In this new Analysis article published by a team of Chilean and European colleagues, we provides a new approach to reconcile apparent contradicting results on the impact of ocean acidification on marine organisms.
Ocean acidification is the global warming evil twin. It is another consequence of human carbon dioxide (CO2) emissions. A quarter of the CO2 we have emitted to the atmosphere is captured by the ocean and significantly modifying its chemistry, including a decrease in the seawater pH. Since the industrial revolution, the oceanic pH has dropped from 8.2 to 8.1, which does not seem like much, but if we even consider that the pH scale is logarithmic, this apparently small change means that the ocean is becoming 30% more acidic. Further, recent studies suggest that acidification rate could accelerate from now to the end of the century, resulting in a potentially catastrophic impact on marine ecosystems. Understanding the impact of ocean acidification is now a priority and one of the targets of then UN Sustainable Development Goals for 2030.
During the last years, the scientific community has been really active, and scientists have been conducting many ocean acidification experiments, where a variety of marine species are usually exposed to pH or CO2 levels expected for the near future. Through this experimental approach, it has been achieved an unequivocal consensus that if we keep emitting CO2 at the same rate, ocean acidification will significantly impact marine species, ecosystems and services that we are depending on. However, they struggle with apparently contradicting results.
We have been studying the impact of ocean acidification for many years but I was puzzled by the fact that making the same experiment on two different populations of the same animal could lead to very different results.
To solve this issue, I’ve gathered a team of scientists from different institutions in Chile (Universidad Santo Tomás, Universidad Adolfo Ibáñez, Universidad Andrés Bello, Universidad de Antofagasta, and the Center for Advanced Studies in Arid Zones), and European institutions in England (Plymouth Marine Laboratory, PML) and Sweden (Göteborg University). The idea was to think out of the box.
At the time, the best practices in the field of ocean acidification recommended to use the same concentration of CO2 for all our experiments, based on potential scenarios for the open ocean in 50 or 100 years more. But it was neglecting the fact that animals living in different coastal regions experience different environments and CO2 levels and seems to be adapted to these conditions.
Many times we discussed with my colleague and friend, Sam Dupont (University of Göteborg) and he agreed that we were comparing pears and apples. Sam did the analogy that by testing the same CO2 all over the world would be the same as keeping Grizzly and Polar bears at the same temperature and expect them to respond in the same way.
The bad news, is that by using potential scenarios for the open ocean, a large set of published experiments conducted with coastal marine species might have significantly underestimated the impact of ocean acidification on marine organisms.
In this new Analysis Article published in Nature Ecology and Evolution, the natural CO2environment from all the tested species in a wide latitudinal range along the Chilean coast was measured and took into account to calculate an index aimed to estimate how much the experimental treatment deviates from these condition. This index is able to reconcile the apparent contradicting results.
In a way, it is quite straightforward: the more they are deviating from what they know, the more stressed and negatively impacted animals are.
This new approach will allow using the growing observation data on chemical changes (pH or pCO2) measured by deployed instruments in the ocean to infer on potential biological impacts. This will allow working toward a better management of marine resources and be better prepared for the future changes to come.
The paper in Nature Ecology & Evolution is here.
Cristian A. Vargas, Nature Ecology and Evolution Community, 15 March 2017. Article.
The paper by Vargas et al. (2017) is an important contribution to the field of ocean acidification research. I would like to correct a wrong information that I have heard several times and am now reading in this paper. I hope that the following will help stop propagating a recurring error.
First, an incorrect reference is provided concerning the guidelines the paper refers to (reference 10). It should actually be:
Barry J. P., Tyrrell T., Hansson L. & Gattuso J.-P., 2010. Atmospheric CO2 targets for ocean acidification perturbation experiments. In: Riebesell U., Fabry V. J., Hansson L. & Gattuso J.-P. (Eds.), Guide to best practices for ocean acidification research and data reporting, pp. 53-66. Luxembourg: Publications Office of the European Union.
Second, it is often mentioned, also in Vargas et al. (2017), that these guidelines recommend using the same future levels of seawater pCO2 all over the world to facilitate comparison between studies. This is incorrect. As the title of this article mentions, it addresses **atmospheric** CO2 targets. It argues that “Comparison of results among ocean acidification studies will be easier by using common atmospheric CO2 targets, even though ocean carbonate chemistry parameters may differ”.
Furthermore:
– The paper by Vargas et al. (2017) overlooks the fact that the guidelines do highlight the temporal and spatial variability of the carbonate system: “Unlike atmospheric p(CO2), which is relatively homogeneous over the Earth, aqueous p(CO2) and other ocean carbonate system parameters can vary greatly over space and time…”.
– The guidelines note that “Key atmospheric p(CO2) values can be defined and used as guidelines, but their corresponding values for ocean carbonate system parameters are the primary measurements for ocean acidification experiments, and should also be reported.”
– Then the authors of the guidelines ask the question “How can investigators convert key atmospheric p(CO2) values to the in situ p(CO2), pH, Ωa, Ωc, or other carbonate system parameters of interest for specific ocean acidification experiments?”. Ways to do that are proposed, specifically addressing coastal zones, inland seas, oxygen minimum zones, and deep-sea environments.
The comments above do not diminish the merit of the authors. The use of “∆pCO2 exposure” as a descriptor is clever and will help analyze the large amount of data on the biological response to ocean acidification collected in recent years.
Nevertheless, it would be nice if the mistakes above as well as the incorrect description of the literature could be corrected in print.
We agree that our article could have been clearer in stating that Barry et al. (2010) targets referred to atmospheric CO2 and not oceanic CO2. Unfortunately it is not possible to modify our manuscript at this stage. However, your message does give us an opportunity to clarify a few additional points. As you know, several of us contributed to various chapters in the “Guide to best practices for ocean acidification research and data reporting”. This has been and still is a fantastic tool for the ocean acidification community. However, we believe that chapter 3 on “Atmospheric CO2 targets for ocean acidification perturbation experiments” (Barry et al. 2010) would benefit from some clarification.
This chapter was carefully written and is fundamentally correct in what it says but sometimes the devil hides in the detail. Given what we have learnt since the original publication of this chapter, particularly concerning the complexities of carbonate chemistry in coastal environments, we would argue that some readers are misinterpreting and poorly implementing the recommendations of the chapter leading to sub-optimal experimental designs. This is evidenced by Barry et al. (2010) having often been cited as a reference to justify the use of atmospheric CO2 values to define the seawater pH treatments in perturbation experiments, completely neglecting the tested species niche and factors that significantly contribute to the natural variability in seawater pH. Of course, we are not suggesting that the authors of Chapter 3 are responsible for these misunderstandings but it does highlight the need to provide clearer guidance for defining both appropriate experimental controls and future scenarios, as well as for determining what is actually driving biological response to ocean acidification. It is only to be expected that as our understanding of particular fields advances we will be able to provide more specific details in our guidance.
Barry et al. (2010) recommends experimentalists to use “common atmospheric CO2 targets” (summarized in Table 3.3) clearly stating that this could lead to different local carbonate system parameters. It acknowledges that the conversion of atmospheric CO2 to corresponding in situ ocean chemistry can be a challenge and several parameters influencing the local carbonate system are mentioned (e.g. physico-chemical parameters such as temperature and salinity, photosynthesis, respiration, mixing, etc.).
In particular, section 3.2.2 provides some suggestions on how to convert atmospheric CO2 levels into in situ local chemistry. These include using “published predictions of future ocean carbonate system values” (often missing for coastal regions) or “assume that surface water is in equilibrium with the atmosphere and use software such as co2sys”. This later approach has been wrongly and extensively used by the scientific community to define experimental scenarios but unfortunately researchers have often neglected the modulating role of biology and oceanographic dynamics critical in coastal zones, despite warnings provided in the chapter (making the case for deep sea waters).
The chapter goes on to discuss the number of treatments in ocean acidification experiments and suggests some target atmospheric values (Table 3.3), in particular when “logistical considerations limit the number of treatment levels and replication”. It is then suggested to “compare one or more treatments simulating future atmospheric CO2 levels to a baseline control treatment. (…) Guidelines presented here (Table 3.3) are based on the number of treatment levels that can be supported both technically and financially.” Barry et al. (2010) refer to atmospheric CO2 targets but it may give the wrong impression that one atmospheric CO2 target corresponds to one seawater CO2 target (one control or “present day” scenario as often presented in the literature). However, a marine organism is often experiencing a wide range of pH/pCO2 through its life due to natural variation in the environment, migration and manipulation of its environment. Consequently, present day atmospheric CO2 levels can generate a wide range of “normal” or “control” conditions that are often not adequately replicated when organisms are transferred to laboratory conditions. This has led to inappropriate assumptions of whether organisms are actually being exposed to stress or not.
Vargas et al. (2017) highlights the key role of this variability in shaping species sensitivity to ocean acidification through selection, adaptation and acclimation. We demonstrate that the more a CO2 treatment deviates from the present range of natural variability, the more negatively an organism is impacted. This results have strong implications for re-evaluation of the literature, since there is a significant percentage of published studies where organisms have been exposed to pH/pCO2 values within the same range or even lower than found in their respective habitat; and therefore they do not correspond to ocean acidification scenarios. In consequence, this is also relevant for future experimental design by taking into account this present, and future, natural variability (not only the average pH at a given location, as recommended as an indicator for the SDG 14.3).
Including the natural variability into an experimental design is not an easy task. Bubbling with CO2 at the right atmospheric target is not enough since it is extremely difficult to recreate the whole complexity of a given location (including ecology, oceanography, etc.) in the laboratory. Few technologies are currently available which allow researchers to simulate pH/pCO2 variability in the laboratory so most experiments are still performed with constant pH/pCO2. In that context, it is critical to consider what are the appropriate levels to be used as experiment controls (targets within the range of natural variability) and what is a future scenario (targets outside the range of natural variability). As demonstrated by Vargas et al. (2017), this requires a proper monitoring of the environment experienced by the tested species or ecosystem and can not simply be deduced from current or predicted atmospheric CO2 values
We believe that chapter 3 of the Best Practice Guide could be improved to make these points much more clearly and thus avoid misinterpretation and the use of inappropriate controls and scenarios:
– One atmospheric CO2 corresponds to a wide range of seawater CO2 and pH levels. As a consequence, there are a wide range of controls that should be considered.
– Converting atmospheric CO2 into the natural variability at a given location is extremely challenging in coastal zones where physico-chemistry, biology and oceanography play a key role. Putting Barry et al. (2017) into practice is then often very challenging or even utopic. An easier strategy is to monitor the present variability at the right spatio-temporal scale relevant for an organism (capturing the high “weather” spatio-temporal resolution) to define targets that are within and outside the natural range of variability. Many recently published articles are demonstrating that different processes are involved if you expose an organism to conditions that are deviating from their present range of natural variability.
We hope that Vargas et al. (2017) will contribute to a better understanding of what is driving biological response to ocean acidification and improve future experimental designs. We would also be more than happy to contribute to a new version of Barry et al. (2010), or even a new product that we could make available to the scientific community.
Cristian A. Vargas, Nelson A. Lagos, Marco A. Lardies, Cristian Duarte, Patricio H. Manríquez, Victor M. Aguilera, Bernardo Broitman, Steve Widdicombe and Sam Dupont
I am glad that you do not disagree with any of my comments.
I am not sure why obvious mistakes, such as the wrong reference to the citation to the chapter by Barry et al. (2011), could not be corrected in print. Surely, Nature Ecology and Evolution would be happy to print an erratum rather than leaving an unidentified error which leads the readership to a wrong paper, adding further confusion.
What could be done in the future?
Even though Barry et al. (2011) clearly and repeatedly mentions that they address **atmospheric** CO2 targets, the SOLAS-IMBER Working Group on Ocean Acidification (SIOA) noticed that several papers, like yours, made the mistake of interpreting the targets as **seawater** pCO2. As early as 2013, SIOA identified this chapter as a candidate for a revision to avoid future mistakes. The Advisory Board of the Ocean Acidification International Coordination Centre agreed but, unfortunately, despite repeated efforts, Ulf Riebesell and I were unable to generate interest and no one took the opportunity to revise the chapter. As a member of the Advisory Board, your coauthor Sam Dupont is familiar with the issue and could provide more background.
We do not disagree with your comments.
Unfortunately, a chapter that gives the wrong impression that one atmospheric CO2 target corresponds to one seawater CO2 target, leading to mistakes not only for us, but probably to authors of more than 50-70% of the literature that exists, and that forms part of OA databases, can not be perfectly elaborated, , so maybe it’s not only mistake of researchers, since again… it seems a common error in literature, not only exceptional cases.
In consequence, this induce confussion for a large scientific community, such as that today constitutes the Ocean Acidification community, but even worst has significant scientific implications. A significant fraction of studies included in OA data bases have probably explored phenotypic plasticity or the environmentally-induced variation in selected phenotypic traits under natural pH/pCO2 variability, and not their responses under the extreme scenarios posed by OA in coastal ocean in the future.
Now, Vargas et al. (2017), clearly demonstrate that we requires a proper monitoring of the environment experienced by the tested species or ecosystem and we can not simply deduce it from current or predicted atmospheric CO2 values
I offer myself, and probably some of my co-authors, to improve this chapter so that it is more useful to the community. I think this point deserves more urgency than even fixing a minor mistake in the literature referenced in a paper in Nature.
Let us know !
Cristian
I thoroughly disagree with your statement that authors of more than 50-70% of the literature that exists were misled with the Guide to best Practices. First, this range is not based on a serious analysis of the literature. I am convinced such analysis would show that this range is a gross exaggeration. Second, to my knowledge, no paper misrepresented the guide and its content as is done in the second paragraph (and especially its first sentence) of your paper.
In any case, Ulf Riebesell and I believe that the misrepresentation of the guide must be corrected. We will contact you offline about that.
As mentioned before, a revised version of this chapter would be most welcome. Ulf and I were unsuccessful getting it done. Note that Kristy Kroeker has taken over from us. Your coauthors Nelson Lagos and Sam Dupont are familiar with this and could provide more information.