the light attenuation coefficient (Morel, 1988) using global satellite derived chlorophyll from NASA’s Aqua MODIS satellite.

Across all three time periods a seasonal variability is evident, with higher growth rates in the northern hemisphere during summer, and lower growth rates during winter (Fig. 7). During the LGM Trichodesmium IMS101 would have been constrained to a 30° N/S latitudinal band, with peaks in growth no higher than ~0.15 d-1. In the present, the distribution has increased further north and south in the Atlantic and Indian Oceans, and the peaks in growth rates are much higher at ~0.35 d-1. The slight increase in aerial distribution is a temperature-driven response, while the major increase in the peak growth rate is CO2 driven. In the year 2100 scenario, the maximum growth rates of Trichodesmium IMS101 are comparable to current rates, confirming some of the suggestions in the literature that the carbon concentrating mechanism (CCM) is fully saturated, or very close to saturation under present CO2 concentrations. What is different is the further increase in aerial distribution in the Atlantic and Indian Oceans, and the sudden decline around the equatorial regions. The poleward shift is driven by the increase in SST’s which has potentially opened up new areas of the ocean for growth. However, the SST’s at the equator have risen above the maximum temperature for growth (Tmax), changing the current single latitudinal aerial distribution into two distinct bands. It is this trade-off between reductions in the equatorial latitudes (15° N/S) versus gains in the marginalised latitudes (~25 - 35° N/S) which is of most importance given the direction and speed at which climate change is occurring. A potentially perturbing consideration are Trichodesmium’s physiological needs for oligotrophic conditions, as found in tropical areas. Although these potentially available higher latitudes may well have some of physical requirements for growth (i.e. SST, light exposure, etc.), other factors not considered in this model (e.g. competition, grazing pressure, iron limitation, etc.) will all play a critical role in defining Trichodesmium’s distribution. Here we report on Trichodesmium’s fundamental niche moving ever polewards, but in reality the distribution could be constrained to a much smaller aerial distribution than depicted here. Further work is required for other Trichodesmium species, as well as other key dizotrophic cyanobacteria. Other factors should also be considered within the model in order to develop from a fundamental niche to a realised niche for Trichodesmium. Using the same experimental setup as described above; we have collected data regarding the growth response of Trichodesmium IMS101 under varying Fe, photon flux densities and CO2 concentrations (~30 treatments). Current efforts are being put into developing the growth rate function to incorporate Fe limitation to gain a better understanding of what the distribution of Trichodesmium spp. may be in the coming decades.


Many thanks go to the Natural Environmental Research Council (NERC) for funding the research, and to IMBER for a travel grant to enable the author to attend the Future Oceans conference.


  • Antonov, J., Seidov, D., Boyer, T., Locarnini, R., Mishonov, A., Garcia, H., Baranova, O., Zweng, M. and Johnson, D. (2010). World Ocean Atlas 2009, Volume 2: Salinity. S. Levitus, Ed. NOAA Atlas NESDIS, 69.
  • Barcelos e Ramos, J., Biswas, H., Schulz, K.G., LaRoche, J. and Riebesell, U. (2007). Effect of rising atmospheric carbon dioxide on the marine nitrogen fixer Trichodesmium. Global biogeochemical cycles, 21, GB2028, doi: 10.1029/2006GB002898.   
  • Breitbarth, E., Oschlies, A. and Laroche, J. (2007). Physiological constraints on the global distribution of Trichodesmium - effect of temperature on diazotrophy. Biogeosciences, 4 (1): 53-61.
  • Capone, D.G., Zehr, J.P., Paerl, H.W., Bergman, B. and Carpenter, E.J. (1997). Trichodesmium, a Globally Significant Marine Cyanobacterium. Science, 276 (5316): 1221-1229.
  • Carpenter, E.J. and Capone, D.G. (1992). Nitrogen fixation in Trichodesmium blooms. Marine pelagic cyanobacteria: Trichodesmium and other diazotrophs, 211-217.
  • Carpenter, E.J., Harvey, H.R., Fry, B. and Capone, D.G. (1997). Biogeochemical tracers of the marine cyanobacterium Trichodesmium. Deep Sea Research Part I: Oceanographic Research Papers, 44 (1): 27-38.
  • Forsythe, W.C., Rykiel Jr, E.J., Stahl, R.S., Wu, H.-I. and Schoolfield, R.M. (1995). A model comparison for daylength as a function of latitude and day of year. Ecological Modelling, 80 (1): 87-95.
  • Garcia, N.S., Fu, F.-X., Breene, C.L., Bernhardt, P.W., Mulholland, M.R., Sohm, J.A. and Hutchins, D.A. (2011). Interactive effects of Irradiance and CO2 on CO2 fixation and N2 fixation in the Diazotroph Trichodesmium erythraeum (Cyanobacteria). Journal of Phycology, 47 (6): 1292-1303.
  • Hansen, J., Russell, G., Rind, D., Stone, P., Lacis, A., Lebedeff, S., Ruedy, R. and Travis, L. (1983). Efficient three-dimensional global models for climate studies: Models I and II. Monthly Weather Review, 111 (4): 609-662.
  • Hutchins, D., Fu, F. X., Zhang, Y., Warner, M., Feng, Y., Portune, K., Bernhardt, P. and Mulholland, M. (2007). CO2 control of Trichodesmium N2 fixation, photosynthesis, growth rates, and elemental ratios: implications for past, present, and future ocean biogeochemistry. Limnology and Oceanography, 52 (4): 1293-1304.
  • Knutson, T.R., Sirutis, J.J., Garner, S.T., Held, I.M. and Tuleya, R.E. (2007). Simulation of the recent multidecadal increase of Atlantic hurricane activity using an 18-km-grid regional model. Bulletin of the American Meteorological Society, 88 (10): 1549-1565.
  • Kranz, S., Wolf-Gladrow, D., Nehrke, G., Langer, G. and Rost, B. (2010). Calcium carbonate precipitation induced by the growth of the marine cyanobacteria Trichodesmium. Limnology and Oceanography, 55 (6): 2563-2569.
  • Levitan, O., Rosenberg, G., Setlik, I., Setlikova, E., Grigel, J., Klepetar, J., Prasil, O. and Berman-Frank, I. (2007). Elevated CO2 enhances nitrogen fixation and growth in the marine cyanobacterium Trichodesmium. Global Change Biology, 13 (2): 531-538.
  • Mills, M.M., Ridame, C., Davey, M., La Roche, J. and Geider, R. J. (2004). Iron and phosphorus co-limit nitrogen fixation in the eastern tropical North Atlantic. Nature, 429 (6989): 292-294.
  • Monterey, G.I. and Levitus, S. (1997). Seasonal Variability of Mixed Layer Depth for the World Ocean. National Oceanic and Atmospheric Administration, National Environmental Satellite, Data, and Information Service.
  • Moore, J.K., Doney, S.C., Glover, D.M. and Fung, I.Y. (2001). Iron cycling and nutrient-limitation patterns in surface waters of the World Ocean. Deep Sea Research Part II: Topical Studies in Oceanography, 49 (1–3): 463-507.
  • Morel, A. (1988). Optical modeling of the upper ocean in relation to its biogenous matter content (case I waters). Journal of Geophysical Research: Oceans (1978–2012), 93 (C9): 10749-10768.
  • Pierrot, D., Lewis, D. and Wallace, D. (2006). CO2SYS. EXE-MS Excel Program Developed for CO2 system calculations. ORNL/CDIAC-105a. Oak Ridge, TN: Carbon Dioxide Information Center, Oak Ridge National Laboratory, US Department of Energy,
  • Post, A.F., de Wit, R. and Mur, L.R. (1985). Interactions between temperature and light intensity on growth and photosynthesis of the cyanobacterium Oscillatoria agardhii. Journal of Plankton Research, 7 (4): 487-495.
  • Stavn, R.H. (1988). Lambert-Beer law in ocean waters: optical properties of water and of dissolved/suspended material, optical energy budgets. Applied optics, 27 (2): 222-231.
  • Tyrrell, T., Marañón, E., Poulton, A.J., Bowie, A.R., Harbour, D.S. and Woodward, E.M.S. (2003). Large-scale latitudinal distribution of Trichodesmium spp. in the Atlantic Ocean. Journal of Plankton Research, 25 (4): 405-416.
banner 2nd transition NL

Phaeocystis pouchetii bloom from the perspective of heterotrophic bacteria

Bernadette Pree1*,  Aud Larsen2, Tron Frede Thingstad1, Marc E. Frischer3, Diane Stoecker4, Andrey F. Sazhin5, Hans Henrik Jakobsen6, Paolo Simonelliand Jens C. Nejstgaard7

1Department of Biology, University of Bergen, Norway

*corresponding author, e-mail: bernadette.pree@bio.uib.no

2Uni Environment, Uni Research, Bergen, Norway

3Skidaway Institute of Oceanography, Savannah, Georgia, USA

4University of Maryland Center for Environmental Science, Horn Point Laboratory, Cambridge, USA

5P.P. Shirshov Institute of Oceanology RAS, Moscow, Russia

6Aarhus University, Bioscience, Roskilde, Denmark

7Leibnitz-Institute of Freshwater Ecology and Inland Fisheries (IGB), Stechlin, Germany


The colony-forming algae Phaeocystis pouchetii seasonally forms dense blooms at higher latitudes and can dominate the pelagic biomass at certain periods (reviewed by Verity et al. 2007). The carbohydrate-rich mucus from the senescing blooms of P. pouchetii represents a substantial input of C-rich organic matter to the planktonic food web (reviewed by Alderkamp et al. 2007).  To what extent this material is degraded within the microbial part of the food web or transported up the food chain is not well understood (see review Nejstgaard et al. 2007).

This question was tackled as a component of the Norwegian Research Council funded program “PHAEONIGMA”.  In May 2013, aboard the research vessel Håkon Mosby, the PHAEONIGMA research team explored a chronological gradient of P. pouchetii blooms in the Porsangerfjord (ca. 70°32’N 26°31’E). At the time of the expedition, the inner section of the fjord was experiencing the early phases of a P. pouchetii bloom, while a more developed bloom was present in the outer part of the fjord. Taking advantage of this natural experiment, we quantified the abundance, production, growth rate and grazing mortality of heterotrophic bacteria associated with each bloom phase.


Figure 8. The RV Håkon Mosby in the Porsangerfjord, northern Norway and the research team - eager to solve the enigma of the haptophyte Phaeocystis pouchetii (Photo by: Jessica L. Ray, Uni Research).

Different blooming conditions in the Porsangerfjord

The status of the bloom was assessed based on several criteria. Live colonies of P. pouchetii were imaged using a FlowCAM® instrument, enabling us to determine their size and shape in real time (Fig. 9). Colonies in the inner Porsangerfjord were more compact and spherical than those in the outer Porsangerfjord. Additionally, the FlowCAM images from the outer section of the fjord revealed overgrowth of the colonies by diatoms, a sign of progression of the Phaeocystis bloom (Fig. 9b, Sazhin et al. 2007). Microscopy showed that solitary, motile cells of P. pouchetii from the inner fjord were slightly smaller (2-4 µm) than those in the outer fjord (3-4 µm), indicating an earlier stage of the bloom (personal observations). The average phaeophytin concentrations were higher in the inner (6.2 µg Chl a L-1, 4.6 µg phaeophytin L-1) than in the outer fjord (6.0 µg Chl L-1, 0.6 µg phaeophytin L-1) indicating decay of the bloom. The inner section also exhibited a greater diversity of phytoplankton communities than the outer fjord. It is thus, likely that the Phaeocystis bloom in the inner Porsangerfjord started soon after a diatom bloom, and was at an earlier stage than the almost monospecific bloom in the outer Porsangerfjord. Nutrient concentrations of nitrate, phosphate and silicate support this bloom-status description. Nutrients were depleted in the inner Porsangerfjord, where the early-stage Phaeocystis bloom occurred simultaneously with the decaying diatom bloom, but were available in the outer Porsangerfjord, where the Phaeocystis bloom was more advanced.


Figure 9. FlowCAM images during the cruise of an early (a) and later (b) stage of blooming colonies of P. pouchetii. Early stages are characterized by dense, compact, spherical shapes, whereas later stages start to be overgrown by diatoms and the P. pouchetii cells less compact.


Bacterial response to blooming colonies of P. pouchetii

Bacterial abundance and productivity differed between the two stations and stages of the blooms. During the early phase of the Phaeocystis bloom in the inner Porsangerfjord, bacterial concentration at the Chl a maximum (20 m) was 7.5 x 105 cells ml-1 and bacterial production (BP) was 0.3 µg C l-1h-1. In the outer Porsangerfjord where the bloom was more developed, bacteria concentration and production at Chl a max were 3x lower (2.0 x 105cells ml-1 and 0.1 µg C l-1 h-1, respectively).

Dilution experiments suggested that heterotrophic nanoflagellates and microzooplankton consumed 13% of the standing stock of heterotrophic bacteria each day during the early phase of the P. pouchetii bloom. In the late bloom phase, the bacterivory rate was higher, and grazers removed 36% of the standing stock per day. However, measurements of bacterial production showed that BP cell-1 (specific growth) varied between the different dilutions following 24 hours incubation, and was significantly higher in diluted rather than undiluted samples (Fig. 10). This violates an important assumption of the dilution technique, namely that dilution does not affect specific growth of cells (Landry and Hassett 1982). Consequently, bacterial grazing rates derived from studies during the later phase of the Phaeocystis bloom should be treated with caution. Nevertheless, it is interesting that dilution experiments with water from early and late bloom stages, show different results in terms of specific growth. One explanation is that the release of organic nutrients during handling of water for grazing experiments in older P. pouchetii colonies, may stimulate bacterial growth. An alternative hypothesis is that dilution of inhibitory substances released by Phaeocystis colonies could provide a substrate for bacteria, increasing their productivity. Lastly, the nutritional value of Phaeocystis in the colonial stage is low due to the high C:N ratio compared to other phytoplankters. Chemical analysis of the samples is still ongoing and may help to determine the correct hypothesis, explaining the different outcomes of the two dilution experiments.

Methodologically, our findings suggest the need to modify the dilution method for bacterial grazing by including measurements of BP to verify that dilution and incubation do not affect bacteria specific growth rates.


Figure 10. Bacterial production per cell at the beginning and end of the experiment in the inner (a) and outer (b) Porsangerfjord.

Preliminary conclusions and outlook

This study suggests that bacteria may respond to different phases of a P. pouchetii bloom. This potentially has implications for the carbon dynamics and trophic structure of the system. Bacteria are more abundant and productive during the early phase of the bloom, possibly because of the decline of a previous diatom bloom. Bacterial abundance and production were 3x less during a later and almost monospecific P. pouchetii bloom. There was an increase in the specific growth of bacteria in the diluted water samples from the later phase of the bloom. This is puzzling, but underlines the importance of bloom age when considering phytoplankton dynamics, and needs further investigation.

Our results suggest that for reasons not yet clearly understood, the colonial stage of P. pouchetii blooms provide unfavourable substrates for bacterial growth and production. Data from natural blooms and from mesocosm experiments are available and will be further investigated.


The study is supported by MINOS (ERC grant 250254) and PHAEONIGMA (NFR project 204479/F20, A novel cross-disciplinary approach to solve an old enigma: the food-web transfer of the mass-blooming phytoplankter Phaeocystis).



  • Alderkamp, A. C., A. G. J. Buma, and M. Van Rijssel. (2007). The carbohydrates of Phaeocystis and their degradation in the microbial food web. Biogeochemistry, 83: 99-118.
  • Landry, M. R., and R. P. Hassett. (1982). Estimating the grazing impact of marine micro-zooplankton. Marine Biology, 67: 283-288.
  • Nejstgaard, J. C. and others (2007). Zooplankton grazing on Phaeocystis: a quantitative review and future challenges. Biogeochemistry, 83: 147-172.
  • Sazhin, A. F., L. F. Artigas, J. C. Nejstgaard, and M. E. Frischer. (2007). The colonization of two Phaeocystis species (Prymnesiophyceae) by pennate diatoms and other protists: a significant contribution to colony biomass. Biogeochemistry, 83: 137-145.
  • Verity, P. G., S. J. Whipple, J. C. Nejstgaard, and A. C. Alderkamp. (2007). Colony size, cell number, carbon and nitrogen contents of Phaeocystis pouchetii from western Norway. Journal of Plankton Research, 29: 359-367.
banner 2nd transition NL

Should we shift towards collaborative management? Case study of the Asturian (northern Spain) gooseneck barnacle fishery.

Antonella Rivera1, Lucia García-Florez2, Stefan Gelcich3 and José Luis Acuña1

1Departamento de Biología de Organismos y Sistemas, Universidad de Oviedo, Oviedo, Spain.

2Centro  de Experimentación Pesquera. Consejería de Agroganadería y Recursos Autóctonos, Gijón, Spain.

Center of Applied Ecology and Biodiversity, Facultad de Ciencias Biológicas, Pontificia Universidad Católica de Chile, Chile.

Nearly 60 years ago, Hardin (1968) stated that “Ruin is the destination toward which all men rush, each pursuing his own best interest…”. Regrettably, with 87% of world’s fish stocks currently overfished (FAO 2012), ruin appears to be where global fisheries are heading. During the 1970s, in an ambitious attempt to address this, the European Union (EU) launched the Common Fisheries Policy (CFP), a legal framework to achieve sustainable fishing in the EU. This has been criticized for its limited success following 30 years of enforcement (Commission of the European Communities 2009). The criticism is largely due to its open-access nature, which promotes the race for fish (Hentrich and Salomon 2006), the difficulty of incorporating scientific knowledge into its policies (Daw and Gray 2005), the lack of approval from the resource users (Khalilian et al. 2010) and its bias towards large-scale fishing fleets, which are socially and ecologically less sustainable than their small-scale, artisanal counterparts (OCEAN2012 April 3, 2010).

In this context, a paradigm shift is occurring where humans are no longer seen as antagonists to ecosystems, but rather as part of a linked social-ecological system (Castilla and Defeo 2005, Gelcich, et al. 2010). The shift involves a transition from the classical, top-down, to collaborative, bottom-up management systems. Between the two extremes, co-management represents a compromise between centralized governmental management and community self-governance (Sen and Raakjaer Nielsen 1996), where all stakeholders work together and share the rights and responsibilities (Jentoft et al. 1998). Co-management has shown potential in incorporating scientific and fishers’ knowledge (Castilla and Defeo 2001), preserving fish stocks (Defeo and Castilla 2005), integrating communities (Jentoft 2000) and adapting to complex social-ecological systems such as the fisheries (Armitage, Plummer et al. 2009).

In Asturias (northern Spain) the gooseneck barnacle (Pollicipes pollicipes) fishery exhibits most of the positive attributes present in co-management systems and represents an interesting case study with potential implications for small-scale benthic European fisheries. Although the resource reaches first sale market prices of 266 Euros/kg, gooseneck barnacles were traditionally considered as a complementary resource in Asturias (Rivera et al. 2014). However, in the early 90s, the Dirección General de Pesca Marítima - a fisheries administration service - saw the opportunity to exploit this resource and carried out stock assessments in the area. The data indicated that the fishery was viable and a tentative 8kg daily Total Allowable Catch (TAC) and a fishing ban during the reproductive period from April to September were proposed. In 1994 they offered Territorial User Rights for Fishing (TURFs) to the local fishers’ associations and played an active role in setting management guidelines in exchange for thorough data gathering and adherence to the agreed regulations. The system was well received and by 2001 had expanded to comprise seven management regions - known as plans - covering most of the western Asturias coast (Fig. 11). Management guidelines are almost homogeneous across all plans and involve a fishing campaign from September to April to protect the species’ recruitment during the summer. An exception is the Cabo Peñas Plan, where the fishery is open all year to meet the fishers’ requirement for a de-seasonalized income. Since 2004, the Cabo Peñas Plan has harvested a third of their area during summer and the remaining area during the regular fishing campaign from October to April, thus harvesting year-round without affecting the recruitment of the entire plan. Such arrangements require the collaboration of all stakeholders and the incorporation of any particularities of the user groups, which is difficult to achieve in top-down management.


Figure 11. a) Map of Asturian coast showing the seven co-management plans. b) Fine-scale map of the Luarca plan. c) Detailed view of 7 fishing zones in the Luarca plan. 

In spite of all the measures taken into consideration in the co-management system, decreasing catches were noted by both the government agency and the resource users. In the 2004-2005 fishing campaign, the daily TAC was reduced from 8 to 6 kg per fisher, except over the Christmas period, when the price for gooseneck barnacles increases abruptly. Following this change, the system showed signs of recovery, with an increasing trend in catch per unit effort (CPUE) for the seasonal plans. In Cabo Peñas, fishing effort was limited by decreasing the number of fishing days, not the daily TAC as in the other plans, allowing for a stable CPUE. Furthermore, despite the global economic crisis that shook the Spanish market, the mean daily price (Euros per kg of gooseneck barnacle per day) showed an increasing trend. We can conclude that co-management in the gooseneck barnacle fishery enabled adaptation, through trial and error, leading to the sustainability of the resource.

One of the particularities of this fishery is its fine spatial scale. This can only be obtained through collaborative management, where the resource users assist with the data collection. The plans encompass 267 fishing zones that range in size from 3 m to 3 km (Fig. 11). Maintaining such an exhaustive database would be impossible in an open-access regime, due to the high enforcement costs and large workforce that would be required. In addition, delimitation of the fishing zones is based on the fishers’ knowledge of the area. The zones are further classified depending on the commercial quality of barnacles they yield, and are managed according to different levels of protection, agreed to by the stakeholders at the beginning of the fishing season. The levels are determined from the status of the resource in each zone during the past campaign as revealed by the fisher's reports, and usually involve temporal limitations to maximize the economic return. An example of the kind of information provided by the fishermen is given in Fig. 12, which shows the temporal evolution of catches and protection levels in a high quality zone in the Tapia-Figueras plan. It became apparent in 2005, that productivity was decreasing despite the partial fishing bans that were in place. A different management regime was put in place, with alternating partial and total bans, an adaptation strategy that appears to be more effective (Fig. 12).


Figure 12.  Total seasonal landings in Percebosas, a good quality zone in the Tapia-Figueras plan. Trend lines (solid line) and standard errors (dotted line) are shown for the two types of management in the zone, either partial bans or alternating partial and total bans.

An added value of the co-management system is its capacity to create social capital. Fishers become empowered and are transformed from resource users to stewards of the fishery. Their investment in the fishery has promoted compliance and increased protection of the resource, and in some plans the fishers volunteer to help with surveillance of the fishing zones (Rivera et al. 2014). Furthermore, the incorporation of TURFs has alleviated the race for fish, leading to greater cooperation between fishers, who tend to work together throughout the entire harvesting process (Fig. 13). Although these social aspects are often overlooked, they are essential for the success of a fishery.


Figure 13.  Cooperative harvesting in the Asturian gooseneck barnacle co-management system.

Since the beginning, the gooseneck co-management system in Asturias has successfully assimilated all underlying aspects of the complex social-ecological system through a learning by doing approach. It has incorporated both scientific advice and fishers’ knowledge and promoted collaboration among stakeholders while generating social capital. Its adaptive nature has been essential to withstand struggles and to generate a sustainable fishery. Being a local solution, it should not be blindly exported to different areas, cultures or socioeconomic setups. However, this system demonstrates the potential of co-management to address the decline of small-scale benthic European fisheries.


  • Armitage, D.R., R. Plummer, F. Berkes, R.I. Arthur, A.T. Charles, I.J. Davidson-Hunt, A.P. Diduck, N.C. Doubleday, D.S. Johnson and M. Marschke (2009). Adaptive co-management for social–ecological complexity. Front Ecol Environ, 7(2): 95-102.
  • Castilla, J.C. and O. Defeo (2001). Latin American benthic shellfisheries: emphasis on co-management and experimental practices. Reviews in Fish Biology and Fisheries, 11(1): 1-30.
  • Castilla, J.C. and O. Defeo (2005). Paradigm shifts needed for world fisheries. Science, 309(5739): 1324-1325.
  • Commission of the European Communities (2009). Green Paper: Reform of the common fisheries policy [COM (2009) 163 final]. Brussels, The European Commission.
  • Daw, T. and T. Gray (2005). Fisheries science and sustainability in international policy: a study of failure in the European Union's Common Fisheries Policy. Marine Policy, 29(3): 189-197.
  • Defeo, O. and J.C. Castilla (2005). More than one bag for the world fishery crisis and keys for co-management successes in selected artisanal Latin American shellfisheries. Reviews in Fish Biology and Fisheries, 15(3): 265-283.
  • FAO (2012). The State of the World Fisheries and Aquaculture: 209.
  • Gelcich, S., T.P. Hughes, P. Olsson, C. Folke, O. Defeo, M. Fernández, S. Foale, L.H. Gunderson, C. Rodríguez-Sickert and M. Scheffer (2010). Navigating transformations in governance of Chilean marine coastal resources. Proceedings of the National Academy of Sciences, 107(39): 16794-16799.
  • Hardin, G. (1968). The Tragedy of the Commons. Science, 162(3859): 1243-1248.
  • Hentrich, S. and M. Salomon (2006). Flexible management of fishing rights and a sustainable fisheries industry in Europe. Marine Policy, 30(6): 712-720.
  • Jentoft, S. (2000). The community: a missing link of fisheries management. Marine policy, 24(1): 53-60.
  • Jentoft, S., B.J. McCay and D.C. Wilson (1998). Social theory and fisheries co-management. Marine Policy, 22(4): 423-436.
  • Khalilian, S., R. Froese, A. Proelss and T. Requate (2010). Designed for failure: A critique of the Common Fisheries Policy of the European Union. Marine Policy, 34(6): 1178-1182.
  • OCEAN2012. (April 3, 2010). A Coruña Declaration: Placing Sustainable Artisanal Coastal Fisheries at the Heart of the CFP Reform.
  • Rivera, A., S. Gelcich, L. García-Florez, J.L. Alcázar and J.L. Acuña (2014). Co-management in Europe: Insights from the gooseneck barnacle fishery in Asturias, Spain. Marine Policy, 50(A): 300–308, DOI: 10.1016/j.marpol.2014.07.011.
  • Sen, S. and J. Raakjaer Nielsen (1996). Fisheries co-management: a comparative analysis. Marine Policy, 20(5): 405-418.
banner 2nd transition NL

Time of emergence of trends in ocean biogeochemistry

Kathrin Keller1,2

1Climate and Environmental Physics, Physics Institute, University of Bern, Bern, Switzerland

2Oeschger Centre for Climate Change Research, University of Bern, Bern, Switzerland

E-mail: keller@climate.unibe.ch


This is a time of change. The steady increase of anthropogenic CO2 emissions leads to a rise in global temperatures and pushes the oceans towards acidification and, potentially, deoxygenation (Bopp et al. 2013; Cocco et al. 2013). One way to quantify these changes is to estimate the magnitude of a trend. Another relevant characteristic is the time period needed for the trend signal to first be detected. The latter has important implications for society, as it sets the pace for adaption and mitigation measures. However, it is not always clear if an observed trend is representative of the long-term evolution of the climate-carbon system, or if it is just an imprint of short-term fluctuations attributable to natural variability; an illustrative example is the current debate about the "hiatus".

The detection of forced trends in biogeochemical cycles and ecosystems is challenging. This is particularly so in the ocean, where observations are scarce and limited in either time or space. A major issue is the presence of natural variability related to, e.g., volcanic eruptions (Frölicher et al. 2011) or climate modes such as the North Atlantic Oscillation (Keller et al.2012) or El Niño-Southern Oscillation (ENSO; Keller et al. in prep). This variability has the potential to enhance or mask trends over decadal timescales. Successful detection of trend signals is thus a signal-to-noise (S/N) issue, i.e., the signal has to be of a magnitude that enduringly exceeds the envelope of background variability. One possible measure to estimate this is the time of emergence (ToE; Hawkins and Sutton, 2012) of a signal, that is, the time at which the S/N ratio exceeds a certain threshold. Thus, ToE is indicative of the time required for the anthropogenic trend to leave the variability band. This must not be confused with the period required to detect this trend in observational or model data.

ToE and related methods have been applied in a number of model studies focusing on ocean acidification indicators such as pH, surface total alkalinity or the saturation state of aragonite (Ilyina et al. 2009; Ilyina and Zeebe, 2012; Friedrich et al. 2012; Mora et al. 2013; Hauri et al. 2013). The implementation of ToE differs between studies; a common approach is the comparison of modelled noise (usually the standard deviation of an unforced control simulation) and observed or modelled trends. Note that the choice of datasets and the definition of thresholds have implications for the robustness of results (see Mora et al. (2013) and the corresponding comment by Hawkins et al. (2014)).

In a recent study (Keller et al. 2014), we use a model ensemble of 17 ESMs to investigate the ToE of trends in surface ocean biogeochemistry. For maximum comparability with the available observations, we focus on three frequently measured carbon cycle variables, dissolved inorganic carbon (DIC), pCO2 and pH, and sea-surface temperature (SST) on the local (grid cell) scale. We use historical simulations covering the years 1870-1999 with annual resolution. ToE is defined as:

            ToE = (2×N)/S

where S is the trend and N a measure of variability. S and N are defined as the linear trend (per year) over the period 1970-1999 and the standard deviation of the (detrended) years 1870-1999, respectively. See Fig. 14 for an illustration. 


Figure 14. An illustration of ToE based on the model CESM1: Annual time series (light blue), corresponding smoothing spline (dark blue), and linear trend (red) of DIC at 22°N, 158°W. The grey bar represents two times the standard deviation of the detrended time series (i.e., annual-spline). The linear trend is based on the years 1970–1999 of annual resolution, as indicated by the grey vertical lines. The intersection of the red vertical line and the upper border of the grey bar (x = 2015) shows when the trend leaves the envelope of background variability and, from then on, is detectable. Consequently, the ToE at this location is 16 years (2015–1999).


Fig. 15 shows the ensemble mean ToE patterns of DIC, pCO2, pH and SST, all variables at the surface. We find that trend signals in the three carbon cycle variables emerge on much shorter timescales than the physical climate variable SST. The ToE pattern of SST is very noisy, varying typically between 45 and 90 years (yr). The exceptions are areas around the equator in the Atlantic, Indian and western Pacific Oceans, that have values of approximately 35 yr. A large coherent area with ToE> 80 yr in the (eastern) equatorial Pacific can be attributed to high levels of variability linked to ENSO. The linear trend in DIC appears in large parts of the global oceans after approximately 10–30 yr; higher values are found at high latitudes, especially in the Arctic Ocean (≈ 50 yr), and localized in the equatorial Pacific (up to ≈ 70 yr). ToE of pCO2 and pH show a very similar pattern. However, the trends emerge much faster for pCO2 and pH than for DIC: after ≈ 12 yr for the majority of the global ocean area, 14–18 yr in the Arctic Ocean and ≈ 20 yr in the equatorial Pacific. A likely reason for these different timescales of DIC and pH/pCO2 are nonlinear processes in ocean chemistry described by the buffer factor (or Revelle factor; Revelle and Suess, 1957), which result in increases of pCO2 of approximately 10 times the magnitude of the corresponding relative increases in DIC. In contrast to DIC, relatively high ToE values are found for both pCO2 and pH in the Southern Ocean and in the upwelling region off Peru and Chile (in both regions, localized > 30 yr).


Figure 15. Ensemble mean ToE (years) of DIC, SST, pCO2, and total pH - all variables at the surface level. Note the different scales for DIC/SST and pCO2/pH. The trend signal is detectable much earlier on the grid cell scale for pCO2 and pH than for SST and DIC.

We find that, in general, the standard deviation is of greater importance in determining ToE than the strength of the linear trend. In areas with high natural variability, even strong trends in both the physical climate and carbon cycle system are masked by variability over decadal timescales. This explains inconsistencies in trends based on time series of insufficient length and illustrates the necessity for long-term observations. Considering the changes since the beginning of industrialization, the rapid emergence of trend signals implies that anthropogenic trends in the surface ocean carbon cycle are already detectable in large parts of the global oceans. This finding is even more relevant as the highest rates of ocean acidification are measured (Bates, 2012; Dore et al. 2009) and modelled (Resplandy et al. 2013) in subsurface waters. A further finding of the study is that, in contrast to the trend, standard deviation is affected by the seasonal cycle. This has important implications for the use of scarse observations. In some parts of the global oceans, there are hints that statements based on irregularly sampled seasonal data are representative for the whole year. In large areas however, especially in the high latitudes, intra-annual variability could interfere with such a generalization.

The study clearly illustrates the need for more long-term measurements with sufficient seasonal data coverage. In particular, global data sets describing the space-time variability of biogeochemical variables over adequate timescales are largely absent. DIC is a very important variable and crucial for our understanding of biogeochemical processes. Solely for the detection of anthropogenic trends, however, pCO2 and pH seem to be a better choice. Further, not only observations are necessary for the correct detection of trends. Independent data sets are also key for the realistic forcing and evaluation of climate models which are, due to the current scarcity of observations, the measure of choice for many research questions. 



  • Bates, N.R. (2012). Multi-decadal uptake of carbon dioxide into subtropical mode water of the North Atlantic Ocean, Biogeosciences, 9: 2649–2659. DOI:10.5194/bg-9-2649-2012.
  • Bopp, L., L. Resplandy, J.C. Orr, S.C. Doney, J.P. Dunne, M. Gehlen, P. Halloran, C. Heinze, T. Ilyina, R. Séférian, J. Tjiputra and M. Vichi (2013). Multiple stressors of ocean ecosystems in the 21st century: projections with CMIP5 models, Biogeosciences, 10: 6225-6245.
  • Cocco, V., F. Joos, M. Steinacher, T. L. Frölicher, L. Bopp, J. Dunne, M. Gehlen, C. Heinze, J. Orr, A. Oschlies, B. Schneider, J. Segschneider and J. Tjiputra (2013). Oxygen and indicators of stress for marine life in multi-model global warming projections, Biogeosciences, 10: 1849-1868.
  • Dore, J. E., Lukas, R., Sadler, D.W., Church, M.J. and Karl, D.M. (2009). Physical and biogeochemical modulation of ocean acidification in the central North Pacific, P. Natl. Acad. Sci. USA, 106: 12235–12240.
  • Friedrich, T., Timmermann, A., Abe-Ouchi, A., Bates, N. R., Chikamoto, M. O., Church, M. J., Dore, J. E., Gledhill, D. K., Gonzalez-Davila, M., Heinemann, M., Ilyina, T., Jungclaus, J. H., McLeod, E., Mouchet, A., and Santana-Casiano, J. M. (2012). Detecting regional anthropogenic trends in ocean acidification against natural variability, Nat. Clim. Change, 2: 167–171.
  • Frölicher, T.L., F. Joos and C.C. Raible. (2011). Sensitivity of atmospheric CO2 and climate to explosive volcanic eruptions, Biogeosciences, 8: 2317-2339.
  • Hauri, C., Gruber, N., Vogt, M., Doney, S.C., Feely, R.A., Lachkar, Z., Leinweber, A., McDonnell, A.M.P., Munnich, M. and Plattner, G.-K. (2013). Spatiotemporal variability and long-term trends of ocean acidification in the California Current System, Biogeosciences, 10: 193–216. DOI:10.5194/bg-10-193-2013.
  • Hawkins, E., B. Anderson, N. Diffenbaugh, I. Mahlstein, R. Betts, G. Hegerl, M. Joshi, R. Knutti, D. McNeall, S. Solomon, R. Sutton, J. Syktus and G. Vecchi (2014). Uncertainties in the timing of unprecedented climates, Nature, 511: E3-E5.
  • Hawkins, E. and Sutton, R. (2012). Time of emergence of climate signals, Geophys. Res. Lett., 39: L01702. DOI:10.1029/2011GL050087.
  • Ilyina, T. and Zeebe, R.E. (2012). Detection and projection of carbonate dissolution in the water column and deep-sea sediments due to ocean acidification, Geophys. Res. Lett., 39: L06606. DOI:10.1029/2012GL051272.
  • Ilyina, T., Zeebe, R.E., Maier-Reimer, E. and Heinze, C. (2009). Early detection of ocean acidification effects on marine calcification, Global Biogeochem. Cy., 23: GB1008, DOI:10.1029/2008GB003278.
  • Keller, K.M., Joos, F. and Raible, C.C. (2014). Time of emergence of trends in ocean biogeochemistry, Biogeosciences, 11: 3647–3659.
  • Keller, K.M., Joos, F. Raible, C.C., Cocco, V., Frölicher, T., Dunne, J., Gehlen, M., Bopp, L., Orr, J., Tjiputra, J., Heinze, C., Segschneider, J., Roy, T. and Metzl, N. (2012). Variability of the ocean carbon cycle in response to the North Atlantic Oscillation, Tellus B, 64: 18738. DOI:10.3402/tellusb.v64i0.18738.
  • Mora, C., Frazier, A.G., Longman, R.J., Dacks, R.S., Walton, M.M., Tong, E.J., Sanchez, J.J., Kaiser, L.R., Stender, Y.O., Anderson, J.M., Ambrosino, C.M., Fernandez-Silva, I., Giuseffi, L.M. and Giambelluca, T.W. (2013). The projected timing of climate departure from recent variability, Nature, 502: 183–187.
  • Resplandy, L., Bopp, L., Orr, J. C., and Dunne, J. P. (2013). Role of mode and intermediate waters in future ocean acidification: analysis of CMIP5 models, Geophys. Res. Lett., 40: 3091–3095.
  • Revelle, R. and Suess, H. (1957). Carbon dioxide exchange between atmosphere and ocean and the question of an increase of atmospheric CO2 during the past decades, Tellus, 9: 18–27. 
banner 2nd transition NL

Recent climatic changes enhance ongoing ocean acidification in the California Current System

Giuliana Turi, Zouhair Lachkar, Matthias Münnich, Nicolas Gruber and Damian Loher

Environmental Physics, Institute of Biogeochemistry and Pollutant Dynamics, ETH Zürich, Switzerland

Contact: giuliana.turi@env.ethz.ch


Eastern Boundary Upwelling Systems (EBUS) experience a wide range of natural fluctuations in carbon fluxes due to the upwelling of cool subsurface waters that are rich in nutrients and dissolved inorganic carbon. These upwelled waters are also naturally low in pH and in the saturation state of aragonite (Ωarag), a mineral form of calcium carbonate, which is used by many calcifying organisms to build and sustain their shells and skeletons (e.g., Orr et al. 2005). Although the ecosystems of these coastal upwelling regions are accustomed to strong seasonal fluctuations, future environmental impacts, such as ocean acidification, changes in the rate and depth of upwelling or changes in temperature, could act as additional strains, leading to unforeseen consequences.

The California Current System (CalCS) - one of the four major EBUS - has recently been the subject of several studies investigating the past and present states of ocean acidification, as well as its future evolution (e.g., Feely et al. 2008; Gruber et al. 2012; Hauri et al. 2013; Bednarsek et al. 2014). Aragonite undersaturation events have been detected and pH values as low as 7.6 have been measured in various places over the shelf (Feely et al. 2008). Such events could become more frequent and widespread in the near future: Hauri et al. (2013) project that by 2030, nearshore surface Ωarag will have left its present-day variability envelope, and Gruber et al. (2012) suggest that by 2050, more than half the waters of the coastal CalCS will be undersaturated with respect to aragonite. These modelling studies assume that the climatic forcing of the CalCS remains constant throughout the duration of the simulations. Several studies, however, suggest that alongshore, upwelling-inducing winds have increased in the CalCS over the last few decades (e.g., Demarcq, 2009; García-Reyes and Largier, 2010; Narayan et al. 2010). Furthermore, in a modelling sensitivity study, Lachkar (2014) demonstrated that a doubling of alongshore wind speeds in the CalCS led to a tripling of the volume of aragonite-undersaturated water.

Atmospheric CO2 increase vs. changes in upwelling-favourable winds

The aim of our study is to quantify trends in ocean acidification over the past three decades and determine the relative importance of the increase in atmospheric CO2 versus changes in wind-driven upwelling for these trends. We use a 5km resolution setup of the regional oceanic model ROMS (Marchesiello et al. 2003; Shchepetkin and McWilliams 2005), coupled to a simple nitrogen-based ecosystem model with an integrated carbon module for the period 1979 to 2012. At the atmospheric boundary of the model, we prescribe observation-based products of CO2, wind stress, heat and freshwater fluxes. At the lateral boundaries, we force the model with anomalies from a CCSM3 hindcast simulation (Graven et al. 2012). To separate the effect of changes in the climatic forcing from the effect of increasing anthropogenic CO2, we perform two simulations with different atmospheric and lateral boundary conditions. The first simulation - which we term “full hindcast” - is forced with fully transient conditions at the atmospheric and lateral boundaries. In the second simulation - called “constant climate” - only atmospheric CO2 and lateral DIC are allowed to vary during the 34 years of the simulation. The difference between the two simulations is thus the effect of a variable climate alone - without the effect of increasing CO2.

In our full hindcast simulation, we model a domain-wide decrease in the depth of the aragonite saturation horizon (Ωarag < 1; Fig. 16 a), with a maximum shoaling of roughly 100m over the last three decades in the regions offshore of Cape Mendocino, California (~38°N). Throughout the simulation, the volume of water which is undersaturated with respect to aragonite also increases, both in the top 200m of the water column and in the underlying 200-400m layer (Fig. 16 b). Towards the end of our simulation, waters in the 200-400m layer of the northern coastal region (Region 1 in Fig. 16 b), are completely undersaturated with respect to aragonite. Similarly in the other two regions, waters with Ωarag < 1 expand by roughly 30-40% over the 34 years of the simulation. These results support the findings of Gruber et al. (2012) and Hauri et al. (2013) that if ocean acidification continues at this rate in the CalCS, undersaturation events such as those measured by Feely et al. (2008) could become the norm in regions close to the coast in the near future.


Figure 16. (a) Change per decade in the aragonite saturation horizon (depth where Ωarag < 1) from 1979 to 2012, calculated as the least squares linear trend. The red stippling indicates the area where the saturation horizon remained on average within the top 200m of the water column during the last five years of the simulation (2008-2012). (b) Temporal evolution of annual mean volume fraction of water undersaturated with respect to aragonite (Ωarag < 1) for the top 200m of the water column (dotted lines) and for the 200-400m layer (dashed lines) from 1979 to 2012. The inset in (b) indicates the three regions for which the volume fractions were computed (color-coded).

Furthermore, our model results suggest that the recent increase in atmospheric CO2 is not the only driving factor and that climatic changes also have a significant impact on trends in ocean acidification (Fig. 17): the difference between trends from the full hindcast simulation and from the constant climate simulation reveals that the strong increase in pCO2 and decreases in pH and Ωarag in the coastal regions along Oregon and northern California (centered roughly around Cape Mendocino) can be attributed to changes in the climatic forcing of the CalCS. These changes can account for up to 30% of the total modelled trends in pCO2, pH and Ωarag. We also see an increase in upwelling-favourable winds in our atmospheric forcing product (ERA-Interim; Dee et al., 2011), suggesting that this is the main mechanism responsible for the local acceleration in ocean acidification in this region (Fig. 18). An analysis of trends in surface ocean temperature and net primary production reveals an opposing effect on trends in ocean acidification, hence eliminating these mechanisms as the possible driving forces (not shown here).


Figure 17. Changes per decade in pCO2, pH and Ωarag from 1979 to 2012 for the “full hindcast” (top row) and the “constant climate” (centre row) simulations, calculated as the least squares linear trends and averaged over the top 60m of the water column. The bottom row shows the difference between the trends from the full hindcast and from the constant climate simulations.


Figure 18. (a) Anomalies in ERA-Interim sea level pressure (SLP; background shading) and wind speed (vectors) calculated as 2008-2012 minus 1979-1983. (b) Change per decade in upwelling index (UWI) from 1979 to 2012, calculated as the least-squares linear trend. The UWI was computed from the alongshore wind stress component of our forcing product and integrated for increments of 100m along the coast, and represents the magnitude of offshore (positive values) or onshore (negative values) water transport.

Climate change or natural variability?

At present, it remains debatable whether this increase in upwelling-favourable winds along the coast of Oregon and northern California from 1979 to 2012 can be attributed to a strengthening of the thermal land-sea gradient due to climate warming - as proposed by Bakun (1990) - or whether it is a consequence of natural, longer-term fluctuations in the large-scale climatic forcing patterns of the CalCS. In a meta-analysis of literature related to upwelling-favourable wind intensification, Sydeman et al. (2014) concluded that Bakun's hypothesis holds true for the CalCS, provided the trend analysis focuses only on the upwelling season. Considering the short simulation time and the fact that we analyze annual mean trends, this explanation of natural variability seems more plausible. A composite analysis of upwelling strength for extreme negative and positive years of the Pacific Decadal Oscillation (PDO) - one of the climate variability patterns with an influence on the North Pacific - highlights the strong control that these variability patterns have over the CalCS (not shown here). During extreme positive years, which prevail at the beginning of our simulation, upwelling is suppressed, particularly in the area around Cape Mendocino. In extreme negative years on the other hand, such as at the end of our simulation, upwelling is enhanced, suggesting that the influence of large-scale climate variability, such as the PDO, on our upwelling trends should not be overlooked. However, to adequately determine the relative importance of climate warming versus natural, large-scale climate variability for the changes in upwelling-favourable winds in the CalCS, a longer simulation and temporally and spatially more comprehensive observational data sets would be crucial.


  • Bakun, A. (1990). Global climate change and intensification of coastal ocean upwelling. Science, 247(4939): 198-201.
  • Bednarsek, N., Feely, R., Reum, J., Peterson, B., Menkel, J., Alin, S., and Hales, B. (2014). Limacina helicina shell dissolution as an indicator of declining habitat suitability owing to ocean acidification in the California Current Ecosystem. Proceedings of the Royal Society B, 281(1785): 20140123.
  • Dee, D., Uppala, S., Simmons, A., Berrisford, P., Poli, P., Kobayashi, S., Andrae, U., Balmaseda, M., Balsamo, G., Bauer, P., Bechtold, P., Beljaars, A., van de Berg, L., Bidlot, J., Bormann, N., Delsol, C., Dragani, R., Fuentes, M., Geer, A., Haimberger, L., Healy, S., Hersbach, H., Hólm, E., Isaksen, L., Kållberg, P., Köhler, M., Matricardi, M., McNally, A., Monge-Sanz, B., Morcrette, J.-J., Park, B.-K., Peubey, C., de Rosnay, P., Tavolato, C., Thépaut, J.-N. and Vitart, F. (2011). The ERA-Interim reanalysis: configuration and performance of the data assimilation system. Quarterly Journal of the Royal Meteorological Society, 137(656): 553-597.
  • Demarcq, H. (2009). Trends in primary production, sea surface temperature and wind in upwelling systems (1998-2007). Progress in Oceanography, 83(1-4): 376-385.
  • Feely, R.A., Sabine, C.L., Hernandez-Yon, J.M., Manson, D. and Hales, B. (2008). Evidence for upwelling of corrosive “acidified” water onto the continental shelf. Science, 320(5882): 1490-1492.
  • García-Reyes, M. and Largier, J. (2010). Observations of increased wind-driven coastal upwelling off central California. Journal of Geophysical Research, 115: C04011.
  • Graven, H.D., Gruber, N., Key, R., Khatiwala, S. and Giraud, X. (2012). Changing controls on oceanic radiocarbon: New insights on shallow-to-deep ocean exchange and anthropogenic CO2 uptake. Journal of Geophysical Research, 117: C10005.
  • Gruber, N., Hauri, C., Lachkar, Z., Loher, D., Frölicher, T. L. and Plattner, G.-K. (2012). Rapid progression of ocean acidification in the California Current System. Science, 337(6091): 220-223.
  • Hauri, C., Gruber, N., McDonnell, A.M.P. and Vogt, M. (2013). The intensity, duration, and severity of low aragonite saturation state events on the California continental shelf. Geophysical Research Letters, 40(13): 3424-3428.
  • Lachkar, Z. (2014). Effects of upwelling increase on ocean acidification in the California and Canary Current systems. Geophysical Research Letters, 41(1): 90-95.
  • Marchesiello, P., McWilliams, J.C. and Shchepetkin, A. (2003). Equilibrium structure and dynamics of the California Current System. Journal of Physical Oceanography, 33(4): 753-783.
  • Narayan, N., Paul, A., Mulitza, S. and Schulz, M. (2010). Trends in coastal upwelling intensity during the late 20th century. Ocean Science, 6(3): 815-823.
  • Orr, J.C., Fabry, V.J., Aumont, O., Bopp, L., Doney, S.C., Feely, R.A., Gnanadesikan, A., Gruber, N., Ishida, A., Joos, F., Key, R.M., Lindsay, K., Maier-Reimer, E., Matear, R., Monfray, P., Mouchet, A., Najjar, R.G., Plattner, G.-K., Rodgers, K.B., Sabine, C.L., Sarmiento, J.L., Schlitzer, R., Slater, R.D., Totterdell, I.J., Weirig, M.-F., Yamanaka, Y. and Yool, A. (2005). Anthropogenic ocean acidification over the twenty-first century and its impact on calcifying organisms. Nature, 437(7059): 681-686.
  • Shchepetkin, A.F. and McWilliams, J.C. (2005). The regional oceanic modeling system (ROMS): a split-explicit, free-surface, topography-following-coordinate oceanic model. Ocean Modelling, 9(4): 347-404.
  • Sydeman, W., García-Reyes, M., Schoeman, D., Rykaczewski, R., Thompson, S., Black, B. and Bograd, S. (2014). Climate change and wind intensification in coastal upwelling ecosystems. Science, 345(6192): 77-80.
banner 2nd transition NL

Diving depth of elephant seals influences mercury bioaccumulation in the north Pacific

Sarah Peterson1, Josh Ackerman2 and Dan Costa1;

'1University of California, Santa Cruz,

2U.S. Geological Survey, Western Ecological Research Center, Dixon, California


Mercury in marine ecosystems is increasing (Sunderland et al. 2009), with an associated increase in risk for both humans and wildlife. Although mercury enters the environment via both natural and anthropogenic sources, the increased levels can be attributed primarily to human sources, such as coal-fired power plants (Streets et al. 2011). Mercury levels will continue to increase, even if anthropogenic emissions are halted, because of the time lag required for mercury to equilibrate between the atmosphere and the ocean (Sunderland and Mason 2007, Mason et al. 2012). Environmental mercury is of concern because it can cause neurological problems in vertebrates. The effects of mercury accumulation can be acute or sub-clinical (not visually detectable) over a wide range of concentrations and lengths of exposure (Das et al. 2003, Basu et al. 2009). Furthermore, mercury has no known beneficial role in organisms. Higher trophic level predators are generally at greatest risk of mercury contamination due to their propensity for biomagnification (Atwell et al. 1998), however, mercury bioaccumulation is poorly understood for predators in remote marine ecosystems.

Northern elephant seals (Mirounga angustirostris) are top marine predators in the northeast Pacific Ocean. They dive deeply and forage within the mesopelagic zone (200-1000 m depths). Elephant seals travel thousands of kilometers during their biannual migrations - moving between beach haul-outs and at-sea foraging areas (Fig. 19). Regional differences in mercury concentrations in surface-feeding (epipelagic zone) marine species have been observed in the north Pacific, but mercury concentrations in predators that use the deeper, mesopelagic food webs have not been well described. Deep-diving predators should encounter higher mercury concentrations as the biogeochemistry associated with these depths makes mercury more bioavailable in the mesopelagic than the epipelagic zone (Sunderland et al. 2009). For example, peak mercury concentrations within the water column were observed in the mesopelagic zone during a latitudinal transect study in the northeast Pacific Ocean (Laurier et al. 2004, Sunderland et al. 2009). Therefore, elephant seals are exposed to the region where the highest concentrations of mercury were measured (Robinson et al. 2012), making them particularly vulnerable to mercury accumulation. Additionally, elephant seals may be good indicators of regional mesopelagic contamination because satellite tracking has revealed that individual seals demonstrate strong regional fidelity during foraging migrations (Simmons 2008, Costa et al. 2012), but the species as a whole forages across an extensive portion of the northeast Pacific Ocean (Robinson et al. 2012).


Figure 19. Satellite tracks of 75 northern elephant seal females that were sampled for mercury in blood and muscle. Satellite tags were deployed on seals at the Año Nuevo colony in California, USA, although foraging occurs across a wide range of the northeast Pacific Ocean.

We coupled animal behaviour, including movements and dive behaviour in the water column, with contaminant research to gain insight into mercury accumulation in predators from a relatively inaccessible marine ecosystem. Mesopelagic predators are, in general, difficult to sample due to the limitations of deep ocean research. However, elephant seals consistently travel back from their open-ocean foraging grounds, at which point they are accessible for sampling. Additionally, long-term demographic studies on the breeding colony at Año Nuevo State Reserve (California, USA) provided us with the opportunity to incorporate animal age and other demographic information that is typically unknown for free-ranging marine predators. The power of our study lies in the fact that we were able to overcome some of the pervasive challenges in contaminant research by being able to link animal behaviour and demographics directly with contaminant accumulation.

We deployed satellite tags and time-depth recorders on more than 70 known-age adult female elephant seals (Fig. 20) during 2010-2013, for either the two- (short) or six- (long) month foraging migration (Fig 21; Peterson et al. 2014). During the deployment and subsequent recovery procedures, we collected (non-lethally) blood and muscle samples.


Figure 20. Satellite tags and time depth recorders were deployed on adult female northern elephant seals for the duration of their foraging migrations (D. Costa; NMFS Permit 14636).


Figure 21. Blood and muscle samples were collected from adult female elephant seals for mercury analysis at the start and end (1A/1B and 2A/2B) of their biannual foraging trips, shown in blue. Seals were sampled 2010-2013 at the Año Nuevo colony in California, USA.

Our preliminary findings indicate that elephant seals have relatively high mercury concentrations compared with many marine predators, and suggest that other marine predators foraging in the mesopelagic may be at an increased risk for mercury bioaccumulation. The average mercury concentration in elephant seals was higher than those measured in their coastal and epipelagic-foraging northeast Pacific counterparts, including harbour seals (Phoca vitulina) (Brookens et al. 2007, McHuron et al. 2014) and California sea lions (Zalophus californianus) (unpublished data S. Peterson and L. McHuron). Furthermore, for comparison, mercury concentrations in elephant seals were more than eight times higher than those in yellowfin tuna (Ordiano-Flores et al. 2011), swordfish (Storelli et al. 2005), and bluefin tuna muscle (Storelli et al. 2005). Although no overt impairments have been observed in elephant seals that could be attributed to mercury contamination, sub-clinical effects may be present and currently undetected in the population.

To our knowledge, this is the first study to combine mercury concentrations in free-ranging, deep-diving marine mammals from the north Pacific with demographic and individual foraging behaviour information. We found that foraging ecology, including foraging location and diving depth, influenced mercury concentrations in adult female elephant seals. Although our results are preliminary, they indicate that predators foraging within the mesopelagic zone may be at greater risk of mercury accumulation than previously assumed. Further, our results provide insight into the potential for mercury bioaccumulation in more elusive and vulnerable species. Our findings align with previous ocean research on mercury in the water column and provide support for cross-disciplinary research and the integration of biogeochemistry and ecology.

This research was presented at the 2014 IMBER Open Science Conference (Bergen, Norway) with the support of a generous travel grant to S. Peterson. The full manuscript describing these results in more detail is currently in preparation for submission.


  • Atwell, L., K. A. Hobson, and H. E. Welch. 1998. Biomagnification and bioaccumulation of mercury in an arctic marine food web: insights from stable nitrogen isotope analysis. Canadian Journal of Fisheries and Aquatic Sciences 55: 1114-1121.
  • Basu, N., A. M. Scheuhammer, C. Sonne, R. J. Letcher, E. W. Born, and R. Dietz. 2009. Is dietary mercury of neurotoxicological concern to wild polar bears (Ursus maritimus)? Environmental Toxicology and Chemistry 28: 133-140.
  • Brookens, T. J., J. T. Harvey, and T. M. O'Hara. 2007. Trace element concentrations in the Pacific harbor seal (Phoca vitulina richardii) in central and northern California. Science of The Total Environment 372: 676-692.
  • Costa, D. P., G. A. Breed, and P. W. Robinson. 2012. New insights into pelagic migrations: implications for ecology and conservation. Annual Review of Ecology, Evolution, and Systematics 43: 73-96.
  • Das, K., V. Debacker, S. Pillet, and J.-M. Bouquegneau. 2003. Heavy metals in marine mammals. Pages 135-167 in J. G. Vos, G. D. Bossart, M. Fournier, and T. J. O'Shea, editors. Toxicology of marine mammals. Taylor & Francis, New York.
  • Laurier, F. J. G., R. P. Mason, G. A. Gill, and L. Whalin. 2004. Mercury distributions in the North Pacific Ocean - 20 years of observations. Marine Chemistry 90: 3-19.
  • Mason, R. P., A. L. Choi, W. F. Fitzgerald, C. R. Hammerschmidt, C. H. Lamborg, A. L. Soerensen, and E. M. Sunderland. 2012. Mercury biogeochemical cycling in the ocean and policy implications. Environmental Research 119: 101-117.
  • McHuron, E. A., J. T. Harvey, J. M. Castellini, C. A. Stricker, and T. M. O’Hara. 2014. Selenium and mercury concentrations in harbor seals (Phoca vitulina) from central California: Health implications in an urbanized estuary. Marine Pollution Bulletin 3(1): 48-57.
  • Ordiano-Flores, A., F. Galván-Magaña, and R. Rosiles-Martínez. 2011. Bioaccumulation of mercury in muscle tissue of yellowfin tuna, Thunnus albacares, of the Eastern Pacific Ocean. Biological Trace Element Research 144: 606-620.
  • Peterson, S. H., J. L. Hassrick, A. Lafontaine, J.-P. Thomé, D. E. Crocker, C. Debier, and D. P. Costa. 2014. Effects of Age, Adipose Percent, and Reproduction on PCB Concentrations and Profiles in an Extreme Fasting North Pacific Marine Mammal. PLoS ONE 9: e96191.
  • Robinson, P. W., D. P. Costa, D. E. Crocker, J. P. Gallo-Reynoso, C. D. Champagne, M. A. Fowler, C. Goetsch, K. T. Goetz, J. L. Hassrick, L. A. Hückstädt, C. E. Kuhn, J. L. Maresh, S. M. Maxwell, B. I. McDonald, S. H. Peterson, S. E. Simmons, N. M. Teutschel, S. Villegas-Amtmann, and K. Yoda. 2012. Foraging behavior and success of a mesopelagic predator in the northeast Pacific Ocean: Insights from a data rich species, the northern elephant seal. PLoS ONE 7: e36728.
  • Simmons, S. E. 2008. Environmental and individual effects on the foraging success of an apex predator, the northern elephant seal (Mirounga angustirostris). University of California Santa Cruz, Santa Cruz.
  • Storelli, M. M., R. Giacominelli-Stuffler, A. Storelli, and G. O. Marcotrigiano. 2005. Accumulation of mercury, cadmium, lead and arsenic in swordfish and bluefin tuna from the Mediterranean Sea: A comparative study. Marine Pollution Bulletin 50: 1004-1007.
  • Streets, D. G., M. K. Devane, Z. Lu, T. C. Bond, E. M. Sunderland, and D. J. Jacob. 2011. All-time releases of mercury to the atmosphere from human activities. Environmental Science & Technology 45: 10485-10491.
  • Sunderland, E. M., D. P. Krabbenhoft, J. W. Moreau, S. A. Strode, and W. M. Landing. 2009. Mercury sources, distribution, and bioavailability in the North Pacific Ocean: Insights from data and models. Global Biogeochemical Cycles 23(2), DOI: 10.1029/2008GB003425.
  • Sunderland, E. M. and R. P. Mason. 2007. Human impacts on open ocean mercury concentrations. Global Biogeochem. Cycles 21 (4), DOI: 10.1029/2006GB002876.


banner transition NL

IMBER-related Research and Activities 

GLODAPv2 – A new and updated global ocean carbon data product

S.K. Lauvset1,2

with contributions from:

A. Olsen1,2,3, R.M. Key4, X. Lin4, T. Tanhua5, M. Hoppema6, S. Jutterström7, R. Steinfeldt8, E. Jeansson2,3, S. van Heuven5,9, M. Ishii10, T. Suzuki11, A. Velo12, C. Schirnick5,  A. Kozyr13 and B. Pfeil1,2

1University of Bergen, Bergen, Norway

2Bjerknes Center for Climate Research, Bergen, Norway

3Uni Research, Bergen, Norway

4Princeton University, USA

5GEOMAR, Germany

6Alfred Wegener Institute, Bremerhaven, Germany

7Swedish Institute for Environmental Research, Sweden

8University of Bremen, Germany

9University of Groningen, Netherlands

10Meteorological Research Institute, Japan

11Marine Information Research Center, Japan

12Instituto de Investigaciones Marinas - CSIC, Spain

13Carbon Dioxide Information Analysis Center, USA


The Global Ocean Data Analysis Project (GLODAP, Fig. 22) was originally published (Key et al. 2004) 10 years ago. This effort to create a fully quality controlled and internally consistent global product of all available interior ocean carbon chemistry data has proven to be of immense value to the scientific community.  At the time of writing, the Key et al. paper has been cited 459 times (according to Web of Science), reflecting the wide use of the GLODAP data product.  Prominent examples include the global ocean inventory for anthropogenic carbon (Sabine et al. 2004) and the validation of several biogeochemical ocean models (e.g. Bopp et al. 2013).  GLODAP has clearly proven to be an invaluable resource for both observational- and model-based ocean carbon cycle science.  GLODAP includes data collected from 1972-1999, and is dominated by the data from the WOCE/JGOFS surveys of the 1990s.  As much more data have been collected in the past ten years, in particular within the framework of CLIVAR (Feely et al. 2014), the time is ripe for an update: GLODAPv2 is coming!


Figure 22.  Map showing all the hydrographic stations in the original GLODAP, which involved 122 cruises undertaken from 1972 - 1999. 

GLODAPv2 combines the original GLODAP with data from two recent regional synthesis products, CARINA (Key et al. 2010) and PACIFICA (Suzuki et al. 2013), as well as from 168 cruises not previously included in any of these. GLODAPv2 is consequently a much larger product than GLODAP (Fig. 23) with data from 775 individual cruises undertaken globally from 1972-2013. It is particularly noteworthy that GLODAPv2 includes many cruises in the Arctic Ocean - a region of great interest for ocean biogeochemical research. 


Figure 23.  Map of all the hydrographic stations in GLODAPv2, which included 775 cruises during the period 1972-2013.


While preparing GLODAPv2 we did not simply concatenate data from the existing products. New quality control tools and the larger data inventory enabled us to re-evaluate all the bias corrections derived during the previous efforts.  It has, therefore taken more than two years of hard work, multiple setbacks, many frustrations, and the occasional argument to complete GLODAPv2.  This data product is truly the result of a team effort by an international group of scientists and data managers.  But let us not forget that it was made possible by the support provided by funding agencies for data collection and the maintenance of large-scale observational networks, and the many people who ventured out to sea to collect data, which they kindly agreed to contribute to GLODAPv2. The creation of the GLODAPv2 data product has been made possible through funding received via the International Ocean Carbon Coordination Project (IOCCP), the EU FP7 project CarboChange, the Norwegian Research Council project DECApH, the FRAM-High North Research Centre for Climate and the Environment, and a NASA grant for participants from the USA, amongst others.

The GLODAPv2 data will be publically available through the Carbon Dioxide Information Analysis Center (CDIAC) later this year. The data will be available in three forms:

  1. Individual WOCE exchange formatted files containing the unadjusted data from each cruise;
  2. Regional synthesis files containing the bias-corrected data product; and
  3. A mapped product at a resolution of 1ºx1º. 

Details of each product will be published in Earth System Science Data (Key et al. in prep; Lauvset et al. in prep). Information about each cruise, including the chief scientists and principal investigators will be presented in a cruise summary table. This will also contain links to the original data files, the cruise metadata, and the online adjustment table, with the reasoning behind any bias correction that has been applied. 

GLODAPv2 is a global ocean carbon data product. It has been thoroughly quality controlled and will hopefully be as widely used as the original GLODAP and contribute to more great science in the years to come.  It has already inspired new and interesting work, some of which was presented at the IMBER Future Oceans Open Science Conference in June.  We hope that it will encourage both scientists and funding agencies to continue this important work.  So, go forth and publish!



  • Bopp, L., Resplandy, L., Orr, J.C., Doney, S.C., Dunne, J.P., Gehlen, M., Halloran, P., Heinze, C., Ilyina, T., Seferian, R., Tjiputra, J., and Vichi, M. (2013). Multiple stressors of ocean ecosystems in the 21st century: projections with CMIP5 models, Biogeosciences, 10: 6225-6245.
  • Feely, R., Talley, L., Bullister, J.L., Carlson, C.A., Doney, S., Fine, R.A., Firing, E., Gruber, N., Hansell, D.A., Johnson, G.C., Key, R., Langdon, C., Macdonald, A., Mathis, J., Mecking, S., Millero, F.J., Mordy, C., Sabine, C., Smethie, W.M., Swift, J.H., Thurnherr, A.M., Wanninkhof, R. and Warner, M. (2014). The US Repeat Hydrography CO2/Tracer Program (GO-SHIP): Accomplishments from the first decadal survey, 47 pp.
  • Key, R.M., Kozyr, A., Sabine, C.L., Lee, K., Wanninkhof, R., Bullister, J.L., Feely, R.A., Millero, F.J., Mordy, C. and Peng, T.-H. (2014). A global ocean carbon climatology: Results from Global Data Analysis Project (GLODAP), Global Biogeochemical Cycles, 18.
  • Key, R.M., Tanhua, T., Olsen, A., Hoppema, M., Jutterström, S., Schirnick, C., Van Heuven, S., Kozyr, A., Lin, X., Velo, A., Wallace, D.W.R. and Mintrop, L. (2010). The CARINA data synthesis project: introduction and overview, Earth Syst. Sci. Data, 2: 105-121.
  • Sabine, C.L., Feely, R.A., Gruber, N., Key, R.M., Lee, K., Bullister, J.L., Wanninkhof, R., Wong, C.S., Wallace, D.W.R., Tilbrook, B., Millero, F.J., Peng, T.-H., Kozyr, A., Ono, T. and Ríos, A.F. (2004). The Oceanic Sink for Anthropogenic CO2, Science, 305: 367-371.
  • Suzuki, T., Ishii, M., Aoyama, M., Christian, J.R., Enyo, K., Kawano, T., Key, R.M., Kosugi, N., Kozyr, A., Miller, L., Murata, A., Nakano, T., Ono, T., Saino, T., Sasaki, K.-I., Sasano, D., Takatani, Y., Wakita, M. and Sabine, C. (2013). PACIFICA Data Synthesis Project, Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, Tennessee.
banner 2nd transition NL

The 2014 Community Event of the Surface Ocean CO2 Atlas


Dorothee Bakker1, Kevin O’Brien2,3, Are Olsen4,5,6, Benjamin Pfeil4,5,  Ute Schuster7, Maciej Telszewski8, Bronte Tilbrook9, Rik Wanninkhof10 and all SOCAT contributors 

1Centre for Ocean and Atmospheric Sciences, School of Environmental Sciences, University of East Anglia, Norwich Research Park, Norwich, United Kingdom; d.bakker@uea.ac.uk

2Pacific Marine Environmental Laboratory, National Oceanic and Atmospheric Administration, Seattle, Washington, USA

3Joint Institute for the Study of the Atmosphere and Ocean, University of Washington, Seattle, Washington, USA

4Geophysical Institute, University of Bergen, Bergen, Norway

5Bjerknes Centre for Climate Research, Bergen, Norway

6Uni Climate, Uni Research, Bergen, Norway

7College of Life and Environmental Sciences, University of Exeter, Exeter, United Kingdom

8International Ocean Carbon Coordination Project, Institute of Oceanology of the Polish Academy of Sciences, Sopot, Poland

9CSIRO Oceans and Atmosphere, Hobart, Australia

10Atlantic Oceanographic and Meteorological Laboratory, National Atmospheric and Oceanographic Administration, Miami, Florida, USA

Ground-breaking data products for the global oceans and coastal seas enable quantification of the ocean uptake of the greenhouse gas CO(carbon dioxide) by a variety of methods. The Surface Ocean COAtlas (SOCAT; http://www.socat.info/) regularly provides quality controlled, synthesis and gridded fugacity of CO2 (fCO2) data products. Here, the fugacity (fCO2) is the partial pressure (pCO2) corrected for the slightly non-ideal behaviour of the gas. SOCAT was first made public in 2011 (Pfeil et al. 2013; Sabine et al. 2013). Two years later version 2, with 10.1 million fCO2 values spanning the years 1967 to 2011, was released (Bakker et al. 2014a).

SOCAT is an activity of the international marine carbon research community. A SOCAT Community Event took place on 23 June 2014 at the IMBER (Integrated Marine Biogeochemistry and Ecosystem Research) Future Oceans Open Science Conference in Bergen, Norway. Thirty seven scientists discussed SOCAT progress and strategy for the next two years (Fig. 24). The participants included SOCAT data providers, data managers, quality controllers, regional group leaders (Arctic, North Atlantic, Southern Ocean), members of the global, automation and sensor groups, and SOCAT users. IMBER and IOCCP (International Ocean Carbon Coordination Project) sponsored the meeting, and IOCCP will publish a meeting report soon. Below is an overview of topics that were discussed at the Community Event.


Figure 24. Participants of the SOCAT Community Event on 23 June 2014 in Bergen, Norway. Photo by Fang Zuo and Peter Landschützer.

Plans for automating SOCAT were initiated in 2011 and the design of a SOCAT automation system was approved in 2012. The system will be launched in early 2015 (in time for version 4), and will enable data providers to upload and check their data, using the SOCAT quality control tools, prior to data submission. Kevin O’Brien showed videos of the prototype automation system at the Community Event and the concept was well received.

Work on SOCAT version 3 is well underway. About 1850 data sets collected between 1957 and 2013 have been acquired (Fig. 25). Half of these are new data, while the remainder are updates of data sets already available through SOCAT, or that were previously suspended during quality control (QC). Version 3 data submissions include calibrated CO2 measurements from new sensors and alternative platforms. All the files have now been uploaded to the QC system. As part of the automation effort, the QC system has been slightly modified from earlier SOCAT versions (Fig. 26). For example, the data files are now subject to range checking during uploading. Experienced quality controllers will test this streamlined system before general QC opens. Regional groups will carry out quality control from late September to December 2014. The revised QC flags (Wanninkhof et al. 2013a) will be used, with comments included to justify the QC flag. Quality control is open to any marine carbon scientist (contact Dorothee Bakker). If all goes as planned, there will be a pre-release for testing purposes, and then version 3 will be released at the SOLAS (Surface Ocean Lower Atmosphere Study) Open Science Conference (Kiel, Germany) in September 2015.


Figure 25. The temporal distribution of new and updated data files acquired for SOCAT version 3 from 1957 to 2013. 


Figure 26.  SOCAT version 3 quality control tools, including the main Live Access Server screen, the correlation viewer, the cruise thumbnail viewer and the cruise quality control table.

These decisions were taken at the Community Event:

  • Only good quality, fully documented data should be submitted to SOCAT.
  • Ideally  all sea surface temperature (SST) and salinity data, submitted as part of the surface water CO2 data set, are fully calibrated and high quality.
  • SOCAT only conducts quality control for surface water fCO2. Other parameters in SOCAT data products, such as SST and salinity, are not  fully quality controlled.
  • Additional surface water parameters will be accepted by SOCAT from version 4 onwards. However, only directly measured, either discrete or continuous, parameters will be included. Any additional parameters will not be quality controlled. There will be two types of output files, one with quality controlled surface water fCO2 values, including SST and salinity, and one with non-quality controlled, ancillary data.

Acknowledgements and credits for SOCAT data providers and funding agencies are essential for sustaining surface water fCOobservations and future SOCAT releases. The SOCAT data policy asks users: "To generously acknowledge the contribution of SOCAT investigators, regional group leaders, quality controllers and data providers in the form of invitation to co-authorship, and/or reference to relevant scientific articles by data contributors". The data policy is currently being revised to state more explicitly that data providers should be consulted and considered for co-authorship, if SOCAT data products are used for regional analyses. To date, SOCAT has been cited or named by at least 49 peer-reviewed scientific publications, three book chapters and several PhD theses. The number of peer-reviewed citations or mentions of SOCAT is increasing rapidly with 20 publications in 2013, 22 in the first eight months of 2014 and a further two in press.

SOCAT surface water fCO2 data are used in a diversity of studies, with topics ranging from a comparison of newly collected data, trend analysis, impacts of ocean acidification on king scallops, and gene expression of marine micro-organisms (Table 2). Frequent applications of SOCAT data products include estimates of the shelf sea and global ocean carbon sink and its seasonal, year-to-year and decadal variation. Quantification of the ocean carbon sink is an active field of research with a number of new methods in development for interpolating surface water fCO2 in space and time. The Surface Ocean pCO2 Mapping Intercomparison (SOCOM), led by Christian Rödenbeck of the Max Planck Institute, Jena, compares at least 12 data-based mapping products of surface ocean pCO2, many of them SOCAT-based. One high profile application is the use of SOCAT-based data products for the quantification of the global ocean carbon sink for the 2013 and 2014 Global Carbon Budgets (http://www.globalcarbonproject.org/) (Le Quéré et al. 2014). SOCAT data products are also used for mapping surface ocean pH. Another important application is the validation of regional and global ocean carbon models.

Table 2. Diverse applications of SOCAT in 49 peer-reviewed, scientific publications (2009 – 1, 2010 – 1; 2011 -1; 2012 – 2; 2013 – 20; 2014 – 22; 2 in press) and three book chapters (2014 - 3). Here fCO2 also refers to pCO2 (partial pressure of CO2), as used in some publications. References will be provided on the SOCAT website (http://www.socat.info/).

The Community Event was a rewarding day with demonstrations of the automation system, open discussions, lively debate and constructive feedback for SOCAT regional and global group leaders and the automation team. SOCAT is in good health, thanks to the efforts of the enthusiastic and dedicated SOCAT data providers, data managers, quality controllers, the automation team and group leaders. The coming year will be exciting with version 3 being quality controlled and the automated data upload for version 4 going live. 


  • Bakker, D.C.E., and 80 co-authors (2014a). An update to the Surface Ocean CO2 Atlas (SOCAT version 2). Earth Syst. Sci. Data, 6: 69-90. DOI:10.5194/essd-6-69-2014.
  • Pfeil, B., and 81 co-authors (2013). A uniform, quality controlled Surface Ocean CO2 Atlas (SOCAT). Earth Syst. Sci. Data, 5: 125-143, DOI:10.5194/essd-5-125-2013. 
  • Sabine, C.L., and 75 co-authors (2013). Surface Ocean CO2 Atlas (SOCAT) gridded data products. Earth Syst. Sci. Data, 5: 145-153, DOI:10.5194/essd-5-145-2013. 
  • Wanninkhof, R. et al. (2013a). Incorporation of alternative sensors in the SOCAT database and adjustments to dataset quality control flags. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, US Department of Energy, Oak Ridge, Tennessee. DOI:10.3334/CDIAC/OTG.SOCAT_ADQCF.
banner transition NL


  • Benway H. M., Hofmann E. and St. John M. (2014). Building international research partnerships in the North Atlantic-Arctic System: An international planning workshop for a North Atlantic-Arctic science program. Eos Trans. AGU 95(35): 317. DOI: 10.1002/2014EO350007. Article
  • Brander L.M., Narita D., Rehdanz K & Tol R.S. (2014). The economic impacts of ocean acidification, In: Nunes, P.A.L.D., Kumar, P. & Dedeurwaerdere, T. (Eds.), Handbook on the economics of ecosystem services and biodiversity, pp. 78-92. Cheltenham: Edward Elgar Publishing Limited Article
  • Cripps G., Lindeque P. & Flynn K. (in press). Parental exposure to elevated pCO2 influences the reproductive success of copepods. Journal of Plankton ResearchArticle 
  • Cossarini, G., Lazzari, P. & Solidoro, C. (2014). Space-time variability of alkalinity in the Mediterranean Sea, Biogeosciences Discuss. 11: 12871-12893. doi:10.5194/bgd-11-12871-2014. Article
  • Endres S., Galgani L., Riebesell U., Schulz K.-G. & Engel A. (2014). Stimulated bacterial growth under elevated pCO2: results from an off-shore mesocosm study. PLoS ONE 9(6): e99228. Article
  • Feely R.A., Talley L.D., Bullister  J.L., Carlson C.A., Doney S.C., Fine R.A., Firing E., Gruber N., Hansell D.A., Johnson G.C., Key R.M., Langdon C., Macdonald A., Mathis J.T., Mecking S., Millero F.J., Mordy C.W., Sabine C.L., Smethie W.M., Swift J.H., Thurnherr A M., Wanninkhof R.H. & Warner M. J. (2014). The US Repeat Hydrography CO2/Tracer Program (GO-SHIP): Accomplishments from the first decadal survey. A US CLIVAR and OCB Report, 2014-5, US CLIVAR Project Office, 47 pp. Report
  • Meier K.J.S., Beaufort L., Heussner S. & Ziveri P. (2014). The role of ocean acidification in Emiliania huxleyi coccolith thinning in the Mediterranean Sea. Biogeosciences 11: 2857-2869. Article
  • Orr J.C., Epitalon J.-M. and Gattuso J.-P. (2014). Interactive comment on “Comparison of seven packages that compute ocean carbonate chemistry” by J.C. Orr et al. Biogeosciences Discuss. 11: C4420–C4431. Article
  • Pope E.C., Ellis R.P., Scolamacchia M., Scolding J.W.S., Keay A., Chingombe P., Shields R. J., Wilcox R., Speirs D.C., Wilson R.W., Lewis C. & Flynn K.J. (2014). European sea bass, Dicentrarchus labrax, in a changing ocean. Biogeosciences 11:2519-2530. Article
  • Roy T., Lombard F., Bopp L. & Gehlen M. (2014). Projected impacts of climate change and ocean acidification on the global biogeography of planktonic foraminifera. Biogeosciences Discussions 11: 10083-10121. Article
  • Ziveri P., Passaro M., Incarbona A., Milazzo M., Rodolfo-Metalpa R. and Hall-Spencer J. M. (2014). Decline in coccolithophore diversity and impact on coccolith morphogenesis along a natural CO2 gradient. Biol. Bull. 226 (3): 282-290. Article
  • Zuo F., Hu L. and Zhang J. (2014). Capacity Building Assessment for Integrated Marine Biogeochemistry and Ecosystem Research in the Asia-Pacific Region. APN Science Bulletin 4: 35-39. Article
banner transition NL

Meeting Calendar



banner transition NL